Dif Lecturenotes

Di↵erential Equations - Lecture Notes
Maksat Ashyraliyev
Associate Professor, Dr.
MAT2062
Bahcesehir University
Differansiyel denklemin çözümü fonks yon veya fonks

kümes d r
Diff d klemler genel olarak k gruba ayırab l r z
Ord nary D ff Eq CODE

2 Part al D ff
B r d ff denklemde tek b r tane bağımsız

değ şken varsa bu ODE olur
EÜİİĞL5y O
B rden fazla
1
bağımsız değ şken varsa Part al
yaya t
Iff_k 2
l
1 Introduction, Elementary Concepts

The laws of the universe are written in the language of mathematics.
Algebraic equations are sufficient to describe many static problems,
but the most interesting natural phenomena involve change and are
described by equations that relate changing quantities.
dy
Since the derivative = f 0 (t) of the function f is the rate at
dt
which the quantity y = f (t) is changing with respect to the indepen-
dent variable t, it is natural that equations involving derivatives are
frequently used to describe the changing universe.
Definition 1.1. An equation involving at least one derivative of an

unknown function is called a di↵erential equation.
Example 1.1. The following equations are some examples of di↵eren-

tial equations:
dy d2 y
a) = cos x b) 2
+ k2y = 0
dx dx
✓ 2 2
◆ ✓ ◆3
@u @ u @ u d2 y dy
c) = h2 + d) + 8y = 0
@t @x2 @y 2 dx2 dx
Definition 1.2. An equation involving derivatives with respect to a

single independent variable is called an Ordinary Di↵erential Equa-
tion (ODE). In general, if y is a function of x only and its derivatives
2
dy 0 d y 00 dn y
are =y, 2
= y , ..., n
= y (n) , then
dx dx dx
⇣ ⌘
0 00 (n)
F x, y, y , y , . . . , y =0
is an ODE.
2
Lecture Notes - Di↵erential Equations M.Ashyraliyev
Definition 1.3. An equation involving partial derivatives with respect

to two or more independent variables is called a Partial Di↵erential
Equation (PDE).
Remark 1.1. First, second and fourth equations in Example 1.1 are
examples of ODEs. Third equation in Example 1.1 is the PDE.
Definition 1.4. The order of an Ordinary Di↵erential Equation is

the order of the highest derivative that appears in it.
Example 1.2.
a) 3y 0 + xy 2 = 0 is a first order ODE.
b) xy 00 + 5y = 0 is a second order ODE.
Remark 1.2. Note that the most general form of an n-th order ODE
is ⇣ ⌘
0 00 (n)
F x, y, y , y , . . . , y = 0. (1.1)
Definition 1.5. An n-th order ODE (1.1) is called linear if F is a

linear function of variables y, y 0 , y 00 , . . . , y (n) . Hence the general form
of a linear n-th order ODE is
a0 (x)y + a1 (x)y 0 + a2 (x)y 00 + . . . + an (x)y (n) = b(x). (1.2)
An n-th order ODE (1.1) is called nonlinear if it is not linear, i.e. it
cannot be written in the form (1.2).
Example 1.3.
a) 3y 0 + xy 2 = 0 is a nonlinear first order ODE.
b) x2 y 00 + 5y = 0 is a linear second order ODE.
c) xy 000 + x3 y 0 + 2y = 0 is a linear third order ODE.
d) y 0 + y 3 = 0 is a nonlinear first order ODE.
3
Definition 1.6. A continuous function u = u(x) is called a solution

of ordinary di↵erential equation on an interval I, if the derivatives
of u up to the order of ODE exist on I and u satisfies the ordinary
di↵erential equation on I.
Example 1.4. Verify that y = e2x is a solution of the equation

d2 y dy
+ 6y = 0.
dx2 dx
Solution: Note that function y = e2x is continuous. First and second

derivatives of y = e2x exist for all values of x. We substitute this
function into the equation and find that for all values of x
d2 y dy
+ 6y = 4e2x + 2e2x 6e2x ⌘ 0.
dx2 dx
Thus, y = e2x is indeed a solution of given equation.
dy
Definition 1.7. A first order di↵erential equation = f (x, y) to-
dx
gether with an initial condition y(x0 ) = y0 is called an initial value
problem (IVP). To solve the initial value problem
8
< dy = f (x, y),
>
dx (1.3)
>
: y(x ) = y
0 0
means to find a di↵erentiable function y = y(x) that satisfies both

conditions in (1.3) on some interval containing x0 .
Example 1.5. First verify that y(x) = Ce x is a solution of the equa-

tion y 0 + y = 0 for any value of constant C. Then determine the value
of the constant C so that y(x) = Ce x satisfies the initial condition
y(0) = 2.
4
x
Solution: Substituting y(x) = Ce into the given equation, we have
y0 + y = Ce x
+ Ce x
⌘ 0.
Thus, y(x) = Ce x is a solution of given equation for any value of

constant C. Now, using the initial condition y(0) = 2, we get C = 2.
So, y(x) = 2e x is a solution of the IVP
( 0
y + y = 0,
y(0) = 2.
1.1 Self-study Problems

Problem 1.1. For each of the following, state whether the equation
is ODE or PDE. If it is ODE then give its order and state whether it
is linear or nonlinear.
d2 x @ 2w 2
2@ w
a) 2 + k 2 x = 0 b) =a
dt @t2 @x2
0 x2 + y 2
c) y = d) y 0 + P (x)y = Q(x)
2xy
e) y 000 3y 0 + 2y = 0 f) yy 00 = x
✓ 3 ◆2 ✓ ◆4
@ 2u @ 2u @ 2u dw dw
g) + + =0 h) 2 + xw = 0
@x2 @y 2 @z 2 dx3 dx
3 4
i) x (y 00 ) + (y 0 ) y=0 k) y 00 + 2y 0 8y = x2 + cos x
Problem 1.2. In the following examples, verify by substitution that

each given function is a solution of the given di↵erential equation.
a) y 0 = 3x2 ; y = x3 + 7
b) y 0 + 2y = 0; y = 3e 2x
5
c) y 00 + 4y = 0; y1 = cos 2x, y2 = sin 2x
d) y 00 2y 0 + 2y = 0; y1 = ex cos x, y2 = ex sin x
1
e) y 0 + 2xy 2 = 0; y=
1 + x2
1
f) x2 y 00 + xy 0 y = ln x; y1 = x ln x, y2 = ln x
x
g) x2 y 00 xy 0 + 2y = 0; y1 = x cos (ln x), y2 = x sin (ln x)
Problem 1.3. In the following examples, substitute y = erx into the

given di↵erential equation to determine all values of the constant r for
which y = erx is a solution of the equation.
a) 3y 0 = 2y b) 4y 00 = y
c) y 00 + y 0 2y = 0 d) 3y 00 + 3y 0 4y = 0
Problem 1.4. In the following examples, first verify that y(x) satisfies
the given di↵erential equation. Then determine a value of the constant
C so that y(x) satisfies the given initial condition.
a) y 0 = y + 1; y(x) = Cex 1, y(0) = 5

x5 C
b) xy 0 + 3y = 2x5 ; y(x) = + 3, y(2) = 1
4 x
c) xy 0 3y = x3 ; y(x) = x3 (C + ln x) , y(1) = 17
d) y 0 = 3x2 y 2 + 1 ; y(x) = tan (x3 + C), y(0) = 1
e) y 0 + y tan x = cos x; y(x) = (x + C) cos x, y(⇡) = 0
6
2 Integrals as General and Particular So-
lutions
dy
The first order di↵erential equation = f (x, y) takes an especially
dx
simple form if the function f is independent of variable y, i.e. it has
the form:
dy
= f (x). (2.1)
dx
Theorem 2.1. The general solution of the first order di↵erential equa-
tion (2.1) has the form:
Z
y(x) = f (x)dx = F (x) + C, (2.2)
where F (x) is an antiderivative of f (x) and C is an arbitrary constant.
Remark 2.1. (2.2) is a general solution of di↵erential equation (2.1),

meaning that it involves an arbitrary constant C. Hence, (2.2) is a
solution of equation (2.1) for every value of C. If we take particular
values of constant C, we obtain so-called particular solutions of
di↵erential equation (2.1).
Example 2.1. Find the general solution of the first order di↵erential
equation:
dy
= 2x + 3.
dx
Solution: Integrating both sides of the equation, we get the general
solution Z
y(x) = (2x + 3)dx = x2 + 3x + C.
Note that y1 (x) = x2 + 3x + 5 and y2 (x) = x2 + 3x 1 are examples of

particular solutions of given di↵erential equation.
7
Remark 2.2. The graph of general general solution (2.2) contains

infinitely many graphs which are ”parallel” to each other. If we take a
particular graph which corresponds to a particular value of the constant
C, we obtain the graph of the solution of the initial value problem:
8
< dy = f (x),
dx
:
y(x0 ) = y0 .
Example 2.2. Solve the initial value problem:

8
< dy = 2x + 1,
dx
ya 2tl dx
:
y(0) = 3.
Mx t C
dy
Solution: Integrating both sides of the equation = 2x + 1, we get
dx
the general solution
Z C 3
y(x) = (2x + 1)dx = x2 + x + C.
Atxt 3
Solution of given initial value problem is a particular solution that
satisfies the initial condition y(0) = 02 + 0 + C = 3. It follows that
C = 3 and the desired particular solution is:
y(x) = x2 + x + 3.
Remark 2.3. (Second order equations)

Result of Theorem 2.1 for first order di↵erential equation (2.1) can be
extended to second order di↵erential equation of the special form:
d2 y
= g(x). (2.3)
dx2
Integrating both sides of equation, we get
Z
dy
= g(x)dx = G(x) + C1 ,
dx
8
where G(x) is an antiderivative of g(x) and C1 is an arbitrary constant.

Then another integration yields
Z Z
y(x) = [G(x) + C1 ] dx = G(x)dx + C1 x + C2 , (2.4)
where C2 is a second arbitrary constant. (2.4) is a general solution of

second order equation (2.3).
Example 2.3. Find the general solution of the second order di↵erential
equation:
Sat
d2 y
dx2
= 2x + 5. Cy
Solution: Integrating both sides of the equation, we obtain
dy
Z ENG
= (2x + 5)dx = x2 + 5x + C1 .
dx Cz
Then another integration gives us the general solution of given equation
Z
⇥ 2 ⇤ x3 5x2
y(x) = x + 5x + C1 dx = + + C1 x + C2 ,
3 2
where C1 and C2 are arbitrary constants.

8 2
>
> dy
< dx2 = 2x + 5,
>
>
> y(0) = 1,
>
: 0
y (0) = 2.
Solution: In the Example 2.3 we have obtained the general solution

d2 y
of equation 2 = 2x + 5 as following:
dx
x3 5x2
y(x) = + + C1 x + C2 .
3 2
9
Now, by using initial conditions y(0) = 1 and y 0 (0) = 2, we find the

values of constants C1 and C2 . Namely,
y(0) = C2 = 1 and y 0 (0) = C1 = 2.
So the solution of given IVP is
x3 5x2
y(x) = + + 2x + 1.
3 2
Remark 2.4. (Velocity and Acceleration)

Direct integration method illustrated for the second order equation
(2.3) allows us to solve a number of important problems concerning
the motion of a particle. The motion of a particle along a straight line
(x-axis) is described by the position function
x = f (t)
giving its x-coordinate at time t. The velocity of the particle is defined
to be
dx
v(t) = f 0 (t) or v= .
dt
The acceleration of the particle is defined to be
0 d2 x
a(t) = v (t) or a= 2.
dt
Example 2.5. Find the position function x(t) of a moving particle

with given acceleration a(t) = 50, initial position x(0) = 20 and initial
velocity v(0) = 10.
Solution: Position function x(t) satisfying given conditions is a solu-
tion of the following initial value problem:
8 2
>
> dx
< dt2 = 50,
>
>
> x(0) = 20,
>
: 0
x (0) = 10.
10
Integrating both sides of the equation, we obtain

Z
dx
= 50dt = 50t + C1 .
dt
Then another integration gives us
Z
x(t) = [50t + C1 ] dt = 25t2 + C1 t + C2 .
Now, by using initial conditions x(0) = 20 and x0 (0) = 10, we find the
values of constants C1 and C2 . Namely,
x(0) = C2 = 20 and x0 (0) = C1 = 10.
So, the solution of the given problem is
x(t) = 25t2 + 10t + 20.

Problem 2.1. Solve the following initial value problems
8 8
< dy
= (x 2) , 2 < dy = px,
a) dx b) dx
: :
y(2) = 1. y(4) = 0.
8 8
< dy = 1 , < dy = p 1 ,
>
c) dx x2 d) dx x+2
: >
: y(2) = 1.
y(1) = 5.
8 p 8
< dy = x x2 + 9, < dy = 10 ,
e) dx f) dx x2 + 1
: :
y( 4) = 0. y(0) = 0.
11
8 8
< dy = cos 2x, < dy = p 1
>
,
g) dx h) dx 1 x2
: >
: y(0) = 0.
y(0) = 1.
8 8
< dy = xe x , < dy = x ln x,
i) dx k) dx
: :
y(0) = 1. y(1) = 1.
Problem 2.2. In the following examples, find the position function

x(t) of a moving particle with given acceleration a(t), initial position
x(0) = x0 and initial velocity v(0) = v0 .
a) a(t) = 20, x0 = 5, v0 = 15
b) a(t) = 3t, x0 = 0, v0 = 5
c) a(t) = 2t + 1, x0 = 4, v0 = 7
d) a(t) = 4(t + 3)2 , x0 = 1, v0 = 1
1
e) a(t) = p , x0 = 1, v0 = 1
t+4
1
f) a(t) = , x0 = 0, v0 = 0
(t + 1)3
g) a(t) = 50 sin 5t, x0 = 8, v0 = 10
Problem 2.3. A diesel car gradually speeds up so that for the first 10
seconds its acceleration is given by
3t2 3t
a(t) = + (f t/sec2 )
25 5
If the car starts from rest (x0 = 0, v0 = 0), find the distance it has
traveled in 10 seconds.
12
3 Separable Equations
Definition 3.1. The first order di↵erential equation
dy
= f (x, y)
dx
is called separable if function f (x, y) can be written as the product
(or ratio) of a function of x and a function of y. So, separable equation
has the form:
dy
= g(x)h(y). (3.1)
dx
Theorem 3.1. For the solution of separable equation (3.1) we have:
Z Z
dy
= g(x)dx. (3.2)
h(y)
Therefore, the solution procedure for separable equation (3.1) consists

of the following two informal steps:
(i) Equation (3.1) can be rewritten as

dy
= g(x)dx (Separation of variables)
h(y)
(ii) Integrating both sides, we get

Z Z
dy
= g(x)dx (Integration step)
h(y)
Example 3.1. Solve the following initial value problem:

8
< dy = 6xy,
dx
:
y(0) = 7.
13
dy
Solution: Obviously, first order di↵erential equation = 6xy is
dx
separable. Separating the variables, we obtain
dy
= 6xdx.
y
Next, integrating both sides gives us
Z Z
dy 3x2
= ( 6x)dx =) ln |y| = 3x2 + ln C =) y = Ce .
y
Now, using the initial condition y(0) = 7, we obtain C = 7. So, the
solution of given IVP is
3x2
y(x) = 7e .
Example 3.2. Solve the following di↵erential equation:

dy 4 2x
= 2 .
dx 3y 5
Solution: It is easy to see that given equation is separable. Separating

the variables, we obtain
3y 2 5 dy = (4 2x) dx.

Z Z
2
3y 5 dy = (4 2x) dx =) y3 5y = 4x x2 + C.
Note that from obtained relation it is not easy to express y as an

explicit function of x. We call y 3 5y = 4x x2 + C an implicit
solution of given di↵erential equation.
14

8
< dy = 4 2x ,
>
dx 3y 2 5
>
: y(1) = 3.
Solution: In the previous example we have already shown that di↵er-

dy 4 2x
ential equation = 2 has the general solution
dx 3y 5
y3 5y = 4x x2 + C.
Then, by using initial condition y(1) = 3 we can find the value of
constant C. If we substitute x = 1 and y = 3, we obtain C = 9. So,
the solution of given IVP is
y3 5y = 4x x2 + 9.
Remark 3.1. In the step of separating the variables, we multiply or

divide both sides of equation (3.1) by some algebraic factors. Solutions
of a given di↵erential equation can be either gained or lost in this
process. The following example illustrates this point.
Example 3.4. Find all solutions of the di↵erential equation:

dy
= 6x(y 1)2/3 .
dx
Solution: Separation of variables gives
dy
= 6xdx.
(y 1)2/3
Integrating both sides, we obtain
Z Z
dy
= 6xdx =) 3(y 1)1/3 = 3x2 + 3C =)
(y 1)2/3
3
=) (y 1)1/3 = x2 + C =) y(x) = 1 + x2 + C .
15
3
Obtained solution y(x) = 1 + x2 + C is a general solution of given
equation, meaning that it is a solution of equation for every value of C.
Any particular choice of a specific value for C yields a single particular
solution of the equation. Now, note that there is a particular solution
of given equation, y(x) = 1, which cannot be obtained from general
solution by selecting a value for C. This particular solution is called
a singular solution of given equation. This solution was lost at
the process of separation of variables when we divided both sides of
given equation by factor (y 1)2/3 . In conclusion, given equation has a
3
general solution y(x) = 1 + x2 + C and a singular solution y(x) = 1.
Example 3.5. Find the general solution of the di↵erential equation:

dy
y = (y 2 + 1) tan x.
dx
Solution: Separation of variables gives
y
dy = tan xdx.
y2 + 1
Integrating both sides, we obtain
Z Z
y 1 1
dy = tan xdx =) ln (y 2 + 1) = ln | cos x| + ln C
y2 + 1 2 2
✓ ◆
2 C C
=) ln (y + 1) = ln =) y2 + 1 = .
cos2 x cos2 x

Problem 3.1. Find general solutions (implicit or explicit) and singular
solutions (if exist) of the following di↵erential equations:
dy dy
a) + 2xy = 0 b) + 2xy 2 = 0
dx dx
dy dy
c) = y sin x d) (1 + x) = 4y
dx dx
16
p dy p dy p
e) 2 x = 1 y2 f) = 3 xy
dx dx
dy p dy
g) = 4 3 xy h) = 2x sec y
dx dx
dy dy
i) 1 x2 = 2y k) (1 + x)2 = (1 + y)2
dx dx
l) yy 0 = x y 2 + 1 m) y 0 = xy 3
p
dy dy 1+ x
n) y 3 = y 4 + 1 cos x o) = p
dx dx 1+ y
dy (x 1)y 5
p) = q) x2 + 1 (tan y)y 0 = x
dx x2 (2y 3 y)
r) y 0 = 1 + x + y + xy s) x2 y 0 = 1 x2 + y 2 x2 y 2
y
0 xy 0 e
t) y = p u) yy =
x2 16 x3 + x
Problem 3.2. Find explicit particular solutions of the following initial

value problems:
8 8
< dy = yex , < dy = 3x2 y 2 + 1 ,
a) dx b) dx
: :
y(0) = 2e. y(0) = 1.
8 8
< 2y dy = p x
>
, < dy = 4x3 y y,
c) dx x 2 16 d) dx
>
: y(5) = 2. :
y(1) = 3.
17
8 8
> dy
< dy + 1 = 2y, < (tan x) = y,
e) dx f) ⇣ ⌘ dx
:
y(1) = 1. : y ⇡ = ⇡.
>
2 2
8 8
< x dy y = 2x2 y, < dy = 2xy 2 + 3x2 y 2 ,
g) dx h) dx
: :
y(1) = 1. y(1) = 1.
8 8
> p dy
< dy = 6e2x y , <2 x = cos2 y,
i) dx k) dx
:
y(0) = 0. : y(4) = ⇡ .
>
4
8 8
> dy 2xe x2 > dy
< = , < tan x = y 2 4,
l) dx cos y m) ⇣ ⇡ ⌘dx
>
: >
:y
y(0) = 0. = 1.
2
18
4 Existence and Uniqueness of Solutions
Before one spends much time attempting to solve a given initial value
problem 8
< dy = f (x, y),
dx (4.1)
:
y(a) = b
it is wise to know whether solutions of this problem actually exist. In
other words, do all initial value problems of type (4.1) have solutions?
The following example illustrates that some initial value problems do
not have solutions!
Example 4.1. (Failure of Existence)

Consider the following initial value problem:
8 2
< dy = y ,
dx x (4.2)
:
y(0) = 1.
y2
Separating the variables in the equation y 0 = , we get
x
Z Z
dy dx 1 1
= =) = ln |x| + C =) y(x) = .
y2 x y ln |x| + C
Solution of given initial value problem, if it exists, should satisfy the
1
initial condition y(0) = 1. However, solutions y(x) = are
ln |x| + C
not defined at x = 0. Therefore, given initial value problem does not
have any solutions.
Another important question is the question of uniqueness of the so-

lution of IVP. In other words, if the initial value problems of type (4.1)
has a solution, is it always unique? The following example illustrates
that some initial value problems have more than one solution!
19
Example 4.2. (Failure of Uniqueness)

Consider the following initial value problem:
8
< dy = 3y 2/3 ,
dx (4.3)
:
y(0) = 0.
It is easy to verify (do it yourself) that this IVP has two di↵erent
solutions y1 (x) = x3 and y2 (x) = 0.
Theorem 4.1. (Existence and Uniqueness of Solutions)

Suppose that both functions f (x, y) (for existence) and its partial
@f (x, y)
derivative (for uniqueness) are continuous on some rectangle
@y
D in the xy-plane that contains the point (a, b) in its interior. Then, for
some open interval I containing the point a, the initial value problem:
8
< dy = f (x, y),
dx (4.4)
:
y(a) = b
has one and only one solution defined on the interval I.
Remark 4.1. From the Theorem 4.1 it follows that continuity of

f (x, y) itself on rectangle D is sufficient to guarantee the existence
of solutions of IVP (4.4). Additionally, continuity of both functions
@f (x, y)
f (x, y) and on rectangle D is sufficient to guarantee the
@y
uniqueness of the solution of IVP (4.4).
Remark 4.2. In Example 4.1 we have shown that the initial value
y2
problem (4.2) has no solutions. Note that the function f (x, y) =
x
is discontinuous when x = 0, and hence at the point (0, 1). So by
Theorem 4.1, existence of solutions of the IVP (4.2) is not guaranteed.
20
Remark 4.3. In Example 4.2 we have shown that the initial value
problem (4.3) has two di↵erent solutions. Note that the function
f (x, y) = 3y 2/3 is continuous everywhere, but the partial derivative
@f (x, y) 2
= p is discontinuous when y = 0, and hence at the point
@y 3 y
(0, 0). So by Theorem 4.1, the existence of the solution of IVP (4.3) is
guaranteed but the uniqueness of solution is not guaranteed. In fact,
it has two di↵erent solutions y1 = x3 and y2 = 0.
Example 4.3. Consider the following initial value problem:

8
< dy = y,
dx
:
y(0) = 3.
@f (x, y)
Both functions f (x, y) = y and = 1 are continuous every-
@y
where in xy-plane and therefore they are continuous on some rectan-
gular region D in xy-plane which contains point (0, 3) in its interior.
Therefore, by Theorem 4.1 existence of unique solution of given IVP
is guaranteed. Although Theorem 4.1 ensures existence of unique so-
lution only on some open interval containing x = 0, solution of this
IVP, y(x) = 3e x , is defined actually for all x.

8
< dy = x2 + y 2 ,
dx
:
y(a) = b.
@f (x, y)
Both functions f (x, y) = x2 + y 2 and = 2y are continuous
@y
everywhere in xy-plane and therefore they are continuous on some
rectangular region D in xy-plane which contains point (a, b) for any
values of a and b. Therefore, given IVP has a unique solution for any
initial data (a, b).
21

8
< dy = y 2 ,
dx
:
y(0) = 1.
@f (x, y)
Both functions f (x, y) = y 2 and = 2y are continuous every-
@y
where in xy-plane and in particular they are continuous on rectangular
region D = {(x, y) | 2 < x < 2, 0 < y < 2} which contains the point
(0, 1). Theorem 4.1 guarantees a unique solution of this IVP on some
open x-interval containing x = 0. It is easy to verify that the unique
1
solution is y(x) = . But this function is discontinuous at x = 1,
1 x
so our unique solution is not defined on the entire interval 2 < x < 2.
Thus, the solution interval I of Theorem 4.1 may not be as wide as the
@f
rectangle D where f and are continuous.
@y

8
< dy = p
> y
,
dx 2 x
>
: y(1) = 2.
y @f (x, y) 1
Both functions f (x, y) = p and = p are continuous
2 x @y 2 x
on rectangular region D = {(x, y) | 0.5 < x < 1.5, 1 < y < 3} which
contains the point (1, 2). Theorem 4.1 guarantees a unique solution of
this IVP on some open x-interval containing x = 1.
Example 4.7. Now, consider the following initial value problem:

8
< dy = p
> y
,
dx 2 x
>
: y(0) = 2.
22
y @f (x, y) 1
Both functions f (x, y) = p and = p are discontinuous
2 x @y 2 x
when x = 0 and hence at the point (0, 2). Thus, we cannot conclude
from Theorem 4.1 that this IVP has a solution. We are NOT saying
that this IVP does not have a solution. Theorem 4.1 simply gives NO
information one wayp or the other. In fact, given IVP does have a
solution: y(x) = 2e x .

