Short Notes On Lyapunov Stability: 1.1 Basic Definitions For Stability and Invariance

Short notes on Lyapunov stability
by Sanand D
1 Introduction
All the discussion here is borrowed predominantly from Haddad, [2], Sastry and Khalil. We consider
the following form for a general non linear dynamical system (continuous/discrete)
ẋ(t) = f (t, x, u), x(t0 ) = x0 , t ∈ [t0 , t1 ]. (1)

x(t + 1) = f (x(k), u(k)), k = 0, 1, . . . (2)
where f for continuous time systems may assumed to be continuous in t and x, x ∈ Rn , u ∈ Rm , x

being the state of the system and u being the input. In the absence of inputs or if inputs are fixed
to some specified value, then one obtains
ẋ(t) = f (t, x), x(t0 ) = x0 , t ∈ [t0 , t1 ]. (3)
1.1 Basic definitions for stability and invariance

In the following definitions, properties are said to be true
• locally if they are true for all x ∈ B (0) for some .
• globally if they are true for all x ∈ Rn .
• uniformly if they are true for all t0 ≥ 0.
• semi globally if they are true for all x ∈ B (0) for an arbitrary .
We may replace s(t, t0 , x0 ) with x(t) in all the following definitions.
Local stability definitions:
Definition 1.1 (stability). The equilibrium point 0 is said to be stable if, for every > 0, there
exists a δ = δ(, t0 ) such that kx0 k < δ(, t0 ) ⇒ ks(t, t0 , x0 )k < for all t ≥ t0 .
Definition 1.2 (uniform stability). The equilibrium point 0 is said to be uniformly stable if, for
every > 0, there exists a δ = δ() such that kx0 k < δ() ⇒ ks(t, t0 , x0 )k < for all t ≥ t0 .
An equilibrium point is unstable if it is not stable. For autonomous systems stability and
uniformly stability is the same.
Definition 1.3 (attractivity). The equilibrium point 0 is attractive, if for each t0 ∈ R+ , there
exists η(t0 ) > 0 such that kx0 k < η(t0 ) ⇒ s(t + t0 , t0 , x0 ) → 0 as t → ∞. It is said to be uniformly
attractive if there exists η > 0 such that kx0 k < η ⇒ s(t + t0 , t0 , x0 ) → 0 as t → ∞ uniformly in
t0 , x0 .
Definition 1.4 (asymptotic stability). The equilibrium point 0 is asymptotically stable if it is

stable and attractive. It is uniformly asymptotically stable if it is uniformly stable and uniformly
attractive.
Definition 1.5 (exponential stability). The equilibrium point 0 is exponentially stable there exists
constants c, γ, such that ks(t0 + t, t0 , x0 )k ≤ ckx0 ke−γt , ∀t, t0 ≥ 0, ∀x0 ∈ B (0). The constant γ
is called an estimate of the rate of convergence.
All the definitions above are local since they are concerned with neighborhoods of the equilibrium
point.
Global stability definitions:
Definition 1.6 (Global asymptotic stability). The equilibrium point 0 is globally asymptotically
stable if it is stable and limt→∞ s(t, t0 , x0 ) = 0 for all x0 ∈ Rn .
Definition 1.7. The equilibrium point 0 is said to be globally uniformly asymptotically stable if it is
uniformly stable and for each pair of positive numbers M, with M arbitrarily large and arbitrarily
small, there exists a finite number T = T (M, ) such that kx0 k < M, t0 ≥ 0 ⇒ ks(t + t0 , t0 , x0 )k <
, ∀t ≥ T (M, ).
Definition 1.8. The equilibrium point 0 is said to be globally exponentially stable if there exists
constants c, γ such that ks(t0 + t, t0 , x0 )k ≤ ckx0 ke−γt , ∀t, t0 ≥ 0, ∀x0 ∈ Rn .
For an equilibrium point to be either globally uniformly asymptotically stable or globally ex-
ponentially stable, a necessary condition is that it is the only equilibrium point.
Energy-like functions:
Definition 1.9 (class K, KR, L, KL, KRL functions). A function φ : R+ → R+ is of class
K if it is continuous, strictly increasing, and φ(0) = 0. It is said to belong to class KR if it is in
class K and in addition, φ(x) → ∞ as x → ∞.
It is of class L if it is continuous on [0, ∞], strictly decreasing, φ(0) < ∞ and φ(x) → 0 as
x → ∞. A continuous function β : [0, ∞) × [0, ∞) → [0, ∞) is said to belong to class KL if, for a
fixed s, β(r, s) ∈ K with respect to r, and for each fixed r, β(r, s) ∈ L with respect to s. Similarly,
we define class KRL functions.
In the definition of class K functions, if we change the domain from [0, ∞) to [0, a), we get
class Ka functions. Then we can define class Ka R, Ka L, Ka RL analogously.
Definition 1.10 (locally positive definite function (lpdf)). A continuous function V : Rn × R+ →
R+ is called locally positive definite if for some > 0 and some α(.) of class K functions,
V (0, t) = 0 and V (x, t) ≥ α(kxk) ∀x ∈ B , t ≥ 0. (4)
Definition 1.11 (positive definite function (pdf)). A continuous function V : Rn × R+ → R+ is

called positive definite if (4) holds ∀x ∈ Rn . It is said to be radially unbounded if for some α(.) of
class KR functions,
V (0, t) = 0 and V (x, t) ≥ α(kxk) ∀x ∈ Rn , t ≥ 0. (5)
Definition 1.12 (decrescent functions). A continuous function V : Rn × R+ → R+ is called

decrescent if there exists some β(.) of class KR functions and an > 0, such that
V (x, t) ≤ β(kxk) ∀x ∈ B (0), t ≥ 0. (6)
Example 1.13. V (t, x) = (t + 1)kxk2 is a pdf but not decrescent.

Derivative along a trajectory: Let V (x(t)) be a real valued function on the solution trajectories
of (3). Then by chain rule, V̇ = ∇V (x).ẋ = ∇V (x).f . This is also called the Lie derivative of V
along f .
Theorem 1.14 (Basic equivalence). The equilibrium point 0 of (3) is
1. stable ⇔ for each t0 ∈ R+ , there exists d(t0 ) > 0 and φt0 ∈ K such that
ks(t, t0 , x0 )k ≤ φt0 (kx0 k), ∀t ≥ t0 , ∀x0 ∈ Bd(t0 ) (0).

2. uniformly stable ⇔ there exists d > 0 and φ ∈ K such that
ks(t, t0 , x0 )k ≤ φ(kx0 k), ∀t ≥ t0 , ∀x0 ∈ Bd (0).
3. attactive ⇔ for each t0 ∈ R+ , there exists r(t0 ) > 0, and for each x0 ∈ Br(t0 ) (0), there exists
a function σt0 ,x0 ∈ L such that
ks(t, t0 , x0 )k ≤ σt0 ,x0 (t), ∀t ≥ 0, ∀x0 ∈ Br(t0 ) (0).
4. uniformly attactive ⇔ there exists r > 0 and a function σt0 ,x0 ∈ L such that
ks(t, t0 , x0 )k ≤ σ(t), ∀t, t0 ≥ 0, ∀x0 ∈ Br(t0 ) (0).
5. asymptotically stable ⇔ there exists a number r(t0 ) > 0, a function φt0 ∈ K and for each
x0 ∈ Br(t0 ) (0), there exists a function σt0 ,x0 ∈ L such that
ks(t, t0 , x0 )k ≤ φt0 (kx0 k)σt0 ,x0 (t), ∀t, t0 ≥ 0, ∀x0 ∈ Br(t0 ) (0).
6. i. uniformly asymptotically stable ⇔ there exists a number r > 0, a function φ ∈ K and a

function σ ∈ L such that
ks(t, t0 , x0 )k ≤ φ(kx0 k)σ(t), ∀t, t0 ≥ 0, ∀x0 ∈ Br (0).
ii. uniformly asymptotically stable ⇔ there exists a class KL function β and a positive
constant c independent of t0 such that
kx(t)k ≤ β(kx(t0 )k, t − t0 ), ∀t ≥ t0 ≥ 0, ∀kx(t0 )k ≤ c.
Proof. ([2] and Khalil) 1. (⇐) Given > 0, t0 ∈ R+ , choose δ(, t0 ) = min{d(t0 ), φ−1
t0 ()}.
(⇒) Fix t0 and > 0. There exists δ(t0 , ) > 0 such that
kx(t0 )k < δ ⇒ kx(t)k < , ∀t ≥ t0 .
Let δ̄(t0 , ) be the supremum of all applicable δ. The function δ̄ is positive and non decreasing, but
need not be continuous. Choose θt0 () ∈ K such that θt0 () ≤ δ̄() for all > 0 (need justification
for existence of such function to be rigorous). Now choose φt0 = θt−1 0
.
2. For uniform stability, we apply similar arguments without any t0 dependence.
The proof of remaining cases is based on similar arguments. Intuitively, the first case says that
the trajectories remain bounded where the bound is given by a level surface of a class K function.
Same for the second case where the level surface is independent of the initial time. For the third case,
the criteria says that the magnitude of the trajectories (“energy” inside the trajectory) decreases,
in other words, it is bounded by a decreasing function which goes to zero at infinity. This means
the magnitude of the trajectory goes to zero as t → ∞ i.e., it approaches the origin. This explains
attractivity of the origin. For asymptotic stability, the trajectory must be bounded and the origin
must be attractive too. This explains the fifth and the sixth (i) case. The case 6(ii) (Khalil) gives
an alternate formulation of uniform asymptotic stability using class KL functions. The trajectory
is bounded depending up on the first argument of β and it decreases to zero since β → 0 as the
second argument in β goes to infinity. Refer Khalil/[2] for rigorous arguments of all cases.
Definition 1.15 (Domain of attraction). Suppose that 0 is an asymptotically stable equilibrium

point of ẋ = f (x). Then the domain of attraction D0 is given by
D0 := {x0 ∈ S ⊂ Rn | if x(0) = x0 , then limt→∞ x(t) = 0}.

Table 1: Basic Lyapunov stability theorems for autonomous systems (Theorem 2.1)
Conditions on V (x) Conditions on −V̇ (x) Conclusion

V (x) > 0, x 6= 0 locally ≥ 0 locally stable
V (x) > 0, x 6= 0 locally > 0, ∀x 6= 0 locally asymptotically stable
V (x) > 0, x 6= 0, V (x) → ∞ as kxk → ∞ > 0, ∀x 6= 0 globally asymptotically stable
αkxkp ≤ V (x) ≤ βkxkp locally ≥ V (x) locally locally exponentially stable
αkxkp ≤ V (x) ≤ βkxkp globally ≥ V (x) globally globally exponentially stable
2 Stability theory for Autonomous/time invariant systems

stability, asymptotic stability, exponential stability, global asymptotic stability, global exponential
stability Lyapunov stability, invariant set stability, constructing Lyapunov functions, Converse
Lyapunov theorems, instability theorems, linear systems and Lyapunov’s linearization
We consider non linear dynamical systems of the form
ẋ = f (x(t)). (7)
2.1 Basic stability theorems using Lyapunov’s direct method

Theorem 2.1. Consider an autonomous system ẋ = f (x(t)), x(0) = x0 where x(t) ∈ S ⊂ Rn . Let
V : S → R. Let α, β, > 0 and p ≥ 1. Then, Table 1 holds.
Proof. Let > 0 such that B (0) ⊂ S. Observe that since ∂B (0) is compact and V (x) is continuous,
V (∂B (0)) is compact hence, α = minx∈∂B (0) V (x) exists and α > 0 since 0 ∈
/ ∂B (0) and V (x) > 0
for x ∈ S, x 6= 0. (Draw a picture of S and B inside S.)
Consider β ∈ (0, α) where α is the minimum value of V on ∂B (0). Let Ωβ = {x ∈ S | V (x) ≤
β}. Then, Ωβ ⊂ B (0) \ ∂B (0). This is true because if there exists y ∈ Ωβ such that y ∈ ∂B (0),
then V (y) ≥ α > β which is a contradiction. (Draw this “ellipsoid” inside B .)
Since V̇ ≤ 0, V is non increasing and V (x(t)) ≤ V (x(0)) ≤ β. Therefore, x(t) ∈ Ωβ for all
x(0) ∈ Ωβ . Consider Bδ (0) ⊂ Ωβ for some δ() > 0. Therefore, for all x(0) ∈ Bδ (0), V (x(0)) < β
and V (x(t)) ≤ V (x(0)) ⇒ x(t) ∈ Ωβ ⊂ B (0). Thus, x(t) ∈ B (0) ⇒ kx(t)k < . This proves
Lyapunov stability.
For the second case, V̇ < 0. Therefore, V is decreasing and it is bounded from below by 0. We
need to show that x(t) → 0 as t → ∞ i.e., for every > 0, there exists T > 0 such that kx(t)k <
for all t > T . By previous arguments, for every > 0, we can choose β > 0 such that Ωβ ⊂ B .
Therefore, it is enough to show that V (x(t)) → 0 as t → ∞.
Since V is decreasing and it is bounded from below by 0, let V (x(t)) → c ≥ 0 as t → ∞.
To show c = 0, we use contradiction. Suppose c > 0. By continuity of V , there exists d > 0
such that Bd ⊂ Ωc . Since V (x(t)) → c ≥ 0 as t → ∞, x(t) lies outside Bd for all t ≥ 0. Let
−γ =maxd≤kxk≤ V̇ (x). Since V̇ < 0, −γ < 0. Moreover,
ˆ t
V (x(t)) = V (x(0)) + V̇ (x(τ ))dτ ≤ V (x(0)) − γt.
0
Observe that the rhs eventually becomes negative contracting the assumption that c > 0. Therefore,
c = 0 which proves asymptotic stability.
The third case is exactly similar to the second case. For any x ∈ Rn , let a = V (x). Since
kxk → ∞ ⇒ V (x) → ∞, for any a > 0, there exists > 0 such that V (x) > a when kxk > .
Therefore, Ωa ⊂ B (0) which implies that Ωa is bounded. Remaining arguments are exactly similar
as to the second case.
Since V̇ ≤ −V (x), V (x(t)) ≤ V (x(0))e−t , t ≥ 0. By assumption, V (x(0)) ≤ βkx(0)kp and
αkx(t)kp ≤ V (x(t)). Therefore,
β 1 ( − )t
αkx(t)kp ≤ βkx(0)kp e−t , t ≥ 0 ⇒ kx(t)k ≤ ( ) p kx(0)ke p , t ≥ 0
α
which proves local exponential stability.
The last case follows similarly.
Example 2.2 (Simple pendulum). Consider a simple pendulum where the constants are chosen
such that θ̈ + sin(θ) = 0. Choosing x1 = θ and x2 = θ̇, ẋ1 = x2 , ẋ2 = − sin(x1 ). The total
energy is 21 θ̇2 − cos(θ) = 21 x22 − cos(x1 ). Consider V (x1 , x2 ) = 1 + 21 x22 − cos(x1 )0. Since this is
continuously differentiable and lpdf, it is a Lyapunov function candidate. Moreover, V̇ (x1 , x2 ) =
sin(x1 )ẋ1 + x2 ẋ2 = sin(x1 )x2 + x2 (− sin x1 ) = 0. Therefore, the equilibrium point is stable.
Consider a simple pendulum with damping i.e., θ̈ + sin(θ) + bθ̇ = 0 (Khalil). Hence, the state
equations are ẋ1 = x2 , ẋ2 = − sin(x1 ) − bx2 . Let V (x1 , x2 ) be the same function as above. Thus,
V̇ = sin(x1 )ẋ1 + x2 ẋ2 = sin(x1 )x2 + x2 (− sin x1 ) − bx22 = −bx22 ≤ 0. Thus, the equilibrium
point is stable for the damped pendulum. Consider a different Lyapunov function V (x1 , x2 ) =
xT P x − cos(x1 ) where P > 0. Need to choose P such that V̇ < 0. Choose p11 = bp12 , p12 =
b/2, p22 = 1. Now, V̇ = − 12 abx1 sin(x1 ) − 21 bx22 and x1 sin(x1 ) > 0 for −π < x1 < π. Thus,
choosing {x ∈ R2 | − π < x1 < π} as an open set around 0, V̇ < 0 on this set ⇒ asymptotic
stability.
Example 2.3 (Non linear series RLC circuit/Non linear mass spring damper). Non linear series
RLC circuit (Sastry)/Mass-spring-damper. Suppose the inductor is linear but the resistor and
capacitor are non linear (or mass-spring- damper system with non linear spring and damping). Let
x1 be the charge on the capacitor and x2 be the current through the inductor. Therefore,
ẋ1 = x2 , ẋ2 = −f (x2 ) − g(x1 ) (8)
where f is a continuous function modelling resistor current-voltage characteristics and g models