8
< dy = 2y ,
dx x
:
y(a) = b.
2y @f (x, y) 2
Both functions f (x, y) = and = are continuous when-
x @y x
ever x 6= 0. Therefore, Theorem 4.1 guarantees a unique solution of
this IVP for any (a, b) with a 6= 0. If a = 0, we cannot conclude any-
thing from Theorem 4.1 about existence or uniqueness of solution of
this IVP. In fact, given IVP
• does not have any solutions if a = 0 but b 6= 0;
• has infinitely many solutions y(x) = Cx2 if a = b = 0.

Problem 4.1. In the following examples, determine whether the The-
orem 4.1 does or does not guarantee existence of a solution of the given
initial value problem. If existence is guaranteed, determine whether the
23
Theorem 4.1 does or does not guarantee uniqueness of that solution.

8 8
< dy < dy = x ln y,
= 2x2 y 2 ,
a) dx b) dx
: :
y(1) = 1. y(1) = 1.
8 8
< dy = p 3
y, < dy = p 3
y,
c) dx d) dx
: :
y(0) = 1. y(0) = 0.
8 8
< dy = px y, < dy = px y,
e) dx f) dx
: :
y(2) = 2. y(2) = 1.
8 8
< y dy = x 1, < y dy = x 1,
g) dx h) dx
: :
y(0) = 1. y(1) = 0.
8 8
< dy < dy = x2 y 2 ,
= ln (1 + y 2 ),
i) dx k) dx
: :
y(0) = 0. y(0) = 1.
8 8
< dy 2 >
< dy y2
= x sin y, = ,
l) dx m) dx x 2
: >
: y(1) = 0.
y(1) = 2.
8 8
2 p
< dy = sin x + y ,
> < dy = x + x3 + 4y,
n) dx x + 2y x 2 o) dx
>
: y(1) = 2. :
y(2) = 2.
24
5 Linear First Order Equations
Definition 5.1. The di↵erential equation in the form:
dy
+ P (x)y = Q(x) (5.1)
dx
is called linear first order di↵erential equation. We assume that
functions P (x) and Q(x) are continuous on some interval.
Definition 5.2. An integrating factor for a di↵erential equation

is a function µ(x, y) such that the multiplication of each side of the
di↵erential equation by µ(x, y) yields an equation in which each side
is recognizable as a derivative.
Example 5.1. Consider the linear first order di↵erential equation

equation:
dy
+ y = 2.
dx
If we multiply both sides of given equation by ex then we get
dy d x d
ex + ex y = 2ex =) (e y) = (2ex ) .
dx dx dx
We see that each side of resulting equation is a derivative. So, function
µ(x) = ex is an integrating factor for given equation.
Theorem 5.1. An integrating factor for the linear first order di↵eren-
tial equation (5.1) has the form:
R
P (x)dx
µ(x) = e . (5.2)
Proof. Multiplying both sides of equation (5.1) by µ(x) gives

dy
µ(x) + µ(x)P (x)y = µ(x)Q(x). (5.3)
dx
25
If function µ(x) is an integration factor for equation (5.1) then by

definition each side of (5.3) should be a derivative. To make a left-hand
side of (5.3) a derivative of some function it is sufficient to choose µ(x)
such that µ(x)P (x) = µ0 (x) is satisfied. Indeed, if µ(x)P (x) = µ0 (x)
then the left-hand side of (5.3) becomes
dy ⇣ ⌘0
0 0
µ(x) + µ(x)P (x)y = µ(x)y (x) + µ (x)y(x) = µ(x)y(x) .
dx
Now, let us find µ(x). We have
dµ
µ(x)P (x) = µ0 (x) =) = P (x)dx =)
µ
Z R
P (x)dx
=) ln µ = P (x)dx =) µ(x) = e
which completes the proof.
Remark 5.1. Using the result of Theorem 5.1, we can now describe
the method to solve the linear first order di↵erential equation (5.1). It
consists of following steps:
R
P (x)dx
1. Calculate the integrating factor µ(x) = e
2. Multiply both sides of the equation (5.1) by µ(x)
3. Recognize the left-hand side of the resulting equation as the

derivative: ⇣ ⌘0
µ(x)y(x) = µ(x)Q(x)
4. Integrate both sides to get

Z
µ(x)y(x) = µ(x)Q(x)dx + C
5. Divide both sides by µ(x) to express y(x)
26
Example 5.2. Find a general solution of the di↵erential equation:

dy 2y
+ = 4x.
dx x
Solution: It is easy to see that given equation is a linear first order
2
di↵erential equation with P (x) = and Q(x) = 4x. We first find the
x
integrating factor:
R R 2 2
P (x)dx x dx
µ(x) = e =e = e2 ln x = eln (x ) = x2 .
Multiplying both sides of given equation by µ(x) = x2 yields
dy
x2 + 2xy = 4x3
dx
which we recognize as ⇣ ⌘0
2
xy = 4x3 .
Then, integration of both sides gives
Z
x2 y = 4x3 dx = x4 + C.
Finally, dividing both sides by x2 allows us to obtain the general solu-

tion of given equation
C
y(x) = x2 + 2 .
x

dy y
= x cos x.
dx x
Solution: It is easy to see that given equation is a linear first order
1
di↵erential equation with P (x) = and Q(x) = x cos x. We first
x
find the integrating factor:
R R 1 1 1
P (x)dx x dx ln x
µ(x) = e =e =e = eln ( x ) = .
x
27
1
Multiplying both sides of given equation by µ(x) = yields
x
1 dy y
· = cos x
x dx x2
which we recognize as ⇣1 ⌘0
· y = cos x.
x
Z
y
= cos xdx = sin x + C.
x
Finally, multiplying both sides by x allows us to obtain the general
solution of given equation
y(x) = x(sin x + C).

dy
x2 + 1 + 3xy = 6x.
dx
Solution: We first divide both sides of given equation by x2 + 1 to
bring it to the form (5.1). We get
dy 3xy 6x
+ 2 = 2 . (5.4)
dx x + 1 x + 1
We now see that this equation is a linear first order di↵erential equation
3x 6x
with P (x) = 2 and Q(x) = 2 . We find the integrating factor:
x +1 x +1
R R 3x
P (x)dx dx 3 2 3/2
µ(x) = e =e x2 +1 = e 2 ln (x +1)
= x2 + 1 .
3/2
Multiplying both sides of (5.4) by µ(x) = x2 + 1 yields
3/2 dy 1/2 1/2
x2 + 1 + 3x x2 + 1 y = 6x x2 + 1
dx
28
which we recognize as
⇣ ⌘0
3/2 1/2
x2 + 1 y = 6x x2 + 1 .

Z
3/2 1/2 3/2
x2 + 1 y = 6x x2 + 1 dx = 2 x2 + 1 + C.
3/2
Finally, dividing both sides by x2 + 1 allows us to obtain the gen-
eral solution of given equation
3/2
y(x) = 2 + C x2 + 1 .

8
< dy y = 11 e x/3 ,
dx 8
:
y(0) = 1.
Solution: The equation in a given IVP is a linear first order di↵erential

11 x/3
equation with P (x) = 1 and Q(x) = e . We first find the
8
integrating factor:
R R
P (x)dx ( 1)dx
µ(x) = e =e = e x.
x
Multiplying both sides of equation by µ(x) = e yields
x dy 11
e e xy = e 4x/3
dx 8
⇣ ⌘0 11
x 4x/3
e y = e .
8
29

Z
y 11 4x/3 33 4x/3
= e dx = e + C.
ex 8 32
Now, multiplying both sides by ex allows us to obtain the general
solution
33 x/3
y(x) = e + Cex .
32
Finally, by using initial condition y(0) = 1 we can find the value of
1
constant C. If we substitute x = 0 and y = 1, we obtain C = .
32
So, the solution of given IVP is
33 x/3 1 x
y(x) = e + e .
32 32

dy
x + (x + 2)y = ex .
dx
Solution: We first divide both sides of given equation by x to bring
it to the form (5.1). We get
dy x + 2 ex
+ y= . (5.5)
dx x x
We now see that this equation is a linear first order di↵erential equation
x+2 ex
with P (x) = and Q(x) = . We find the integrating factor:
x x
R R R
= e (1+ x )dx = ex+2 ln x = x2 ex .
x+2 2
P (x)dx x dx
µ(x) = e =e
Multiplying both sides of (5.5) by µ(x) = x2 ex yields

dy
x2 e x + x(x + 2)ex y = xe2x
dx
30
2 x
xe y = xe2x .
Z
xe2x e2x
x2 ex y = xe2x dx = + C.
2 4
Finally, dividing both sides by x2 ex allows us to obtain the general
xe2x e2x
2 4 +C
y(x) = .
x ex
2
Remark 5.2. In some cases it becomes easier to solve the di↵erential

equation if we consider y as the independent variable rather than x.
The following example illustrates this point.
Example 5.7. Solve the following di↵erential equation

dy
1 4xy 2 = y3
dx
by regarding y as the independent variable rather than x.
Solution: We first rewrite given equation in the form:
dx 4 1
+ x = 3. (5.6)
dy y y
We can see that this equation is a linear first order di↵erential equation
with respect to x = x(y). We find the integrating factor:
R 4
µ(y) = e y dy = e4 ln y = y 4 .
Multiplying both sides of (5.6) by µ(y) = y 4 yields

dx
y4 + 4y 3 x = y
dy
31
d⇣ 4 ⌘
y · x(y) = y.
dy
Then, integration of both sides with respect to y gives
Z
4 y2
y x(y) = ydy = + C.
2
Finally, by dividing both sides by y 4 , we get

1 C
x(y) = + .
2y 2 y 4

Problem 5.1. Find general solutions of the following di↵erential equa-
tions:
2
a) y 0 + 3y = 2xe 3x
b) y 0 2xy = ex
p
c) 2xy 0 + y = 10 x d) 3xy 0 + y = 12x
e) 2xy 0 3y = 9x3 f) xy 0 = 2y + x3 cos x
g) y 0 + y cot x = cos x h) xy 0 + (2x 3)y = 4x4
y
i) y 0 + 2xy = x k) y 0 = (x2 1)ex
x+1
y 1
l) xy 0 + (2x + 1)y = xe 2x
m) y 0 + =
x2 x2
32
Problem 5.2. Find solutions of the following initial value problems:

( 0 ( 0
y + y = 2, y 2y = 3e2x ,
a) b)
y(0) = 0. y(0) = 0.
( (
xy 0 + 2y = 3x, xy 0 + 5y = 7x2 ,
c) d)
y(1) = 5. y(2) = 5.
( (
xy 0 y = x, xy 0 + y = 3xy,
e) f)
y(1) = 7. y(1) = 0.
( (
xy 0 + 3y = 2x5 , y 0 + y = ex ,
g) h)
y(2) = 1. y(0) = 1.
( (
xy 0 3y = x3 , y 0 + 2xy = x,
i) k)
y(1) = 10. y(0) = 2.
( (
y 0 = (1 y) cos x, (1 + x)y 0 + y = cos x,
l) m)
y(⇡) = 2. y(0) = 1.
( (
y 0 = 1 + x + y + xy, xy 0 = 3y + x4 cos x,
n) o)
y(0) = 0. y(2⇡) = 0.
( (
0
y = 2xy + 3x e , 2 x2 x2 + 4 y 0 + 3xy = x,
p) q)
y(0) = 5. y(0) = 1.
33
( (
x2 + 1 y 0 + 4xy = x,
3 2
x2 + 1 y 0 + 3x3 y = 6xe 2x ,
r) s)
y(0) = 1. y(2) = 1.
Problem 5.3. Solve the following di↵erential equations by regarding

y as the independent variable rather than x.
dy dy
a) (x + yey ) =1 b) (1 + 2xy) = 1 + y2
dx dx
34
6 Substitution Methods, Homogeneous Equa-
tions, Bernoulli Equations
The first order di↵erential equations we have solved so far have all been
either separable or linear. But many applications involve di↵erential
equations that are neither separable nor linear. In this section we illus-
trate that substitution methods sometimes can be used to transform a
given di↵erential equation into either separable or linear equation.
The first order di↵erential equation

dy
= f (x, y), (6.1)
dx
with dependent variable y and independent variable x, may contain a
conspicuous combination
v = ↵(x, y) (6.2)
of x and y that suggests itself as a new dependent variable v. If the

substitution relation in (6.2) can be solved for
y = (x, v), (6.3)
then application of the chain rule, regarding v as an unknown function

of x, yields
dy @ dx @ dv @ @ dv
= + = + . (6.4)
dx @x dx @v dx @x @v dx
dy
If we substitute the right-hand side of (6.4) for in equation (6.1)
dx
dv
and then solve for , the result is a new di↵erential equation of the
dx
form
dv
= g(x, v) (6.5)
dx
with new dependent variable v. If this new equation is either separable
or linear, then we can solve it. If v = v(x) is a solution of (6.5), then
35
y = (x, v(x)) will be a solution of the original equation (6.1). The

trick is to select a substitution such that the transformed equation
(6.5) is one we can solve. Even when possible, this is not always easy;
it may require a fair amount of ingenuity or trial and error.
dy
= (x + y + 3)2 .
dx
Solution: We make a substitution v = x + y + 3, so that y = v x 3.
dy dv
Then = 1. Putting these in a given equation we get
dx dx
dv dv
1 = v 2 or = 1 + v2
dx dx
which is a separable equation. Separating the variables, we obtain
dv
= dx.
1 + v2
Integrating both sides gives us
Z Z
dv
= dx =) arctan v = x+C =) v(x) = tan (x + C).
1 + v2
Finally, making a back substitution we obtain a general solution of
given di↵erential equation
y(x) = v(x) x 3 = tan (x + C) x 3.

dy
2xe2y = 3x4 + e2y .
dx
dv dy
Solution: We make a substitution v = e2y , = 2e2y . Putting
dx dx
these in a given equation we get
dv dv 1
x = 3x4 + v or v = 3x3
dx dx x
36
1
which is a first order linear di↵erential equation with P (x) = and
x
Q(x) = 3x3 . We find the integrating factor:
R R dx 1
P (x)dx ln x
µ(x) = e =e x =e = .
x
1
Therefore, multiplying both sides by µ(x) = yields
x
1 dv 1
2
v = 3x2
x dx x
which we recognize as ⇣ 1 ⌘0
v = 3x2 .
x
v
= x3 + C.
x
Multiplying both sides by x gives us
v = x4 + Cx.
Since v = e2y , making a back substitution we obtain a general solution
of given di↵erential equation
1
e2y = x4 + Cx =) y(x) = ln |x4 + Cx|.
2
6.1 Homogeneous Equations

Definition 6.1. If the right-hand side of equation (6.1) can be written
y
as a function of v only, where v = , then first order di↵erential equa-
x
tion is called homogeneous. So, a homogeneous first order di↵erential
equation has the form:
dy ⇣y ⌘
=F . (6.6)
dx x
37
y dy dv
We use the substitution v = , so that y = xv. Then = v+x .
x dx dx
Putting these in equation (6.6) we get
dv dv F (v) v
su
v+x = F (v) or =
dx dx x
which is a separable equation.
Thus, every homogeneous equation can be transformed to
y
separable equation with the use of substitution v = .
x
dy
2xy = 4x2 + 3y 2 .
dx
Solution: We first divide both sides of given equation by 2xy. We get
dy 4x2 + 3y 2 x 3 y
= =2· + ·
dx 2xy y 2 x
We now see that this equation is a homogeneous first order di↵erential
y
equation. We use the substitution v = . Then
x
y dy dv
v= =) y = xv, =v+x .
x dx dx
Putting these in equation we get
dv 2 3v
v+x = +
dx v 2
and hence
dv 2 v dv v2 + 4
x = + or x =
dx v 2 dx 2v
2vdv dx
= .
v2 + 4 x
38

Z Z
2vdv dx
2
= =) ln (v 2 + 4) = ln |x| + ln C =)
v +4 x
⇣ y ⌘2
2
=) v + 4 = Cx =) + 4 = Cx =) y 2 + 4x2 = Cx3 .
x
So, the general solution of given di↵erential equation is y 2 +4x2 = Cx3 .

8 p
< x dy = y + x2 y 2 ,
dx
:
y(1) = 0.
Solution: We divide both sides of equation in a given IVP by x and

find that r
dy y ⇣ y ⌘2
= + 1 .
dx x x
We see that this equation is a homogeneous first order di↵erential
y
x
y dy dv
v= =) y = xv, =v+x .
x dx dx
dv p
v+x =v+ 1 v2
dx
and hence
dv p
x = 1 v2
dx
dv dx
p = .
1 v2 x
39

Z Z
dv dx
p = =) arcsin v = ln |x| + C =)
1 v2 x
=) v = sin (ln |x| + C) =) y(x) = x sin (ln |x| + C).

constant C. If we substitute x = 1 and y = 0, we obtain C = 0. So,
y(x) = x sin (ln |x|).

dy
x2 xy = y 2 4x2 .
dx
Solution: We first divide both sides of given equation by x2 . We get

dy y ⇣ y ⌘2
= 4
dx x x
We can see that this equation is a homogeneous first order di↵erential
y
x
y dy dv
v= =) y = xv, =v+x .
x dx dx
dv
v+x v = v2 4
dx
and hence
dv
x = v2 4
dx
40

dv dx
= .
v2 4 x
Z Z Z ✓ ◆ Z
dv dx 1/4 1/4 dx
= =) dv = =)
v2 4 x v 2 v+2 x
1 1 1
=) ln |v 2| ln |v + 2| = ln |x| + ln C =)
4 4 4
v 2
=) ln |v 2| ln |v + 2| = ln x4 + ln C =) = Cx4 =)
v+2
y 2x 2x(1 + Cx4 )
=) = Cx4 =) y(x) = .
y + 2x 1 Cx4
2x(1 + Cx4 )
So, the general solution of given di↵erential equation is y(x) = .
1 Cx4
6.2 Bernoulli Equations

Definition 6.2. A first order di↵erential equation of the form
dy
+ P (x)y = Q(x)y n (6.7)
dx
is called a Bernoulli equation. We assume that functions P (x) and
Q(x) are continuous on some interval.
Remark 6.1. If either n = 0 or n = 1, then equation (6.7) is linear.

We consider here only nonlinear cases when n 6= 0 and n 6= 1.
41
Theorem 6.1. Substitution
v = y1 n
(6.8)
transforms Bernoulli equation (6.7) into the linear equation

dv
+ (1 n)P (x)v = (1 n)Q(x). (6.9)
dx
Proof. We first divide both sides of (6.7) by y n and find that
n dy
y + P (x)y 1 n
= Q(x). (6.10)
dx
Di↵erentiating (6.8) we have
dv n dy 1 dv n dy
= (1 n)y or =y .
dx dx 1 n dx dx
Therefore, substituting (6.8) into equation (6.10) yields
1 dv
+ P (x)v = Q(x).
1 n dx
Finally, multiplying both sides by 1 n we obtain (6.9), which com-
pletes the proof.
Thus, every Bernoulli equation can be transformed to linear

equation with the use of substitution (6.8).

dy
x + 6y = 3xy 4/3 .
dx
Solution: We first divide both sides of given equation by x to bring
it to the form (6.7). We get
dy 6y
+ = 3y 4/3 . (6.11)
dx x
42
4
We now see that this equation is a Bernoulli equation with n = ,
3
6
P (x) = and Q(x) = 3. Next, we divide both sides by y 4/3 to get
x
4/3 dy6 1/3
y + y = 3.
dx x
Now, making a substitution
dv 1 4/3 dy
v = y1 n
= y1 4/3
=y 1/3
, = y
dx 3 dx
transforms the equation into
dv 6 dv 2
3 + v = 3 or v= 1
dx x dx x
which is a linear equation. We find the integrating factor:
R 2 1
µ(x) = e x dx =e 2 ln x
= .
x2
1
x2
1 dv 1 1
v=
x2 dx x3 x2
which we recognize as ⇣ 1 ⌘0 1
v = .
x2 x2
v 1
= + C.
x2 x
Multiplying both sides by x2 gives us
v = x + Cx2 .
43
Since v = y 1/3 , making a back substitution we obtain a general solu-

tion of given di↵erential equation
1/3 1
y = x + Cx2 =) y(x) = .
(x + Cx2 )3

8
< dy + y = xy 3 ,
dx
:
y(0) = 1.
Solution: We see that equation in a given IVP is a Bernoulli equation

with n = 3, P (x) = 1 and Q(x) = x. We divide both sides by y 3 to
get
1 dy 1
+ = x.
y 3 dx y 2
1 dv 1 dy
v = y1 n
= y1 3
= , = 2
y2 dx y 3 dx
1 dv dv
+ v = x or 2v = 2x
2 dx dx
R
( 2)dx 2x
µ(x) = e =e .
2x
Therefore, multiplying both sides by µ(x) = e yields
2x dv 2x 2x
e 2e v= 2xe
dx
⇣ ⌘0
2x 2x
e v = 2xe .
44

Z ✓ ◆
2x 2x 1 2x
e v = 2 xe dx = x + e + C.
2
Multiplying both sides by e2x gives us
1
v =x+ + Ce2x .
2
1 1
Since v = y 2 , making a back substitution we obtain 2
= x+ +Ce2x .
y 2
1
constant C. If we substitute x = 0 and y = 1, we obtain C = . So,
2
1 1 1 2x
= x + + e .
y2 2 2

dy y x2 + 1
= .
dx x y2
Solution: We observe that given equation is a Bernoulli equation with
1
n = 2, P (x) = and Q(x) = x2 + 1. We first multiply both sides
x
by y 2 to get
2 dy y3
y = x2 + 1.
dx x
dv dy
v = y1 n
= y 1+2 = y 3 , = 3y 2
dx dx
1 dv v dv 3
= x2 + 1 or v = 3x2 + 3
3 dx x dx x
45

R 3 1
µ(x) = e x dx =e 3 ln x
= .
x3
1
x3
1 dv 3 3 3
v = +
x3 dx x4 x x3
⇣ 1 ⌘0 3 3
v = + .
x3 x x3
v 3
= 3 ln |x| + C.
x3 2x2
Multiplying both sides by x3 gives us
✓ ◆
3 3
v = x 3 ln |x| +C .
2x2
Since v = y 3 , making a back substitution we obtain a general solution
of given di↵erential equation
✓ ◆ r
3 3
y 3 = x3 3 ln |x|
3
+ C =) y(x) = x 3 ln |x| + C.
2x2 2x2

Problem 6.1. Find general solutions of the following homogeneous
di↵erential equations:
a) (x + y)y 0 = x y b) 2xyy 0 = x2 + 2y 2
p
c) xy 0 = y + 2 xy d) (x y)y 0 = x + y
e) x(x + y)y 0 = y(x y) f) (x + 2y)y 0 = y

46
g) xy 2 y 0 = x3 + y 3 h) x2 y 0 = xy + x2 ey/x
i) x2 y 2 y 0 = 2xy j) x2 y 0 = xy + y 2
p
k) xyy 0 = x2 + 3y 2 0 2
l) xyy = y + x 4x2 + y 2
p p
0 0
m) xy = y + x2 + y 2 n) yy + x = x2 + y 2
⇣y ⌘
o) x(x + y)y 0 + y(3x + y) = 0 0
p) xy = y + x tan
x
0 2xy + 3y 2 y 3
q) y = r) xy 0 = 2
2xy + x2 y x2
p p p
3 2
0 x + y x2 + y 2 0 x+y+ x y
s) y = p t) y = p p
xy x2 + y 2 x+y x y
Problem 6.2. Find general solutions of the following Bernoulli di↵er-

ential equations:
a) x2 y 0 + 2xy = 5y 3 b) y 2 y 0 + 2xy 3 = 6x
c) y 0 = y + y 3 d) x2 y 0 + 2xy = 5y 4
y y2
e) y 0 = f) 2xy 0 + y 3 e 2x
= 2xy
x x
p
0
2
g) y (xy + y) 1 + x4 = x h) 3y 2 y 0 + y 3 = e x
i) 3xy 2 y 0 = 3x4 + y 3 j) x2 y 0 + xy = xy 3
8x
k) xy 0 + y = 2x6 y 4 l) y 0 + 4xy =
y3
(x + 1)y x+1
m) y 0 + = n) xy 0 + y = y 2 x2 ln x
2x xy
47
Problem 6.3. Find general solutions of the following di↵erential equa-

tions by making appropriate substitution.
p
a) y 0 = x + y + 1 b) y 0 = (4x + y)2
c) (x + y)y 0 = 1 d) xey y 0 = 2 ey + x3 e2x
e) (x sin 2y) y 0 = 4x2 + sin2 y f) (x + ey ) y 0 = xe y

1

8 8
>
< 2xy dy < (x + y) dy = x y,
>
= x2 + 3y 2 ,
a) dx b) dx
>
: y(2) = 6. >
: y(1) = 0.
8 8
< dy + y = x ,
> < x dy + y = (xy)3/2 ,
>
c) dx 2x y 3 d) dx
>
: >
: y(1) = 4.
y(1) = 2.
8 8 2 2
< dy = 2x 5y ,
> < dy = 3x + 9xy + 5y ,
>
e) dx y 4x f) dx 6x2 + 4xy
>
: >
:
y(1) = 4. y(2) = 6.
48
7 Exact Di↵erential Equations
The first order di↵erential equation:
dy
= f (x, y) (7.1)
dx
can be written in the di↵erent form:
M (x, y)dx + N (x, y)dy = 0. (7.2)
(7.1) is called derivative form and (7.2) is called di↵erential form.
dy x2 + y 2
Example 7.1. For instance, the equation = can be written
dx x y
in the di↵erential form:
x2 + y 2 dx + (y x)dy = 0.
And other way around, the equation
(sin x + y) dx + (x + 3y)dy = 0
dy sin x + y
may be written in the derivative form as = .
dx x + 3y
Definition 7.1. Let F (x, y) be a function which has continuous first

partial derivatives in open region D of xy-plane. The total di↵eren-
tial dF of the function F is defined by the formula
@F @F
dF (x, y) = dx + dy (7.3)
@x @y
for all (x, y) 2 D.
Example 7.2. Let F (x, y) = xy 2 + 2x3 y for all real (x, y). Then the
total di↵erential of this function is defined by
@F @F
dF = dx + dy = y 2 + 6x2 y dx + 2xy + 2x3 dy.
@x @y
49
Definition 7.2. Di↵erential equation (7.2) is called exact di↵eren-

tial equation if there exists some function F (x, y) such that
@F @F
= M (x, y) and = N (x, y). (7.4)
@x @y
Theorem 7.1. If the di↵erential equation (7.2) is an exact di↵erential

equation, i.e. there exists some function F (x, y) that satisfies condi-
tions (7.4), then its general solution has the form
F (x, y) = C,
where C is an arbitrary constant.
Proof. Indeed, from (7.2) using (7.3) and (7.4) we have

@F @F
M (x, y)dx + N (x, y)dy = 0 =) dx + dy = 0 =)
@x @y
=) dF (x, y) = 0 =) F (x, y) ⌘ C.
Example 7.3. Consider the di↵erential equation
y 2 dx + 2xydy = 0.
This equation is an exact di↵erential equation, since there exists func-

tion F (x, y) which satisfies conditions (7.4). Indeed, for F (x, y) = xy 2
@F @F
we have = y 2 = M (x, y) and = 2xy = N (x, y).
@x @y
Therefore, using Theorem 7.1 we can conclude that the general solution
of given di↵erential equation has the form
xy 2 = C.
50
Theorem 7.2. Consider the di↵erential equation
M (x, y)dx + N (x, y)dy = 0. (7.5)
where functions M (x, y) and N (x, y) have continuous first partial deriva-
tives. Equation (7.5) is an exact di↵erential equation if and only if
@M @N
= . (7.6)
@y @x
Proof. If equation (7.5) is exact then by definition there exists some
function F (x, y) such that
@F @F
= M (x, y) and = N (x, y).
@x @y
Then
@ 2F @M @ 2F @N
= and = .
@y@x @y @x@y @x
@ 2F @ 2F
Since, both and are continuous, we have
@y@x @x@y
@ 2F @ 2F @M @N
= =) = .
@y@x @x@y @y @x
We leave out the proof of converse for self-study.
Remark 7.1. Using the results of Theorem 7.1 and Theorem 7.2, we
can now describe the method to solve equation (7.5) if it is exact. It
consists of following steps:
1. Check the condition (7.6) for exactness. If it is satisfied then

proceed with next steps.
2. Find function F (x, y) that satisfies equalities (7.4).
51
3. Function F (x, y) obtained in the previous step allows us to write

the general solution of equation (7.5) in the form
F (x, y) = C.
Example 7.4. Find the general solution of the di↵erential equation

2xy 3 dx + 3x2 y 2 dy = 0.
Solution: We first check whether or not the equation is exact. Here

@M @N
M (x, y) = 2xy 3 , N (x, y) = 3x2 y 2 =) = 6xy 2 , = 6xy 2 .
@y @x
Thus condition (7.6) is satisfied and therefore given equation is exact.
Now we must find function F (x, y) such that
8
> @F
>
< = M (x, y) = 2xy 3 ,
@x
(7.7)
>
> @F 2 2
: = N (x, y) = 3x y .
@y
Integrating both sides of first equality in (7.7) with respect to x we
obtain
F (x, y) = x2 y 3 + b(y),
where b(y) is an arbitrary function of y. Plugging this into second
equality in (7.7) we get
@F
= 3x2 y 2 + b0 (y) = 3x2 y 2 .
@y
Hence, b0 (y) = 0 and therefore b(y) = C1 , where C1 is constant. Thus,
function F (x, y) = x2 y 3 + C1 satisfies both equalities in (7.7). Then
the general solution of given equation is
x2 y 3 = C.
Note that we have absorbed the constant C1 into the constant C.
52
Example 7.5. Find the general solution of the di↵erential equation

✓ ◆ ✓ ◆
1 1
y+ dx + x + 2 dy = 0.
x y
Solution: We first check whether or not the equation is exact. Here

1 1 @M @N
M (x, y) = y + , N (x, y) = x + 2 =) =1= .
x y @y @x
8
> @F 1
>
< @x = M (x, y) = y + ,
x
(7.8)
>
> @F 1
: = N (x, y) = x + 2 .
@y y
obtain
F (x, y) = xy + ln |x| + b(y).
Plugging this into second equality in (7.8) we get
@F 1
= x + b0 (y) = x + 2 .
@y y
1 1
Hence, b0 (y) = and therefore b(y) = + C1 . Thus, function
y2 y
1
F (x, y) = xy + ln |x| + C1 satisfies both equalities in (7.8). Then
y
the general solution of given equation is
1
xy + ln |x| = C.
y
53
Example 7.6. Solve the initial value problem

(
2x cos y + 3x2 y dx + x3 x2 sin y y dy = 0,
y(0) = 2.
Solution: We first observe that the di↵erential equation in a given

IVP is exact, since
@M @N
= 2x sin y + 3x2 = .
@y @x
8
> @F 2
>
< @x = M (x, y) = 2x cos y + 3x y,
(7.9)
>
> @F
: = N (x, y) = x3 x2 sin y y.
@y
obtain
F (x, y) = x2 cos y + x3 y + b(y).
Plugging this into second equality in (7.9) we get
@F
= x2 sin y + x3 + b0 (y) = x3 x2 sin y y.
@y
y2
Hence, b0 (y) = y and therefore b(y) = + C1 . Thus function
2
2 3 y2
F (x, y) = x cos y + x y + C1 satisfies both equalities in (7.9).
2
Then the general solution of equation is
2 3 y2
x cos y + x y = C.
2
54
Finally, applying the initial condition y(0) = 2, we find C = 2. Thus

the solution of the given IVP is
2 3 y2
x cos y + x y + 2 = 0.
2
Remark 7.2. Finding function F (x, y) that satisfies equalities (7.4)

is only possible if the equation is exact. Consider for instance the
following di↵erential equation
2xdx + x2 + y 2 dy = 0.
We cannot find any function F (x, y) that satisfies both
@F @F
= 2x and = x2 + y 2 .
@x @y
This is due to the fact that equation is not exact. Indeed, we have
@M @N
M (x, y) = 2x, N (x, y) = x2 + y 2 =) = 0, = 2x
@y @x
@M @N
and therefore 6= .
@y @x
Example 7.7. Determine the value of constant A such that the dif-
ferential equation
✓ ◆ ✓ 2 ◆
2x Ax
+ y cos (xy) 3 dx + + x cos (xy) dy = 0
y2 y3
is exact and solve the resulting exact equation.
Solution: We first find the value of constant A for which the given
di↵erential equation is exact.
2x @M 4x
M (x, y) = + y cos (xy) 3 =) = + cos (xy) xy sin (xy)
y2 @y y3
Ax2 @N 2Ax
N (x, y) = 3 + x cos (xy) =) = 3 + cos (xy) xy sin (xy)
y @x y
55
Then condition (7.6) is satisfied when A = 2. Therefore, given equa-

tion is exact if and only if A = 2.
Now we solve the resulting exact equation. We must find function
F (x, y) such that
8
>
> @F 2x
>
< @x = + y cos (xy) 3,
y2
(7.10)
>
> @F 2x 2
>
: = + x cos (xy).
@y y3

obtain
x2
F (x, y) = 2 + sin (xy) 3x + b(y).
y
Integrating both sides of second equality in (7.10) with respect to y we
obtain
x2
F (x, y) = 2 + sin (xy) + a(x).
y
2
So a(x) = 3x and b(y) = 0. Thus, function F (x, y) = xy2 +sin (xy) 3x
satisfies both equalities in (7.10). Then the general solution of given
equation is
x2
+ sin (xy) 3x = C.
y2

Problem 7.1. In the following examples, first verify that the given
equation is exact and then solve it.
a) (2x + 3y)dx + (3x + 2y)dy = 0
b) (4x y)dx + (6y x)dy = 0
c) 3x2 + 4xy dx + 2x2 + 2y dy = 0
56
d) 2y 2 + 6xy 5 dx + 3x2 + 4xy 6 dy = 0
e) 3x2 + 2y 2 dx + 4xy + 6y 2 dy = 0
f) 2xy 2 + 3x2 dx + 2x2 y + 4y 3 dy = 0

⇣ y⌘
3
g) x + dx + y 2 + ln x dy = 0
x
h) (1 + yexy ) dx + (2y + xexy ) dy = 0
✓ ◆
x
i) (cos x + ln y) dx + + ey dy = 0
y
x+y
j) (x + arctan y) dx + dy = 0
1 + y2
k) 3x2 y 3 + y 4 dx + 3x3 y 2 + y 4 + 4xy 3 dy = 0
l) (ex sin y + tan y) dx + ex cos y + x sec2 y dy = 0

✓ ◆ ✓ ◆
2x 3y 2 2y x2 1
m) dx + + p dy = 0
y x4 x3 y 2 y
✓ 5/2 ◆ ✓ 5/3 ◆
2x 3y 5/3 3y 2x5/2
n) dx + dy = 0
2x5/2 y 2/3 3x3/2 y 5/3
o) y sec2 x + sec x tan x dx + (tan x + 2y) dy = 0

✓ ◆ ✓ ◆
2x 1 x x2
p) dx + dy = 0
y y2

(
y sin 2x + y 2 sin x dx + sin2 x 2y cos x dy = 0
a)
y(0) = 3
57
8✓ ◆ ✓ ◆
< 3
> y y 2 2x
dx + dy = 0
b) x2 xy 2
>
:
y( 1) = 2
Problem 7.3. In the following examples, first find the value of con-
stant A such that the given equation is exact and then solve the re-
sulting exact equation:
a) x2 + 3xy dx + Ax2 + 4y dy = 0
✓ ◆ ✓ ◆
1 1 Ax + 1
b) + dx + dy = 0
x2 y 2 y3
58
8 Population Models
Di↵erential equations have wide range of applications in various fields
of science and engineering. Here we will present an application for
population models.
8.1 Natural Growth Equation

Suppose that P (t) is the number of individuals in a population (of
humans, or insects, or bacteria) having constant birth and death rates
and (in births and deaths per individual per unit of time). Sup-
pose that the population changes only by the occurrence of births and
deaths, so there is no migration. Then, during a short time interval
4t, approximately P (t)4t births and P (t)4t deaths occur, so the
change in P (t) is given approximately by
P (t + 4t) P (t) ⇡ P (t)4t P (t)4t
and therefore
P (t + 4t) P (t)
⇡( ) P (t).
4t
Taking the limit as 4t ! 0 from both sides, we get
dP
= kP, (8.1)
dt
where k = .
Equation (8.1) can be easily solved by separating the variables and
integrating. Its solution is
P (t) = Cekt .
So, the particular solution of (8.1) with the initial condition P (0) = P0
is simply
P (t) = P0 ekt . (8.2)
59
Definition 8.1. Because of the presence of the natural exponential

function in its solution, the di↵erential equation (8.1) is often called
the natural growth equation.
Remark 8.1. Note that for solution (8.2) of the natural growth equa-
tion (8.1) we have three di↵erent possibilities:
• P (t) ! 1 as t ! 1 when k = >0
• P (t) ! 0 as t ! 1 when k = <0
• P (t) ⌘ P0 when k = =0
So, population explodes if the birth rate is larger than the death rate
and it dies out if the death rate is larger than the birth rate.
Example 8.1. According to data listed at www.census.gov, the

world’s total population reached 6 billion persons in mid-1999, and
was then increasing at the rate of about 212 thousand persons each
day. Assuming that natural population growth at this rate continues,
we want to answer these questions:
a) What is the annual growth rate k?
b) What will be the world population at the middle of the 21st
century?
c) How long will it take the world population to increase tenfold -
thereby reaching the 60 billion that some demographers believe
to be the maximum for which the planet can provide adequate
food supplies?
Solution: (a) We measure the world population P (t) in billions and
measure time in years. We take t = 0 to correspond to (mid) 1999, so
P0 = 6. The fact that P is increasing by 212000, or 0.000212 billion,
persons per day at time t = 0 means that
P 0 (0) = 0.000212 · 365.25 ⇡ 0.07743
60
billion per year. From the natural growth equation (8.1) with t = 0
we now obtain
P 0 (0) 0.07743
k= ⇡ ⇡ 0.0129.
P (0) 6
Thus the world population was growing at the rate of about 1.29%
annually in 1999. This value of k gives the world population function
P (t) = 6e0.0129t .
(b) With t = 51 we obtain the prediction
P (51) = 6e0.0129·51 ⇡ 11.58 (billion)
for the world population in mid-2050.

(c) The world population should reach 60 billion when
ln 10
60 = 6e0.0129t =) t= ⇡ 178
0.0129
and thus in the year 2177.
Note: Actually, the rate of growth of the world population is ex-
pected to slow somewhat during the next half-century, and the best
current prediction for the 2050 population is about 9.1 billion. A sim-
ple mathematical model (8.1) cannot be expected to mirror precisely
the complexity of the real world.
8.2 General Population Equation

We have shown that equation (8.1) with solution (8.2) serves as a
mathematical model for natural population growth that occurs as a
result of constant birth and death rates. Here we present a more
general population model that accommodates birth and death rates
that are not necessarily constant.
Suppose that P (t) is the number of individuals in a population
at time t. As before, suppose that the population changes only by
61
the occurrence of births and deaths, so there is no migration. It is

customary to track the growth or decline of a population in terms of
its birth rate and death rate functions defined as follows:
• (t) is the number of births per unit of population per unit of

time at time t;
• (t) is the number of deaths per unit of population per unit of

time at time t.
Then the number of births and deaths that occur during the time in-
terval [t, t+4t] is given approximately by (t)P (t)4t and (t)P (t)4t,
respectively. Hence the change in the population during the time in-
terval [t, t + 4t] is
P (t + 4t) P (t) ⇡ (t)P (t)4t (t)P (t)4t
so
P (t + 4t) P (t)
⇡ ( (t) (t)) P (t).
4t
Taking the limit as 4t ! 0 from both sides, we get
dP
= (t) (t) P (t). (8.3)
dt
Equation (8.3) is called the general population equation. If func-
tions (t) and (t) are constant, then equation (8.3) reduces to the
natural growth equation (8.1). But it also includes the possibility that
(t) and (t) are variable functions of t.
Remark 8.2. The birth and death rates need not be known in advance;
they may well depend on the unknown function P (t).
Example 8.2. Suppose that an alligator population numbers 100 ini-

tially, and that its death rate is (t) ⌘ 0 (so none of the alligators is
62
dying). If the birth rate is (t) = 0.0005P (t) - and thus increases as the
population does - then equation (8.3) gives the initial value problem
8
< dP = 0.0005P 2 ,
dt
:
P (0) = 100.
Then upon separating the variables and integrating both sides we get
Z Z
dP 1
= 0.0005dt =) = 0.0005t + C.
P2 P
Substitution of t = 0 and P = 100, gives C = 1/100; and then we
readily solve for
2000
P (t) = .
20 t
For instance, P (10) = 200, so after 10 years the alligator population
has doubled. But we see that P (t) ! 1 as t ! 20, so a real ”popula-
tion explosion” occurs in 20 years.
8.3 Logistic Equation

In situations as diverse as human population of a nation and a fruit
fly population in a closed container, it is often observed that the birth
rate decreases as the population itself increases. The main reason for
that is a limited food supply. Suppose, for example, the birth rate
is a linear decreasing function of the population size P , so that
(t) = 0 1 P (t), where 0 and 1 are positive constants. If the
death rate (t) ⌘ 0 remains constant, then the general population
equation (8.3) takes the form
dP
= 0 1P 0 P
dt
or
dP
= aP bP 2 , (8.4)
dt
63
where a = 0 0 and b = 1 . If the coefficients a and b are both

positive, then equation (8.4) is called the logistic equation. For the
purpose of relating the behavior of the population P (t) to the values
of the parameters in the equation, it is useful to rewrite the logistic
equation in the form
dP
= kP (M P ), (8.5)
dt
where k = b and M = a/b are positive constants.
Remark 8.3. It is easy to show that the solution of logistic equation

(8.5) satisfying the initial condition P (0) = P0 has the form
M P0
P (t) = kM t
. (8.6)
P0 + (M P0 )e
Then it follows that
M P0
lim P (t) = = M.
t!1 P0 + 0
Thus a population that satisfies the logistic equation does not grow
without bound like a naturally growing population modeled by natu-
ral growth equation (8.1). Instead, it approaches the finite limiting
population M as t ! 1. Sometimes M is called the carrying ca-
pacity of the environment.
Example 8.3. Consider a population that is modeled by the logistic

equation
dP
= 0.0004P (150 P )
dt
with initial condition P (0) = P0 .
To solve this di↵erential equation, we first separate the variables
dP
= 0.0004dt
P (150 P )
64
and then integrate both sides

Z ✓ ◆ Z
1 1 P
+ dP = 0.06dt =) = Ce0.06t .
P 150 P 150 P
P0
Now using initial condition P (0) = P0 , we obtain C = . Hence
150 P0
P P0
= e0.06t .
150 P 150 P0
Solving it for P (t) we get the solution of given problem
150P0
P (t) = 0.06t
. (8.7)
P0 + (150 P0 )e
Note that for solution (8.7) we have lim P (t) = 150, whatever the
t!1
initial value P0 > 0.

Problem 8.1. The time rate of change of a rabbit population P is
proportional to the square root of P . At time t = 0 the population
numbers 100 rabbits and is increasing at the rate of 20 rabbits per
month. How many rabbits will be there one year later?
Problem 8.2. Suppose that the fish population P (t) in a lake is at-
tacked by a disease at time t = 0, with the result that the fish cease
to reproduce (so that the birth rate is = 0) and the p death rate
(deaths per week per fish) is thereafter proportional to 1/ P . If there
were initially 900 fish in the lake and 441 were left after 6 weeks, how
long will it take all the fish in the lake to die?
65
Problem 8.3. The time rate of change of an alligator population P

in a swamp is proportional to the square of P . The swamp contained
12 alligators at time t = 0 and 24 alligators 10 years later. How long
will it take an alligator population to reach the number 48?
Problem 8.4. Suppose that the population P (t) of a country satisfies

the logistic di↵erential equation
dP
= kP (200 P ),
dt
where k is constant. Its population in 1940 was 100 million and was
then growing at the rate of 1 million per year. Predict this country’s
population for the year 2000.
66
9 Reducible Second Order Di↵erential Equa-
tions
A second order di↵erential equation involves the second derivative of
unknown function y(x), and thus has the general form
F (x, y, y 0 , y 00 ) = 0. (9.1)
Definition 9.1. Equation (9.1) is called reducible second order

di↵erential equation if either the dependent variable y or the inde-
pendent variable x is missing in it.
Remark 9.1. Reducible second order di↵erential equation can be eas-

ily reduced by a simple substitution to a first order di↵erential equa-
tion.
9.1 Case I - dependent variable y missing

If y is missing, then equation (9.1) takes the form
F (x, y 0 , y 00 ) = 0. (9.2)
Then the substitution
dy dv
v = y0 = , y 00 = (9.3)
dx dx
results in the first order di↵erential equation:
F (x, v, v 0 ) = 0.
If we can solve this equation for a general solution v(x, C1 ) involving
an arbitrary constant C1 , then we only need to write
Z Z
y(x) = y 0 (x)dx = v(x, C1 )dx + C2
to get a solution of the equation (9.2) that involves two arbitrary con-
stants C1 and C2 (as is to be expected in the case of second order
di↵erential equation).
67
Example 9.1. Solve the second order di↵erential equation
xy 00 + 2y 0 = 6x.
Solution: Observe that in this di↵erential equation dependent vari-

able y is missing. Then the substitution (9.3) gives the first order
equation
xv 0 + 2v = 6x
or
2
v 0 + v = 6,
x
which is a linear first order equation. We find the integrating factor:
R 2 2
µ(x) = e x dx = e2 ln x = eln (x ) = x2 .
Multiplying both sides of the equation by µ(x) = x2 yields

dv
x2 + 2xv = 6x2
dx
2
xv = 6x2 .
Z
x2 v = 6x2 dx = 2x3 + C1 .
Dividing both sides by x2 , we find

C1
v(x) = 2x + .
x2
Since v = y 0 , a final integration with respect to x yields the general
Z ✓ ◆
C1 C1
y(x) = 2x + 2 dx = x2 + C2 .
x x
68
equation
2
y 00 = y 0 .
Solution: Given equation is reducible since dependent variable y is
missing in it. Then the substitution (9.3) gives the first order separable
di↵erential equation:
dv
= v2.
dx
Separation of variables gives
Z Z
dv 1 1
2
= dx =) = x + C1 =) v = .
v v x + C1
dy
Since v = y 0 = , we get
dx
dy 1
= .
dx x + C1
Finally, integrating both sides of last equality we obtain the general
solution of given second order di↵erential equation as following:
Z
dx
y(x) = = ln |x + C1 | + C2 .
x + C1
9.2 Case II - independent variable x missing

If x is missing, then equation (9.1) takes the form
F (y, y 0 , y 00 ) = 0. (9.4)
Then the substitution

dy dv dv dy dv
v = y0 = , y 00 = = =v (9.5)
dx dx dy dx dy
69
results in the first order di↵erential equation:

dv
F (y, v, v )=0
dy
for v as a function of y. If we can solve this equation for a general
solution v(y; C1 ) involving an arbitrary constant C1 , then (assuming
that y 0 6= 0) we only need to write
Z Z Z Z
dx 1 1 dy
x(y) = dy = dy = dy = + C2 ,
dy dy/dx v v(y, C1 )
which gives us an implicit solution of the equation (9.4).
Example 9.3. Solve the second order di↵erential equation:
yy 00 = (y 0 )2 ,
where y 0 and y 0 0.
Solution: Observe that in this di↵erential equation the independent
variable x is missing. Then the substitution (9.5) gives the first order
di↵erential equation
dv
yv = v2.
dy
Then separation of variables gives
Z Z
dv dy
= =) ln v = ln y + ln C1 =) v = C1 y.
v y
dy
Since v = , we have
dx
Z Z
dy dy
= C1 y =) = C1 dx =) ln y = C1 x + ln C2 ,
dx y
so that the general solution of given equation is
y(x) = C2 eC1 x .
70
equation
2
yy 00 + y 0 = yy 0 .
Solution: Given equation is reducible since the independent variable
x is missing in it. Then the substitution (9.5) gives the first order
di↵erential equation:
dv
yv + v 2 = yv.
dy
Dividing both sides by yv gives
dv v
+ = 1,
dy y
which is the first order linear equation for v(y). We find the integrating
factor: R dy
µ(y) = e y = eln y = y.
Multiplying both sides of the equation by µ(y) = y yields
dv
y +v =y
dy
or
d⇣ ⌘
yv = y.
dy
Then, integration of both sides with respect to y gives
Z
y2
yv = ydy = + C1 .
2
Dividing both sides by y, we obtain
y C1 y 2 + 2C1
v= + = .
2 y 2y
dy
Since v = y 0 = , we get
dx
dy y 2 + 2C1
= ,
dx 2y
71
which is the separable equation. Separation of variables gives

Z Z
2ydy
2
= dx =) ln (y 2 + 2C1 ) = x + C2 ,
y + 2C1
which is the general solution of given second order equation.