capacitor charge-voltage characteristics. Suppose both f, g model locally passive elements i.e., there
exists σ0 such that
σf (σ) ≥ 0, σg(σ) ≥ 0, ∀σ ∈ [−σ0 , σ0 ].
A Lyapunov function candidate is the total energy of the system i.e.,
ˆ x1
x22
V (x) = + g(σ)dσ
2 0
where the first term represents the energy stored in the inductor (kinetic energy) and the second
term represents the energy stored in the capacitor (potential energy). By passivity of g, V is lpdf.
Moreover,
V̇ (x) = x2 (−f (x2 ) − g(x1 )) + g(x1 )x2 = −x2 f (x2 ) ≤ 0 (9)
when |x2 | < σ0 . Thus, this shows local stability of the origin.
Example 2.4 (Rigid body and rotational motion). Consider a rotational motion of a rigid body
in 3 − D space. Let ω be the angular velocity and I be the Inertia matrix. Then, in the absence of
external torques, the motion is described by
I ω̇ + ω × Iω = 0 (10)
where × dentoes cross product in R3 . Do a change of basis such that I is a diagonal matrix. Let
ωx , ωy , ωz be the components of the angular velocity. Equation (10) reduces to
Ix ω̇x = −(Iz − Iy )ωy ωz , Iy ω̇y = −(Ix − Iz )ωx ωz , Iz ω̇z = −(Iy − Ix )ωx ωx . (11)
Without loss of generality assume that Ix ≥ Iy ≥ Iz > 0 and replace ωx , ωy , ωz by x, y, z. Define
I −I I −I
a = yIx z , b = IxI−I
y
z
, c = xIz y . Therefore, (11) becomes
ẋ = ayz, ẏ = −bxz, ż = cxy.
For simplicity, suppose Ix > Iy > Iz ⇒ a, b, c > 0. For equilibrium, at least two quantities among
x, y, z must be zero. Thus, the equilibria are union of the three axes. Therefore, none of the
equilibria is isolated. Consider the origin as an equilibrium point. Define a Lyapunov function
candidate V (x, y, z) = px2 + qy 2 + rz 2 where p, q, r > 0. Then V is an lpdf.
V̇ = 2(pxẋ + qy ẏ + rz ż) = 2xyz(ap − bq + cr).
Choose p, q, r such that ap − bq + cr = 0. Hence, the origin is stable.
Example 2.5 (Non linear RC circuit). non linear resistive and linear capacitive circuit. Consider
a bank of capacitors (linear characteristics) with capacitances C1 , . . . , Cn connected to a bank of
resistors (non linear). Let x = [x1 · · · , xn ] denote the voltages accross each capacitor. Then the
current through the capacitors is C ẋ where C =diag(C1 , . . . , Cn ). Let i(x) denote the current vector
in the resistive network when the voltage vector x is applied accross its terminals such that i(0) = 0.
Let i(x) = G(x)x where G(.) is the non linear version of the conductance matrix. Therefore,
C ẋ = G(x)x ⇒ ẋ = C −1 G(x)x. (12)
Clearly, 0 is an equilibrium point. Consider the total energy stored in the capacitors i.e., V (x) =
1 T
2 x Cx. Therefore,
1 1
V̇ (x) = (ẋT Cx + xT C ẋ) = − xT [GT (x) + G(x)]x.
2 2
Let M (x) := GT (x)+G(x). If M (x) is lpdf, then 0 is asymptotically stable. Since M is continuous,
M is lpdf if M (0) > 0.
Suppose M (0) > 0. Therefore, by continuity, there exists > 0 such that M (x) > 0 for all
x ∈ B (0). Let d := infx∈B (0) λmin M (x) and choose small enough such that d > 0. Then,
V̇ = −xT M (x)x ≤ −dkxk2 , ∀x ∈ B (0). Since, V (x) = 12 xT Cx,
−d
λmax (C)kxk2 ≥ V (x) ⇒ −dkxk2 ≤ V (x).
λmax (C)
−d
Therefore, V̇ ≤ λmax (C) V (x) and from Theorem 2.1 case 4, 0 is locally exponentially stable. If
d = infx∈Rn λmin M (x) > 0, then 0 is globally exponentially stable. It is also globally asymptotically
stable since, V̇ < 0.
Example 2.6. not radially unbounded and finite escape time: Ex. on p.174 of [2]. It shows
that if V is not radially unbounded and satisfies all the remaining properties, then trajectories
starting sufficiently far from the origin may have a finite escape time and the origin is not globally
exponentially stable.
Example 2.7 (Phase locked loop). [2] p.181.
2.2 Invariant sets and stability theorems

Lemma 2.8 (limit sets). If a solution x(t) of ẋ = f (x) is bounded and belongs to S ⊂ Rn for t ≥ 0,
then its limit set L is nonempty, compact and invariant. Moreover, x(t) → L as t → ∞.
Proof. The proof is slightly technical and we refer the reader to Haddad Theorem 2.41.
Relaxing strict negative definite condition on V̇ :
Theorem 2.9 (LaSalle’s invariance principle (Barbashin-Krasovskii-LaSalle Theorem)). Let Ω ⊂ S
be a compact invariant set of ẋ = f (x). Let V : S → R be a continuously differentiable function
such that V̇ ≤ 0 in Ω. Let E be the set of points where V̇ (x) = 0. Let M be the largest invariant
set in E. If x(0) ∈ Ω, then x(t) → M as t → ∞.
Proof. Let x(t) be a solution of ẋ = f (x) starting in Ω. Since V̇ ≤ 0 in Ω, V is a decreasing
function of t. Furthermore, since V is continuous on the compact set Ω, it is bounded from below
on Ω. Therefore, limt→∞ V (x(t)) = a exists. The limit set L of the trajectory x(t) lies in Ω since
Ω is invariant and closed. For any p ∈ L, there exists a sequence {tn } such that tn → ∞ and
x(tn ) → p as n → ∞. By continuity of V (x), V (p) = limn→∞ V (x(tn )) = a. Hence, V (x) = a on
L. By the previous lemma, L is an invariant set and since V is a constant function on L, V̇ = 0
on L. Therefore, L ⊂ M ⊂ E ⊂ Ω. Since x(t) is bounded, x(t) approaches L as t → ∞. Hence,
x(t) → M as t → ∞.
Observe that there is no positiveness assumption on V in the above theorem. However, invari-
ance of Ω is assumed an the initial condition is assumed to belong to this invariant set. One can
construct invariant sets using level sets of V (x) > 0 with V̇ ≤ 0. So one can observe that one can
have any of these two hypotheses to obtain invariance.
Corollary 2.10 (asymptotic stability using LaSalle). Let x = 0 be an equilibrium point of ẋ = f (x).
Let V : S → R be a continuously differentiable positive definite function on S (i.e., V is lpdf )
containing x = 0 such that V̇ ≤ 0 on S. Let E be the set of points where V̇ (x) = 0 and suppose
that no solution can stay identically in E other than the trivial solution x = 0. Then, the origin is
asymptotically stable.
Proof. Follows as a consequence of the previous theorem.
Corollary 2.11 (global asymptotic stability using LaSalle). Let x = 0 be an equilibrium point of
ẋ = f (x). Let V : Rn → R be a continuously differentiable radially unbounded positive definite
function on Rn \ {0} such that V̇ ≤ 0 on Rn . Let E be the set of points where V̇ (x) = 0 and
suppose that no solution can stay identically in E other than the trivial solution x = 0. Then, the
origin is globally asymptotically stable.
Proof. Again follows from the theorem.
Example 2.12. Consider the case of damped simple pendulum in Example 2.2. Since V̇ = −bx22 ,
V̇ = 0 on the line x2 = 0. To maintain V̇ = 0, the trajectory must be confined to the line x2 = 0.
Now x2 = 0 ⇒ ẋ2 = 0 ⇒ sin(x1 ) = 0 ⇒ x1 = nπ for n = 0, ±1, ±2, . . . . Therefore, on the segment
−π < x1 < π, the largest invariant set is the origin and this implies asymptotic stability by LaSalle.
Example 2.13. Non linear RLC circuit (Example 2.3), asymptotic stability of the origin using
LaSalle: Recall from Example 2.3 that V̇ (x1 , x2 ) = −x2 f (x2 ) ≤ 0 for x2 ∈ [−σ0 , σ0 ]. Let c =
min(V (−σ0 , 0), V (σ0 , 0)). Then, V̇ ≤ 0 for x ∈ Ωc := {(x1 , x2 ) | V (x1 , x2 ) ≤ c}. By LaSalle’s in-
variance principle, the trajectory approaches the largest invariant set in Ωc ∩{(x1 , x2 ), | V̇ (x1 , x2 ) =
0} = Ωc ∩ {x1 , 0} (since f (x2 ) = 0 only when x2 = 0). Note that x2 = 0 in the largest invariant
set which implies that ẋ1 = 0 ⇒ x1 = x10 . Furthermore x2 = 0 ⇒ ẋ2 = 0 = −f (0) − g(x10 ) ⇒
g(x10 ) = 0 ⇒ x10 = 0 (since, g(x1 ) = 0 only when x1 = 0). Therefore, the origin is the largest
invariant set in Ωc ∩ {(x1 , x2 ), | V̇ (x1 , x2 ) = 0}. Hence, it is locally asymptotically stable.
Example 2.14 (Stabilization of a rigid robot). [2] p.183. Consider a rigid robot where components
of q represent generalized co-ordinates of the robot. Let u represent generalized forces. Suppose the
system of equations are
M (q)q̈ + C(q, q̇)q̇ = u
Let x = q and y = q̇. Therefore, ẋ = y and ẏ = [M (x)]−1 [u − C(x, y)y]. Let qd represent the
desired value of the generalized co-ordinates. Suppose
u = −Kp (q − qd ) − Kd q̇ = −Kp (x − qd ) − Kd y,
where Kp , Kd > 0. Now the system equations are
ẋ = y, ẏ = −[M (x)]−1 [Kp (x − qd ) + Kd y + C(x, y)y]
The equilibrium point is (x, y) = (qd , 0). Consider a Lyapunov function

1 T
V (x, y) = [y M (x)y + (x − qd )T Kp (x − qd )]
2
1
⇒ V̇ = yT M (x)ẏ + yT Ṁ (x)y + ẋT Kp (x − qd )
2
1
T
= −y [Kp (x − qd ) + Kd y + C(x, y)y] + yT Ṁ (x)y + yT Kp (x − qd )
2
1
= −yT Kd y + yT [Ṁ (x, y) − 2C(x, y)]y
2
It turns out that Ṁ − 2C is skew symmetric. Therefore, V̇ ≤ 0 which implies stability. Applying
Krasovskii-LaSalle principle, for the set of points where V̇ = 0 i.e., to points (x, y) ∈ Rn × 0, the
largest invariant set is given by the set of points in Rn × 0 where (ẋ, ẏ) = (0, 0). Now ẋ = 0 implies
that y = 0 and ẏ = 0 implies [M (x)]−1 [Kp (x − qd ) + Kd y + C(x, y)y] = 0. Since y = 0 and M
is invertible, this implies that Kp (x − qd ) = 0 and since Kp > 0, x = qd . Therefore, (qd , 0) forms
the largest invariant set and by LaSalle (Corollary 2.11), it is globally asymptotically stable.
Example 2.15 (Stability of a limit cycle using the invariance principle). Consider a system
ẋ1 = x2 + x1 (r2 − x21 − x22 )