Problem 9.1. Find general solutions of the following reducible second
order di↵erential equations (you can assume that x, y and y 0 are all
positive).
2
a) xy 00 = y 0 b) yy 00 + (y 0 ) = 0
c) xy 00 + y 0 = 4x d) x2 1 y 00 + 2y 0 = 0
⇣ 1⌘ 0 2
e) x2 y 00 + 3xy 0 = 2 00
f) y = 2y (y )
y
2 3
g) y 00 = (x + y 0 ) h) y 00 = 2y (y 0 )
i) y 3 y 00 = 1 j) y 00 = 2yy 0
2
k) yy 00 = 3 (y 0 ) l) (tan x)y 00 = y 0
Problem 9.2. Find solution of the following initial value problem

8 00 0 2
< y = x (y )
>
y(0) = 1
>
: 0
y (0) = 2
72
10 Basic Theory of Higher Order Linear
Di↵erential Equations
Linear di↵erential equation of order n has the most general form:
a0 (x)y (n) + a1 (x)y (n 1)
+ . . . + an 1 (x)y 0 + an (x)y = F (x). (10.1)
Unless otherwise noted, we will always assume that coefficients func-

tions ai (x) and F (x) are continuous on some open interval I where we
wish to solve the equation.
Definition 10.1. If the function F (x) is identically zero on I, then
equation (10.1) is called a homogeneous linear di↵erential equation.
Thus, a homogeneous linear di↵erential equation of order n has the
most general form:
a0 (x)y (n) + a1 (x)y (n 1)
+ . . . + an 1 (x)y 0 + an (x)y = 0. (10.2)
If F (x) is not identically zero on I, then equation (10.1) is called a
nonhomogeneous linear di↵erential equation.
For instance, the second order linear equation
x2 y 00 + 2xy 0 + 3y = cos x
is nonhomogeneous, while
x2 y 00 + 2xy 0 + 3y = 0
is a homogeneous equation.
Under the assumption that a0 (x) 6= 0 at each point of interval I, we
can divide both sides of nonhomogeneous equation (10.1) by a0 (x) and
write it in the form:
y (n) + p1 (x)y (n 1)
+ . . . + pn 1 (x)y 0 + pn (x)y = f (x), (10.3)
ai (x) F (x)
where pi (x) = and f (x) = .
a0 (x) a0 (x)
73
The homogeneous linear di↵erential equation associated with equation

(10.3) has the form:
y (n) + p1 (x)y (n 1)
+ . . . + pn 1 (x)y 0 + pn (x)y = 0.
Theorem 10.1. Suppose that the functions p1 , p2 , . . . , pn and f are

continuous on the open interval I containing the point a. Then, for
any values of b0 , b1 , . . . , bn 1 the initial value problem
8
>
> y (n) + p1 (x)y (n 1) + . . . + pn 1 (x)y 0 + pn (x)y = f (x),
>
>
>
> y(a) = b0 ,
>
>
>
< y 0 (a) = b ,
1
>
>
> y 00 (a) = b2 ,
>
>
>
> .........
>
>
: (n 1)
y (a) = bn 1 .
has a unique solution on the entire interval I.
Example 10.1. Determine the longest interval in which the following

initial value problem
8 00
< (x 1)y
> 3xy 0 + 4y = sin x,
y( 2) = 2,
>
: 0
y ( 2) = 1
is certain to have a unique solution.

Solution: We first rewrite the equation in the form
3x 4 sin x
y 00 y0 + y= .
x 1 x 1 x 1
3x sin x4
Functions p1 (x) = , p2 (x) = and f (x) =
are con-
x 1 x 1 x 1
tinuous everywhere except the point x = 1. Therefore, by using the
74
Theorem 10.1, given initial value problem has a unique solution on the
open interval I which contains point x = 2 and does not contain
point x = 1. The longest interval that satisfies these conditions is
I = ( 1, 1).
10.1 Homogeneous Linear Equations

Theorem 10.2. (Principle of Superposition)
Let y1 , y2 , . . . , yn be solutions of the homogeneous linear di↵erential
equation
y (n) + p1 (x)y (n 1)
+ . . . + pn 1 (x)y 0 + pn (x)y = 0. (10.4)
on the interval I. Then for any values of constants c1 , c2 , . . . , cn the
linear combination
y = c1 y1 + c2 y2 + . . . + cn yn
is also a solution of equation (10.4) on the interval I.
Proof. If y1 , y2 , . . . , yn are solutions of the homogeneous linear equation
(10.4) on the interval I, then we have
(n) (n 1)
y1 + p1 (x)y1 + . . . + pn 1 (x)y10 + pn (x)y1 ⌘ 0,
(n) (n 1)
y2 + p1 (x)y2 + . . . + pn 1 (x)y20 + pn (x)y2 ⌘ 0,
................................................
yn(n) + p1 (x)yn(n 1)
+ . . . + pn 1 (x)yn0 + pn (x)yn ⌘ 0.
for all x 2 I. Then for y = c1 y1 + c2 y2 + . . . + cn yn we have
y (n) + p1 (x)y (n 1)
+ . . . + pn 1 (x)y 0 + pn (x)y =
(n) (n)
= c1 y1 + c2 y2 + . . . + cn yn(n) +
(n 1) (n 1)
+ p1 (x) c1 y1 + c2 y2 + . . . + cn yn(n 1)
)+
+ ... +
+ pn 1 (x) c1 y10 + c2 y20 + . . . + cn yn0 +
+ pn (x) c1 y1 + c2 y2 + . . . + cn yn =
75
(n) (n 1)
= c1 y1 + p1 (x)y1 + . . . + pn 1 (x)y10 + pn (x)y1 +
(n) (n 1)
+ c2 y2 + p1 (x)y2 + . . . + pn 1 (x)y20 + pn (x)y2 +
+ ...... +
+ cn yn(n) + p1 (x)yn(n 1)
+ . . . + pn 1 (x)yn0 + pn (x)yn =
= c1 · 0 + c2 · 0 + . . . + cn · 0 ⌘ 0
for all x 2 I. So, y = c1 y1 + c2 y2 + . . . + cn yn satisfies the homogeneous
linear equation (10.4) on the whole interval I.
Example 10.2. Consider the following homogeneous second order lin-

ear di↵erential equation
y 00 + y = 0.
By inspection one can easily show that y1 (x) = cos x and y2 (x) = sin x
are solutions of this equation. Theorem 10.2 tells us that any linear
combination of these solutions, such as
y(x) = 3y1 (x) 2y2 (x) = 3 cos x 2 sin x
is also a solution of given equation.
Example 10.3. Consider the following homogeneous third order linear

y 000 2y 00 y 0 + 2y = 0.
By inspection one can easily show that y1 (x) = ex , y2 (x) = e x and

y3 (x) = e2x are solutions of this equation. Theorem 10.2 tells us that
any linear combination of these solutions, such as
2 2
y(x) = 2y1 (x) 3y2 (x) + y3 (x) = 2ex 3e x
+ e2x
3 3
is also a solution of given equation.
76
Remark 10.1. The principle of superposition generally holds only for

homogeneous linear di↵erential equations. Consider the following two
examples.
Example 10.4. One can easily show that both y1 (x) = 1 + cos x and
y2 (x) = 1 + sin x are solutions of the second order linear di↵erential
equation
y 00 + y = 1.
However,
y(x) = y1 (x) + y2 (x) = 2 + cos x + sin x
is not a solution. The principle of superposition fails here, since the
given linear equation is not homogeneous.
p
Example 10.5. One can easily show that y1 (x) = 1 and y2 (x) = x
are solutions of the second order di↵erential equation
yy 00 + (y 0 )2 = 0.
However, p
y(x) = y1 (x) + y2 (x) = 1 + x
is not a solution. The principle of superposition fails here, since the
given equation is not linear.
Example 10.6. Verify that the functions y1 (x) = ex and y2 (x) = xex
are solutions of the second order linear di↵erential equation
y 00 2y 0 + y = 0
and then find a solution satisfying the initial conditions y(0) = 3 and
y 0 (0) = 1.
Solution: We have
y1 = e x =) y10 = ex =) y100 = ex =)
=) y100 2y10 + y1 = ex 2ex + ex ⌘ 0.
77
So, y1 (x) = ex is indeed a solution of given equation. Similarly,
y2 = xex =) y20 = (x + 1)ex =) y200 = (x + 2)ex =)

=) y200 2y20 + y2 = (x + 2)ex 2(x + 1)ex + xex ⌘ 0.
So, y2 (x) = xex is also a solution of given equation.

Then, by using the principle of superposition
y(x) = c1 y1 (x) + c2 y2 (x) = c1 ex + c2 xex
is also a solution of given linear homogeneous equation for any values

of constants c1 and c2 . Now, by using the given initial conditions we
obtain
y(0) = c1 = 3 and y 0 (0) = c1 + c2 = 1
which results in c1 = 3 and c2 = 2. Hence the solution of the original
initial value problem is
y(x) = 3ex 2xex .
Note that from the Theorem 10.1 we know that this solution is unique.
Definition 10.2. (Linear Dependence of Functions)

The n functions f1 , f2 , . . . , fn are called linearly dependent on inter-
val I if there exist constants c1 , c2 , . . . , cn , not all zero, such that
c1 f1 (x) + c2 f2 (x) + . . . + cn fn (x) = 0 (10.5)
holds for all x 2 I.
Example 10.7. Functions
f1 (x) = sin 2x and f2 (x) = sin x cos x
are linearly dependent on any interval I because
1 · f1 (x) + ( 2) · f2 (x) = sin 2x 2 sin x cos x ⌘ 0.
78

f1 (x) = sin 2x, f2 (x) = sin x cos x, f3 (x) = ex
1 · f1 (x) + ( 2) · f2 (x) + 0 · f3 (x) = sin 2x 2 sin x cos x ⌘ 0.

f1 (x) = 2x, f2 (x) = 3x2 , f3 (x) = 5x 8x2
15
· f1 (x) + 8 · f2 (x) + 3 · f3 (x) = 15x + 24x2 + 3(5x 8x2 ) ⌘ 0.
2
Remark 10.2. If not all the coefficients in (10.5) are zero, then clearly
we can solve for at least one of the functions as a linear combination
of the others. Thus the functions f1 , f2 , . . . , fn are linearly dependent
if and only if at least one of them is a linear combination of the others.
Definition 10.3. (Linear Independence of Functions)

The n functions f1 , f2 , . . . , fn are said to be linearly independent on
the interval I if they are not linearly dependent there. Equivalently,
they are linearly independent on I if the equality (10.5) holds on I
only in the trivial case c1 = c2 = . . . = cn = 0.

f1 (x) = ex and f2 (x) = e2x
are linearly independent on any interval I. We can verify it in the
following way. We di↵erentiate both sides of c1 ex + c2 e2x = 0 to obtain
c1 ex + 2c2 e2x = 0 which must also hold on I. Subtracting from this
equality the initial one we get c2 e2x = 0, which can hold only in the
case c2 = 0. But if c2 = 0 then c1 = 0. So, c1 f1 (x) + c2 f2 (x) = 0 holds
only in the trivial case c1 = c2 = 0 and therefore by definition f1 and
f2 are linearly independent.
79

f1 (x) = 1, f2 (x) = x, f3 (x) = x2
are linearly independent on any interval I. We can verify it in the
following way. We di↵erentiate both sides of c1 + c2 x + c3 x2 = 0 to
obtain c2 + 2c3 x = 0 which must also hold on I. Di↵erentiating again
both sides gives 2c3 = 0, which can hold only in the case c3 = 0. If
c3 = 0 then from c2 +2c3 x = 0 we get c2 = 0. Finally, if c2 = c3 = 0 then
from c1 +c2 x+c3 x2 = 0 we get c1 = 0. So, c1 f1 (x)+c2 f2 (x)+c3 f3 (x) = 0
holds only in the trivial case c1 = c2 = c3 = 0 and therefore by
definition function f1 , f2 , f3 are linearly independent.
Definition 10.4. Given n functions f1 , f2 , . . . , fn , the Wronskian of

them, denoted by W (f1 , f2 , . . . , fn ), is the determinant
f1 f2 ... fn
f10 f20 ... fn0
W (f1 , f2 , . . . , fn ) = f100 f200 ... fn00
... ... ... ...
(n 1) (n 1) (n 1)
f1 f2 . . . fn
Theorem 10.3. If functions f1 , f2 , . . . , fn are linearly dependent on

the interval I then W (f1 , f2 , . . . , fn ) ⌘ 0 on I.
Proof. If the functions f1 , f2 , . . . , fn are linearly dependent on I then
by definition there exist constants c1 , c2 , . . . , cn , not all zero, such that
(10.5) holds for all x 2 I. Di↵erentiating (10.5) n 1 times in succes-
sion, we obtain the n equations
c1 f1 (x) + c2 f2 (x) + . . . + cn fn (x) = 0,
c1 f10 (x) + c2 f20 (x) + . . . + cn fn0 (x) = 0,
c1 f100 (x) + c2 f200 (x) + . . . + cn fn00 (x) = 0, (10.6)
........................
(n 1) (n 1) (n 1)
c1 f 1 (x) + c2 f2 (x) + . . . + cn fn (x) = 0,
80
which must all hold for all x 2 I. We recall from linear algebra that
a system of n linear homogeneous algebraic equations in n unknowns
has a nontrivial solution if and only if the determinant of coefficients
is zero. In (10.6) the unknowns are the constants c1 , c2 , . . . , cn and the
determinant of coefficients is simply the Wronskian W (f1 , f2 , . . . , fn )
evaluated at the point x 2 I. Because we know that the ci are not all
zero, it follows that W (f1 , f2 , . . . , fn ) ⌘ 0.
Theorem 10.4. If W (f1 , f2 , . . . , fn ) 6= 0 for at least one point of in-

terval I, then functions f1 , f2 , . . . , fn are linearly independent on I.
Proof. The proof of this theorem directly follows from the previous
theorem.
Example 10.12. Let f1 (x) = cos x and f2 (x) = sin x. Then

f1 f2 cos x sin x
W (f1 , f2 ) = = = cos2 x + sin2 x = 1 6= 0.
f10 f20 sin x cos x
Therefore, functions f1 (x) = cos x and f2 (x) = sin x are linearly inde-
pendent on any interval I.
Example 10.13. Let f1 (x) = ex and f2 (x) = e2x . Then

f1 f2 ex e2x
W (f1 , f2 ) = = = 2e3x e3x = e3x 6= 0.
f10 f20 e x
2e 2x
Therefore, functions f1 (x) = ex and f2 (x) = e2x are linearly indepen-

dent on any interval I.
Example 10.14. Let f1 (x) = ex and f2 (x) = xex . Then

f1 f2 ex xex
W (f1 , f2 ) = = = e2x 6= 0.
f10 f20 e x
(x + 1)e x
Therefore, functions f1 (x) = ex and f2 (x) = xex are linearly indepen-

dent on any interval I.
81
Example 10.15. Let f1 (x) = 1, f2 (x) = x and f3 (x) = x2 . Then
f1 f2 f3 1 x x2
W (f1 , f2 , f3 ) = f10 f20 f30 = 0 1 2x = 2 6= 0.
f100 f200 f300 0 0 2
Therefore, functions f1 (x) = 1, f2 (x) = x, f3 (x) = x2 are linearly

independent on any interval I.
3x
Example 10.16. Let f1 (x) = e , f2 (x) = cos 2x and f3 (x) = sin 2x.
Then
3x
f1 f2 f3 e cos 2x sin 2x
3x
W (f1 , f2 , f3 ) = f10 f20 f30 = 3e 3x
2 sin 2x 2 cos 2x = 26e 6= 0.
f100 f200 f300 9e 3x
4 cos 2x 4 sin 2x
Therefore, functions f1 (x) = e 3x , f2 (x) = cos 2x, f3 (x) = sin 2x are

linearly independent on any interval I.
Remark 10.3. If W (f1 , f2 , . . . , fn ) ⌘ 0 on interval I then we cannot

say anything about linear dependence and linear independence of func-
tions f1 , f2 , . . . , fn on interval I. Consider the following two examples.
f1 (x) = x2 and f2 (x) = 2x2
are linearly dependent on the whole real line because
( 2) · f1 (x) + 1 · f2 (x) = 2x2 + 2x2 ⌘ 0.
Note that for these functions W (f1 , f2 ) ⌘ 0 on the entire real line.
82

(
2
x2 , if x 0
f1 (x) = x and f2 (x) =
x2 , if x < 0
are linearly independent on the whole real line. We can verify it in the
following way. The equality
c1 f1 (x) + c2 f2 (x) = 0
must hold on entire real line. Putting x = 1 and x = 1 we get

c1 + c2 = 0 and c1 c2 = 0, respectively. Last two equalities can hold
simultaneously only if c1 = c2 = 0. So, c1 f1 (x) + c2 f2 (x) = 0 holds only
in the trivial case c1 = c2 = 0 and therefore by definition f1 and f2 are
linearly independent. Note that for these functions W (f1 , f2 ) ⌘ 0 on
the entire real line.
Theorem 10.5. Suppose that y1 , y2 , . . . , yn are solutions of the homo-

geneous n-th order linear di↵erential equation
y (n) + p1 (x)y (n 1)
+ . . . + pn 1 (x)y 0 + pn (x)y = 0
on an open interval I on which all coefficient functions pi (x) are con-

tinuous. Then
(i) y1 , y2 , . . . , yn are linearly dependent on I if and only if
W (y1 , y2 , . . . , yn ) = 0 for all x 2 I
(ii) y1 , y2 , . . . , yn are linearly independent on I if and only if
W (y1 , y2 , . . . , yn ) 6= 0 for all x 2 I
83
Theorem 10.6. Suppose that coefficient functions pi (x) are continu-

ous on an interval I. Then the homogeneous n-th order linear di↵er-
ential equation
y (n) + p1 (x)y (n 1)
+ . . . + pn 1 (x)y 0 + pn (x)y = 0 (10.7)
has n linearly independent solutions on interval I.

Proof. Let a be an arbitrary point of interval I. By the Theorem 10.1,
equation (10.7) has a unique solution, say y1 (x), that satisfies initial
conditions
(n 2) (n 1)
y1 (a) = 1, y10 (a) = 0, y100 (a) = 0, . . . , y1 (a) = 0, y1 (a) = 0.
Next, by using the Theorem 10.1, equation (10.7) has a unique solution,
say y2 (x), that satisfies initial conditions
(n 2) (n 1)
y2 (a) = 0, y20 (a) = 1, y200 (a) = 0, . . . , y2 (a) = 0, y2 (a) = 0.
Next, by using the Theorem 10.1, equation (10.7) has a unique solution,
say y3 (x), that satisfies initial conditions
(n 2) (n 1)
y3 (a) = 0, y30 (a) = 0, y300 (a) = 1, . . . , y3 (a) = 0, y3 (a) = 0.
In this way, we continue obtaining solutions y4 (x), . . . , yn (x). The

Wronskian of these solutions, W (y1 , y2 , . . . , yn ), evaluated at x = a
is
y1 (a) y2 (a) ... yn (a) 1 0 ... 0
y10 (a) y20 (a) ... yn0 (a) 0 1 ... 0
= = 1 6= 0
... ... ... ... ... ... ... ...
(n 1) (n 1) (n 1)
y1 (a) y2 (a) . . . yn (a) 0 0 ... 1
and therefore from the Theorem 10.5 it follows that these solutions
y1 (x), y2 (x), . . . , yn (x) are linearly independent on I.
84
Theorem 10.7. Let y1 , y2 , . . . , yn be any n linearly independent solu-

tions of the homogeneous linear di↵erential equation
y (n) + p1 (x)y (n 1)
+ . . . + pn 1 (x)y 0 + pn (x)y = 0 (10.8)
on an open interval I where all coefficients functions pi (x) are con-

tinuous. Then every solution Y (x) of the equation (10.8) can be ex-
pressed as a linear combination of y1 , y2 , . . . , yn ; so there exist constants
c1 , c2 , . . . , cn such that
Y (x) = c1 y1 (x) + c2 y2 (x) + . . . + cn yn (x)
for all x 2 I.
Proof. Since y1 , y2 , . . . , yn are linearly independent solutions of equa-

tion (10.8), from the Theorem 10.5 we have W (y1 , y2 , . . . , yn )(x) 6= 0
at each point of interval I. Let Y (x) be any solution of equation (10.8)
and let a be an arbitrary point of interval I. We first show that there
exist constants c1 , c2 , . . . , cn such that the following equalities
c1 y1 (a) + c2 y2 (a) + . . . + cn yn (a) = Y (a),

c1 y10 (a) + c2 y20 (a) + . . . + cn yn0 (a) = Y 0 (a),
(10.9)
.......................................
(n 1) (n 1)
c1 y1 (a) + c2 y2 (a) + . . . + cn yn(n 1)
(a) = Y (n 1)
(a)
hold simultaneously. Indeed, (10.9) can be considered as a system

of linear algebraic equations in the unknowns c1 , c2 , . . . , cn . The de-
terminant of the coefficients in this system is simply the Wronskian
W (y1 , y2 , . . . , yn ) evaluated at x = a. Since a 2 I, we have
W (y1 , y2 , . . . , yn )(a) 6= 0 and therefore from the linear algebra it fol-
lows that system (10.9) can be solved for c1 , c2 , . . . , cn . Now, with these
values of c1 , c2 , . . . , cn , we define the function
G(x) = c1 y1 (x) + c2 y2 (x) + . . . + cn yn (x)
85
which is a solution of equation (10.8) by the principle of superposition.

Obviously we have
G(a) = Y (a), G0 (a) = Y 0 (a), . . . , G(n 1)

(a) = Y (n 1)
(a).
Thus we have two solutions of equation (10.8), Y (x) and G(x), both
satisfying the same initial conditions. But from the Theorem 10.1 we
have the uniqueness of a solution of corresponding initial value problem
and therefore we get
Y (x) ⌘ G(x) = c1 y1 (x) + c2 y2 (x) + . . . + cn yn (x).
Remark 10.4. From the Theorem 10.7 it follows that having n linearly
independent solutions y1 , y2 , . . . , yn , every solution y(x) of the n-th
order homogeneous linear di↵erential equation (10.8) can be written
in the form
y(x) = c1 y1 (x) + c2 y2 (x) + . . . + cn yn (x). (10.10)
We therefore say that (10.10) is a general solution of equation (10.8).

So, to find the general solution of the n-th order homogeneous linear
di↵erential equation (10.8) it is enough to find its n linearly indepen-
dent solutions and form the linear combination of them.
Definition 10.5. If y1 , y2 , . . . , yn are linearly independent solutions of

the n-th order homogeneous linear di↵erential equation, then the set
{y1 , y2 , . . . , yn } is called the fundamental set of solutions of this
di↵erential equation.
Example 10.19. Show first that functions y1 (x) = x, y2 (x) = x ln x

and y3 (x) = x2 are linearly independent solutions of the third order
linear di↵erential equation
x3 y 000 x2 y 00 + 2xy 0 2y = 0
86
on the open interval I = (0, 1). Then write the general solution of
this equation.
Solution: We have
y1 = x =) y10 = 1 =) y100 = y1000 = 0 =)
=) x3 y1000 x2 y100 + 2xy10 2y1 = 2x 2x ⌘ 0.
So, y1 (x) = x is indeed a solution of given equation. Similarly,

1 1
y2 = x ln x =) y20 = 1 + ln x =) y200 = =) y2000 =
x x2
=) x3 y2000 x2 y200 + 2xy20 2y2 = x x + 2x(1 + ln x) 2x ln x ⌘ 0.
So, y2 (x) = x ln x is also a solution of given equation. Finally,
y3 = x2 =) y30 = 2x =) y300 = 2 =) y3000 = 0 =)

=) x3 y3000 x2 y300 + 2xy30 2y3 = 2x2 + 4x2 2x2 ⌘ 0.
So, y3 (x) = x2 is also a solution of given equation.

We have
y1 y2 y3 x x ln x x2
W (y1 , y2 , y3 ) = y10 y20 y30 = 1 1 + ln x 2x = x 6= 0
1
y100 y200 y300 0 x 2
for x > 0. Then, from the Theorem 10.5 it follows that y1 (x) = x,
y2 (x) = x ln x, y3 (x) = x2 are linearly independent solutions of given
equation. Hence the general solution of given equation is
y(x) = c1 y1 (x) + c2 y2 (x) + c3 y3 (x) = c1 x + c2 x ln x + c3 x2 .
87
10.2 Nonhomogeneous Linear Equations

Theorem 10.8. Let yp (x) be a particular solution of the nonhomoge-
neous linear di↵erential equation
y (n) + p1 (x)y (n 1)
+ . . . + pn 1 (x)y 0 + pn (x)y = f (x) (10.11)
on an open interval I, where all coefficient functions pi (x) and f (x)
are continuous. Let yh (x) be a general solution of the associated ho-
mogeneous linear di↵erential equation
y (n) + p1 (x)y (n 1)
+ . . . + pn 1 (x)y 0 + pn (x)y = 0 (10.12)
on the interval I. Then the general solution of nonhomogeneous equa-
tion (10.11) on the interval I has the form
y(x) = yh (x) + yp (x).
Remark 10.5. From the Theorem 10.8 it follows that in order to find
the general solution of the n-th order nonhomogeneous linear di↵eren-
tial equation (10.11), we first have to find the general solution of the
associated homogeneous equation (10.12) and then we have to find one
particular solution of equation (10.11).
Example 10.20. Consider the following second order linear nonho-

mogeneous equation
y 00 + 4y = 12x.
It is obvious that yp (x) = 3x is a particular solution of this equation.
The associated homogeneous equation
y 00 + 4y = 0
has two linearly independent solutions y1 = cos 2x and y2 = sin 2x (ver-
ify yourself). Therefore, the general solution of homogeneous equation
has the form
yh (x) = c1 y1 + c2 y2 = c1 cos 2x + c2 sin 2x.
88
Then the general solution of given nonhomogeneous equation is

y(x) = yh (x) + yp (x) = c1 cos 2x + c2 sin 2x + 3x.