ẋ2 = −x1 + x2 (r2 − x21 − x22 ), r > 0
which has an equilibrium point at the origin. Observe that if x21 + x22 = r2 , then, ẋ1 = x2 and
ẋ2 = −x1 and the evolution remains on the circle x21 + x22 = r2 and it forms an invariant set. Let
V (x) = 41 (x21 + x22 − r2 )2 . Thus, V (x) ≥ 0 on R2 . Note that V̇ (x) = ∇V.f = −(x21 + x22 )(x21 +
x22 − r2 )2 ≤ 0. Note that V̇ = 0 either at the origin or on the circle x21 + x22 = r2 . Let c > r and
Mc := {x ∈ R2 | V (x) ≤ c}. Since V̇ ≤ 0 on Mc and V ≥ 0, Mc is an invariant set containing the
origin and the circle x21 + x22 = r2 . On both these sets, V̇ = 0 and the largest invariant set in Mc is
the union of the origin with the circle (they are not comparable as neither is a subset of the other).
Therefore, all trajectories converge to either one of them by the invariance principle.
4 4 4
Observe that at the origin, V (0) = r4 . Suppose r < r4 and let r < c < r4 . Let Mc be as
defined before. Note that Mc excludes the origin in this case but contains the circle x21 + x22 = r2 .
Now applying the invariance principle, all trajectories converge to the circle implying a stable limit
cycle.
2.3 Constructing Lyapunov functions

Observe that by chain rule, V̇ = ni=1 ∂x
∂V
ẋi = ni=1 ∂x
∂V
fi (x). Let g(x) = ∂V
P P
i i ∂x . Therefore, V̇ =
g(x).f (x). We need to construct g such that it is a gradient of a positive definite function and
V̇ (x) = g(x).f (x) < 0. Note that
ˆ x ˆ n
xX
V (x) = g(s)ds = gi (s)dsi . (13)
0 0 i=1
Since line integral of a gradient is path independent, take a path from 0 to x formed by lines
parallel to the co-ordinate axes. Therefore, (13) becomes
ˆ x1 ˆ x2 ˆ xn
V (x) = g1 (s1 , 0, . . . , 0)ds1 + g2 (x1 , s2 , 0, . . . , 0)ds2 + . . . + gn (x1 , x2 , . . . , xn−1 , sn )dsn .(14)
0 0 0
Alternatively, choosing a straight line path joining x to 0 with parametrization s = σx, where
σ ∈ [0, 1], (13) can also be written as
ˆ 1 ˆ n
1X
V (x) = g(σx).xdσ = gi (σx)xi dσ. (15)
0 0 i=1
Proposition 2.16. A function g : Rn → Rn is the gradient vector of a scalar valued function

∂gi ∂g
V : Rn → R if and only if ∂x j
= ∂xji , i, j = 1, . . . , n.
∂gi ∂gj
Proof. (⇒) trivial. (⇐) Suppose ∂xj = ∂xi , i, j = 1, . . . , n. Let
ˆ 1 ˆ n
1X
V (x) = g(σx)xdσ = gj (σx)xj dσ (16)
0 0 j=1
Let y = σx, therefore,

ˆ n
1X ˆ 1
∂V ∂gj
= (σx)xj dσ + gi (σx)dσ
∂xi 0 j=1 ∂xi 0
ˆ 1Xn ˆ 1
∂gi
= (σx)xj dσ + gi (σx)dσ
0 j=1 ∂xj 0
ˆ 1Xn ˆ 1
∂gi
= (y)σxj dσ + gi (σx)dσ
0 j=1 ∂yj 0
ˆ 1
d(σgi (σx))
= dσ (17)
0 dσ
= gi (x), i = 1, . . . , n
Proposition 2.17. Let f , g : Rn → Rn be continuously differentiable functions such that f (0) = 0.

Then for every x ∈ Rn , there exists α ∈ [0, 1] such that
∂f
g(x).f (x) = gT (x) (αx)x. (18)
∂x
Proof. (MVT for vector valued functions (Theorem 2.16 of Haddad): Let D ⊂ Rm and f : D → Rn
which is continuously differentiable. Let x, y ∈ D such that L := {z | z = µx + (1 − µ)y, µ ∈
(0, 1)} ⊂ D. Then, for every v ∈ Rn , there exists z ∈ L such that vT [f (y) − f (x)] = vT [f 0 (z)(y −
x)].) Let p ∈ Rn . Then by mean value theorem, for every x ∈ Rn , there exists α ∈ [0, 1] such that
gT (p)f (x) = gT (p)[f (x) − f (0)]

∂f
= gT (p)[ (αx)x].
∂x
∂f
Hence, there exists α ∈ [0, 1] such that g(p).f (p) = gT (p) ∂p (αp)p. Since p is arbitrary, the result
follows.
Theorem 2.18 (Krasovskii’s theorem). Let 0 be an equilibrium point of ẋ = f (x(t)), x(0) = x0 ,
t ≥ 0 where f : S → Rn is continuously differentiable and S is an open set containing 0. Let
P, R ∈ Rn×n , P, R > 0 such that
∂f (x) T ∂f (x)
[ ] P + P[ ] ≤ −R, x ∈ S, x 6= 0. (19)
∂x ∂x
Then the zero solution is a unique asymptotically stable equilibrium point with Lyapunov function
V (x) = f T (x)P f (x). If S = Rn , then the zero solution is a unique globally asymptotically stable
equilibrium.
Proof. Suppose there exists xe ∈ S, xe 6= 0 and f (xe ) = 0. By the previous proposition, for every
xe ∈ S, there exists α ∈ [0, 1] such that 0 = xT T ∂f (αxe ) x . Hence,
e P f (xe ) = xe P ∂x e
∂f (αxe ) T ∂f (αxe )
xT
e {[ ] P + P[ ]}xe = 0
∂x ∂x
which contradicts (19). Therefore, S does not contain any other equilibrium point of f . Note that
V (x) = f T (x)P f (x) ≥ λmin (P )kf (x)k22 ≥ 0, x ∈ S. This implies that V (x) = 0 if and only if
f (x) = 0 i.e., x = 0. Therefore, V > 0 on S \ {0}. Moreover,
∂f (x) ∂f (x) T ∂f (x)

V̇ = V 0 (x)f (x) = 2f T (x)P f (x) = f T (x){[ ] P + P[ ]}f (x)
∂x ∂x ∂x
≤ −f T (x)Rf (x) ≤ −λmin (R)kf (x)k22 ≤ 0, x ∈ S.
Since f (x) = 0 iff x = 0 in S, V̇ < 0 is S \ {0} which proves that 0 is asymptotically stable
equilibrium point with Lyapunov function V (x) = f T (x)P f (x).
Now suppose S = Rn . We need to show that V (x) → ∞ as kxk → ∞. By the previous
proposition, for every x ∈ Rn and some α ∈ (0, 1),
∂f (x) 1 ∂f (αx) T ∂f (αx)

xT P f (x) = xT P (αx)x = xT {[ ] P + P[ ]}x
∂x 2 ∂x ∂x
1 1
≤ − xT Rx ≤ − λmin (R)kxk2 . (20)
2 2
This implies that xT P f (x) < 0 since rhs above is negative. Therefore, |xT P f (x)| ≥ 12 λmin (R)kxk2 .
|xT P f (x)| λmin (R)
Moreover, |xT P f (x)| ≤ λmax (P )kxkkf (x)k ⇒ kf (x)k ≥ λmax (P )kxk ≥ 2λmax (P ) kxk. This implies by
definition of V that V (x) → ∞ as kxk → ∞.
Example 2.19. Example 3.9 Haddad.
The following theorem allows us to find the domain of attraction as well.
Theorem 2.20 (Zubov’s theorem). Let 0 be an equilibrium point of ẋ = f (x(t)). Let S ⊂ Rn be

bounded and suppose there exists a continuously differentiable function V : S → R and a continuous
function h : Rn → R such that V (0) = 0, h(0) = 0 and
0 < V (x) < 1, x ∈ S, x 6= 0 (21)

V (x) → 1 asx → ∂S (22)
n
h(x) > 0, x ∈ R , x 6= 0 (23)
0
V (x)f (x) = −h(x)[1 − V (x)]. (24)
Then, 0 is asymptotically stable with domain of attraction S.

Proof. It follows from (21), (23) and (24) that in the neighborhood B (0) of the origin, V (x) > 0
and V̇ < 0. Hence, 0 is locally asymptotically stable. To show that S is the domain of attraction,
we need to show that x(0) ∈ S implies that x(t) → 0 as t → ∞ and x(0) ∈ / S implies that x(t) 6→ 0
as t → ∞.
Let x(0) ∈ S. Then from (21), V (x(0)) < 1. Let β > 0 be such that V (x(0)) ≤ β < 1 and
define Ωβ := {x ∈ S | V (x) ≤ β}. Since Ωβ ⊂ S, Ωβ is bounded. Since V̇ < 0 on Ωβ , it is clear
that Ωβ forms an invariant set. If V̇ = 0, then h(x) = 0 which further implies that x = 0. Now by
Theorem 2.9, x(t) → 0 as t → ∞.
Let x(0) ∈ / S and suppose x(t) → 0 as t → ∞. Thus, x(t) → S for some t ≥ 0. Therefore, there
exists t1 , t2 such that x(t1 ) ∈ ∂S and x(t) ∈ S for all t ∈ (t1 , t2 ]. Let W (x) = 1 − V (x). Hence,
Ẇ = h(x)W (x). Therefore,
ˆ W (x(t)) ˆ t ´t
dW − h(x(s))ds
= h(x(s)) ⇒ 1 − V (x(t0 )) = [1 − V (x(t))]e t0 . (25)
W (x(t0 )) W t0
Let t = t2 ´and t0 → t1 . Using (22), it follows that limt0 →t1 [1 − V (x(t0 ))] = 0 and limt0 →t1 [1 −
t
− h(x(s))ds
V (x(t))]e t0 > 0 which is a contradiction. Therefore, for x(0) ∈
/ S, x(t) 6→ 0 as t → ∞.
Example 2.21. Haddad: Example 3.10
Constants of motion or first integrals or integrals of motion or dynamic invariants or Casimir

functions. A function C : S → R is an integral of motion if it is conserved along the flow of the
dynamical system i.e., C 0 (x).f (x) = 0. Let Ci : S → R i = 1, . . . , r be twice differentiable Casimir
functions. Define
r
X
E(x) := µi Ci (x) (26)
i=1
for µi ∈ R, i = 1, . . . , r.
Theorem 2.22 (Energy-Casimir theorem). Consider a non linear dynamical system ẋ = f (x)
where f : S → Rn is Lipschitz. Let xe be an equilibrium point and let Ci : S → R, i = 1, . . . , r be
Casimir functions of the dynamical system. Suppose Ci0 (xe ) i = 2, . . . , r are linearly independent
and suppose there exists µ = [µ1 , . . . , µr ]T ∈ Rr such that µ1 6= 0, E 0 (xe ) = 0 and xT E 00 (xe )x > 0
for x ∈ M where M = {x ∈ S | Ci0 (xe )x = 0, i = 2, . . . , r}. Then, there exists α ≥ 0 such that
r
X ∂Ci (xe ) T ∂Ci (xe )
E 00 (xe ) + α ( ) ( ) > 0. (27)
∂x ∂x
i=2
Furthermore, the equilibrium point xe is Lyapunov stable with Lyapunov function

r
αX
V (x) = E(x) − E(xe ) + [Ci (x) − Ci (xe )]2 . (28)
2
i=2
Proof. Note that since Ci are Casimir functions, Ci0 (x).f (x) = 0. Moreover,
V̇ (x) = V 0 (x).f (x)

r
X
0
= E (x).f (x) + α [Ci (x) − Ci (xe )]Ci0 (x).f (x)
i=2
r
X r
X
= µi Ci0 (x).f (x) +α [Ci (x) − Ci (xe )]Ci0 (x).f (x) = 0 (29)
i=1 i=2
for x ∈ S. We need to show that V (xe ) = 0 and V (x) > 0, x ∈ S \ xe . Clearly, V (xe ) = 0 and
r
X
V 0 (x) = E 0 (x) + α [Ci (x) − Ci (xe )]Ci0 (x) ⇒ V 0 (xe ) = 0.
i=2
r
X
V 00 (x) = E 00 (x) + α {[Ci0 (x)]T Ci0 (x) + [Ci (x) − Ci (xe )]Ci00 (x)}
i=2
r
X ∂Ci (xe ) T ∂Ci (xe )
⇒ V 00 (xe ) = E 00 (xe ) + α ( ) ( ). (30)
∂x ∂x
i=2
Recall that by hypothesis, if x ∈ M, then Ci0 (xe ).x = 0, i = 2, . . . , r. Consider a subspace V1

formed by linearly independent elements of M which is a r − 1 dimensional subpace and extend
the basis of this subspace to a basis of Rn by taking linearly independent vectors in the orthogonal
complement V1⊥ . With respect to the new basis, elements of V1 can be written as [∗ 0]T and
elements of V1⊥ can be written as [0 ∗]T . Therefore, there exists an invertible matrix T ∈ Rn×n
such that [0 Ir−1 ]T x = 0 when x ∈ M and [In−r−1 0]T x = 0 when x ∈ M ⊥ . Hence, for xe ∈ S,

00 T E1 E12
E (xe ) = T T T (31)
E12 E2
where E1 > 0 and

r
X ∂Ci (xe ) T ∂Ci (xe ) T 0 0
( ) ( )=T T (32)
∂x ∂x 0 N
i=2
where N > 0. Substituting (31) and (32) in (30),

00 T E1 E12
V (x) = T T T.
E12 E2 + αN
By choosing appropriate α, V 00 (xe ) > 0. Since V is twice differentiable and V (xe ) = V 0 (xe ) = 0, it
follows from V 00 (xe ) > 0 that V > 0 in the neighborhood of xe .
Thus, using integrals of motions, one can construct a Lyapunov function.