Problem 10.1. In each of the following problems determine the longest
interval in which the given initial value problem is certain to have a
unique solution. Do not attempt to find the solution.
8 00 8 00
>
< xy + 3y = x, < (x 1)y
> 3xy 0 + 4y = sin x,
a) y(1) = 1, b) y( 2) = 2,
>
: 0 >
: 0
y (1) = 2. y ( 2) = 1.
8 00 0
8 00 0
< x(x 4)y + 3xy + 4y = 2,
> < y + (cos x)y + 3(ln |x|)y = 0,
>
c) y(3) = 0, d) y(2) = 3,
>
: 0 >
: 0
y (3) = 1. y (2) = 1.
8 8
>
> x3 y 000 x2 y 00 + xy 0 + 3y = 0, >
> (x 1)y 000 xy 00 + 3y = cos 2x,
>
> >
>
< y(1) = 0, < y(0) = 1,
e) f)
>
>
> y 0 (1) = 1, >
>
> y 0 (0) = 1,
>
: y 00 (1) = 2. >
: y 00 (0) = 2.
Problem 10.2. Show that y1 = x2 and y2 = x3 are two di↵erent

solutions of the initial value problem
8 2 00
<xy
> 4xy 0 + 6y = 0,
y(0) = 0,
>
: 0
y (0) = 0.
Explain why these facts do not contradict Theorem 10.1 (with respect
to the guaranteed uniqueness).
89
Problem 10.3. Determine whether the the following pairs of functions

are linearly independent or linearly dependent on the whole real line.
a) f1 (x) = ⇡, f2 (x) = sin2 x + cos2 x
b) f1 (x) = x + 1, f2 (x) = x2
c) f1 (x) = x, f2 (x) = |x|
d) f1 (x) = x3 , f2 (x) = x2 |x|
e) f1 (x) = 1 + x, f2 (x) = 1 + |x|
f) f1 (x) = xex , f2 (x) = |x|ex
g) f (x) = sin2 x, f2 (x) = 1 cos 2x
h) f1 (x) = ex sin x, f2 (x) = ex cos x
i) f1 (x) = 2 cos x + 3 sin x, f2 (x) = 3 cos x 2 sin x
Problem 10.4. In the following problems determine whether the given

functions are linearly independent or linearly dependent on the indi-
cated interval I.
a) f1 (x) = 0, f2 (x) = sin x, f3 (x) = ex , I = ( 1, 1)
b) f1 (x) = 17, f2 (x) = cos2 x, f3 (x) = cos 2x, I = ( 1, 1)
c) f1 (x) = ex , f2 (x) = e2x , f3 (x) = e3x , I = ( 1, 1)
d) f1 (x) = ex , f2 (x) = cos x, f3 (x) = sin x, I = ( 1, 1)
e) f1 (x) = x, f2 (x) = cos (ln x), f3 (x) = sin (ln x), I = (0, 1)
90
Problem 10.5. First show that y1 = x3 and y2 = |x|3 are linearly

independent solutions on the whole real line of the equation
x2 y 00 3xy 0 + 3y = 0
and then verify that W (y1 , y2 ) ⌘ 0. Why do these facts do not contra-
dict Theorem 10.5?
Problem 10.6. Show that y1 = sin (x2 ) and y2 = cos (x2 ) are linearly
independent functions, but that their Wronskian vanishes (becomes
zero) at x = 0. Why does this imply that there is no di↵erential
equation of the form y 00 + p(x)y 0 + q(x)y = 0, with both p and q
continuous everywhere, having both y1 and y2 as solutions?
Problem 10.7. In the following problems verify that given functions

y1 and y2 are linearly independent solutions of the given second order
linear di↵erential equation on the given interval I. Then write the
general solution of the given equation.
a) y 00 + 4y = 0, y1 = cos 2x, y2 = sin 2x, I = ( 1, 1)
b) y 00 2y 0 + y = 0, y1 = ex , y2 = xex , I = ( 1, 1)
c) x2 y 00 x(x + 2)y 0 + (x + 2)y = 0, y1 = x, y2 = xex , I = (0, 1)
d) (1 x cot x)y 00 xy 0 + y = 0, y1 = x, y2 = sin x, I = (0, ⇡)
Problem 10.8. In the following problems verify that given functions

y1 , y2 and y3 are linearly independent solutions of the given third order
linear di↵erential equation on the given interval I. Then write the
general solution of the given equation.
a) y 000 + 2y 00 y0 2y = 0, y1 = ex , y2 = e x , y3 = e 2x
, I=R
b) y 000 + 9y 0 = 0, y1 = 1, y2 = cos 3x, y3 = sin 3x, I = R
91
Problem 10.9. Consider the following third order linear di↵erential

equation
x3 y 000 + 6x2 y 00 + 4xy 0 4y = 0.
1 ln x
a) Show that functions y1 (x) = x, y2 (x) = 2 and y3 (x) = 2
x x
are linearly independent solutions of the given equation on the
interval I = (0, 1).
b) Write the general solution of the given equation.
c) Find the solution of the initial value problem

8
>
> x3 y 000 + 6x2 y 00 + 4xy 0 4y = 0,
>
>
< y(1) = 1,
>
>
> y 0 (1) = 5,
>
: y 00 (1) = 11.
92
11 Homogeneous Linear Equations
We have already mentioned that a general solution of an n-th order
homogeneous linear di↵erential equation is a linear combination of n
linearly independent solutions. In this section we illustrate how to find
those linearly independent solutions. Note that it is only possible for
some specific types of linear equations.
11.1 Homogeneous linear equations with constant

coefficients
Consider the following homogeneous n-th order linear di↵erential equa-
tion
an y (n) + an 1 y (n 1)
+ . . . + a2 y 00 + a1 y 0 + a0 y = 0, (11.1)
where coefficients a0 , a1 , . . . , an are real constants with an 6= 0. We
search the solution of equation (11.1) in the form y = erx . Di↵erenti-
ating this function n times in succession we have
y = erx =) y 0 = rerx =) y 00 = r2 erx =) . . . =) y (n) = rn erx .
Putting these in equation (11.1), we have
an rn erx + an 1 rn 1 erx + . . . + a2 r2 erx + a1 rerx + a0 erx = 0.
Dividing both sides by erx , we get
an r n + an 1 r n 1
+ . . . + a2 r2 + a1 r + a0 = 0. (11.2)
Algebraic equation (11.2) is called the characteristic equation of
the di↵erential equation (11.1). If r is the root of algebraic equation
(11.2) then the corresponding function y = erx is the solution of the
homogeneous di↵erential equation (11.1).
Remark 11.1. According to fundamental theorem of algebra, every
n-th degree polynomial, such as the one in equation (11.2), has n zeros,
though not necessarily distinct and not necessarily real.
93
11.1.1 Distinct real roots

Theorem 11.1. If the roots r1 , r2 , . . . , rn of the characteristic equation
(11.2) are real and distinct, then
y(x) = c1 er1 x + c2 er2 x + . . . + cn ern x (11.3)
is a general solution of di↵erential equation (11.1).
Example 11.1. Find the general solution of di↵erential equation
2y 00 7y 0 + 3y = 0.
Solution: Given di↵erential equation has a characteristic equation
2r2 7r + 3 = 0.
It can be solved by factoring:
(r 3)(2r 1) = 0
and so the characteristic equation has two distinct real roots r1 = 3

and r2 = 1/2. Then the general solution of given di↵erential equation
is
y(x) = c1 e3x + c2 ex/2 .
Example 11.2. Solve the following initial value problem

8
>
> y 000 + 3y 00 10y 0 = 0,
>
>
< y(0) = 7,
> 0
>
> y (0) = 0,
>
: y 00 (0) = 70.
r3 + 3r2 10r = 0.
94
r(r + 5)(r 2) = 0
and so the characteristic equation has three distinct real roots r1 = 0,

r2 = 5 and r3 = 2. Then the general solution of given di↵erential
equation is
y(x) = c1 + c2 e 5x + c3 e2x .
Applying given initial conditions yields the linear equations
y(0) = c1 + c2 + c3 = 7,
y 0 (0) = 5c2 + 2c3 = 0,
y 00 (0) = 25c2 + 4c3 = 70
in the coefficients c1 , c2 , c3 . Solving this linear system we obtain c1 = 0,

c2 = 2, c3 = 5. Thus the solution of given initial value problem is
5x
y(x) = 2e + 5e2x .
11.1.2 Repeated real roots

Theorem 11.2. If the characteristic equation (11.2) has a repeated
real root r of multiplicity k, then the part of a general solution of the
di↵erential equation (11.1) corresponding to this root r has the form
c1 + c2 x + c3 x2 + . . . + ck xk 1
erx . (11.4)
y 00 + 2y 0 + y = 0.
r2 + 2r + 1 = 0
95
or
(r + 1)2 = 0.
So, the characteristic equation has one repeated real root r = 1 of
multiplicity 2. Then the general solution of given di↵erential equation
is
y(x) = (c1 + c2 x)e x .
Example 11.4. Find the general solution of the fifth order di↵erential
equation
9y (5) 6y (4) + y (3) = 0.
9r5 6r4 + r3 = 0
or
r3 (3r 1)2 = 0.
So, the characteristic equation has one repeated real root r = 0 of
multiplicity 3 and another repeated real root r = 13 of multiplicity 2.
Then the general solution of given di↵erential equation is
y(x) = c1 + c2 x + c3 x2 + (c4 + c5 x)ex/3 .
11.1.3 Unrepeated complex roots

Theorem 11.3. If the characteristic equation (11.2) has an unrepeated
pair of complex roots r1 = ↵ + i and r2 = ↵ i, then the corre-
sponding part of a general solution of the di↵erential equation (11.1)
has the form
c1 cos x + c2 sin x e↵x . (11.5)
96
y 00 + 2y 0 + 5y = 0.
r2 + 2r + 5 = 0.
Completion of the square yields
(r + 1)2 + 4 = 0.
Thus we have r + 1 = ±2i and we obtain two unrepeated complex

roots r1 = 1 + 2i and r2 = 1 2i. Then the general solution of
given di↵erential equation is
y(x) = (c1 cos 2x + c2 sin 2x)e x .

8 00
<y
> 4y 0 + 5y = 0,
y(0) = 1,
>
: 0
y (0) = 5.
r2 4r + 5 = 0.
(r 2)2 + 1 = 0.
Thus we have r 2 = ±i and we obtain two unrepeated complex roots

r1 = 2 + i and r2 = 2 i. Then the general solution of the di↵erential
equation is
y(x) = (c1 cos x + c2 sin x)e2x .
97
Applying given initial conditions yields

y(0) = c1 = 1,
y 0 (0) = 2c1 + c2 = 5,
which results in c1 = 1 and c2 = 3. Thus the solution of given initial
value problem is
y(x) = (cos x + 3 sin x)e2x .

y (4) + 4y = 0.
r4 + 4 = 0.
This characteristic equation has one pair of unrepeated complex roots
r1 = 1 + i, r2 = 1 i and another pair of unrepeated complex roots
r3 = 1 + i, r4 = 1 i. Then the general solution of given di↵erential
equation is
y(x) = c1 cos x + c2 sin x ex + c3 cos x + c4 sin x e x .

y 000 + y 0 10y = 0.
r3 + r 10 = 0.
(r 2)(r2 + 2r + 5) = 0
and so the characteristic equation has one unrepeated real root r1 = 2
and a pair of unrepeated complex roots r2 = 1 + 2i, r3 = 1 2i.
Then the general solution of given di↵erential equation is
y(x) = c1 e2x + c2 cos 2x + c3 sin 2x e x .
98
11.1.4 Repeated complex roots

Theorem 11.4. If the characteristic equation (11.2) has a repeated
pair of complex roots r1 = ↵ + i and r2 = ↵ i of multiplicity
k, then the corresponding part of a general solution of the di↵erential
equation (11.1) has the form
c1 + c2 x + c3 x2 + . . . + ck xk 1
e↵x cos x+
(11.6)
2 k 1 ↵x
+ ck+1 + ck+2 x + ck+3 x + . . . + c2k x e sin x.

y (4) + 18y 00 + 81y = 0.

r4 + 18r2 + 81 = 0.
(r2 + 9)2 = 0
and so the characteristic equation has a pair of repeated complex roots
r1 = 3i and r2 = 3i of multiplicity 2. Then the general solution of
given di↵erential equation is
y(x) = c1 + c2 x cos 3x + c3 + c4 x sin 3x.
11.2 Euler Equations

Definition 11.1. The following homogeneous n-th order linear di↵er-
ential equation
an xn y (n) + an 1 xn 1 y (n 1)
+ . . . + a2 x2 y 00 + a1 xy 0 + a0 y = 0 (11.7)
is called n-th order Euler equation. Here coefficients a0 , a1 , . . . , an
are real constants with an 6= 0.
99
For example,
ax2 y 00 + bxy 0 + cy = 0
is a second order Euler equation and
ax3 y 000 + bx2 y 00 + cxy 0 + ky = 0
is a third order Euler equation.
Theorem 11.5. If x > 0 then the substitution t = ln x transforms the

second order Euler equation
ax2 y 00 + bxy 0 + cy = 0 (11.8)
into the following linear equation with constant coefficients
d2 y dy
a 2 + (b a) + cy = 0. (11.9)
dt dt
Proof. Using the chain rule, we have
dy dy dt 1 dy
y0 = = · = ·
dx dt dx x dt
and
✓ ◆ ✓ ◆ ✓ ◆
00 d dy d 1 dy 1 dy 1 d dy
y = = · = · + · =
dx dx dx x dt x2 dt x dx dt
✓ ◆ ✓ ◆
1 dy 1 d dy dt 1 d2 y dy
= · + · · = .
x2 dt x dt dt dx x2 dt2 dt
0 dy 2 00 d2 y dy
So, xy = and x y = 2 . Putting these in equation (11.8),
dt dt dt
we immediately obtain the equation (11.9).
Example 11.10. Find the general solution of the second order Euler
equation
x2 y 00 + xy 0 y = 0, x > 0.
100
Solution: The substitution t = ln x transforms given Euler equation

d2 y
y=0
dt2
with associated characteristic equation
r2 1 = 0.
Two distinct real roots r = ±1 of the characteristic equation allow us

to write the general solution
y = c1 e t + c2 e t .
Now, making a back substitution et = x, we obtain the general solution

of given Euler equation
c2
y(x) = c1 x + .
x
Theorem 11.6. If x > 0 then the substitution t = ln x transforms the

third order Euler equation
ax3 y 000 + bx2 y 00 + cxy 0 + ky = 0 (11.10)
d3 y d2 y dy
a 3 + (b 3a) 2 + (c b + 2a) + ky = 0. (11.11)
dt dt dt
Proof. Prove yourself.
Example 11.11. Find the general solution of the third order Euler
equation
x3 y 000 + 6x2 y 00 + 7xy 0 + y = 0, x > 0.
101
Solution: The substitution t = ln x transforms given Euler equation

d3 y d2 y dy
+ 3 + 3 +y =0
dt3 dt2 dt
with associated characteristic equation
r3 + 3r2 + 3r + 1 = 0.
This equation has one repeated real root r = 1 of multiplicity 3.

Then the general solution has the form
y = c1 + c2 t + c3 t 2 e t .
Now, making a back substitution et = x, we obtain the general solution

of given Euler equation
c1 + c2 ln x + c3 ln2 x
y(x) = .
x
11.3 Reduction of Order

We have seen so far that in some cases we can find the general solution
of the homogeneous second order linear di↵erential equation
a(x)y 00 + b(x)y 0 + c(x)y = 0. (11.12)
Solving this equation in general without any prior knowledge is not

possible. However, if we knew one solution of the equation (11.12) we
could find its general solution by using the method of reduction of
order.
Suppose that one solution y1 (x) of the homogeneous second order linear
di↵erential equation (11.12) is known. We search the second linearly
independent solution of equation (11.12) in the form:
y2 (x) = v(x)y1 (x).
102
Then
y20 = v 0 y1 + vy10 and y200 = v 00 y1 + 2v 0 y10 + vy100 .
Putting these in equation (11.12) we get
a(x)y200 + b(x)y20 + c(x)y2 = 0,
a(x) v 00 y1 + 2v 0 y10 + vy100 + b(x) v 0 y1 + vy10 + c(x)vy1 = 0,
a(x)y1 v 00 + 2a(x)y10 + b(x)y1 v 0 + a(x)y100 + b(x)y10 + c(x)y1 v = 0,
a(x)y1 v 00 + 2a(x)y10 + b(x)y1 v 0 = 0.
Resulting equation is reducible second order di↵erential equation. With
substitution z = v 0 and z 0 = v 00 we obtain the first order separable
equation
p(x)z 0 + q(x)z = 0,
where p(x) = a(x)y1 (x) and q(x) = 2a(x)y10 (x) + b(x)y1 (x). Solving
this equation for z and then obtaining v from z = v 0 allows us to find
the second linearly independent solution y2 of equation (11.12).
Example 11.12. Find the general solution of the second order di↵er-
ential equation
(x + 1)y 00 (x + 2)y 0 + y = 0, x> 1
if one solution y1 (x) = ex is given.
Solution: We search a second linearly independent solution y2 in the
form:
y2 (x) = v(x)y1 (x) = v(x)ex .
Then
y20 = (v 0 + v)ex and y200 = (v 00 + 2v 0 + v)ex .
Putting these in given equation we get
(x + 1)y200 (x + 2)y20 + y2 = 0,
(x + 1) v 00 + 2v 0 + v ex (x + 2) v 0 + v ex + vex = 0,
(x + 1)ex v 00 + xex v 0 = 0,
103
(x + 1)v 00 + xv 0 = 0.
Substituting z = v 0 and z 0 = v 00 gives us
(x + 1)z 0 + xz = 0,
which is a first order separable equation. Separation of variables gives
dz x
= dx.
z x+1
Integrating both sides we obtain
Z Z ✓ ◆
dz 1
= 1 dx =) ln z = ln (x + 1) x + C̃.
z x+1
We take C̃ = 0. So, z(x) = (x + 1)e x . Since z = v 0 , we have
Z Z
v(x) = z(x)dx = (x + 1)e x dx = (x + 2)e x + C̃. ˜
We take C̃˜ = 0. Then

y2 (x) = v(x)y1 (x) = (x + 2)e x ex = (x + 2).
Then the general solution of given equation is
y(x) = c1 y1 (x) + c2 y2 (x) = c1 ex c2 (x + 2).

Problem 11.1. Find general solutions of the following second order
di↵erential equations.
a) y 00 + 2y 0 15y = 0 b) 2y 00 + 3y 0 = 0
c) 4y 00 + 8y 0 + 3y = 0 d) 9y 00 12y 0 + 4y = 0
e) 35y 00 y0 12y = 0 f) y 00 + 9y = 0
g) y 00 3y 0 + 2y = 0 h) y 00 + 6y 0 + 9y = 0
104
i) y 00 6y 0 + 13y = 0 j) y 00 + 8y 0 + 25y = 0
k) y 00 4y = 0 l) y 00 + 3y 0 10y = 0
Problem 11.2. In each of the following problems a general solution

y(x) of a homogeneous second order di↵erential equation with constant
coefficients is given. Find such an equation.
10x
a) y(x) = c1 + c2 e b) y(x) = c1 e10x + c2 e 10x
10x 10x
c) y(x) = c1 e + c2 xe d) y(x) = c1 e10x + c2 e100x
⇣ p p ⌘
x x 2 x 2
e) y(x) = c1 + c2 x f) y(x) = e c1 e + c2 e
x
g) y(x) = c1 cos 2x + c2 sin 2x h) y(x) = e (c1 cos 3x + c2 sin 3x)
Problem 11.3. Find general solutions of the following di↵erential

equations.
a) y (4) 8y 000 + 16y 00 = 0 b) y (4) + 3y 00 4y = 0
c) y (4) 8y 00 + 16y = 0 d) y 000 + y 00 y0 y=0
e) y (4) 3y 000 + 3y 00 y0 = 0 f) 9y 000 + 12y 00 + 4y 0 = 0
g) y 000 3y 0 2y = 0 h) 6y (4) + 11y 00 + 4y = 0
i) y 000 + 3y 00 4y = 0 j) y (4) 16y = 0
Problem 11.4. In the following problems, one solution of the di↵er-

ential equation is given. Find the general solution of given equation.
a) y 000 + 3y 00 54y = 0, y(x) = e3x
b) 3y 000 2y 00 + 12y 0 8y = 0, y(x) = cos 2x
105
c) 3y 000 + 4y 00 + 11y 0 10y = 0, y(x) = e2x/3
d) 6y (4) + 5y 000 + 25y 00 + 20y 0 + 4y = 0, y(x) = cos 2x

e) 9y 000 + 11y 00 + 4y 0 14y = 0, y(x) = e x
sin x
Problem 11.5. In each of the following problems a general solution

y(x) of a homogeneous linear di↵erential equation with constant coef-
ficients is given. Find such an equation.
a) y(x) = c1 + c2 x + c3 x2 e2x
b) y(x) = c1 e2x + c2 cos 2x + c3 sin 2x

c) y(x) = c1 + c2 x + c3 x2 cos 2x + c4 + c5 x + c6 x2 sin 2x
Problem 11.6. Find the general solutions of the following Euler equa-
tions for x > 0.
a) 4x2 y 00 + 8xy 0 3y = 0 b) x2 y 00 3xy 0 + 4y = 0
c) x2 y 00 + 2xy 0 12y = 0 d) x2 y 00 + xy 0 = 0
e) x2 y 00 5xy 0 + 9y = 0 f) x2 y 00 + xy 0 9y = 0
g) x2 y 00 + xy 0 + 9y = 0 h) x2 y 00 + 7xy 0 + 25y = 0
i) x3 y 000 + 6x2 y 00 + 4xy 0 = 0 j) x3 y 000 x2 y 00 + xy 0 = 0
k) x3 y 000 3x2 y 00 + xy 0 = 0 l) x3 y 000 + 6x2 y 00 + 4xy 0 4y = 0
Problem 11.7. Solve the following initial value problems.

8 00 0
8 00
>
< y 4y + 3y = 0, <y
> 6y 0 + 25y = 0,
a) y(0) = 7, b) y(0) = 3,
>
: 0 >
: 0
y (0) = 11. y (0) = 1.
106
8 00 8 00
<y
> 2y 0 + 2y = 0, 0
< y + 6y + 13y = 0,
>
c) y(0) = 0, d) y(0) = 2,
>
: 0 >
: 0
y (0) = 5. y (0) = 0.
8 2 00 8 2 00
<xy
> 2xy 0 + 2y = 0, <xy
> xy 0 + y = 0,
e) y(1) = 3, f) y(1) = 7,
>
: 0 >
: 0
y (1) = 1. y (1) = 2.
8 2 00 8 00
0
< x y + xy + y = 0,
> <y
> 2y 0 + y = 0,
g) y(1) = 2, h) y(0) = 1,
>
: 0 >
: 0
y (1) = 3. y (0) = 1.
8 8
>
> 2y 000 3y 00 2y 0 = 0, >
> 3y 000 + 2y 00 = 0,
>
> >
>
< y(0) = 1, < y(0) = 1,
i) j)
>
>
> y 0 (0) = 1, >
>
> y 0 (0) = 0,
>
: y 00 (0) = 3. >
: y 00 (0) = 1.
8 8
>
> y 000 + 10y 00 + 25y 0 = 0, >
> y 000 5y 00 + 9y 0 5y = 0,
>
> >
>
< y(0) = 3, < y(0) = 0,
k) l)
>
>
> y 0 (0) = 4, >
>
> y 0 (0) = 1,
>
: y 00 (0) = 5. >
: y 00 (0) = 6.
Problem 11.8. In the following problems, a di↵erential equation and

one solution y1 are given. Use the method of reduction of order to find
a second linearly independent solution y2 and then write the general
solution of the given equation.
a) 1 x2 y 00 + 2xy 0 2y = 0, 1 < x < 1; y1 = x
107
b) 1 x2 y 00 2xy 0 + 2y = 0, 1 < x < 1; y1 = x
c) x2 y 00
x(x + 2)y 0 + (x + 2)y = 0, x > 0; y1 = x
✓ ◆
2 00 0 2 1 cos x
d) x y + xy + x y = 0, x > 0; y1 = p
4 x
108
12 Nonhomogeneous Linear Equations
The general n-th order nonhomogeneous linear di↵erential equation
has the form
an (x)y (n) + an 1 (x)y (n 1)

+ . . . + a1 (x)y 0 + a0 (x)y = f (x). (12.1)
A general solution of equation (12.1) has the form
y(x) = yh (x) + yp (x),
where yh (x) is a general solution of the associated homogeneous linear

an (x)y (n) + an 1 (x)y (n 1)

+ . . . + a1 (x)y 0 + a0 (x)y = 0 (12.2)
and yp (x) is one particular solution of the equation (12.1).

In this section we will study the methods to find the particular
solution yp (x) of the equation (12.1). Note that it is only possible in
some specific cases.
12.1 Method of undetermined coefficients for non-

homogeneous linear equations with constant
coefficients
The general nonhomogeneous n-th order linear di↵erential equation
with constant coefficients has the form
an y (n) + an 1 y (n 1)
+ . . . + a1 y 0 + a0 y = f (x), (12.3)
where coefficients a0 , a1 , a2 , . . . , an are real constants with an 6= 0. We

already know how to find the general solution yh (x) of the associated
homogeneous equation
an y (n) + an 1 y (n 1)
+ . . . + a1 y 0 + a0 y = 0. (12.4)
109
The method of undetermined coefficients is used to find the

particular solution of the equation (12.3) whenever the function f (x)
in (12.3) is a linear combination of (finite) products of functions of
following four types:
1. A power function xk , where k is a nonnegative integer
2. An exponential function erx
3. Trigonometric function cos kx
4. Trigonometric function sin kx
Rule 1: Suppose that no term appearing either in f (x) or in any

of its derivatives satisfies the associated homogeneous equation (12.4).
Then, take as a trial solution for yp a linear combination of all linearly
independent such terms and their derivatives. Then determine the co-
efficients by substitution of this trial solution into the nonhomogeneous
equation (12.3).
y 00 + 4y = 3x3 .
Solution: Associated homogeneous equation
y 00 + 4y = 0
has a characteristic equation
r2 + 4 = 0
with complex roots r = ±2i. Then the general solution of homogeneous

equation is
yh (x) = c1 cos 2x + c2 sin 2x.
The right-hand side of given equation f (x) = 3x3 and its derivatives
are constant multiples of linearly independent functions x3 , x2 , x, 1.
110
Since none of these satisfy the associated homogeneous equation, we

apply the Rule 1 and search yp in the form
yp (x) = Ax3 + Bx2 + Cx + D.
For this function we have
yp0 = 3Ax2 + 2Bx + C and yp00 = 6Ax + 2B
Then substitution of yp and its derivatives into given nonhomogeneous

equation yields
yp00 + 4yp = 6Ax + 2B + 4(Ax3 + Bx2 + Cx + D) = 3x3
or
4Ax3 + 4Bx2 + (6A + 4C)x + (2B + D) = 3x3
We equate coefficients of the same powers of x in both sides of the last
equality to get 8
>
> 4A = 3,
>
>
< 4B = 0,
>
> 6A + 4C = 0,
>
>
: 2B + D = 0.
3 9
Solving this system we get A = , B = 0, C = , D = 0. Hence a
4 8
particular solution of given nonhomogeneous equation is
3 9
yp (x) = x3 x
4 8
and therefore its general solution is
3x3 9x
y(x) = yh (x) + yp (x) = c1 cos 2x + c2 sin 2x + .
4 8
111

8 00
<y
> 3y 0 + 2y = 3e x 10 cos 3x,
y(0) = 1,
>
: 0
y (0) = 2.
y 00 3y 0 + 2y = 0
r2 3r + 2 = 0
with two real roots r1 = 1 and r2 = 2. Then the general solution of

homogeneous equation is
yh (x) = c1 ex + c2 e2x .
The right-hand side of given equation f (x) = 3e x 10 cos 3x and its

derivatives involve linearly independent terms e x , cos 3x, sin 3x. Since
none of these satisfy the associated homogeneous equation, we apply
the Rule 1 and search yp in the form
x
yp (x) = Ae + B cos 3x + C sin 3x.

yp0 = Ae x
3B sin 3x + 3C cos 3x,
yp00 = Ae x
9B cos 3x 9C sin 3x.
After we substitute yp and its derivatives into given nonhomogeneous

equation and collect the terms, we get
x x
6Ae + ( 7B 9C) cos 3x + (9B 7C) sin 3x = 3e 10 cos 3x.
112
We equate coefficients of the terms involving e x , terms involving

cos 3x and terms involving sin 3x in both sides of the last equality
to get 8
< 6A = 3,
>
7B + 9C = 10,
>
:
9B 7C = 0.
1 7 9
Solving this system we get A = , B = , C = . Hence a particular
2 13 13
solution of given nonhomogeneous equation is
1 x 7 9
yp (x) = e + cos 3x + sin 3x
2 13 13
1 7 9
y(x) = yh (x) + yp (x) = c1 ex + c2 e2x + e x
+ cos 3x + sin 3x.
2 13 13
Finally, if we impose the initial conditions y(0) = 1 and y 0 (0) = 2 we
get 8
> 1 7
< y(0) = c1 + c2 + + = 1,
2 13
: y 0 (0) = c1 + 2c2 1 + 27 = 2,
>
2 13
1 6
with solution c1 = and c2 = . So, the solution of given initial
2 13
value problem is
1 x 6 1 7 9
y(x) = e + e2x + e x
+ cos 3x + sin 3x.
2 13 2 13 13
Example 12.3. Find the general form of the particular solution of

y 000 + 9y 0 = x2 e2x + x sin x.
113
y 000 + 9y 0 = 0
r3 + 9r = 0
with one real root r1 = 0 and two complex roots r2,3 = ±3i. Then the
general solution of homogeneous equation is
yh (x) = c1 + c2 cos 3x + c3 sin 3x.

The right-hand side of given equation f (x) = x2 e2x + x sin x and its
derivatives involve the linearly independent terms
x2 e2x , xe2x , e2x , x sin x, x cos x, sin x, cos x.
Since none of these terms satisfy the associated homogeneous equation,

we apply the Rule 1 and yp should have the form
yp (x) = Ax2 e2x +Bxe2x +Ce2x +Dx sin x+Ex cos x+F sin x+G cos x.
Rule 2: Suppose that f (x) in equation (12.3) is a linear combination

of functions f1 (x), f2 (x), . . . , fm (x), where each function fi (x) is a
product of functions of above mentioned four types. Suppose that
some term appearing either in fi (x) or in any of its derivatives satisfies
the associated homogeneous equation (12.4). Then, we multiply all
the terms appearing either in fi (x) or in any of its derivatives by xk ,
where k is the smallest positive integer such that none of the new terms
satisfy the associated homogeneous equation (12.4). After that take as
a trial solution for yp a linear combination of all linearly independent
terms and then determine the coefficients by substitution of this trial
solution into the nonhomogeneous equation (12.3).
y 00 4y = 2e2x .
114
y 00 4y = 0
r2 4=0
with two real roots r1,2 = ±2. Then the general solution of homoge-
neous equation is
yh (x) = c1 e2x + c2 e 2x .
The right-hand side of given nonhomogeneous equation f (x) = 2e2x
and its derivatives involve only one term e2x . Since e2x satisfies the
homogeneous equation, we apply Rule 2 and multiply this term by x.
New term xe2x does not satisfy the homogeneous equation. Then we
search yp in the form
yp (x) = Axe2x .
yp0 = A(1 + 2x)e2x and yp00 = 4A(1 + x)e2x .

equation yields
yp00 4yp = 4A(1 + x)e2x 4Axe2x = 2e2x
1
with solution A = . Hence a particular solution of given nonhomo-
2
geneous equation is
1
yp (x) = xe2x
2
1
y(x) = yh (x) + yp (x) = c1 e2x + c2 e 2x
+ xe2x .
2
115

y 000 + y 00 = 3ex + 4x2 .
y 000 + y 00 = 0
r3 + r2 = 0
with one repeated real root r1 = 0 of multiplicity 2 and one unrepeated
real root r2 = 1. Then general solution of homogeneous equation is
yh (x) = c1 + c2 x + c3 e x .
The right-hand side of given equation f (x) = 3ex + 4x2 is a linear
combination of two functions f1 (x) = ex and f2 (x) = x2 . Function
f1 (x) = ex and its derivatives involve only one term ex , which does not
satisfy the homogeneous equation. Function f2 (x) = x2 and its deriva-
tives involve terms x2 , x, 1. Since x and 1 satisfy the homogeneous
equation, we apply Rule 2 and multiply the terms x2 , x, 1 by x2 . New
terms are x4 , x3 , x2 and they do not satisfy homogeneous equation.
ex ex
x2 x4
x x3
1 x2
Then we search yp in the form

yp (x) = Aex + Bx4 + Cx3 + Dx2 .
yp0 = Aex + 4Bx3 + 3Cx2 + 2Dx,
yp00 = Aex + 12Bx2 + 6Cx + 2D,
yp000 = Aex + 24Bx + 6C.
116

equation yields
yp000 + yp00 = 2Aex + 12Bx2 + (24B + 6C)x + (6C + 2D) = 3ex + 4x2 .
3
Equating the coefficients of the terms in last equality, we get A = ,
2
1 4
B = , C = and D = 4. Hence a particular solution of given
3 3
nonhomogeneous equation is
3 1 4 3
yp (x) = ex + x4 x + 4x2
2 3 3
x 3 1 4 3
y(x) = yh (x) + yp (x) = c1 + c2 x + c3 e + e x + x4 x + 4x2 .
2 3 3
y 000 + y 00 = 3e x
+ 4x.
y 000 + y 00 = 0
has a general solution
yh (x) = c1 + c2 x + c3 e x .
The right-hand side of given equation f (x) = 3e x + 4x is a linear

combination of two functions f1 (x) = e x and f2 (x) = x. Function
f1 (x) = e x and its derivatives involve only one term e x , which sat-
isfies the homogeneous equation. So, we apply Rule 2 and multiply
the term e x by x. New term xe x does not satisfy the homogeneous
equation. Function f2 (x) = x and its derivatives involve terms x, 1.
Since both terms satisfy the homogeneous equation, we apply Rule 2
117
x x
e xe
x x3
1 x2
and multiply the terms x and 1 by x2 . New terms are x3 , x2 and they
do not satisfy homogeneous equation.
Then we search yp in the form
x
yp (x) = Axe + Bx3 + Cx2 .
yp0 = A(1 x)e x
+ 3Bx2 + 2Cx,
yp00 = A(x 2)e x
+ 6Bx + 2C,
yp000 = A(3 x)e x
+ 6B.
equation yields
yp000 + yp00 = Ae x
+ 6Bx + (6B + 2C) = 3e x
+ 4x.
Equating the coefficients of the terms in last equality, we get A = 3,
2
B = and C = 2. Hence a particular solution of given nonhomoge-
3
neous equation is
x 2
yp (x) = 3xe + x3 2x2
3
x x 2
y(x) = yh (x) + yp (x) = c1 + c2 x + c3 e + 3xe + x3 2x2 .
3
Example 12.7. Find the general form of the particular solution of

y 00 + 6y 0 + 13y = e 3x
cos 2x.
118
y 00 + 6y 0 + 13y = 0
r2 + 6r + 13 = 0
with two complex roots r = 3 ± 2i. Then general solution of homo-

geneous equation is
3x 3x
yh (x) = c1 e cos 2x + c2 e sin 2x.
The right-hand side of given equation f (x) = e 3x cos 2x and its deriva-
tives involve two terms e 3x cos 2x and e 3x sin 2x. Since both of these
terms satisfy the homogeneous equation, we apply Rule 2 and multi-
ply these terms by x. New terms xe 3x cos 2x and xe 3x sin 2x do not
satisfy the homogeneous equation. Then yp has the form
3x 3x
yp (x) = Axe cos 2x + Bxe sin 2x.
12.2 Method of variation of parameters for second

order linear equations
Theorem 12.1. Consider the second order nonhomogeneous linear
y 00 + p(x)y 0 + q(x)y = f (x). (12.5)
Assume that associated homogeneous equation
y 00 + p(x)y 0 + q(x)y = 0 (12.6)
yh (x) = c1 y1 (x) + c2 y2 (x).
119
Then nonhomogeneous equation (12.5) has a particular solution

Z Z
y2 (x)f (x) y1 (x)f (x)
yp (x) = y1 (x) dx + y2 (x) dx, (12.7)
W (x) W (x)
where W (x) = W (y1 , y2 ) is the Wronskian of y1 and y2 .
Proof. We search the particular solution of nonhomogeneous equation
(12.5) in the form
yp (x) = v1 (x)y1 (x) + v2 (x)y2 (x), (12.8)
where v1 and v2 are unknown functions. Di↵erentiating (12.8) we get
yp0 (x) = v10 (x)y1 (x) + v1 (x)y10 (x) + v20 (x)y2 (x) + v2 (x)y20 (x). (12.9)
Let
v10 (x)y1 (x) + v20 (x)y2 (x) = 0. (12.10)
Then (12.9) takes the form
yp0 (x) = v1 (x)y10 (x) + v2 (x)y20 (x). (12.11)
Di↵erentiating it again, we obtain
yp00 (x) = v10 (x)y10 (x) + v1 (x)y100 (x) + v20 (x)y20 (x) + v2 (x)y200 (x). (12.12)
We put (12.8), (12.11) and (12.12) into nonhomogeneous equation
(12.5) to get
yp00 + p(x)yp0 + q(x)yp = f (x) =)
v10 y10 + v1 y100 + v20 y20 + v2 y200 + p(x) v1 y10 + v2 y20 + q(x) v1 y1 + v2 y2 = f (x),
v1 y100 + p(x)y10 + q(x)y1 + v2 y200 + p(x)y20 + q(x)y2 + v10 y10 + v20 y20 = f (x),
v10 y10 + v20 y20 = f (x).
Combining the last equality with (12.10) we have
( 0
v1 y1 + v20 y2 = 0,
(12.13)
v10 y10 + v20 y20 = f.
120
Solving the system (12.13) for v10 and v20 , we get

Z
y2 f y2 f y2 (x)f (x)
v10 = = =) v1 (x) = dx
y1 y20 y10 y2 W (y1 , y2 ) W (x)
and
Z
y1 f y1 f y1 (x)f (x)
v20 = 0 0 = =) v2 (x) = dx.
y1 y2 y1 y2 W (y1 , y2 ) W (x)
Finally, putting the expressions which we found for v1 (x) and v2 (x)
into (12.8) we obtain the formula (12.7).

y 00 + y = tan x.

y 00 + y = 0
yh (x) = c1 cos x + c2 sin x.
So, y1 = cos x and y2 = sin x are two linearly independent solutions of
the associated homogeneous equation. Their Wronskian is
y1 y2 cos x sin x
W (y1 , y2 ) = = = cos2 x + sin2 x = 1.
y10 y20 sin x cos x
Then using the formula (12.7), the particular solution of given nonho-
mogeneous equation has the form
Z Z
yp (x) = cos x sin x tan xdx + sin x cos x tan xdx =
Z Z
= cos x sec x cos x dx + sin x sin xdx =
= cos x ln | sec x + tan x|.

121
So, the general solution of given nonhomogeneous equation is
y(x) = yh (x) + yp (x) = c1 cos x + c2 sin x cos x ln | sec x + tan x|.

Problem 12.1. In the following problems, use the method of undeter-
mined coefficients to find a particular solution of the given di↵erential
equation. Then write its general solution.
a) y 00 + y = 3x b) y 00 2y 0 + 2y = 2x
c) y 00 4y = 2e3x d) 3y 00 + y 0 2y = 2 cos x
e) y 00 + 16y = e3x f) y 00 y0 6y = 2 sin 3x
g) y 00 + y 0 + y = sin2 x h) y 00 + 2y 0 3y = 1 + xex
i) y 000 + 4y 0 = 3x 1 j) y 00 + 2y 0 + 5y = ex sin x
k) y 00 + 9y = 5 + 2x2 e3x l) y (4) 2y 00 + y = xex
m) y 000 + y 0 = 2 sin x n) y 000 2y 00 + y 0 = 1 + xex
Problem 12.2. In the following problems, use the method of variation

of parameters to find a particular solution of the given di↵erential
equation. Then write its general solution.
a) y 00 + 3y 0 + 2y = 4ex b) y 00 4y 0 + 4y = 2e2x
c) y 00 + 4y = cos 3x d) y 00 + 9y = 2 sec 3x
e) y 00 + 4y = sin2 x f) y 00 + y = csc2 x
122
13 Laplace Transform
13.1 Laplace Transform and Inverse Laplace Trans-
form
Definition 13.1. Given a function f (t) defined for all t 0, the
Laplace transform of f (t) is the function F (s) defined as follows:
Z1
st
F (s) = L{f (t)} = e f (t)dt (13.1)
0
for all values of s for which the improper integral converges. Table 1
lists the Laplace transforms which we will derive here.
Example 13.1. Find Laplace transform of function f (t) = 1, t 0.

Solution: By definition we have
Z1  st 1
st e 1
L{1} = e dt = = for s > 0.
s 0 s
0
So,
1
L{1} = for s > 0. (13.2)
s
Example 13.2. Find Laplace transform of function f (t) = eat , t 0.

Z1 Z1  1
at st at (s a)t e (s a)t 1
L{e } = e e dt = e dt = = for s > a.
s a 0 s a
0 0
So,
1
L{eat } = for s > a. (13.3)
s a
123
Remark 13.1. It is easy to verify that the formula in (13.3) also holds
when a is a complex number provided that s > Re(a).
Example 13.3. Find Laplace transform of function
f (t) = tn , t 0
for any n 2 N.
Solution: For any n 2 N, using integration by parts we have
Z1  st n 1 Z1
e t n n
L{tn } = e st n
t dt = + e st n 1
t dt = L{tn 1 }
s 0 s s
0 0
for s > 0. Then using it recursively, we obtain

n n(n 1) n(n 1)(n 2)
L{tn } = L{tn 1 } = L{t n 2
} = L{tn 3 } =
s s2 s 3
n(n 1)(n 2) · · · 2 · 1 n!
= ... = n
L{t0 } = L{1}.
s sn
Then, using (13.2), for any n 2 N we obtain
n!
L{tn } = for s > 0. (13.4)
sn+1
In particular, we have
1 2 6
L{t} = , L{t2 } = , L{t3 } = , ...
s2 s3 s4
Theorem 13.1. (Linearity of the Laplace Transform)

If a and b are constants, then
L{af (t) + bg(t)} = aL{f (t)} + bL{g(t)} (13.5)
for all s such that the Laplace transforms of the functions f (t) and
g(t) both exist.
124
Proof. Linearity of the Laplace transform follows from the linearity of

integration. Indeed, by definition we have
Z1
st
L{af (t) + bg(t)} = e af (t) + bg(t) dt =
0
Z1 Z1
st st
=a e f (t)dt + b e g(t)dt = aL{f (t)} + bL{g(t)}.
0 0
Example 13.4. Find Laplace transform of function
f (t) = 3 + 5t t2 + 2e3t , t 0.
Solution: Using (13.2)-(13.5), we have

3 5 2 2
L{f (t)} = 3L{1} + 5L{t} L{t2 } + 2L{e3t } = + 2 + .
s s s3 s 3
Example 13.5. Find the Laplace transform of function
f (t) = cos (kt), t 0.
eikt + e ikt
Solution: Since cos (kt) = , using (13.3) we have
2
1⇣ ikt ikt
⌘ 1⇣ 1 1 ⌘ s
L{cos (kt)} = L{e } + L{e } = + = 2
2 2 s ik s + ik s + k2
for s > 0. So,
s
L{cos (kt)} = for s > 0. (13.6)
s2 + k 2
125
f (t) = sin (kt), t 0.
eikt e ikt
Solution: Since sin (kt) = , using (13.3) we have
2i
1⇣ ⌘ 1⇣ 1 1 ⌘ k
L{sin (kt)} = L{eikt } L{e ikt
} = = 2
2i 2i s ik s + ik s + k2
for s > 0. So,
k
L{sin (kt)} = for s > 0. (13.7)
s2 + k 2
f (t) = 3e2t + 2 sin2 (3t), t 0.
Solution: Using formulas (13.3)-(13.7), we get
L{f (t)} = L{3e2t + 2 sin2 (3t)} = L{3e2t + 1 cos (6t)} =

3 1 s
= 3L{e2t } + L{1} L{cos (6t)} = + 2
s 2 s s + 36
for s > 2.
Definition 13.2. The unit step function u(t) is the piecewise con-
tinuous function defined as following
(
0, if t < 0,
u(t) =
1, if t 0.
The unit step function u(t) has a jump discontinuity at t = 0. We

denote by ua (t) the function which has a jump at t = a rather than at
t = 0. So (
0, if t < a,
ua (t) = u(t a) = (13.8)
1, if t a.
126
Example 13.8. Find L{ua (t)} if a > 0.

Z1 Z1  st 1 as
st st e e
L{ua (t)} = e ua (t)dt = e dt = = for s > 0.
s a s
0 a
So,
as
e
L{ua (t)} = for s > 0, a > 0. (13.9)
s
Example 13.9. Let F (s) be the Laplace transform of f (t). Prove that
the Laplace transform of ua (t)f (t a) is e as F (s).
Z1 Z1
st st
L{ua (t)f (t a)} = e ua (t)f (t a)dt = e f (t a)dt =
0 a
Z1 Z1
s(t+a) as st as
= e f (t)dt = e e f (t)dt = e L{f (t)}.
0 0
So,
as
L{ua (t)f (t a)} = e F (s). (13.10)
Definition 13.3. If F (s) = L{f (t)}, then f (t) is called the inverse
Laplace transform of F (s) and we write
f (t) = L 1 {F (s)}.
⇢
1 1
For instance, from (13.2) it follows that L = 1.
s
Remark 13.2. The inverse Laplace transform is a linear operation.

So, for any constants a and b we have
L 1 {aF (s) + bG(s)} = aL 1 {F (s)} + bL 1 {G(s)}.
127
Example 13.10. We have

⇢ ⇢
1 1 1 1 2! t2
L = L = ,
s3 2 s3 2
⇢ ⇢
3 1
L 1 = 3L 1 = 3e 2t ,
s+2 s ( 2)
⇢ ⇢
2 2 3 2
L 1 2 = L 1 2 = sin (3t),
s +9 3 s + 32 3
⇢ ⇢ ⇢
3s + 1 s 1 2 sin (2t)
L 1 = 3L 1
+ L 1
= 3 cos (2t) + .
s2 + 4 s2 + 2 2 2 s2 + 2 2 2
1
Example 13.11. Find the inverse Laplace transform of F (s) = .
s(s 3)
Solution: We first find a partial fraction decomposition of F (s). We
have
1 A B
= + ,
s(s 3) s s 3
where A and B are some constants. Multiplying both sides by s(s 3)
gives
⇢
A + B = 0,
1 = A(s 3) + Bs =) 1 = (A + B)s 3A =)
3A = 1.
1 1
Thus A = and B = . Then
3 3
⇢ ⇢
1 1 1 1/3 1/3
L =L + =
s(s 3) s s 3
⇢ ⇢
1 1 1 1 1 1
= L + L =
3 s 3 s 3
1 1 3t e3t 1
= + e = .
3 3 3
128
2s + 1
s(s2 + 9)
Solution: We first find a partial fraction decomposition of F (s). We
have
2s + 1 A Bs + C
= + 2 ,
s(s2 + 9) s s +9
where A, B and C are constants. Multiplying both sides by s(s2 + 9)
gives
2s + 1 = A(s2 + 9) + (Bs + C)s =) 2s + 1 = (A + B)s2 + Cs + 9A =)

8
< A + B = 0, 1 1
=) C = 2, =) A = , B = , C = 2.
: 9 9
9A = 1.
Then
⇢ ⇢
1 2s + 1 1 1/9 s/9 + 2
L =L + 2 =
s(s2 + 9) s s +9
⇢ ⇢ ⇢
1 1 1 1 1 s 2 1 3
= L L + L =
9 s 9 s2 + 3 2 3 s2 + 3 2
1 1 2
= cos (3t) + sin (3t).
9 9 3
Definition 13.4. Function f (t) is said to be of exponential order

as t ! 1 if there exist nonnegative constants M , c and T such that
|f (t)|  M ect for t T.
Theorem 13.2. (Existence of Laplace Transforms)

If the function f (t) is piecewise continuous for t 0 and is of expo-
nential order as t ! 1, then its Laplace transform F (s) = L{f (t)}
exists.
129
Theorem 13.3. (Uniqueness of Inverse Laplace Transforms)

Suppose that functions f (t) and g(t) are piecewise continuous for t 0
and are of exponential order as t ! 1, so that their Laplace transforms
F (s) = L{f (t)} and G(s) = L{g(t)} both exist. If F (s) = G(s) for all
s > c (for some c), then f (t) = g(t) for t 0.
Theorem 13.4. Assume that F (s) = L{f (t)} for s > c. Then
L{eat f (t)} = F (s a) for s > a + c.
Proof. By definition we have
Z1 Z1
L{eat f (t)} = e st eat f (t)dt = e (s a)t
f (t)dt = F (s a).
0 0
Remark 13.3. Using result of Theorem 12.4 and (13.4), we get

n!
L{eat tn } = for s > a. (13.11)
(s a)n+1
Similarly, from (13.6) and (13.7) we obtain
s a
L{eat cos (kt)} = for s > a, (13.12)
(s a)2 + k 2
k
L{eat sin (kt)} = for s > a. (13.13)
(s a)2 + k 2

f (t) = te2t + e t cos (3t), t 0.
Solution: Using formulas (13.11)-(13.12), we get
1 s+1
L{f (t)} = L{te2t } + L{e t cos (3t)} = +
(s 2)2 (s + 1)2 + 9
for s > 2.
130
2s 1
s2 + 4s + 13
Solution: We have
⇢ ⇢ ⇢
1 2s 1 1 2s 1 2(s + 2) 5
1
L =L =L =
s2 + 4s + 13 (s + 2)2 + 32 (s + 2)2 + 32
⇢ ⇢
1 s+2 5 1 3
= 2L L =
(s + 2)2 + 32 3 (s + 2)2 + 32
2t 5 2t
= 2e cos (3t) e sin (3t).
3
Theorem 13.5. (Laplace Transform of Derivative)

Suppose that the function f (t) is of exponential order as t ! 1. If f (t)
is continuous and piecewise smooth for t 0, then Laplace transform
of f 0 (t) exists and
L{f 0 (t)} = sL{f (t)} f (0) = sF (s) f (0), (13.14)
where F (s) is the Laplace transform of f (t).