Example 2.23. Example 3.11 Haddad.
2.4 Instability theorems

Theorem 2.24 (instability theorem). Let 0 be an equilibrium point of ẋ = f (x) and let V : S → R
be a continuously differentiable function such that V (0) = 0 and V (x0 ) > 0 for an x0 with arbitrarily
small kx0 k. Let r > 0 and U = {x ∈ Br (0) | V (x) > 0} such that V̇ > 0 on U . Then, x = 0 is
unstable.
Proof. By continuity of V if V (x0 ) > 0, then V > 0 in a neighborhood of x0 hence, x0 lies in the
interior of U . Let V (x0 ) = a > 0. The trajectory starting at x0 must eventually leave U . One
justifies this claim as follows. As long as x(t) is inside U , V (x(t)) ≥ a since, V̇ > 0 in U . Let
γ = min{V̇ (x) | x ∈ U, V (x) ≥ a}
which exists since, the continuous function V̇ has a minimum over the compact set {x ∈ U | V (x) ≥
a}. Clearly, γ > 0. Note that
ˆ t
V (x(t)) = V (x0 ) + V̇ (x(s))ds ≥ a + γt.
0
This inequality shows that x(t) can not stay in U forever because V (x) is bounded on U . Moreover,
x(t) can not leave through the surface {x | V (x) = 0} since V (x(t)) ≥ a. Hence, it must leave
U through the sphere kxk = r. Beause this can happen for arbitrarily small kx0 k, the origin is
unstable.
Theorem 2.25 (second instability theorem). Let 0 be an equilibrium point of ẋ = f (x) and let
V : S → R be a continuously differentiable function, W : S → R and λ, > 0 such that V (0) = 0
and W (x) ≥ 0 for all x ∈ B (0) and V̇ = V 0 (x).f (x) = λV (x) + W (x). Furthermore, assume that
for every sufficiently small δ > 0, there exists x0 ∈ S such that kx0 k < δ and V (x0 ) > 0. Then,
x = 0 is unstable.
Proof. Suppose there exists δ > 0 such that if x0 ∈ Bδ (0), then x(t) ∈ B (0), t ≥ 0. Note that
V (x0 ) > 0. From the hypothesis, V̇ = λV (x) + W (x) and W (x) ≥ 0. This implies that V̇ ≥ λV (x)
for x ∈ B (0) i.e., V̇ − λV (x) ≥ 0 for x ∈ B (0). Therefore,
e−λt V 0 (x(t)).f (x(t)) − λe−λt V (x(t)) ≥ 0, t ≥ 0

d
⇒ [e−λt V (x(t))] ≥ 0, t ≥ 0.
dt
Integrating both sides, e−λt V (x(t)) − V (x(0)) ≥ 0 ⇒ V (x(t)) ≥ eλt V (x(0)), t ≥ 0. Since V (x0 ) >
0, x(t) ∈
/ B (0) as t → ∞ which is a contradiction. Hence, the zero solution is unstable.
Theorem 2.26 (Chetaev’s instability theorem). Let 0 be an equilibrium point of ẋ = f (x) and let
V : S → R be a continuously differentiable function. Let > 0 and an open set O ⊂ B (0) such
that
V (x) > 0 ∀x ∈ O, supx∈O V (x) < ∞, 0 ∈ ∂O (33)

V (x) = 0 ∀x ∈ ∂O ∩ B (0), (34)
0
V (x).f (x) > 0, x ∈ O. (35)
Then, the zero solution is unstable.
Proof. Let P ⊂ O be a closed set such that for x0 ∈ O, x(t) ∈ P ⊂ O ⊂ B (0) for t ≥ 0. Note that
ˆ t
V (x(t)) = V (x(0)) + V̇ (x(s))ds
0
ˆ t
= V (x(0)) + V 0 (x(s)).f (x(s))ds ≥ V (x(0)) + αt (36)
0
where α = minx∈P V 0 (x).f (x) > 0. This implies that V (x(t)) → ∞ as t → ∞ contradicting one of
the hypothesis supx∈O V (x) < ∞. Therefore, there exists T > 0 such that x(T ) ∈ ∂O (i.e., either
x(t) escapes O at some finite time) or x(t) → ∂O as t → ∞ (i.e., x(t) escapes O at infinity). One
of these two cases must be true since we have shown by contradiction that there is no closed set in
O that contains limt→∞ x(t).
Consider the first case i.e., suppose there exists T > 0 such that x(T ) ∈ ∂O. Since V is
strictly increasing on O, V (x(T )) > 0. By hypothesis, V (x) = 0 on ∂O ∩ B (0). This implies that
x(T ) ∈/ ∂O ∩ B (0) hence, x(T ) ∈ ∂O \ B (0). Observe that ∂O = Ō ∩ ∂O and Ō ⊂ B̄ (0). This
implies that
∂O \ B (0) = (Ō ∩ ∂O) \ B (0) ⊂ (B̄ (0) ∩ ∂O) \ B (0) = ∂O ∩ ∂B (0).
Therefore, x(T ) ∈ ∂B (0). Similarly for the second case, if x(t) → ∂O as t → ∞, then x(t) →
∂B (0) as t → ∞. Thus, there does not exists a δ > 0 such that if x0 ∈ Bδ (0), then x(t) ∈ B (0)
for t ≥ 0. This implies that the zero solution is unstable.
Thus, it can be observed that Chetaev’s instability theorem requires milder conditions (milder
in the sense that the properties need to be satisfied on a subset O ⊂ S and not on the whole of S)
than Lyapunov instability theorems to conclude about instability of an equilibrium point.
Example 2.27.
2.5 Linear autonomous/LTI systems, linearization

Consider an autonomous linear system
ẋ = Ax. (37)
Theorem 2.28 (Stability for LTI). The zero solution of (37) is Lyapunov stable if and only if
every eigenvalue of A has a real part strictly less than zero or equal to zero and eigenvalues with
zero real part have trivial Jordan structure (i.e., they are semi-simple).
Proof. Note that eAt is bounded iff the condition on the eigenvalues of A mentioned above is
satisfied.
Theorem 2.29 (Lyapunov asymptotic stability for LTI). Following are equivalent.
1. The system (37) is asymptotically stable.
2. The system (37) is exponentially stable.
3. All eigenvalues of A have strictly negative real parts.
4. For every Q > 0, there exists a unique solution P > 0 to the following Lyapunov equation
AT P + P A = −Q. (38)
5. There exists P > 0 which satisfies the following Lyapunov matrix inequality
AT P + P A < 0. (39)
6. There exists C ∈ Rm×n such that the pair (C, A) is observable, and there exists P > 0 which
satisfies
AT P + P A + C T C = 0. (40)
7. For all C ∈ Rm×n such that the pair (C, A) is observable, and there exists P > 0 which
satisfies (40).
Proof. The equivalence of first five statements is shown in Hespanha. Equivalence of the remaining
two statements with the remaining onces is given in Sastry. (7) ⇒ (6). To show (6) ⇒ (1),
let V (x) = xT P x and −V̇ = xT C T Cx. Then, by LaSalle’s invariance principle, the trajectories
starting from arbitrary initial conditions converge to the largest invariant set in the null space of
C. Since (C, A) is observable, largest invariant set in the null space of C is the origin. Therefore,
the origin is asymptotically stable. ´∞ T
(1) ⇒ (7) Just as in (2) ⇒ (4), let P = 0 eA t C T CeAt dt. Clearly, P ≥ 0. To show that
P > 0, let xT P x = 0. This implies that CeAt x = 0. Differentiating n times at the origin, we obtain
Cx = CAx = . . . = CAn−1 x = 0. Since (C, A) is observable, x = 0 ⇒ P > 0 and P also satisfies
(40).
Theorem 2.30 (Lyapunov stability for LTI). The zero solution of (37) is Lyapunov stable ⇔ there
exists P > 0 and Q ≥ 0 such that (38) holds.
Proof. (⇐) Follows by choosing V (x) = xT P x. (⇒) Suppose the zero solution is stable. From
Theorem 2.28, eigenvalues of A lie in the left half complex plane with those having zero real parts
being semi-simple. Without loss of generality, suppose A is in Jordan canonical form with first r
Jordan blocks having eigenvalues strictly
in the left half plane and the remaining ones with zero
Jl 0
real parts and semi-simple. Let A = . It is clear that J0 is skew symmetric. Since
0 J0
Jl has T
eigenvalues
in theLHP, there
exists P1 > 0, Q1 > 0 such that Jl P1 + P1 Jl = Q1 . Let
P1 0 Q1 0
P = and Q = such that (38) holds.
0 I 0 0
√
Remark 2.31. In the above theorem, if C = Q and (C, A) is observable, then the origin is
2.6 Indirect method of Lyapunov and local linearization

These methods work only locally i.e., they tell about local stability of an equilibrium point. No
conclusion can be made about global stability.
Theorem 2.32 (Lyapunov’s indirect theorem for autonomous systems using linearization). Let
zero be an equilibrium point of (7) where f : S → Rn is continuously differentiable and S is open
∂f
with 0 ∈ S. Let A = ∂x |x=0 . Then following holds:
1. (Local exponential stability of linearization) If Re(λ) < 0 where λ ∈ spec(A), then the zero
solution is exponentially stable.
2. (Unstability of linearization) If there exists λ ∈ spec(A) such that Re(λ) > 0, then the zero
solution is unstable.
Proof. Suppose Re(λ) < 0, λ ∈ spec(A). Therefore, there exists a unique P > 0 which satisfies
(38). Let V (x) = xT P x. Note that using the Taylor expansion of f around 0,
V̇ = V 0 .f = 2xT P [Ax + r(x)]
= xT (AT P + P A)x + 2xT P r(x)
= −xT Qx + 2xT P r(x). (41)
Using Cauchy-Schwarz inequality and using −xT Qx ≤ −λmin (Q)kxk22 ,
V̇ ≤ −λmin (Q)kxk22 + 2λmax (P )kxk2 kr(x)k2 . (42)
Since kr(x)k2 goes to zero faster than kxk, for every γ > 0, there exists > 0 such that kr(x)k2 <
γkxk2 . Therefore, for all x ∈ B (0),
V̇ < −[λmin (Q) − 2γλmax (P )]kxk22 . (43)
Choosing γ ≤ λmin (Q)/2λmax (P ), it follows that V̇ (x) < 0 for all x ∈ B (0) \ 0. This shows
local asymptotic stability. Local exponential stability follows from the inequality λmin (P )kxk22 ≤
V (x) ≤ λmax (P )kxk22 .
For the proof of the second statement we refer the reader to Haddad p. 178.
For the converse of the first statement, refer Hespanha
Example 2.33. Consider damped simple pendulum from Example 2.2. Linearizing around the
origin, it can be checked that the linearization matrix has eigenvalues in the left half plane which
implies asymptotic stability. For a pendulum without damping, eigenvalues of the linearization
matrix are purely imaginary. Therefore, we can not conclude about stability by linearization. But
we have already checked stability for this case using Lyapunov function.
Consider another equilibrium point x1 = π, x2 = 0. It can be checked that the linearization
matrix has an eigenvalue in the right half plane which implies unstable equilibrium point.
Example 2.34 (Van der Pol oscillator). Consider the Van der Pol oscillator ([2]) described by
ẋ1 = x2 , ẋ2 = −µ(1 − x21 )x2 − x1 .
Linearize around 0. It turns out that if µ > 0, then both eigenvalues of the linearized matrix have
positive real parts which implies instability.
Theorem 2.35 (Lyapunov’s indirect theorem for control systems using linearization). Consider a
non linear control system ẋ = F(x, u) such that F(0, 0) = 0 and F is continuously differentiable. Let
A := ∂F ∂F
∂x |(x,u)=(0,0) , B := ∂u |(x,u)=(0,0) such that (A, B) is stabilizable i.e., there exists K ∈ R
m×n
such that spec(A + BK) lies in the left half complex plane. With the linear control law u = Kx,
the zero solution of the closed loop non linear system ẋ = F(x, u) is locally exponentially stable.
Proof. The closed loop system ẋ = F(x, Kx) = f (x) and
∂f ∂F(x, Kx) ∂F(x, Kx)
|x=0 = [ + K]|x=0 = A + BK.
∂x ∂x ∂u
Since A + BK is asymptotically stable, by the first statement of the previous theorem, the result
follows.
One can locally linearize the system and use separation principle to construct a local controller
which locally asymptotically stabilizes the non linear control system.
Applications in singularly perturbed systems ([2])
2.7 Converse theorems

Does there exists a Lyapunov function for stable, asymptotically stable, exponentially stable dy-
namical systems? These are converse theorems. For time varying systems, such functions exist
under some conditions. For time invariant or autonomous systems, Lyapunov stability does not
gurantee existence of a continuously differentiable or continuous time-independent Lyapunov func-
tion. (Existence of lower semi-continuous Lyapunov function Haddad.) However, for asymptotically
stable equilibrium points of autonomous systems, a continuously differentiable time-independent
Lyapunov function exists.
Lemma 2.36 (Massera’s lemma). We refer the reader to Haddad p. 162.
Theorem 2.37 (asymptotic stability converse). Suppose the zero solution of (7) is asymptotically
stable, f : S → Rn is continuously differentiable and let δ > 0 be such that Bδ (0) ⊂ S is contained
in the domain of attraction of (7). Then, there exists a continuously differentiable function V :
Bδ (0) → R such that V (0) = 0, V (x) > 0 for all x ∈ Bδ (0) \ {0} and V 0 (x).f (x) < 0 for all
x ∈ Bδ (0) \ {0}.
Proof. Theorem 3.9 Haddad.
Theorem 2.38 (exponential stability converse). Suppose the zero solution of (7) is exponentially
stable, f : S → Rn is continuously differentiable and let δ > 0 be such that Bδ (0) ⊂ S is contained
in the domain of attraction of (7). Then, for every p > 1, there exists a continuously differentiable
function V : S → R and scalars α, β, > 0 such that αkxkp ≤ V (x) ≤ βkxkp for all x ∈ Bδ (0) and
V 0 (x).f (x) < −V (x) for all x ∈ Bδ (0).
Proof. Theorem 3.10 Haddad.
Corollary 2.39. Suppose the zero solution of (7) is exponentially stable, f : S → Rn is continuously
differentiable and let δ > 0 be such that Bδ (0) ⊂ S is contained in the domain of attraction of (7).
Then, there exists a continuously differentiable function V : S → R and scalars α, β, > 0 such
that αkxk2 ≤ V (x) ≤ βkxk2 for all x ∈ Bδ (0) and V 0 (x).f (x) < −V (x) for all x ∈ Bδ (0).
Table 2: Basic Lyapunov stability theorems for non autonomous systems (Theorem 3.1)
Conditions on V (x, t) Conditions on −V̇ (x, t) Conclusion

lpdf ≥ 0 locally stable
lpdf, decrescent ≥ 0 locally uniformly stable
lpdf lpdf asymptotically stable
lpdf, decrescent lpdf uniformly asymptotically stable
pdf, decrescent pdf globally uniformly asymptotically stable
akxkp ≤ V (t, x) ≤ bkxkp locally ≥ ckxkp locally locally exponentially stable
akxkp ≤ V (t, x) ≤ bkxkp globally ≥ ckxkp globally globally exponentially stable
Proof. Take p = 2 in the previous theorem.
Theorem 2.40 (global exponential stability converse). Suppose the zero solution of (7) is globally
exponentially stable, f : Rn → Rn is continuously differentiable and globally Lipshitz. Then, for
every p > 1, there exists a continuously differentiable function V : S → R and scalars α, β, > 0
such that αkxk2 ≤ V (x) ≤ βkxk2 for all x ∈ Bδ (0) and V 0 (x).f (x) < −V (x) for all x ∈ Bδ (0).
Proof. Theorem 3.11 of Haddad.
3 Stability theory for non autonomous/time varying systems