Proof. By definition we have
Z1 h i1 Z1
L{f 0 (t)} = e st 0
f (t)dt = e st
f (t) +s e st
f (t)dt = f (0)+sF (s).
0
0 0
Remark 13.4. (Laplace Transform of Second Derivative)

Let g(t) = f 0 (t). Then using (13.14), we get
L{f 00 (t)} = L{g 0 (t)} = sL{g(t)} g(0) = sL{f 0 (t)} f 0 (0) =
⇥ ⇤
= s sL{f (t)} f (0) f 0 (0) = s2 L{f (t)} sf (0) f 0 (0)
and thus
L{f 00 (t)} = s2 F (s) sf (0) f 0 (0). (13.15)
131
Remark 13.5. (Laplace Transform of Third Derivative)

Let g(t) = f 00 (t). Then using (13.14)-(13.15), we get
L{f 000 (t)} = L{g 0 (t)} = sL{g(t)} g(0) = sL{f 00 (t)} f 00 (0) =
⇥ ⇤
= s s2 L{f (t)} sf (0) f 0 (0) f 00 (0) =
= s3 L{f (t)} s2 f (0) sf 0 (0) f 00 (0)
and thus
L{f 000 (t)} = s3 F (s) s2 f (0) sf 0 (0) f 00 (0). (13.16)
Remark 13.6. (Laplace Transform of Higher Derivatives)

By induction one can prove that
L{f (n) (t)} = sn F (s) sn 1 f (0) sn 2 f 0 (0) ... f (n 1)
(0). (13.17)

f (t) = t sin (kt), t 0.
Solution: Di↵erentiating f (t) twice, we get

f 0 (t) = sin (kt) + kt cos (kt)
and
f 00 (t) = 2k cos (kt) k 2 t sin (kt) = 2k cos (kt) k 2 f (t).
Then, taking Laplace transform of both sides in the last equality we
obtain
L{f 00 (t)} = 2kL{cos (kt)} k 2 L{f (t)} =)
2ks
=) s2 L{f (t)} sf (0) f 0 (0) = k 2 L{f (t)} =)
s2 + k2
2ks 2ks
=) s2 + k 2 L{f (t)} = =) L{f (t)} =
s2 + k 2 (s2 + k 2 )2
132
for s > 0. So,

2ks
L t sin (kt) = for s > 0. (13.18)
(s2 + k 2 )2
Remark 13.7. In the similar way, one can prove that
s2 k 2
L t cos (kt) = 2 for s > 0. (13.19)
(s + k 2 )2
Theorem 13.6. (Laplace Transform of Integral)

If function f (t) is piecewise continuous for t 0 and is of exponential
order as t ! 1, then
8 t 9
<Z = 1 F (s)
L f (⌧ )d⌧ = L{f (t)} = , (13.20)
: ; s s
0
where F (s) is the Laplace transform of f (t).
Proof. We denote
Zt
g(t) = f (⌧ )d⌧.
0
Then by Fundamental theorem of Calculus we have g 0 (t) = f (t). Tak-

ing Laplace transform of both sides and using (13.14), we obtain
L{g 0 (t)} = L{f (t)} =) sL{g(t)} g(0) = L{f (t)} =)

1
=) sL{g(t)} = L{f (t)} =) L{g(t)} = L{f (t)}.
s
133
Table 1: Laplace Transforms.
f (t) F (s) = L{f (t)}
1
1 (s > 0)
s
n!
tn (n 2 N) (s > 0)
sn+1
1
eat (s > a)
s a
s
cos (kt) (s > 0)
s2 + k 2
k
sin (kt) (s > 0)
s2 + k 2
as
e
ua (t) (a > 0) (s > 0)
s
n!
eat tn (n 2 N) (s > a)
(s a)n+1
s a
eat cos (kt) (s > a)
(s a)2 + k 2
k
eat sin (kt) (s > a)
(s a)2 + k 2
s2 k 2
t cos (kt) (s > 0)
(s2 + k 2 )2
2ks
t sin (kt) (s > 0)
(s2 + k 2 )2
ga
134
t x
xysdy t
13.2 Solution of Initial Value Problems with the

use of Laplace Transform Method
Example 13.16. Solve the following IVP
8 00
<x
> x0 6x = 0,
x(0) = 2,
>
: 0
x (0) = 1
by using the Laplace transform method.

Solution: Let x(t) be the solution of given IVP and X(s) be its
Laplace transform. Taking Laplace transform of both sides di↵erential
equation in given IVP and using the given initial values, we obtain
L{x00 x0 6x} = L{0} =) L{x00 (t)} L{x0 (t)} 6L{x(t)} = 0 =)

⇣ ⌘ ⇣ ⌘
2 0
=) s X(s) sx(0) x (0) sX(s) x(0) 6X(s) = 0 =)
⇣ ⌘ ⇣ ⌘
2
=) s X(s) 2s + 1 sX(s) 2 6X(s) = 0 =)
⇣ ⌘
2
=) s s 6 X(s) = 2s 3 =)
2s 3 2s 3
=) X(s) = = .
s2 s 6 (s 3)(s + 2)
Now, we find a partial fraction decomposition of X(s). We have

2s 3 A B
= + ,
(s 3)(s + 2) s 3 s + 2
where A and B are constants. Multiplying both sides by (s 3)(s + 2)

gives
⇢
A + B = 2,
2s 3 = A(s + 2) + B(s 3) =)
2A 3B = 3.
135
3 7
Solving this system we obtain A = and B = . Then, the solution
5 5
of given IVP is
⇢
2s 3
x(t) = L 1 {X(s)} = L 1 =
(s 3)(s + 2)
⇢
1 3/5 7/5 3 7
=L + = e3t + e 2t .
s 3 s+2 5 5

8 00
< x + 4x = sin 3t,
>
x(0) = 0,
>
: 0
x (0) = 0
Solution: Let x(t) be the solution of given IVP and X(s) be its
Laplace transform. Taking Laplace transform of both sides di↵erential
equation in given IVP and using the given initial values, we obtain
L{x00 + 4x} = L{sin 3t} =) L{x00 (t)} + 4L{x(t)} = L{sin 3t} =)
⇣ ⌘ 3
0
=) s2 X(s) sx(0) x (0) + 4X(s) = =)
s2 + 9
⇣ ⌘ 3 3
2
=) s + 4 X(s) = =) X(s) = .
s2 + 9 (s2 + 4)(s2 + 9)
Now, we find a partial fraction decomposition of X(s). We have
3 As + B Cs + D
= + 2 ,
(s2 + 4)(s2 + 9) s2 + 4 s +9
where A, B, C, D are some constants. Multiplying both sides by
(s2 + 4)(s2 + 9) gives
8
>
> A + C = 0,
>
< B + D = 0,
3 = (As + B)(s2 + 9) + (Cs + D)(s2 + 4) =)
>
> 9A + 4C = 0,
>
: 9B + 4D = 3.
136
3 3
Solving this system we obtain A = C = 0, B = and D = . Then,
5 5
⇢
1 1 3
x(t) = L {X(s)} = L =
(s2 + 4)(s2 + 9)
⇢ ⇢ ⇢
3/5 3/5 3 2 1 3
=L 1 2 = L 1
L 1
=
s + 4 s2 + 9 10 s2 + 2 2 5 s2 + 3 2
3 3
= sin 2t sin 3t.
10 5

8 00 0
< y + 2y + 5y = f (t),
>
y(0) = 0,
>
: 0
y (0) = 0,
where (
1, if 0  t < ⇡
f (t) =
0, if t ⇡
Solution: First of all, using (13.8) we can write

f (t) = 1 u⇡ (t).
Let y(t) be the solution of given IVP and Y (s) be its Laplace transform.
Taking Laplace transform of both sides di↵erential equation in given
IVP and using the given initial values, we obtain
L{y 00 + 2y 0 + 5y} = L{f (t)} =)
=) L{y 00 (t)} + 2L{y 0 (t)} + 5L{y(t)} = L{1} L {u⇡ (t)} =)

⇣ ⌘ 1 e ⇡s
2
=) s + 2s + 5 Y (s) = =)
s s
1 e ⇡s
=) Y (s) = .
s(s2 + 2s + 5) s(s2 + 2s + 5)
137
1
Now, we find a partial fraction decomposition of . We
s(s2 + 2s + 5)
have
1 A Bs + C
Y (s) = = + ,
s(s2 + 2s + 5) s s2 + 2s + 5
where A, B and C are some constants. Multiplying both sides by
s(s2 + 2s + 5) gives
1 = A(s2 + 2s + 5) + (Bs + C)s.
Equating the coefficients of the common terms in both sides of this

1
equality and then solving the resulting system, we obtain A = ,
5
1 2
B= and C = . Then, we have
5 5
⇢
1 1 1 e ⇡s
y(t) = L {Y (s)} = L =
s(s2 + 2s + 5) s(s2 + 2s + 5)
⇢ ⇢ ✓ ◆
1 1 1 s+2 1 1 1 s + 2
= L 2
L e ⇡s =
5 s s + 2s + 5 5 s s2 + 2s + 5
⇢
1 1 1 s+1 1
= L
5 s (s + 1)2 + 4 (s + 1)2 + 4
⇢ ✓ ◆
1 1 1 s + 1 1
L e ⇡s =
5 s (s + 1)2 + 4 (s + 1)2 + 4
✓ ◆ ✓ ◆
1 t 1 t u⇡ (t) t+⇡ 1 t+⇡
= 1 e cos 2t e sin 2t 1 e cos 2t e sin 2t .
5 2 5 2
8 ✓ ◆
>
> 1 t 1 t
>
< 5 1 e cos 2t 2 e sin 2t , if 0  t < ⇡
y(t) = ✓ ◆
>
> 1 1
>
: e t+⇡ e t cos 2t + sin 2t , if t ⇡
5 2
138

8 00
< x + 4x = f (t),
>
x(0) = 0,
>
: 0
x (0) = 0,
where (
cos (2t), if 0  t < 2⇡
f (t) =
0, if t 2⇡
Solution: First of all, using (13.8) we can write
f (t) = 1 u2⇡ (t) cos (2t) = cos (2t) u2⇡ (t) cos (2t).
Let x(t) be the solution of given IVP and X(s) be its Laplace trans-
form. Taking Laplace transform of both sides di↵erential equation in
given IVP and using the given initial values, we obtain
L{x00 + 4x} = L{f (t)} =)
=) L{x00 (t)} + 4L{x(t)} = L{cos (2t)} L u2⇡ (t) cos 2(t 2⇡) =)
⇣ ⌘ s s
2
=) s + 4 X(s) = 2 e 2⇡s 2 =)
s +4 s +4
s 2⇡s s
=) X(s) = 2 e .
(s + 4)2 (s2 + 4)2
Then, we have
⇢
s s
x(t) = L 1 {X(s)} = L 1
e 2⇡s
=
(s2 + 4)2 (s2 + 4)2
⇢ ⇢
1 s 1 s
=L L e 2⇡s 2 =
(s2 + 4)2 (s + 4)2
t t 2⇡
= sin (2t) u2⇡ (t) sin 2(t 2⇡) .
4 4
139

8
> t
>
< sin (2t), if 0  t < 2⇡
4
x(t) =
>
> ⇡
: sin (2t), if t 2⇡
2
Remark 13.8. Laplace Transform method can be also used for solving
the systems of linear di↵erential equations.

8
>
> x0 = x + 2y,
>
>
< y0 = x + e t,
>
> x(0) = 0,
>
>
: y(0) = 0

Solution: Let x(t), y(t) be the solutions of given IVP and X(s), Y (s)
be their Laplace transforms. Taking Laplace transform of both sides
di↵erential equations in given IVP and using the given initial values,
we obtain 8
< sX(s) = X(s) + 2Y (s),
1
: sY (s) = X(s) + .
s+1
Solving this system, we get
2 s 1
X(s) = , Y (s) = .
(s + 1)2 (s 2) (s + 1)2 (s 2)
Then, we have
⇢ ⇢ 2 2 2
1 2 1 9 3 9
x(t) = L =L + + =
(s + 1)2 (s 2) s+1 (s + 1)2 s 2
2 t 2 t 2
= e te + e2t
9 3 9
140
and
⇢ ⇢ 1 2 1
1 s 1 1 9 3 9
y(t) = L =L + + =
(s + 1)2 (s 2) s+1 (s + 1)2 s 2
1 t 2 t 1
= e + te + e2t .
9 3 9

Problem 13.1. Find the Laplace transforms of the following functions
3t
a) f (t) = t 2e b) f (t) = cos 2t + sin 2t
c) f (t) = cos2 2t d) f (t) = sin 3t cos 3t
e) f (t) = (1 + t)3 f) f (t) = tet
Problem 13.2. Find the inverse Laplace transforms of the following

functions
2 1
a) F (s) = 3 b) F (s) =
s s+5
3 5 3s
c) F (s) = d) F (s) =
s 4 s2 + 9
9+s 10s 3
e) F (s) = f) F (s) =
4 s2 25 s2
2 3s 3
g) F (s) = e h) F (s) =
s s(s + 5)
1 1
i) F (s) = j) F (s) =
s(s2 + 4) s2 (s2 + 1)
1 1
k) F (s) = l) F (s) =
s(s2 9) s2 (s2 1)
1 s 1
m) F (s) = n) F (s) =
s(s + 1)(s + 2) (s + 1)3
141
1 3s + 5
o) F (s) = p) F (s) =
s2 + 4s + 4 s2 6s + 25
Problem 13.3. Use the Laplace transform method to solve the initial
value problems
8 00 8 00 0
>
< x + x = sin 2t, < x + 3x + 2x = t,
>
a) x(0) = 0, b) x(0) = 0,
>
: 0 >
: 0
x (0) = 0. x (0) = 2.
8
>
> x000 + x00 6x0 = 0, 8 00
> 0 t
>
< x(0) = 0, < x + 4x + 13x = te ,
>
c) d) x(0) = 0,
>
>
> x0 (0) = 1, >
: 0
>
: x00 (0) = 1. x (0) = 2.
8 (4)
8 > x + 2x00 + x = e2t ,
> x0 = x + 2y, >
>
>
> >
>
>
< y0 = x + e t, < x(0) = 0,
>
e) f) x0 (0) = 0,
>
> x(0) = 0, >
>
>
> >
> x00 (0) = 0,
: y(0) = 0. >
>
: 000
x (0) = 0.
Problem 13.4. Use Laplace transform method to solve the IVP:

(
y 0 + 2y = f (t),
y(0) = 5,
where (
4, if 0  t < 6,
f (t) =
0, if t 6.
142

(
3y 0 5y = f (t),
y(0) = 4,
where (
0, if 0  t < 6,
f (t) =
10, if t 6.

8 00
<y
> 3y 0 + 2y = f (t),
y(0) = 0,
>
: 0
y (0) = 0,
where (
2, if 0  t < 4,
f (t) =
0, if t 4.

8 00 0
< y + 2y + 5y = f (t),
>
y(0) = 0,
>
: 0
y (0) = 0,
where (
1, if 0  t < ⇡,
f (t) =
0, if t ⇡.
143
14 Systems of Linear Di↵erential Equa-
tions
So far, we have been concerned with methods for solving an ordi-
nary di↵erential equation that involves only one dependent variable
(unknown function). Now we shall consider systems of ordinary dif-
ferential equations that involve two or more dependent variables. We
will usually denote the independent variables by x1 , x2 , x3 , . . . or by
x, y, z, . . . Primes will indicate derivatives with respect to independent
variable t. For instance, a system of two first order di↵erential equa-
tions in two unknown functions has the general form
(
f (t, x, y, x0 , y 0 ) = 0,
g(t, x, y, x0 , y 0 ) = 0.
A solution of this system is a pair of functions x(t) and y(t) that
satisfy both equations identically over some interval of values of t.
Consider a system of di↵erential equations that can be solved for
the highest order derivatives of the dependent variables that appear,
as explicit functions of t and lower order derivatives of the dependent
variables. For instance, in the case of a system of two second order
equations, our assumption is that it can be written in the form
(
x001 = f1 (t, x1 , x2 , x01 , x02 ),
x002 = f2 (t, x1 , x2 , x01 , x02 ).
It is of both practical and theoretical importance that any such higher
order system can be transformed into an equivalent system of first order
equations. The general theory of systems and systematic solutions
techniques are more easily and more concisely described for first order
systems than for higher order systems.
To describe how such a transformation is accomplished, we consider
first the n-th order equation
⇣ ⌘
(n) 0 00 (n 1)
x = f t, x, x , x . . . , x . (14.1)
144
We introduce the dependent variables x1 , x2 , x3 , . . . , xn defined as fol-

lows:
x1 = x, x2 = x0 , x3 = x00 , . . . , xn = x(n 1) . (14.2)
Note that, x01 = x0 = x2 , x02 = x00 = x3 , and so on. Hence, the
substitution of (14.2) in equation (14.1) yields the system of n first
order equations
8 0
>
> x1 = x2 ,
>
>
> 0
< x2 = x3 ,
>
..
. (14.3)
>
>
>
> x0n 1 = xn ,
>
>
: 0
xn = f (t, x1 , x2 , x3 , . . . , xn ) .
Obviously, this system is equivalent to the original n-th order equation

(14.1) in the sense that x(t) is a solution of equation (14.1) if and
only if the functions x1 (t), x2 (t), . . . , xn (t) defined in (14.2) satisfy the
system of equations (14.3).
Example 14.1. Consider the third order equation
x000 + 3x00 + 2x0 5x = sin 2t.
The substitutions x1 = x, x2 = x0 , x3 = x00 yield the equivalent system

of three first order equations
8 0
< x1 = x2 ,
>
x02 = x3 ,
>
: 0
x3 = 5x1 2x2 3x3 + sin 2t.
Example 14.2. Consider the system of two second order equations

(
x00 = 3x + y,
y 00 = 2x 2y + 40 sin 3t.
145
The substitutions x1 = x, x2 = x0 , y1 = y, y2 = y 0 yield the equivalent

system of four first order equations
8
>
> x01 = x2 ,
>
>
< x0 = 3x + y ,
2 1 1
>
>
> y10 = y2 ,
>
: y 0 = 2x
2 2y + 40 sin 3t.
1 1
14.1 Systems of Linear First Order Equations

The general linear system of n first order equations is of the form:
8
>
> x01 = a11 (t)x1 + a12 (t)x2 + . . . + a1n (t)xn + f1 (t),
>
>
< x0 = a (t)x + a (t)x + . . . + a (t)x + f (t),
2 21 1 22 2 2n n 2
(14.4)
>
> . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
>
>
: x0 = a (t)x + a (t)x + . . . + a (t)x + f (t).
n n1 1 n2 2 nn n n
Linear system (14.4) is called homogeneous if the functions f1 , f2 , . . . , fn

are all identically zero; otherwise, it is called nonhomogeneous. A
solution of the system (14.4) is an n-tuple of functions x1 (t), x2 (t), . . . , xn (t)
that identically satisfy each of the equations in (14.4) on some interval
of values of t. We will see that the general theory of a system of n
linear equations shares many similarities with the general theory of a
single n-th order linear di↵erential equation.
Theorem 14.1. (Existence and Uniqueness for Linear Systems)

Suppose that the functions a11 , a12 , . . . , ann and the functions f1 , f2 , . . . , fn
are continuous on the open interval I containing the point a. Then,
given the n numbers b1 , b2 , . . . , bn the system in (14.4) has a unique
solution on the entire interval I that satisfies the initial conditions
x1 (a) = b1 , x2 (a) = b2 , . . . , xn (a) = bn .
146
14.1.1 The Method of Elimination

The method of elimination for linear di↵erential systems is similar
to the solution of a linear systems of algebraic equations by a process
of eliminating the unknowns one at a time until only a single equation
with a single unknown remains.
Example 14.3. Find the general solution of the system
(
x0 = y,
y 0 = 2x + y.
Solution: Since x00 = y 0 = 2x + y = 2x + x0 , the elimination method

for given system results in the single second order equation
x00 x0 2x = 0
with characteristic equation
r2 r 2=0
and general solution
t
x(t) = c1 e + c2 e2t .
Then
y(t) = x0 (t) = c1 e t
+ 2c2 e2t .

8
>
> x0 = 4x 3y,
>
>
< y 0 = 6x 7y,
>
> x(0) = 2,
>
>
: y(0) = 1.
Solution: From the first equation in given system we have

4 1 0
y= x x.
3 3
147
So that,
4 1 00
y 0 = x0 x .
3 3
Putting these in the second equation, we get
✓ ◆
4 0 1 00 4 1 0
x x = 6x 7 x x .
3 3 3 3
So, we obtain a second order linear equation
x00 + 3x0 10x = 0
with characteristic equation
r2 + 3r 10 = 0
and general solution
x(t) = c1 e2t + c2 e 5t
.
Then
4 1 0 2 2t 5t
y(t) = x x = c1 e + 3c2 e .
3 3 3
So, the general solution of the system is
x(t) = c1 e2t + c2 e 5t ,
2
y(t) = c1 e2t + 3c2 e 5t .
3
The initial conditions x(0) = 2, y(0) = 1 imply that c1 = 3, c2 = 1.
Hence the solution of given initial value problem is
x(t) = 3e2t e 5t
,
y(t) = 2e2t 3e 5t
.
148
14.1.2 Matrix-Valued Functions

Definition 14.1. A matrix-valued function, or simply matrix func-
tion, is a matrix such as
2 3 2 3
x1 (t) f1 (t)
6 7 6 7
6 x2 (t) 7 6 f2 (t) 7
x(t) = 6 7 6
6 ... 7 or f(t) = 6 ... 7
7 (14.5)
4 5 4 5
xn (t) fn (t)
or 2 3
a11 (t) a12 (t) . . . a1n (t)
6 7
6 a21 (t) a22 (t) . . . a2n (t) 7
A(t) = 6
6 ... .. .. 7 (14.6)
4 . . 7 5
an1 (t) an2 (t) . . . ann (t)
in which each entry is a function of t.
Definition 14.2. We say that matrix function is continuous on an
interval I if each of its elements is continuous on I. We say that
matrix function is di↵erentiable on an interval I if each of its elements
is di↵erentiable on I. The derivative of a di↵erentiable matrix function
is defined by elementwise di↵erentiation.
Example 14.5. Let

2 3
t " #
6 7 sin t 1
x(t) = 4 t2 5 and A(t) = .
t cos t
e t
Then
2 3
1 " #
6 7 cos t 0
x0 (t) = 4 2t 5 and A0 (t) = .
t
1 sin t
e
149
Remark 14.1. Bu using (14.5) and (14.6), the linear system (14.4)
can be written in the matrix form
dx
= A(t)x(t) + f (t). (14.7)
dt
A solution of (14.7) on the open interval I is a column vector func-
tion x(t) = [xi (t)] such that the component functions of x satisfy the
system (14.4) identically on I. If the functions aij (t) and fi (t) are all
continuous on I, then Theorem 13.1 guarantees the existence on I of
a unique solution x(t) satisfying initial conditions x(a) = b.
Example 14.6. The first order linear system
(
x01 = 4x1 3x2 ,
x02 = 6x1 7x2
can be written as the single matrix equation

" # " #
dx x1 (t) 4 3
= Ax, where x(t) = and A = .
dt x2 (t) 6 7
Now, consider the vector functions

" # " #
2t 5t
3e e
x1 (t) = and x 2 (t) = .
2e2t 3e 5t
Since " #" # " #

4 3 3e2t 6e2t
Ax1 = 2t
= 2t
= x01
6 7 2e 4e
and " #" # " #
5t 5t
4 3 e 5e
Ax2 = 5t
= 5t
= x02 ,
6 7 3e 15e
vector functions x1 and x2 are the solutions of given system.
150
14.1.3 Homogeneous Linear Systems

The associated homogeneous linear system has the form
dx
= A(t)x(t). (14.8)
dt
Theorem 14.2. (Principle of Superposition)

Let x1 , x2 , . . . , xn be solutions of the homogeneous linear system (14.8)
on the open interval I. Then for any values of constants c1 , c2 , . . . , cn
the linear combination
x(t) = c1 x1 (t) + c2 x2 (t) + . . . + cn xn (t)
is also a solution of (14.8) on the interval I.
Example 14.7. We have shown that

" # " #
2t 5t
3e e
x1 (t) = and x 2 (t) =
2e2t 3e 5t
are two solutions of the homogeneous linear system discussed in the

Example 13.6. Theorem 13.2 implies that for any values of constants
c1 and c2 the linear combination
" # " #
2t 5t
3e e
x(t) = c1 x1 (t) + c2 x2 (t) = c1 + c 2
2e2t 3e 5t
is also a solution.
Definition 14.3. Let

2 3 2 3 2 3
x11 (t) x12 (t) x1n (t)
6 7 6 7 6 7
6 x21 (t) 7 6 x22 (t) 7 6 x2n (t) 7
x1 (t) = 6 7 6
6 ... 7 , x2 (t) = 6 ... 7 ,
7 ... , xn (t) = 6
6 ...
7
7
4 5 4 5 4 5
xn1 (t) xn2 (t) xnn (t)
151
be n column vector functions, then their Wronskian is the n ⇥ n

determinant
x11 (t) x12 (t) . . . x1n (t)
x21 (t) x22 (t) . . . x2n (t)
W (x1 , x2 , . . . , xn ) =
... ... ... ...
xn1 (t) xn2 (t) . . . xnn (t)
Theorem 14.3. Suppose that x1 , x2 , . . . , xn are n solutions of the ho-

mogeneous linear system (14.8) on an open interval I. Suppose also
that A(t) is continuous on I. Then
(i) x1 , x2 , . . . , xn are linearly dependent on I if and only if
W (x1 , x2 , . . . , xn ) = 0 for all t 2 I
(ii) x1 , x2 , . . . , xn are linearly independent on I if and only if
W (x1 , x2 , . . . , xn ) 6= 0 for all t 2 I
Theorem 14.4. Let x1 , x2 , . . . , xn be n linearly independent solutions

of the homogeneous linear system (14.8) on an open interval I, where
A(t) is continuous. Then a general solution of homogeneous system
(14.8) on interval I has the form
x(t) = c1 x1 (t) + c2 x2 (t) + . . . + cn xn (t),
where c1 , c2 , . . . , cn are arbitrary constants.
Example 14.8. Consider the following homogeneous system

2 3 2 3
x1 (t) 3 2 0
dx 6 7 6 7
= Ax, where x(t) = 4 x2 (t) 5 and A = 4 1 3 2 5.
dt
x3 (t) 0 1 3
152
a) Show that vector functions

2 t3 2 3 2 3
2e 2e3t 2e5t
6 7 6 7 6 7
x1 (t) = 4 2et 5 , x2 (t) = 4 0 5 , x3 (t) = 4 2e5t 5
et e3t e5t
are solutions of given homogeneous system.
b) Use Wronskian to show that x1 , x2 , x3 are linearly independent.
c) Write the general solution of given homogeneous system.
d) Find a particular solution of given system that satisfies the initial
condition 2 3
0
6 7
x(0) = 4 2 5 .
6
Solution:
a) Since
2 32 3 2 3
3 2 0 2et 2et
6 76 7 6 7
Ax1 = 4 1 3 2 5 4 2et 5 = 4 2et 5 = x01 ,
0 1 3 et et
2 32 3 2 3
3 2 0 2e3t 6e3t
6 76 7 6 7 0
Ax2 = 4 1 3 2 54 0 5=4 0 5 = x2 ,
0 1 3 e3t 3e3t
2 32 3 2 3
3 2 0 2e5t 10e5t
6 76 7 6 7
Ax3 = 4 1 3 2 54 2e5t 5 = 4 10e5t 5 = x03 ,
0 1 3 e5t 5e5t
vector functions x1 , x2 , x3 are the solutions of given system.
153
b) Since
2et 2e3t 2e5t

W (x1 , x2 , x3 ) = 2et 0 2e5t = 16e9t 6= 0,
et e3t e5t
it follows that x1 , x2 , x3 are linearly independent.
c) The general solution of given homogeneous system is

2 t3 2 3 2 3
2e 2e3t 2e5t
6 7 6 7 6 7
x(t) = c1 x1 (t)+c2 x2 (t)+c3 x3 (t) = c1 4 2et 5+c2 4 0 5+c3 4 2e5t 5 .
et e3t e5t
d) We have
2 3 2 3 2 3 2 3
2 2 2 0
6 7 6 7 6 7 6 7
x(0) = c1 4 2 5 + c2 4 0 5 + c3 4 25=425
1 1 1 6
which results in the algebraic linear system

8
< 2c1 + 2c2 + 2c3 = 0,
>
2c1 2c3 = 2,
>
:
c1 c2 + c3 = 6.
Solving this system, we obtain c1 = 2, c2 = 3, c3 = 1. So, the

desired particular solution is
2 t3 2 3 2 3 2 t 3
2e 2e3t 2e5t 4e 6e3t + 2e5t
6 7 6 7 6 7 6 7
x(t) = 2 4 2et 5 3 4 0 5+4 2e5t 5 = 4 4et 2e5t 5.
et e3t e5t 2et + 3e3t + e5t
154
14.1.4 Nonhomogeneous Linear Systems

Consider the nonhomogeneous linear system written in the matrix form
dx
= A(t)x(t) + f (t) (14.9)
dt
with the associated homogeneous linear system
dx
= A(t)x(t). (14.10)
dt
Theorem 14.5. Let xh (t) be the general solution of homogeneous

linear system (14.10) and xp (t) be the particular solution of nonhomo-
geneous linear system (14.9). Then the general solution of nonhomo-
geneous linear system (14.9) has the form
x(t) = xh (t) + xp (t).

Problem 14.1. In the following problems, transform the given equa-
tion or system into an equivalent system of first order di↵erential equa-
tions.
a) x00 + 3x0 + 7x = t2 b) x(4) + 6x00 3x0 + x = cos 3t
c) t2 x00 + tx + (t2 1)x = 0 d) t3 x000 2t2 x00 + 3tx0 + 5x = ln t

(
x00 5x + 4y = 0,
e) x000 = (x0 )2 + cos x f)
y 00 + 4x 5y = 0.
8 00
(
x00 + 3x0 + 4x 2y = 0, < x = 3x y + 2z,
>
g) h) y 00 = x + y 4z,
y 00 + 2y 0 3x + y = cos t. >
: 00
z = 5x y z.
155
Problem 14.2. Use the method of elimination to find the general

solution of the linear systems
( (
0
x = x + 3y, x0 = x 2y,
a) b)
y 0 = 2y. y 0 = 2x 3y.
( (
0
x = 3x 4y, x0 = 4x + y + 2t,
c) d)
y 0 = 2x + y. y 0 = 2x + y.
( (
0
x = 2x + y, x0 = 2x 3y + 2 sin 2t,
e) f)
y 0 = x + 2y e2t . y 0 = x 2y cos 2t.
Problem 14.3. Use the method of elimination to solve the initial value
problems
8 8
0
>
> x = 3x + 2y, >
> x0 = 3x y,
>
> >
>
< y 0 = 3x + 4y, < y 0 = 5x 3y,
a) b)
>
> x(0) = 0, >
> x(0) = 1,
>
> >
>
: y(0) = 2. : y(0) = 1.
8 8
>
> x0 = x + 9y, >
> x0 + 2y 0 = 4x + 5y,
>
> >
>
< y 0 = 2x 5y, < 2x0 y 0 = 3x,
c) d)
>
> x(0) = 3, >
> x(0) = 1,
>
> >
>
: y(0) = 2. : y(0) = 1.
Problem 14.4. Consider the following homogeneous system

" # " #
dx x 1 (t) 4 2
= Ax, where x(t) = and A = .
dt x2 (t) 3 1
a) Show that vector functions
" # " #
t 2t
2e e
x1 (t) = and x2 (t) =
3et e2t
156
are solutions of given homogeneous system.
b) Use Wronskian to show that x1 and x2 are linearly independent.
c) Write the general solution of given homogeneous system.
d) Find a particular solution of given system that satisfies the initial

condition " #
0
x(0) = .
1
157
15 The Eigenvalue Method for Homoge-
neous Linear Systems
We now introduce a powerful method for constructing the general so-
lution of a homogeneous first order linear system with constant coeffi-
cients 8
>
> x01 = a11 x1 + a12 x2 + . . . + a1n xn ,
>
>
< x0 = a x + a x + . . . + a x ,
2 21 1 22 2 2n n
(15.1)
>
> . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
>
>
: x0 = a x + a x + . . . + a x .
n n1 1 n2 2 nn n
System (15.1) can be written in the matrix form

dx
= Ax, (15.2)
dt
where
2 3 2 3
x1 (t) a11 a12 . . . a1n
6 7 6 7
6 x2 (t) 7 6 a21 a22 . . . a2n 7
x(t) = 6
6 ...
7
7 and A = 6
6 .. .. .. 7 .
4 5 4 . . . 7
5
xn (t) an1 an2 . . . ann
We know that to find the general solution of the system (15.1) it suffices
to find n linearly independent solution vectors x1 , x2 , . . . , xn . Then the
linear combination
x(t) = c1 x1 (t) + c2 x2 (t) + . . . + cn xn (t)
with arbitrary coefficients c1 , c2 , . . . , cn will be a general solution of

system (15.1).
To search for the n needed linearly independent solution vectors, we
proceed by analogy with the characteristic root method for solving a
single homogeneous system with constant coefficients. It is reasonable
158
to anticipate solution vectors of the form

2 3 2 3 2 3
x1 (t) v1 e t v1
6 7 6 7 6 7
6 x2 (t) 7 6 v2 e t 7 6 v2 7 t
x(t) = 6 7 6
6 ... 7 = 6 ... 7 = 6
7 6 .. 7
7 e = ve t ,
4 5 4 5 4 . 5
t
xn (t) vn e vn
where , v1 , v2 , . . . , vn are appropriate scalar constants. We substitute

the trial solution x(t) = ve t in (15.2), the result is
ve t = Ave t .
Simplifying both sides by the nonzero scalar factor e t , we get
Av = v. (15.3)
This means that x = ve t will be a nontrivial solution of (15.2) pro-

vided that v is a nonzero vector and is a constant such that (15.3)
holds.
Now, to find v and we first rewrite equation (15.3) in the form
(A In )v = 0, (15.4)
where In is an identity matrix of size n. Given , (15.4) is a system of

n homogeneous linear equations in the unknowns v1 , v2 , . . . , vn . By a
standard theorem of linear algebra, it has a nontrivial solution if and
only if the determinant of its coefficient matrix vanishes; that is, if and
only if
a11 a12 ... a1n

a21 a22 ... a2n
|A In | = = 0. (15.5)
... ... ... ...
an1 an2 . . . ann
159
Definition 15.1. The number (either zero or nonzero) is called

eigenvalue of the n ⇥ n matrix A if it satisfies (15.5). An eigen-
vector associated with the eigenvalue is a nonzero vector v that
satisfies (15.4).
Remark 15.1. Note that, if v is an eigenvector associated with the

eigenvalue , then so is any nonzero constant scalar multiple cv of v.
Remark 15.2. Equation (15.5) is called the characteristic equation

of the matrix A; its roots are the eigenvalues of A.
Theorem 15.1. Let be an eigenvalue of the coefficient matrix A of

the first order linear system (15.2). If v is an eigenvector associated
with , then x(t) = ve t is a nontrivial solution of the system (15.2).
The Eigenvalue Method

In outline, this method for solving the n ⇥ n homogeneous constant-
coefficient system (15.2) proceeds as follows:
1. We first solve the characteristic equation in (15.5) for eigenvalues
1 , 2 , . . . , n of the matrix A.
2. Next, we attempt to find n linearly independent eigenvectors

v1 , v2 , . . . , vn associated with these eigenvalues.
3. Step 2 is not always possible, but if it is, we get n linearly inde-
pendent solutions
1t 2t nt
x1 (t) = v1 e , x2 (t) = v2 e , . . . , xn (t) = vn e . (15.6)
In this case, the general solution of system (15.2) is a linear
combination
x(t) = c1 x1 (t) + c2 x2 (t) + . . . + cn xn (t)
of these n solutions.
160
15.1 Distinct Real Eigenvalues

If the eigenvalues 1 , 2 , . . . , n of the matrix A are real and distinct,
then we substitute each of them in equation (15.4) and solve for the
associated eigenvectors v1 , v2 , . . . , vn . In this case it can be proved
that the particular solution vectors given in (15.6) are always linearly
independent.
Example 15.1. Find a general solution of the system

(
x01 = 4x1 + 2x2 ,
x02 = 3x1 x2
by using the Eigenvalue method.

Solution:
Given system can be written as the single matrix equation
" # " #
dx x 1 4 2
= Ax, where x = and A = .
dt x2 3 1
The characteristic equation of the coefficient matrix is
4 2 2
|A I2 | = = (4 )( 1 ) 6= 3 10 = 0,
3 1
so we have the distinct real eigenvalues 1 = 2 and 2 = 5.

" #
a
Let v1 = be an eigenvector associated with eigenvalue 1 = 2.
b
Substitution of 1 and v1 in equation (15.4) yields the system
" #" # " # (
6 2 a 0 6a + 2b = 0,
= =)
3 1 b 0 3a + b = 0.
Since the two scalar equations in this system are equivalent, it has
infinitely many nonzero solutions. We can choose a arbitrarily (but
161
nonzero) and then solve for " b. For# instance, the choice a = 1 yields
1
b = 3, and thus v1 = is an eigenvector associated with
3
eigenvalue 1 = 2. So, we obtain a solution
" # " #
2t
1 e
x1 (t) = v1 e 1 t = e 2t = .
3 3e 2t
" #
c
Let v2 = be an eigenvector associated with eigenvalue 2 = 5.
d
" #" # " # (
1 2 c 0 c + 2d = 0,
= =)
3 6 d 0 3c 6d = 0.
" #
2
The choice d = 1 yields c = 2, and thus v2 = is an eigenvector
1
associated with eigenvalue 2 = 5. So, we obtain a solution
" # " #
5t
2 2e
x2 (t) = v2 e 2 t = e5t = .
1 e5t
The solutions we obtained are linearly independent since their Wron-
skian
e 2t 2e5t
W (x1 , x2 ) = 2t 5t
= 7e3t
3e e
is nonzero. Hence, a general solution of given system is
" # " #
2t 5t
e 2e
x(t) = c1 x1 (t) + c2 x2 (t) = c1 + c 2
3e 2t e5t
or in scalar form
x1 (t) = c1 e2t + 2c2 e5t ,
x2 (t) = 3c1 e2t + c2 e5t .
162
15.2 Complex Eigenvalues

Now, suppose that the characteristic equation (15.5) has a pair of
complex conjugate roots 1 = ↵ + i and 2 = ↵ i . Let v1 be an
eigenvector associated with complex eigenvalue 1 , so that 1 and v1
satisfy (15.4). Obviously, v1 is a complex vector. Let v1 = a + ib,
where a and b are real vectors. It is easy to show that v2 = a ib
will be an eigenvector associated with eigenvalue 2 .
The complex-valued solution of system (15.2) associated with 1 and
v1 is then
1t
x(t) = v1 e = (a + ib)e(↵+i )t
= (a + ib)e↵t (cos t + i sin t)
or
x(t) = e↵t (a cos t b sin t) + ie↵t (b cos t + a sin t) .
Because the real and imaginary parts of a complex-valued solution are

also solutions, we thus get the two real-valued solutions
x1 (t) = Re[x(t)] = e↵t (a cos t b sin t) ,
x2 (t) = Im[x(t)] = e↵t (b cos t + a sin t) .
It is easy to check that the same two real-valued solutions result from
taking real and imaginary parts of v2 e 2 t .

(
x01 = 4x1 3x2 ,
x02 = 3x1 + 4x2

Solution:
" # " #
dx x1 4 3
= Ax, where x = and A = .
dt x2 3 4
163
4 3
|A I2 | = = (4 )2 + 9 = 0,
3 4
so we have the pair of complex conjugate eigenvalues 1 = 4 3i and

2 = 4+" 3i. #
a
Let v1 = be an eigenvector associated with eigenvalue 1 = 4 3i.
b
" #" # " # (
3i 3 a 0 ia b = 0,
= =)
3 3i b 0 a + ib = 0.
" #
1
The choice a = 1 yields b = i, and thus v1 = is an eigenvector
i
associated with eigenvalue 1 = 4 3i. The corresponding complex-
valued solution is
" # " #
1 1
x(t) = v1 e 1 t = e(4 3i)t = e4t (cos 3t i sin 3t) =
i i
" #
cos 3t i sin 3t
= e4t .
sin 3t + i cos 3t
The real and imaginary parts of x(t) are the real-valued functions
" # " #
cos 3t sin 3t
x1 (t) = e4t and x2 (t) = e4t .
sin 3t cos 3t
A real-valued general solution of given system is

" #
c1 cos 3t c2 sin 3t
x(t) = c1 x1 (t) + c2 x2 (t) = e4t .
c1 sin 3t + c2 cos 3t
164
Finally, a general solution of given system in scalar form is

x1 (t) = e4t (c1 cos 3t c2 sin 3t) ,
x2 (t) = e4t (c1 sin 3t + c2 cos 3t) .

8 0
< x1 = 2x1 + x2 x3 ,
>
x02 = 4x1 3x2 x3 ,
>
: 0
x3 = 4x1 + 4x2 + 2x3
Solution:
2 3 2 3
x1 2 1 1
dx 6 7 6 7
= Ax, where x = 4 x2 5 and A = 4 4 3 1 5.
dt
x3 4 4 2
2 1 1
2
|A I3 | = 4 3 1 = + 4 (1 ) = 0,
4 4 2
so we have the pair of complex conjugate eigenvalues 1 = 2i, 2 = 2i
and one real
2 eigenvalue
3 3 = 1.
a
6 7
Let v1 = 4 b 5 be an eigenvector associated with eigenvalue 1 = 2i.
c
2 32 3 2 3 8
2 2i 1 1 a 0 < (2 2i)a + b c = 0,
>
6 76 7 6 7
4 4 3 2i 1 5 4 b 5 = 4 0 5 =) 4a (3 + 2i)b c = 0,
>
:
4 4 2 2i c 0 4a + 4b + (2 2i)c = 0.
165
Adding second and third equations gives b + c = 0. The choice c = 2

yields 2b = 2.3 Then from the first equality we have a = 1 + i. Thus
1+i
6 7
v1 = 4 2 5 is an eigenvector associated with eigenvalue 1 = 2i.
2
The corresponding complex-valued solution is
2 3 2 3
1+i 1+i
6 7 6 7
x(t) = v1 e 1 t = 4 2 5 e2it = 4 2 5 (cos 2t + i sin 2t) =
2 2
2 3
cos 2t sin 2t + i(cos 2t + sin 2t)
6 7
=4 2 cos 2t 2i sin 2t 5.
2 cos 2t + 2i sin 2t
So, the two real-valued solutions associated with the complex eigen-
values ±2i are
2 3 2 3
cos 2t sin 2t cos 2t + sin 2t
6 7 6 7
x1 (t) = 4 2 cos 2t 5 and x2 (t) = 4 2 sin 2t 5 .
2 cos 2t 2 sin 2t
2 3
d
6 7
Let v3 = 4 r 5 be an eigenvector associated with eigenvalue 3 = 1.
q
2 32 3 2 3 8
1 1 1 d 0 < d + r q = 0,
>
6 76 7 6 7
4 4 4 1 5 4 r 5 = 4 0 5 =) 4d 4r q = 0,
>
:
4 4 1 q 0 4d + 4r + q = 0.
Adding first and third equations gives d + r = 0. The choice d = 1

yields r = 1. Then from the second equality we have q = 0. Thus
166
2 3
1
6 7
v3 = 4 1 5 is an eigenvector associated with eigenvalue 3 = 1. So
0
we obtain a solution,
2 3 2 3
1 et
3t 6 7 6 7
x3 (t) = v3 e =4 1 5 et = 4 et 5 .
0 0
Finally, a general solution of given system is
x(t) = c1 x1 (t) + c2 x2 (t) + c3 x3 (t) =

2 3
c1 (cos 2t sin 2t) + c2 (cos 2t + sin 2t) + c3 et
6 7
=4 2c1 cos 2t 2c2 sin 2t c3 et 5
2c1 cos 2t + 2c2 sin 2t
or in scalar form
x1 (t) = c1 (cos 2t sin 2t) + c2 (cos 2t + sin 2t) + c3 et ,
x2 (t) = 2c1 cos 2t 2c2 sin 2t c3 e t ,
x3 (t) = 2c1 cos 2t + 2c2 sin 2t.

Problem 15.1. Use the Eigenvalue method to find the general solution
of the linear systems
( (
x01 = x1 + 2x2 , x01 = 2x1 + 3x2 ,
a) b)
x02 = 2x1 + x2 . x02 = 2x1 + x2 .
( (
x01 = 7x1 5x2 , x01 = 50x1 + 20x2 ,
c) d)
x02 = 4x1 + 3x2 . x02 = 100x1 60x2 .
167
8 0 8 0
< x1 = 4x1 + x2 + 4x3 ,
> < x1 = 5x1 + 5x2 + 2x3 ,
>
e) x02 = x1 + 7x2 + x3 , f) x02 = 6x1 6x2 5x3 ,
>
: 0 >
: 0
x3 = 4x1 + x2 + 4x3 . x3 = 6x1 + 6x2 + 5x3 .
Problem 15.2. Use the Eigenvalue method to solve the initial value
problems
8 8
0
>
> x1 = 3x1 + 4x2 , >
> x01 = 9x1 + 5x2 ,
>
> >
>
< x0 = 3x + 2x , < x0 = 6x 2x2 ,
2 1 2 2 1
a) b)
>
> x1 (0) = 1, >
> x1 (0) = 1,
>
> >
>
: x (0) = 1. : x (0) = 0.
2 2
8 8
>
> x01 = 2x1 5x2 , >
> x01 = x1 2x2 ,
>
> >
>
< x0 = 4x 2x2 , < x0 = 2x + x ,
2 1 2 1 2
c) d)
>
> x1 (0) = 2, >
> x1 (0) = 0,
>
> >
>
: x (0) = 3. : x (0) = 4.
2 2
168

Dif Lecturenotes

Uploaded by

Copyright:

Available Formats

Dif Lecturenotes

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Dif Lecturenotes

Uploaded by

Copyright:

Available Formats

Di↵erential Equations - Lecture Notes

Differansiyel denklemin çözümü fonks yon veya fonks

Diff d klemler genel olarak k gruba ayırab l r z

Ord nary D ff Eq CODE

B r d ff denklemde tek b r tane bağımsız

1 Introduction, Elementary Concepts

Definition 1.1. An equation involving at least one derivative of an

Example 1.1. The following equations are some examples of di↵eren-

Definition 1.2. An equation involving derivatives with respect to a

Definition 1.3. An equation involving partial derivatives with respect

Definition 1.4. The order of an Ordinary Di↵erential Equation is

Definition 1.5. An n-th order ODE (1.1) is called linear if F is a

Definition 1.6. A continuous function u = u(x) is called a solution

Example 1.4. Verify that y = e2x is a solution of the equation

Solution: Note that function y = e2x is continuous. First and second

means to find a di↵erentiable function y = y(x) that satisfies both

Example 1.5. First verify that y(x) = Ce x is a solution of the equa-

Thus, y(x) = Ce x is a solution of given equation for any value of

1.1 Self-study Problems

Problem 1.2. In the following examples, verify by substitution that

c) y 00 + 4y = 0; y1 = cos 2x, y2 = sin 2x

Problem 1.3. In the following examples, substitute y = erx into the

a) y 0 = y + 1; y(x) = Cex 1, y(0) = 5

d) y 0 = 3x2 y 2 + 1 ; y(x) = tan (x3 + C), y(0) = 1

e) y 0 + y tan x = cos x; y(x) = (x + C) cos x, y(⇡) = 0

where F (x) is an antiderivative of f (x) and C is an arbitrary constant.

Remark 2.1. (2.2) is a general solution of di↵erential equation (2.1),

Note that y1 (x) = x2 + 3x + 5 and y2 (x) = x2 + 3x 1 are examples of

Remark 2.2. The graph of general general solution (2.2) contains

Example 2.2. Solve the initial value problem:

Remark 2.3. (Second order equations)

where G(x) is an antiderivative of g(x) and C1 is an arbitrary constant.

where C2 is a second arbitrary constant. (2.4) is a general solution of

Example 2.4. Solve the initial value problem:

Solution: In the Example 2.3 we have obtained the general solution

Now, by using initial conditions y(0) = 1 and y 0 (0) = 2, we find the

Remark 2.4. (Velocity and Acceleration)

Example 2.5. Find the position function x(t) of a moving particle

Integrating both sides of the equation, we obtain

x(0) = C2 = 20 and x0 (0) = C1 = 10.

So, the solution of the given problem is

x(t) = 25t2 + 10t + 20.

2.1 Self-study Problems

Problem 2.2. In the following examples, find the position function

Therefore, the solution procedure for separable equation (3.1) consists

(i) Equation (3.1) can be rewritten as

(ii) Integrating both sides, we get

Example 3.1. Solve the following initial value problem:

Example 3.2. Solve the following di↵erential equation:

Solution: It is easy to see that given equation is separable. Separating

Next, integrating both sides gives us

Note that from obtained relation it is not easy to express y as an

Example 3.3. Solve the following initial value problem:

Solution: In the previous example we have already shown that di↵er-

Remark 3.1. In the step of separating the variables, we multiply or

Example 3.4. Find all solutions of the di↵erential equation:

Example 3.5. Find the general solution of the di↵erential equation:

3.1 Self-study Problems