3.1 Basic stability theorems for time varying systems
Theorem 3.1. Consider a non autonomous system ẋ = f (x(t), t), x(0) = x0 where x(t) ∈ S ⊂ Rn .
Let V : S × R → R. Let a, b, c > 0 and p ≥ 1. Then, Table 2 holds.
Proof. Consider the first case. Since V is an lpdf, there exists α ∈ K such that V (x, t) ≥ α(kxk),
∀x ∈ Bs (0). Moreover, since V̇ ≤ 0 locally, V̇ (x, t) ≤ 0 ∀t ≥ t0 , ∀x ∈ Br (0). To show local
stability of 0, we need to show that for given > 0, t0 ≥ 0, there exists δ = δ(, t0 ) such that for
all kx(t0 )k < δ, kx(t)k < , ∀t ≥ t0 .
Define 1 := min(, s, r). Choose δ > 0 such that
β(t0 , δ) := supkxk≤δ V (x, t0 ) < α(1 ).
Such δ always exists because, α(1 ) > 0 and limδ→0 β(t0 , δ) = limδ→0 supkxk≤δ V (x, t0 ) = 0. Note
that
α(kx(t0 )k) ≤ V (x(t0 ), t0 ) < α(1 ).
Since α is increasing, this implies that kx(t0 )k < 1 . We need to show that kx(t)k < 1 for all
t ≥ t0 . Suppose this is not true. Let t1 > t0 be the first instant such that kx(t)k ≥ 1 . Then
V (x(t1 ), t1 ) ≥ α(1 ) > V (x(t0 ), t0 ).
But this is a contradiction since V̇ (x, t) ≤ 0 for all kxk < 1 . Thus, kx(t)k < 1 for all t ≥ t0 .
For the second case, since V is decrescent, define
β(δ) := supkxk≤δ supt≥t0 V (x, t).
Note that β is non decreasing and there exists d > 0 such that β(δ) < ∞ for 0 ≤ δ ≤ d. Choosing
δ such that β(δ) < α(1 ), using similar arguments used in the previous case, the statement follows.
For the third case, it is clear from the first case that 0 is stable. We need to show that
limt→∞ x(t) = 0 i.e., limt→∞ kx(t)k = 0. To show this, we need to show that for every > 0,
there exists δ = δ(, t0 ) > 0 and T (, t0 ) < ∞ such that kx(t0 )k < δ ⇒ ks(t, t0 , x(t0 ))k < for all
t > T (, t0 ) ≥ t0 . For the sake of simplicity, we avoid giving further details here. Case 4 is proved
in [2] and the proof of case 3 follows on similar lines. For cases 5, 6, 7, we refer the reader to [2].
Converses two first two cases are true ([2] p.160, Remark pt. 1).
Alternate geometric proof (Khalil):
Proof. 1. Since V is an lpdf, there exists α ∈ K such that α(kxk) ≤ V (t, x) for all x ∈ Bs (0).
Moreover, V̇ = ∂V ∂V
∂t + ∂x f (t, x) ≤ 0 for all x ∈ Br (0). Given < min (r, s), B (0) ⊂ S. Choose
c > 0 such that c < min (minkxk= α(kxk), r, s). Then, {x ∈ B (0) | α(kxk) ≤ c} is in the interior
of B (0). Define
Ωt,c := {x ∈ B (0) | V (t, x) ≤ c}.
Since V (t, x) ≤ c ⇒ α(kxk) ≤ c, Ωt,c ⊂ {x ∈ B (0) | α(kxk) ≤ c}. Observe that V̇ ≤ 0 on Ωt,c for
all t ≥ t0 . Therefore, solution starting in Ωt0 ,c , stays in Ωt0 ,c for all t ≥ t0 . Now choose a δ ball
around 0 which lies inside Ωt0 ,c . For all x0 lying in this δ ball, x(t) remains inside Ωt0 ,c ⊂ B (0).
This shows the stability of 0.
2. Now suppose V is decrescent. Hence, α1 (kx(t)k) ≤ V (t, x(t)) ≤ α2 (kx(t)k). Therefore,
kx(t)k ≤ α1−1 (V (t, x(t))) ≤ α1−1 (V (t0 , x(t0 ))) ≤ α1−1 (α2 (kx(t0 )k))
Therefore, given > 0, let δ = α2−1 (α1 ()). This shows uniform stability.
3. Since, V̇ < 0, Ωt,c shrinks as t → ∞. This implies asymptotic stability.
4. Follows from 2 and 3.
5. Since V is a pdf, follows from 4.
6, 7. Follows similarly using extension of arguments used for the time invariant case. Using the
given inequalities, it follows that V̇ ≤ − cb V = V . Now using comparison lemma and similar
arguments used for time invariant case, the statements follow.
Example 3.2.
For Lyapunov stability of periodic systems, refer [2]
Theorem 3.3 (Exponential stability and its converse). Sastry, Theorem 5.17.
Lemma 3.4 (Barbalat’s

´∞ lemma). Suppose that x(t) is bounded, ẋ(t) is bounded, w is uniformly
continuous, and t0 w(x(t))dt < ∞. Then, w(x(t)) → 0 as t → ∞.
Proof. Since w is uniformly continuous, for every > 0, there exists δx > 0 such that kx − yk <
δx ⇒ kw(x) − w(y))k < 2 . Since ẋ is bounded, x(t) is uniformly continuous i.e., given any δx > 0,
there exists δt such that |t1 − t2 | < δt ⇒ kx − yk < δx ⇒ kw(x) − w(y)k < 2 . Therefore, w is
uniformly continuous w.r.t. t.
Suppose w(x(t)) 6→ 0 as t → ∞, i.e., there is an > 0 and a sequence {tk } → ∞ such that
|w(x(tk ))| ≥ for all k. Let t ∈ [tk , tk + δ]. Then,

|w(x(t))| = |w(x(tk )) − (w(x(tk )) − w(x(t)))| ≥ |w(x(tk ))| − |w(x(tk )) − w(x(t))| ≥ − = .
2 2
Therefore,
ˆ tk +δ ˆ tk +δ

| w(x(t))dt| = |w(x(t))|dt ≥ δ > 0 (44)
tk tk 2
´s
where the equality holds since w(x(t)) retains its sign for tk ≤ t ≤ tk + δ. Therefore, t0 w(x(t))dt
cannot converge to a finite limit as s → ∞, a contradiction.
Theorem 3.5 (Generalization of LaSalle (LaSalle-Yoshizawa)). Suppose that the function f of (3)
is Lipshitz continuous in x, uniformly continuous in t in Br (0). Let α1 , α2 ∈ K and V (t, x) such
that
α1 (kxk) ≤ V (t, x) ≤ α2 (kxk).
Suppose there exists a continuous non negative function W (x) such that
∂V ∂V
V̇ (t, x) = + .f (t, x) ≤ −W (x) ≤ 0. (45)
∂t ∂x
Then for all kx(t0 )k ≤ α2−1 (α1 (r)), the trajectories x(.) are bounded and limt→∞ W (x(t)) = 0.
Proof. (Sastry, Theorem 5.27.) Consider V̇ (t, x) ≤ −W (x), and integrate from t0 to t. Therefore,
ˆ t ˆ t
V (t, x) − V (t0 , x0 ) ≤ − W (x(s))ds ⇒ W (x(s))ds ≤ V (t0 , x0 ) − V (t, x) ≤ V (t0 , x0 )
t0 t0
´t
where the last inequality holds because V > 0. Therefore, t0 W (x(s))ds exists and is bounded.
Claim: kx(t0 ))k ≤ α2−1 (α1 (ρ)) ⇒ kx(t)k ≤ ρ for all t ≥ t0 .
Let ρ < min( minkxk=r α1 (r), r). Then, {x ∈ Br (0) | α1 (kxk ≤ ρ)} ⊂ Br (0). Let µ := α2−1 (α1 (ρ)).
Suppose x(t0 ) ∈ Bµ (0). Therefore, α1 (kx(t0 )k) ≤ V (t0 , x(t0 )) ≤ α2 (kx(t0 )k).
Let Ωt,ρ = {x ∈ Br (0) | V (t, x) ≤ ρ}. Therefore, since V̇ is decreasing, if x(t0 ) ∈ Ωt0 ,ρ ,
x(t) ∈ Ωt0 ,ρ . Since V̇ is decreasing,
V (t, x(t)) ≤ V (t0 , x(t0 )).
Moreover, α1 (kx(t)k) ≤ V (t, x(t)) ≤ α2 (kx(t)k). Therefore,
kx(t)k ≤ α1−1 (V (t, x(t))) ≤ α1−1 (V (t0 , x(t0 ))) ≤ α1−1 (α2 (kx(t0 )k))
Therefore, if kx(t0 )k ≤ µ, then kx(t)k ≤ α1−1 (α2 (µ)) = ρ. Thus, x(t) is bounded.
We need to show that W (x(s)) → 0 as s → ∞. Since x(t) is bounded, it is contained in some

compact set D. Since W is continuous, it is uniformly continuous over D (continuous functions
over compact sets are uniformly continuous) i.e., for every > 0, there exists δx > 0 such that
kx − yk < δx ⇒ |W (x) − W (y)| < .
Since f is Lipshitz, kẋk = kf (t, x) − f (t, y)k ≤ L(kx − yk) (y in some neighborhood of x), for
all x ∈ D and for all t. Thus, ẋ is bounded and now by Barbalat’s lemma, theorem follows.
3.2 Instability theorems

Theorem 3.6 (non autonomous). Consider a system (1) such that there exists a continuously
differentiable function V : R+ × Rn → R and a time t0 ≥ 0 such that
• V̇ is lpdf,
• V (0, t) = 0, ∀t ≥ t0 ,
• there exists points x0 in a neighborhood around 0 such that V (t0 , x0 ) ≥ 0.
Then 0 is an unstable equilibrium point.
Proof. [2]
Theorem 3.7. Consider a system (1) such that there exists a continuously differentiable function
V : R+ × Rn → R and a constant r > 0 such that
Table 3: Basic Lyapunov stability theorems for non autonomous systems (Theorem 3.9)
Conditions on Φ(t, t0 ) Conclusion

supt≥t0 Φ(t, t0 ) = m(t0 ) < ∞ stable
supt0 ≥0 supt≥t0 Φ(t, t0 ) = m < ∞ uniformly stable
limt→∞ kΦ(t, t0 )k = 0 asymptotically stable
supt0 ≥0 supt≥t0 Φ(t, t0 ) = m < ∞, limt→∞ kΦ(t, t0 )k = 0 uniformly asymptotically stable
kΦ(t, t0 )k ≤ me−λ(t−t0 ) exponentially stable
• V is decrescent,
• V (0, 0) = 0 and V (., 0) is positive in a neighborhood around 0,
• there exists λ > 0 and a function W : R+ × Rn → R such that V̇ (x, t) = λV (x, t) + W (x, t)
and W (x, t) ≥ 0, ∀t ≥ 0, ∀x ∈ Br (0).
Then 0 is an unstable equilibrium point.
Proof. [2]
Theorem 3.8 (Chetaev’s instability theorem). Consider a system (1) such that there exists a
continuously differentiable function V : R+ × Rn → R an open ball B (0), an open set O ⊂ B (0)
and a function γ ∈ K such that
V (x, t) > 0, ∀t ≥ 0, ∀x ∈ O, supt≥0 supx∈O V (t, x) < ∞, 0 ∈ ∂O (46)

V (t, x) = 0, ∀t ≥ 0, ∀x ∈ ∂O ∩ B (0), (47)
0
V (x).f (x) > γ(kxk), ∀t ≥ 0, ∀x ∈ O. (48)
Then, the zero solution is unstable.
Proof. [2]
3.3 Linear autonomous/LTV systems, linearization

Consider an LTV system
ẋ(t) = A(t)x(t). (49)
Let x(t0 ) = x0 and let Φ(t, t0 ) be the state transition matrix (refer Hespanha or short notes on
linear systems). Then x(t) = Φ(t, t0 )x(t0 ). Note that for linear systems, local and global stability
is the same.
Theorem 3.9. Consider an LTV system (49). Then for the equilibrium point 0, conditions in
Table 3 hold. Moreover, the last two cases are equivalent.

Proof. Suppose supt≥t0 Φ(t, t0 ) = m(t0 ) < ∞. For an > 0 and t0 ≥ 0, define δ(, t0 ) := m(t0 ) .
Then,
kx(t0 )k < δ ⇒ kx(t)k = kΦ(t, t0 )x(t0 )k ≤ kΦ(t, t0 )kkx(t0 )k < m(t0 )δ = .
This shows that 0 is stable.

Suppose kΦ(t, t0 )k is unbounded function of t for some t0 ≥ 0. We need to show that one can
choose x(t0 ) ∈ Bδ (0) such that kx(t)k ≥ for some t ≥ t0 . Select δ1 ∈ (0, δ). Since kΦ(t, t0 )k is
unbounded function of t, there exists t ≥ t0 such that kΦ(t, t0 )k > δ1 . Choose v such that kvk = 1
and kΦ(t, t0 )vk = kΦ(t, t0 )k and let x(t0 ) = δ1 v. Thus, x(t0 ) ∈ Bδ (0). Furthermore,
kx(t)k = kΦ(t, t0 )x(t0 )k = kδ1 Φ(t, t0 )vk = δ1 kΦ(t, t0 )k > .
Hence, 0 is unstable.
We refer the reader to Sastry and [2] for proofs of other statements.
Quadratic Lyapunov functions for LTV systems:
Lemma 3.10. Suppose Q : R+ → Rn×n is continuous and bounded, and that the equilibrium 0 of
(49) is uniformly asymptotically stable. Then,for each t ≥ 0, the matrix
ˆ ∞
P (t) = ΦT (τ, t)Q(t)Φ(τ, t)dτ (50)
t
is well-defined; moreover, P (t) is bounded as a function of t.
Proof. Uniform asymptotic stability is equivalent to exponential stability for LTV systems (Theo-
rem 3.9). Therefore, kΦ(τ, t)k ≤ me−λ(τ −t) , ∀τ ≥ t ≥ 0. The boundedness of Q and exponential
boundedness of Φ implies that P is bounded.
Lemma 3.11. Suppose that, in addition to the hypotheses of the previous lemma, the following
conditions also hold:
1. Q(t) is symmetric and positive definite for each t ≥ 0; moreover, there exists a constant α > 0
such that αkxk2 ≤ xT Q(t)x, ∀t ≥ 0, ∀x ∈ Rn .
2. The matrix A(.) is bounded i.e., m0 = supt≥0 kA(t)k < ∞.
Under these conditions, the matrix P (t) defined in (50) is positive definite for each t ≥ 0; moreover,
there exists a constant β > 0 such that βkxk2 ≤ xT P (t)x, ∀t ≥ 0, ∀x ∈ Rn .
Proof. [2], Lemma 56, p.203.
Theorem 3.12. Suppose Q(.) and A(.) satisfy the hypotheses of Lemmas 3.10 and 3.11. Then,
for each function Q(.) satisfying the hypotheses, the function V (x, t) = xT P (t)x is a Lyapunov
function establishing the exponential stability of the equilibrium point 0.
Proof. [2] Theorem 64, p.216, Sastry, Theorem 5.40, p.213.
periodic systems: [2].
3.4 Indirect method of Lyapunov and local linearization

For a non autonomous system (3), define
∂f (x, t)
A(t) = |x=0 and f1 (x, t) = f (x, t) − A(t)x. (51)
∂x
Then it follows that limx→0 kf1kxk
(x,t)k
= 0. However, it may not be true that
kf1 (x, t)k
limkxk→0 supt≥0 = 0. (52)
kxk
Theorem 3.13. Consider (3) where f is continuously differentiable. Define A(t), f1 (x, t) as in
(51). Suppose (52) holds and A(.) is bounded. If 0 is an exponentially stable equilibrium point of
the linearized system ẋ(t) = A(t)x(t), then it is also an exponentially stable equilibrium point of
(3).
Proof. [2], Theorem 15, Section 5.5, p.211 or Sastry, Theorem 5.41, p.215.
Example 3.14. For LTV systems, exponential stability/uniform asymptotic stability is not char-
acterized by the location of eigenvalues of A(t) for all t, Example 4.22, Khalil.
3.5 Converse theorems
Converse theorems for uniform asymptotic stability, exponential stability and and global exponen-
tial stability are stated in [2] Section 5.7. The essence of these theorems is that they state that the
sufficient conditions obtained before for stability are necessary as well.
3.5.1 Applications of converse theorems

Theorem 3.15. local exponential stability of autonomous systems (Section 5.8.1, [2]). This shows
that the converse of the first statement of Theorem 2.32.
slowly varying systems (Section 5.8.2, [2]), Observer-Controller stabilization for time varying
systems (separation principle): local stabilizing controllers and local separation principle for non
linear systems via linearization (Section 5.8.3, [2]), stabilizing triangular/hierarchical systems (iso-
lated subsystems) (Section 5.8.4, [2]).
4 Stability of periodic solutions and tracking trajectories

Let γ(t) be a periodic solution of ẋ = f (t, x). We want to investigate stability of this periodic
solution for trajectories starting from nearby points. Let y(t) = x(t) − γ(t). Thus, y = 0 is
an equilibrium point of the time varying system ẏ = f (t, y + γ(t)) − f (t, γ(t)). Thus, one can
characterize stability properties of γ(t) (e.g., Lyapunov stability, asymptotic stability etc.) based
on stability properties of y = 0.
Example 4.1.
Suppose we are interested in tracking a trajectory r(t). Again, let y(t) = x(t) − r(t). Now,
one needs to choose an input such that y = 0 is an equilibrium point of the time varying system
ẏ = f (t, y + r(t)) − ṙ(t). To achieve asymptotic tracking, one needs an input which makes y = 0
Example 4.2. Recall the example of stabilization of a rigid robot.
Suppose
one
wants
to track a
z1 x r
trajectory r(t). Thus, one wants (x, y) → (r, ṙ). Let z = = − . One needs to
z2 y ṙ
find u such that 0 is an asymptotically stable equilibrium point of

ż1 y − ṙ
ż = .
ż2 M (x)−1 [u − C(x, y)y] − r̈
Let u = −Kp (x − r) − Kd (y − ṙ) + C(x, y)ṙ + M (x)r̈ where Kp , Kd > 0. Therefore,

y − ṙ z2
ż = =
M (x)−1 [−Kp (x − r) − Kd (y − ṙ)] M (x)−1 [−Kp z1 − Kd z2 − C(x, y)z2 ]
and ż has the required equilibrium point. Consider a Lyapunov function
1 T
V (z1 , z2 ) = [z M (x)z2 + zT
1 Kp z1 ]
2 2
1 T
⇒ V̇ = zT T
2 M (x)z˙2 + z2 Ṁ (x)z2 + z1 Kp z˙1
2
1 T
= zT T
2 (−Kp z1 − Kd z2 − C(x, y)z2 ) + z2 Ṁ (x)z2 + z1 Kp z2
2
= −zT 2 K d z2
since M − 2C is skew symmetric. Now we find the largest invariant set. Observe that V̇ vanishes
on Rn × 0. Note that ż1 = 0 ⇒ z2 = 0 and ż1 = 0 ⇒ M (x)−1 [−Kp z1 − Kd z2 − C(x, y)z2 ] = 0.
Since z2 = 0 and M is invertible, Kp z1 = 0 ⇒ z1 = 0. Thus, (0, 0) forms the largest invariant set
and global asymptotic stability follows from LaSalle.
5 Input to state stability
(Haddad Khalil) Boundedness of solutions
Consider an LTI system ẋ = Ax + Bu where A is Hurwitz. Recall that
ˆ t
x(t) = eA(t−t0 ) x(t0 ) + eA(t−τ ) Bu(τ )dτ.
t0
Using keA(t−t0 ) k ≤ ke−λ(t−t0 ) ,

ˆ t
kx(t)k ≤ ke−λ(t−t0 ) kx(t0 )k + ke−λ(t−τ ) kBkku(τ )kdτ
t0
ˆ t
−λ(t−t0 ) −λt
≤ ke kx(t0 )k + kkBksupt0 ≤τ ≤t ku(τ )ke eλτ dτ
t0
kkBk
= ke−λ(t−t0 ) kx(t0 )k + supt0 ≤τ ≤t ku(τ )k(1 − e−λ(t−t0 ) )
λ
kkBk
≤ ke−λ(t−t0 ) kx(t0 )k + supt0 ≤τ ≤t ku(τ )k.
λ
This shows that the zero input response decays to zero exponentially whereas; the zero state
response is bounded for bounded inputs. Thus, A Hurwitz implies that state remains bounded for
an LTI system if the input is bounded.
We want to understand conditions under which x(t) remains bounded starting from an initial
condition.
Definition 5.1 (Haddad/Khalil). The solutions of (3) are
1. uniformly bounded if there exists γ > 0, independent of t0 ≥ 0 such that for every δ ∈ (0, γ),
there exists = (δ) > 0, independent of t0 , such that
kx(t0 )k < δ ⇒ kx(t)k < , ∀t ≥ t0 . (53)
2. globally uniformly bounded if (53) holds for arbitrarily large δ.
3. uniformly ultimately bounded with an ultimate bound if there exists γ > 0, independent of
t0 ≥ 0, such that for every δ ∈ (0, γ), there exists T = T (δ, ), independent of t0 such that
kx(t0 )k < δ ⇒ kx(t)k < , ∀t ≥ t0 + T. (54)
4. globally uniformly ultimately bounded if (54) holds for arbitrarily large δ.
For solutions of (7), we drop the word uniformly since the solution depends only on t − t0 .
We refer the reader to generalized LaSalle theorem (Theorem 3.5).
Theorem 5.2 (boundedness). Let S ⊂ Rn be an open set containing 0 and V : [0, ∞) × S → R be

a continuously differentiable function such that
α1 (kxk) ≤ V (t, x) ≤ α2 (kxk) (55)

∂V ∂V
+ f (t, x) ≤ −W3 (x), ∀kxk ≥ µ > 0 (56)
∂t ∂x
∀t ≥ 0 and ∀x ∈ S where α1 , α2 ∈ K and W3 is continuous positive definite function. Take r > 0
such that Br ⊂ S and suppose that µ < α2−1 (α1 (r)). Then, there exists a class KL function β and
for every initial state x(t0 ), satisfying kx(t0 )k ≤ α2−1 (α1 (r)), there exists T = T (x(t0 ), µ) such that
the solution to (3) satisfies
kx(t)k ≤ β(kx(t0 )k, t − t0 ), ∀t0 ≤ t ≤ t0 + T (57)

kx(t)k ≤ α1−1 (α2 (µ)), ∀t ≥ t0 + T. (58)
Moreover, if S = Rn , and α1 ∈ KR, then (57) and (58) hold for any initial state x(t0 ) with no
restriction on how large µ is.
Proof. (Sketch): Observe that if µ = 0, then recalling conditions for uniform asymptotic stability of
the origin (Theorem 1.14, Theorem 3.1) and comparing with the conditions of the present theorem,
it is clear that one gets uniform asymptotic stability of the origin. The proof is based on extending
similar ideas to the case where we can have uniform asymptotic stability w.r.t. a set rather than
just a point. We refer the reader to Khalil for the complete proof.
Let ρ = α1 (r), hence, α2 (µ) < α1 (r) = ρ. Since kx(t0 )k ≤ α2−1 (α1 (r)), α2 (kx(t0 )k) ≤ α1 (r) = ρ.
Let η = α2 (µ) and define Ωt,η := {x ∈ Br (0) | V (t, x) ≤ η} and Ωt,ρ := {x ∈ Br (0) | V (t, x) ≤ ρ}.
Note that if x(t0 ) lies outside Bµ (0) and inside Br (0), then for t ≥ t0 ,
α1 (kx(t)k) ≤ V (t, x(t)) ≤ V (t, x(t0 ))α2 (kx(t0 )k) ⇒ kx(t)k ≤ α1−1 α2 (kx(t0 )k).
Note that Bµ ⊂ Ωt,η ⊂ Ωt,ρ ⊂ Br . The fact that these sets are contained in Br follows from
definition. To show that Bµ ⊂ Ωt,η , note that kxk < µ ⇒ α2 (kxk) < α2 (µ). Moreover, V (t, x) ≤
α2 (kxk) < α2 (µ) = η. Thus, Bµ ⊂ Ωt,η . Since η < ρ, Ωt,η ⊂ Ωt,ρ . Sets Ωt,η , Ωt,ρ have a property
that a solution starting inside it can not leave it since V̇ is negative on the boundary. A solution
starting in Ωt,ρ must enter Ωt,η in finite time because in the set Ωt,ρ \ Ωt,η , V̇ satisfies V̇ ≤ −k < 0
where k = minW (x) over the set Bµ ≤ kxk ≤ Br . Therefore,
V (t, x) ≤ V (t0 , x(t0 )) − k(t − t0 ) ≤ ρ − k(t − t0 ) (59)
which shows that V (t, x) reduces to η within the time interval [t0 , t0 + (ρ − η)/k]. For a solution
starting in Ωt,η , (58) is satisfied for all t ≥ t0 . For a solution starting in Ωt,ρ , it enters Ωt,η after
time T . Now (57) can be obtained by applying arguments used for checking uniform asymptotic
stability (Theorem 1.14).
Definition 5.3 (Input to state stability). The system (3) is said to be input-to-state stable if there
exists a class KL function β and a class K function γ such that for any initial state x(t0 ) and
any bounded input u(t), the solution x(t) exists for all t ≥ t0 and satisfies
kx(t)k ≤ β(kx(t0 )k, t − t0 ) + γ(supt0 ≤τ ≤t ku(τ )k). (60)
Theorem 5.4 (ISS for time independent systems). A system ẋ = F(x, u) is input-to-state stable
⇔ there exists a continuously differentiable radially unbounded pdf function V : Rn → R and
continuous functions γ1 , γ2 ∈ K such that for every u ∈ Rm ,
V 0 (x)F(x, u) ≤ −γ1 (kxk), kxk ≥ γ2 (kuk) (61)
Proof. Sketch: (Haddad) (⇐) Let f (t, x) = F(x, u), V (t, x) = V (x) and W (x) = γ1 (kxk) in the
previous theorem. One obtains that there exists t1 such that kxk ≤ γ(kuk), t ≥ t1 (this follows from
the proof of (58)). And kxk ≤ η(kx0 k, t), t ≤ t1 (this follows from the proof of (57)). Therefore,
kxk ≤ η(kx0 k, t) + γ(kuk), t ≥ 0.
(⇒) involves technicalities.

Theorem 5.5 (ISS for time varying systems). Let V : [0, ∞) × Rn → R be a continuously differ-
entiable function such that
α1 (kxk) ≤ V (t, x) ≤ α2 (kxk) (62)

∂V ∂V
+ .f (t, x, u) ≤ −W (x), ∀ kxk ≥ ρ(kuk) > 0 (63)
∂t ∂x
∀(t, x, u) ∈ [0, ∞) × Rn × Rm , where α1 , α2 ∈ KR, ρ ∈ K and W > 0 on Rn . Then the system
ẋ = F(t, x, u) is input-to-state stable with γ = α1−1 ◦ α2 ◦ ρ.
Proof. Sketch: Again using f (t, x) = F(x, u), and using the same method used in the proof of
Theorem 5.2, the proof can be constructed.
6 Center manifold theory

A manifold M is said to be invariant w.r.t. a dynamical system if x(0) ∈ M ⇒ x(t) ∈ M
for all t. As seen in the Introduction of nonlinear control, at an equilibrium x0 of a nonlinear
dynamical system, if one computes the spectrum of the linearization A and finds a number of
eigenvalues in the left half-plane, then there is an invariant manifold (i.e., a manifold that is
invariant under the flow and that is simply the graph of a mapping in this case) that is tangent to
the corresponding (generalized) eigenspace; it is called the local stable manifold. All trajectories
on this stable manifold are asymptotic to the point x0 as t → ∞. Similarly, associated with the
eigenvalues in the right half-plane is an unstable manifold.
If none of the eigenvalues associated with an equilibrium are on the imaginary axis, then the
equilibrium is called hyperbolic. In this case, the tangent spaces to the stable and unstable manifolds
span the whole of Rn . When there are zero eigenvalues or eigenvalues on the imaginary axis, one
needs the notion of the center manifold as one cannot conclude about the local stability from
linearization.
∂f
Consider a time invariant system (7). Let A = ∂x |x=0 and suppose f is twice continuously
differentiable. Then, (7) can be rewritten as
ẋ = Ax + [f (x) − Ax] = Ax + f̂ (x)

∂ f̂
where f̂ (x) = f (x) − Ax such that f̂ (0) = 0 and ∂x (0) = 0. Since we are interested in the case
when the linearization fails, suppose A has k eigenvalues with zero real parts and n − k eigenvalues
with negative real parts. We can assume
without
loss of generality using a similarity transform
A1 0
T that A is in block diagonal form where eigenvalues of A1 has zero real parts and
0 A2

y
eigenvalues of A2 have negative real parts. Let T x = which transforms (7) into the form
z
ẏ = A1 y + g1 (y, z) (64)
ż = A2 z + g2 (y, z) (65)
where g1 and g2 inherit properties of f̂ . They are twice continuously differentiable and
∂gi
gi (0, 0) = 0, (0, 0) = 0 (66)
∂z
for i = 1, 2.
Definition 6.1. If z = h(y) is an invariant manifold for (64) − (65) and h is smooth, then it is
called a center manifold if
∂h
h(0) = 0, (0) = 0.
∂y
Theorem 6.2 (Existence (Khalil, Theorem 8.1)). If gi , i = 1, 2 are twice continuously differentiable
and satisfy (66), all eigenvalues of A1 have zero real parts, and all eigenvalues of A2 have negative
real parts, then there exists a constant δ > 0 and a continuously differentiable function h(y) defined
for all kyk < δ, such that z = h(y) is a center manifold for (64) − (65).
Due to its invariance property, if the initial condition lies on the center manifold, then the
solution remains on the center manifold. Since z(t) = h(y(t)), the evolution of the system in the
center manifold is described by the k−th order differential equation
ẏ = A1 y + g1 (y, h(y)) (67)
which is refered as reduced system.

Suppose z(0) 6= h(y(0)). Then z(t) − h(y(t)) represents the deviation of the trajectory from
the center manifold at any time t. Let w = z − h(y). This transforms (64) − (65) into
ẏ = A1 y + g1 (y, w + h(y)) (68)

∂h
ẇ = A2 (w + h(y)) + g2 (y, w + h(y)) − [A1 y + g1 (y, w + h(y))]. (69)
∂y
In the new coordinates, w = 0 is the center manifold. The dynamics on the center manifold are
characterized by w(t) = 0 ⇒ ẇ = 0. Substituting this in (69),
∂h
0 = A2 h(y) + g2 (y, h(y)) − [A1 y + g1 (y, h(y))]. (70)
∂y
Since the above equation must be satisfied for any solution lying on the center manifold, we conclude
that h(y) must satisfy the pde (70). This provides a condition that the center manifold z = h(y)
must satisfy.
Adding and subtracting g1 (y, h(y)) to rhs of (68) and subtracting (70) from (69),
ẏ = A1 y + g1 (y, h(y)) + N1 (y, w) (71)

ẇ = A2 (w + h(y)) + N2 (y, w) (72)
where N1 (y, w) = g1 (y, w+h(y))−g1 (y, h(y)), N2 (y, w) = g2 (y, w+h(y))−g2 (y, h(y)) ∂h
∂y N1 (y, w).
∂Ni
One can check that N1 , N2 are twice differentiable and Ni (y, 0) = 0, ∂w (0, 0) = 0 for i = 1, 2.
Theorem 6.3 (Stability condition). Under assumptions of Theorem 6.2, if the origin y = 0 of
the reduced order system (67) is asymptotically stable (unstable), then the origin of the full system
(64) − (65) is also asymptotically stable (unstable).
Proof. Theorem 8.2 of Khalil, the proof makes use of the converse Lyapunov theorem for asymptotic
stability to construct a Lyapunov candidate for the reduced order system.
Corollary 6.4. Under assumptions of Theorem 6.2, if the origin y = 0 of the reduced order system
(67) is stable and there is a continuously differentiable Lyapunov function V (y) such that
∂V
[A1 y + g1 (y, h(y))] ≤ 0
∂y
in some neighborhood of y = 0, then the origin of the full system (64) − (65) is stable.
Proof. Note that unlike asymptotic stability, converse theorem does not hold for Lyapunov stability.
Therefore, if the origin of the reduced order system is known to be Lyapunov stable, we need an
existence of Lyapunov candidate as well in our hypothesis to make the previous proof work.
Corollary 6.5. Under assumptions of Theorem 6.2, the origin y = 0 of the reduced order system
(67) is asymptotically stable ⇔ the origin of the full system (64) − (65) is also asymptotically stable.
Proof. (⇒) Follows from the theorem above. (⇐) Stability of the full system implies the stability
of the reduced order system as well, so this implication is trivial.
To use Theorem 6.3, one needs to find the center manifold z = h(y) for which one needs to
solve the pde (70) with boundary conditions h(0) = 0, ∂h
∂y (0) = 0. This is a difficult pde in general
and one uses Taylor series approximation of the solution. We refer the reader to Khalil, Section
8.1 for more details.
7 Control Lyapunov functions and feedback stabilization

Consider non linear controlled dynamical system
ẋ(t) = F(x(t), u(t)), x(t0 ) = x0 , t ≥ 0 (73)
where x(t) ∈ S ⊂ Rn , u(t) ∈ U ⊂ Rm and F : S × U → Rn satisfies F(0, 0) = 0. Suppose F is

Lipshitz continuous in the neighborhood of the origin in S × U . We need to find a feedback control
law u(t) = φ(x(t)) such that the closed loop system is stable.
Definition 7.1 (Control Lyapunov function (Haddad)). Consider the controlled nonlinear dy-
namical system given by (73). A continuously differentiable positive-definite function V : S → R
satisfying
infu∈U V 0 (x).F(x, u) < 0, x ∈ S, x 6= 0 (74)
is called a control Lyapunov function.
If (74) holds, then there exists a feedback control law φ : S → U such that V 0 (x).F(x, φ(x)) < 0,
then by Theorem 2.1 case 2, the origin is asymptotically stable. Conversely, if there exists a feedback
law such that the origin is asymptotically stable, then by Theorem 2.37, there exists a Lyapunov
function V such that V 0 (x).F(x, φ(x)) < 0 which implies the existence of a control Lyapunov
function. Thus, (73) is feedback stabilizable iff there exists a control Lyapunov function satisfying
(74). The analogues of other cases of Theorem 2.1 also hold for control Lyapunov functions.
Consider an affine non linear control system
ẋ(t) = f (x(t)) + G(x(t))u(t), x(t0 ) = x0 , t ≥ 0 (75)
where f : Rn → Rn , G : Rn → Rn×m .
Theorem 7.2 (Haddad). Consider the controlled nonlinear system given by (75). Then a continu-
ously differentiable positive-definite, radially unbounded function V : Rn → R is a control Lyapunov
function of (75) if and only if
V 0 (x)f (x) < 0, x ∈ R (76)
where R = {x ∈ Rn , x 6= 0 : V 0 (x)G(x) = 0}.
Proof. (⇐) Obvious. (⇒) If x ∈ / R, then infu∈U V 0 (x).[f (x(t)) + G(x(t))u(t)] = −∞. When x ∈ R,
(76) must hold if V is a control Lyapunov function.
We now construct an explicit feedback control law which is a function of the control Lyapunov
function V (Haddad). Let α(x) := V 0 (x).f (x) and β(x) := GT (x)(V 0 (x))T and c0 ≥ 0. Let
 √
α(x)+ α2 (x)+(β T (x)β(x))2
−(c0 + )β(x) if β(x) 6= 0,

φ(x) = β T (x)β(x) (77)
0 if β(x) = 0.
In this case, the control Lyapunov function turns out to be the Lyapunov function for the closed
loop system with u = φ(x). Observe that
V̇ (x) = V 0 (x)[f (x) + G(x)φ(x)]

= α(x) + β T (x)φ(x)
( p
−c0 β T (x)β(x) − α2 (x) + (β T (x)β(x))2 if β(x) 6= 0,
=
α(x) if β(x) = 0
< 0, ∀x ∈ Rn \ 0 (78)
provided that Theorem 7.2 is satisfied. This implies that V is indeed a Lyapunov function for the
closed loop system. Since f and G are smooth, the feedback law (77) is smooth everywhere except
the origin.
Theorem 7.3. Consider the nonlinear dynamical system G given by (75) with a radially unbounded
control Lyapunov function V : Rn → R. Then the following statements hold:
1. The control law φ(x) given by (77) is continuous at x = 0 if and only if for every > 0,
there exists δ > 0 such that for all 0 < kxk < δ, there exists u ∈ Rm such that kuk < and
α(x) + β T (x)u < 0.
2. There exists a stabilizing control law φ̂(x) such that α(x) + β T (x)φ̂(x) < 0, x ∈ Rn , x 6= 0
and φ̂(x) is Lipschitz continuous at x = 0 if and only if the control law φ̂(x) given by (77) is
Lipschitz continuous at x = 0.
Proof. We refer the reader to Theorem 6.8 of Haddad.
8 Periodic systems
8.1 Linear periodic systems
The following discussion is from [7]. Consider a linear periodic system ẋ(t) = A(t)x(t) such that
A(t + T ) = A(t), T > 0. Let Φ(t, τ ) be the state transition matrix. Recall that for general linear
systems,
ˆ t ˆ t ˆ s1
Φ(t, t0 ) := I + A(s1 )ds1 + A(s1 ) A(s2 )ds2 ds1 + . . . .
t0 t0 t0
d d
and dt Φ(t, t0 ) = A(t)Φ(t, t0 ). Hence, dt Φ(t + T, t0 ) = A(t + T )Φ(t + T, t0 ). Therefore, for periodic
d
systems, dt Φ(t + T, t0 ) = A(t)Φ(t + T, t0 ) since A(t + T ) = A(t). Furthermore, by the concatenation
property of the state transition matrix, Φ(t + T, t0 ) = Φ(t + T, t)Φ(t, t0 ). Let C(t) = Φ−1 (t, t0 )Φ(t +
T, t0 ). Therefore
Ċ = −(Φ−1 (t, t0 )A(t)Φ(t, t0 )Φ−1 (t, t0 ))Φ(t + T, t0 ) + Φ−1 (t, t0 )A(t)Φ(t + T, t0 ) = 0.
Hence, C is a constant matrix and C = C(t0 ) = Φ(t0 + T, t0 ). Therefore, Φ(t + T, t0 ) = Φ(t, t0 )C.
Define R := T1 ln(C), hence, eRT = C . Let P (t) = Φ(t, 0)e−Rt . Clearly P (t) is non-singular.
P (t + T ) = Φ(t + T, 0)e−R(t+T ) = Φ(t, 0)Ce−R(t+T ) = Φ(t, 0)Ce−RT e−Rt = Φ(t, 0)e−Rt = P (t).
Therefore, Φ(t, 0) = P (t)e−Rt where P (t) is non singular and periodic. This is called Floquet
decomposition.
The state evolution is given by x(t) = Φ(t, 0)x0 = P (t)e−Rt x0 . Since P (t) is periodic and
continuous, it is bounded. Therefore the periodic linear system is globally exponentially stable ⇔
R has eigenvalues in the strict LHP ⇔ C has eigenvalues strictly inside the unit circle. The matrix
C = Φ(t + T, t) is called monodromy matrix and eigenvalues of C are called Floquet multipliers.
Let y(t) := P −1 (t)x(t). Therefore,
ẋ = Ṗ y + P ẏ ⇒ Ax = Ṗ y + P ẏ ⇒ AP y = Ṗ y + P ẏ ⇒ ẏ = P −1 (AP y − Ṗ y).
Using P (t) = Φ(t, 0)e−Rt in the above equation,
ẏ = P −1 (AΦ(t, 0)e−Rt − AΦ(t, 0)e−Rt + Φ(t, 0)Re−Rt )y = eRt Φ(t, 0)−1 (Φ(t, 0)Re−Rt )y = eRt Re−Rt y = Ry.(79)
Therefore the linear periodic system is globally exponentially stable iff the LTI system obtained
above is exponentially stable.
8.2 Nonlinear periodic systems

Consider a nonlinear periodic system ẋ = f (t, x) such that f (t+T, x) = f (t, x) for all x ∈ Rn , T > 0.
d
Let s(t, 0, x0 ) be the flow of this ode. Hence, dt s(t, 0, x0 ) = f (t, x(t)). Consider s(t + T, T, x0 ).
Because the time varying vector field f (t, x) is periodic with period T , by existence and uniqueness,
s(t + T, T, x0 ) = s(t, 0, x0 ).
Consider a fixed point or a periodic solution x∗ (t) of this periodic system. Let z = x − x∗ be
the perturbation from this trajectory. Therefore,
ż = f (t, x) − t, x∗ .
Using linearization, one obtains a periodic LTV system. If the linear system is exponentially stable,
then the nonlinear periodic system is locally exponentially stable at the solution x∗ .
It turns out that for periodic systems, 0 is stable/asymptotically stable iff is is stable/asymptotically
stable ([2]). Lyapunov analysis for stability is applicable to periodic systems as well. One looks
for a periodic lpdf Lyapunov function V with the same period as that of the system such that
V̇ ≤ 0 in some open neighborhood N of 0. Then, one can apply LaSalle-Krasovskii theorem to
conclude convergence of trajectories towards the invariant set M of V where M ⊂ S ⊂ N such
that S := {x ∈ N | V̇ (x) = 0} and M is the largest invariant set of S. If S contains no invariant
trajectory other than the zero trajectory, then 0 is uniformly asymptotically stable. If in addition,
V is a pdf and radially unbounded, then 0 is globally uniformly asymptotically stable equilibrium
point of the periodic system ([2]).
9 Discrete non linear systems

Consider an autonomous discrete time non linear system
x(k + 1) = f (x(k)), x(0) = x0 , k ∈ Z+ (80)
where x ∈ S ⊂ Rn (S open), 0 ∈ S and f (0) = 0. Assume that f is continuous on S. The flow map
is given by s : Z+ × S → S where s(0, x0 ) = x0 , s(1, x0 ) = f (s(0, x0 )) = f (x0 ), s(2, x0 ) = f (s(1, x0 ))
and so on. There is an analogue of existence and uniqueness for discrete non linear system ([1],
Theorem 13.1)
There are similar definitions on stability and asymptotic stability as we had in the continuous
time case. Instead of exponential stability for continuous time systems, we define geometric stability
for discrete systems where kx(k)k ≤ αkx(0)kβ −k , α, β > 1. The discrete time varying systems are
of the form
x(k + 1) = f (k, x(k)), x(0) = x0 , k ∈ Z+ . (81)
There are natural discrete time counterparts of all results stated in the continuous time case. We
refer the reader to Chapter 13 of Haddad ([1]) for complete details.
10 Advanced topics
10.1 Partial stability
Partial stability means stability with respect to part of the system’s state. The state vector x is
divided into two components x1 and x2 and one asks for stability w.r.t. one of the two components.
The sufficient stability conditions given in previous sections can be extended to obtain sufficient
conditions for partial stability (Haddad [1], Section 4.2).
10.2 Finite time stability

In some applications, one needs to reach the equilibrium point in finite time rather than reaching
it asymptotically.
10.3 Semistability
For systems with continuum of equilibria, one need an appropriate notion of stability instead of
asymptotic stability.
10.4 Generalized Lyapunov theorems

Here the differentiability criteria is weakened and one only assumes lower semi-continuity of the
Lyapunov function to obtain sufficient conditions for stability, asymptotic stability, invariance of
sets and so on. Lyapunov function V must be positive definite in the neighborhood of an equilibrium
point but it should be decreasing i.e., V (x(t)) decreases as t increases. These results allow us
to patch different Lyapunov functions over different sets together without being worried about
differentiability conditions to obtain sufficient conditions for stability. We refer the reader to
Haddad ([1]), Section 4.8
10.5 Lyapunov and asymptotic stability of sets and periodic orbits

We can define Lyapunov and asymptotic stability of sets by replacing the equilibrium point in the
corresponding definitions with appropriate sets say S0 . One needs a Lyapunov function V which is
zero on the say S0 , positive on points in an open set containing S0 which lie outside S0 such that
V (x(t)) is decreasing as x(t) evolves in time. For extension of the results stated for equilibrium
points, we refer the reader to Haddad ([1], 4.9).
Poincare’s theorem provides necessary and sufficient conditions for stability of periodic orbits
(Section 4.10, [1]).
10.6 Vector Lyapunov functions

Here one uses vector Lyapunov function which requires less rigid requirements compared to a scalar
Lyapunov function. One can obtain sufficient conditions for stability using the vector Lyapunov
functions (Section 4.11, [1]).
10.7 Applications
Stability of switched systems, hybrid systems..
11 Appendix
11.1 lpdf, pdf and decrescent functions
The discussion here is from [2].
Lemma 11.1. Suppose φ : R+ → R+ is continuous, non decreasing and φ(0) = 0, φ(r) > 0 ∀r > 0.
Then, there is a class K function α such that α(r) ≤ φ(r), ∀r. Moreover, if φ(r) → ∞, as r → ∞,
then α can be chosen to be a class KR function.
Proof. [2], Lemma 1, Section 5.2.
Lemma 11.2. A continuous function V : Rn → R is an lpdf ⇔
• V (0) = 0.
• there exists r > 0 such that V (x) > 0, ∀x ∈ Br \ {0}.
V is a pdf ⇔
• V (0) = 0.
• V (x) > 0, ∀x ∈ Rn \ {0}.
• there exists r > 0 such that infkxk≥r V (x) > 0.
V is radially unbounded ⇔ V (x) → ∞ as kxk → ∞ uniformly in x.
Proof. (⇒) Obvious. (⇐) Define φ(p) := infp≤kxk≤r V (x). Then, φ(0) = 0, φ is continuous and
non decreasing because as p increases, the infimum is taken over a smaller region. Moreover,
φ(p) > 0 when p > 0. Now by the previous lemma, one can chosen a class K function α such that
α(kxk) ≤ φ(kxk) ≤ V (x), ∀x ∈ Br .
Necessary conditions for pdf follow from the definition. The sufficient conditions follows from
the above arguments (using the third condition). The last statement also follows.
x2
A function V (x) = 1+x 4 > 0 on R \ {0} but it is not a pdf. Note that it violates the third
condition in the above lemma. If V is pointwise positive everywhere except the origin and radially
unbounded, then V is a pdf.
Lemma 11.3. A continuous function V : Rn × R+ → R is an lpdf ⇔
1. V (t, 0) = 0, ∀t.
2. there exists an lpdf W : Rn → R and a constant r > 0 such that V (t, x) ≥ W (x), ∀t ≥ 0,
∀x ∈ Br (0).
V is a pdf ⇔
1. V (t, 0) = 0, ∀t.
2. there exists a pdf W : Rn → R such that V (t, x) ≥ W (x), ∀t ≥ 0, ∀x ∈ Rn .
V is radially unbounded ⇔ there exists a radially unbounded function W : Rn → R such that

V (t, x) ≥ W (x), ∀t ≥ 0, ∀x ∈ Rn .
Proof. Consider lpdfs. If V is an lpdf, then there exists a class K function α such that V (t, x) ≥
α(kxk). Define W (x) := α(kxk). Conversely, suppose there exists W satisfying given conditions.
Then using the previous lemma, one can show that V is an lpdf.
The necessary conditions for V being pdf follow from the definition. For sufficiency, we can use
the previous lemma. The last statement also follows similarly.
Example 11.4. Consider a quadratic function V (x) = xT P x where P is a square real symmetric
matrix. V is pdf ⇔ P is positive definite. Thus, the term positive definite matrix and positive
definite function are consistent. Note that V is radially unbounded.
Example 11.5. A function V1 (x1 , x2 ) = x21 +x22 is an example of a radially unbounded pdf. Another
function V2 (t, x1 , x2 ) = (t + 1)(x21 + x22 ) is a pdf since it dominates time invariant V1 . A function
V3 (t, x1 , x2 ) = e−t (x21 + x22 ) is not a pdf but it is decrescent.
A function W1 (x1 , x2 ) = x21 + sin2 (x2 ) is an lpdf but not a pdf. A function W2 (x1 , x2 ) =
x21 + tanh2 (x2 ) is a pdf but not radially unbounded since, tanh2 x2 → 1 as |x2 | → ∞.
References
[1] W. Haddad and V. Chellaboina, Non linear dynamical systems and control, Princeton university
press, 2008.
[2] Mathukumali Vidyasagar, Non linear systems analysis, Dover, 2009.
[3] S. Sastry, Non linear systems, Springer, 1999.
[4] H. Khalil, Non linear systems, Pearson, 2009.
[5] D. Liberzon, Non linear systems analysis, Lecture notes, 2012.
[6] J. Hespanha, Linear systems theory, Princeton university press, 2009.
[7] H. Taha, Geometric control theory, Lecture notes, 2019.

Short Notes On Lyapunov Stability: 1.1 Basic Definitions For Stability and Invariance

Uploaded by

Copyright:

Available Formats

Short Notes On Lyapunov Stability: 1.1 Basic Definitions For Stability and Invariance

Uploaded by

Document Information

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

Short Notes On Lyapunov Stability: 1.1 Basic Definitions For Stability and Invariance

Uploaded by

Copyright:

Available Formats

Short notes on Lyapunov stability

ẋ(t) = f (t, x, u), x(t0 ) = x0 , t ∈ [t0 , t1 ]. (1)

where f for continuous time systems may assumed to be continuous in t and x, x ∈ Rn , u ∈ Rm , x

ẋ(t) = f (t, x), x(t0 ) = x0 , t ∈ [t0 , t1 ]. (3)

1.1 Basic definitions for stability and invariance

• locally if they are true for all x ∈ B (0) for some .

• globally if they are true for all x ∈ Rn .

• uniformly if they are true for all t0 ≥ 0.

We may replace s(t, t0 , x0 ) with x(t) in all the following definitions.

Local stability definitions:

Definition 1.4 (asymptotic stability). The equilibrium point 0 is asymptotically stable if it is

V (0, t) = 0 and V (x, t) ≥ α(kxk) ∀x ∈ B , t ≥ 0. (4)

Definition 1.11 (positive definite function (pdf)). A continuous function V : Rn × R+ → R+ is

V (0, t) = 0 and V (x, t) ≥ α(kxk) ∀x ∈ Rn , t ≥ 0. (5)

Definition 1.12 (decrescent functions). A continuous function V : Rn × R+ → R+ is called

V (x, t) ≤ β(kxk) ∀x ∈ B (0), t ≥ 0. (6)

Example 1.13. V (t, x) = (t + 1)kxk2 is a pdf but not decrescent.

ks(t, t0 , x0 )k ≤ φt0 (kx0 k), ∀t ≥ t0 , ∀x0 ∈ Bd(t0 ) (0).

ks(t, t0 , x0 )k ≤ φ(kx0 k), ∀t ≥ t0 , ∀x0 ∈ Bd (0).

ks(t, t0 , x0 )k ≤ σt0 ,x0 (t), ∀t ≥ 0, ∀x0 ∈ Br(t0 ) (0).

ks(t, t0 , x0 )k ≤ σ(t), ∀t, t0 ≥ 0, ∀x0 ∈ Br(t0 ) (0).

6. i. uniformly asymptotically stable ⇔ there exists a number r > 0, a function φ ∈ K and a

ks(t, t0 , x0 )k ≤ φ(kx0 k)σ(t), ∀t, t0 ≥ 0, ∀x0 ∈ Br (0).

kx(t)k ≤ β(kx(t0 )k, t − t0 ), ∀t ≥ t0 ≥ 0, ∀kx(t0 )k ≤ c.

kx(t0 )k < δ ⇒ kx(t)k < , ∀t ≥ t0 .

Definition 1.15 (Domain of attraction). Suppose that 0 is an asymptotically stable equilibrium

D0 := {x0 ∈ S ⊂ Rn | if x(0) = x0 , then limt→∞ x(t) = 0}.

Conditions on V (x) Conditions on −V̇ (x) Conclusion

2 Stability theory for Autonomous/time invariant systems

2.1 Basic stability theorems using Lyapunov’s direct method

ẋ1 = x2 , ẋ2 = −f (x2 ) − g(x1 ) (8)

where f is a continuous function modelling resistor current-voltage characteristics and g models

ẋ = ayz, ẏ = −bxz, ż = cxy.

V̇ = 2(pxẋ + qy ẏ + rz ż) = 2xyz(ap − bq + cr).

Choose p, q, r such that ap − bq + cr = 0. Hence, the origin is stable.

C ẋ = G(x)x ⇒ ẋ = C −1 G(x)x. (12)

Example 2.7 (Phase locked loop). [2] p.181.

2.2 Invariant sets and stability theorems

where Kp , Kd > 0. Now the system equations are

ẋ = y, ẏ = −[M (x)]−1 [Kp (x − qd ) + Kd y + C(x, y)y]

The equilibrium point is (x, y) = (qd , 0). Consider a Lyapunov function

ẋ1 = x2 + x1 (r2 − x21 − x22 )

2.3 Constructing Lyapunov functions

Proposition 2.16. A function g : Rn → Rn is the gradient vector of a scalar valued function

Let y = σx, therefore,

Proposition 2.17. Let f , g : Rn → Rn be continuously differentiable functions such that f (0) = 0.

gT (p)f (x) = gT (p)[f (x) − f (0)]

∂f (x) ∂f (x) T ∂f (x)

∂f (x) 1 ∂f (αx) T ∂f (αx)

Example 2.19. Example 3.9 Haddad.

The following theorem allows us to find the domain of attraction as well.

Theorem 2.20 (Zubov’s theorem). Let 0 be an equilibrium point of ẋ = f (x(t)). Let S ⊂ Rn be

0 < V (x) < 1, x ∈ S, x 6= 0 (21)

Then, 0 is asymptotically stable with domain of attraction S.

Example 2.21. Haddad: Example 3.10

Constants of motion or first integrals or integrals of motion or dynamic invariants or Casimir

Furthermore, the equilibrium point xe is Lyapunov stable with Lyapunov function

V̇ (x) = V 0 (x).f (x)

Recall that by hypothesis, if x ∈ M, then Ci0 (xe ).x = 0, i = 2, . . . , r. Consider a subspace V1

• locally if they are true for all x ∈ B (0) for some .

V (0, t) = 0 and V (x, t) ≥ α(kxk) ∀x ∈ B , t ≥ 0. (4)

V (x, t) ≤ β(kxk) ∀x ∈ B (0), t ≥ 0. (6)

kx(t0 )k < δ ⇒ kx(t)k < , ∀t ≥ t0 .

β(t0 , δ) := supkxk≤δ V (x, t0 ) < α(1 ).

V (x(t1 ), t1 ) ≥ α(1 ) > V (x(t0 ), t0 ).

kx(t0 )k < δ ⇒ kx(t)k = kΦ(t, t0 )x(t0 )k ≤ kΦ(t, t0 )kkx(t0 )k < m(t0 )δ = .

kx(t0 )k < δ ⇒ kx(t)k < , ∀t ≥ t0 . (53)

kx(t0 )k < δ ⇒ kx(t)k < , ∀t ≥ t0 + T. (54)