Proceedings of the International Conference on Industrial Engineering and Operations Management

Rabat, Morocco, April 11-13, 2017

Comparative study of two sliding surfaces to control a

double boost converter
Fadwa El Otmani, Abdelmajid Abouloifa, Meriem Aourir and Abdellatif Hamdoun
LTI Lab. Faculty of Science Ben M’sik, university HASSAN II of Casablanca, Casablanca,
Morocco
[email protected], [email protected], [email protected],
[email protected]

Ibtissam Lachkar
LISER Lab, ENSEM Casablanca, university HASSAN II of Casablanca, Casablanca, Morocco
[email protected]

Abstract
In this paper, a double cascade boost converter is studied and controlled. The system dynamics have been
described by the averaged 4th order nonlinear state space model (1). Based on such a model and taking account the
non-minimum phase nature of the converter, two robust non -linear controllers are synthesized using the sliding mode
approach, where each controller use a different surface and a different structure. The first controller is elaborate with
two nested loops, the inner currant loop is designed with the sliding mode approach and the outer voltage loop is
designed with the famous PI controller, the second controller is also designed with the sliding mode approach, but this
time with different sliding surface, this surface allows to use a single control loop to regulate the currant and the
voltage in the same time. The simulation results have been performed through Matlab/SimPowerSystems environment
and show that the designed controllers meet their objectives.

Keywords
Sliding mode approach, DC-DC converter, double cascade boost, non-minimum phase.

1.Introduction

To keep up with the progress of the technology in invention’s world and electric installation, there is a vital
part which is needed to be developed and improved, this one is the conversion of energy or more precisely the power
electronics which steel pose many problems in term of stability, response time and circuit’s harmonics etc. To
overcome those problems, the vital aim is to focus on the automatic control of the electronic systems with chosen the
most appropriate methods and parameters. About the methods there is plenty, as backstepping, passivity, flatness and
sliding mode approach, whose give good and satisfying results.
This work focus on the DC-DC converters, those converters have many features as boosting or decreasing the
output voltage, when this paper is interested in improving and increasing the voltage gain, to get this purpose several
strategies are proposed, and between the more interesting strategies there is the cascading of two or more elementary
converters. Some previous works have presented a study of cascaded converters, such as (Rensburg et al., 2008) who
presented a new boost topology which ensures a significan t improvement on boost factor, and(Zhao and Lee, 2003)
who presents a family of high-efficiency, high step-up clamp-mode converters without extreme duty ratios . In this
article, a study of two cascaded boost converters will be presented.
In this paper, a double cascade boost converter is studied and controlled, this system is just a cascading of two
identical elementary simple boosts which allows to improve the voltage gain without using systems with high
frequency. The system dynamics have been described by the averaged 4th order nonlinear state s pace model (1). Based
on such a model and taking account the non-minimum phase nature of the converter, two robust non -linear controllers
are synthesized using the sliding mode approach; this one is suitable to our system thanks to its appropriateness for
the nonlinearity aspect of power system converter and also because of its robustness and capability of system order
reduction (Slotine et al., 1991), (FLOQUET, 2000) ; where each controller use a different sliding surface and a
different structure, and as the converter comprise two stages , each stage will be controlled separately. The first

1350
Proceedings of the International Conference on Industrial Engineering and Operations Management
Rabat, Morocco, April 11-13, 2017

controller is elaborate with the traditional surface proposed by (Utkin, 2013), moreover because of the non-minimu m
phase of the double boost converter the regulator must comprise two nested loops, where the inner currant loop is
designed with the sliding mode approach and the famous PI linear controller is used to design the outer voltage loop,
on the other hand a second controller is also designed with the sliding mode approach, but this time with different
sliding surface, this surface which is proposed by(Hijazi et al., 2009) allows to use a single control loop to regulate
the currant and the voltage in the same time. The simulatio n results have been performed through
Matlab/SimPowerSystems environment and show that the designed controllers meet their objectives.

The paper is organized as follows: in Section 2, the Double Cascade Boost converter is described and modeled;
when the section 3 is devoted to designing the Sliding mode controller with two methods, and whose performances
are illustrated in section 4; in the last the section 5 gives a conclusion and a reference list end the paper.

2. Double Boost Converter Presentation and Modeling

A double boost system is a non-isolated DC-DC converter with high voltage ratio, it comprises two identical
elementary boosts placed in cascades. The figure below illustrates the topology of a double boost converter, wherein
each stage is controlled by a binary signal  .

Figure 1. Double Boost converter

Where:
- iL1 and iL 2 are respectively the currents in inductors L1 and L2 .
- vc1 denotes the voltage in capacitor C1 and vc 2 is the output voltage.
- 1 and  2 are denoted the duties ratio functions:
1 if T j is ON and D j is OFF
 j  j  1,2  
0 if T j is OFF and D j is ON
2.1 Switched model

By applying standard Kirchhoff Laws to the circuit, it can be represented by the following differential
equations:
 ve  1  1 vc1
diL1
L1 (1.a)
dt

 1  1 iL1  iL 2
dvc1
C1 (1.b)
dt

 vc1  1   2 vc 2
diL 2
L2 (1.c)
dt
C2 c 2  1   2 iL 2  c 2
dv v
(1.d)
dt R

2.2 Averaged model

The above equations represent the instantaneous model of a double boost converter; although the control
signal u (the switch position) is a binary variable changing at a high frequency and some variables of the boost
converter have discontinuous derivative. Therefore, it is convenient to establish the control objective and some other
relationships in terms of average variables.

1351
Proceedings of the International Conference on Industrial Engineering and Operations Management
Rabat, Morocco, April 11-13, 2017

The averaged model of such a converter is shown to be the following:

ve 1  u1  (2.a)
x1   x2
L1 L1

x2 
1  u1  x 
x3 (2.b)
1
C1 C1

x2 1  u2  (2.c)
x3   x4
L2 L2

x4 
1  u2  x 
x4 (2.d)
3
C2 RC2

where x1 , x 2 , x 3 and x 4 denote the average values over cutting periods, of the signals il1 , vc1 , il 2 and vc 2 . Th e
controls inputs for the above model is the functions u1 , u 2 called duty ratio function, which is the average values
over cutting periods of signals 1 ,  2 .

2.3 Equilibrium points

The equilibrium point can be obtained by forcing the time derivative of the average model (2) to be null x  0 .

ve
x1e  (3.a)
R1  u10 1  u20 
ve
x2e  (3.b)
1  u10 
ve (3.c)
x3e 
R1  u10 1  u20 2
ve (3.d)
x4e 
R1  u10 2 1  u20 2
3. Controller design

Double Boost converter being a variable structure system, where, switching over changes the structure of the
system. Sliding mode control is the most suitable control strategy for such systems, because of its low sensitivity to
the system parameters variations and disturbances which makes the control robust.

3.1 Control objectives

 Controlling the voltage x 2 to make it track a given reference signal x*2 .

 Regulating the output voltage x 4 to pursuit a reference value x*4 .

x*2 and x*4 denote the corresponding voltage’s references signals.

3.2 Controller design

The double cascade boost converter has a non-minimum phase nature (Byrnes and Isidori, 1988), this nature
causes a stability and control problems because when we regulate the output voltage directly the current phase grow
unstably which causes a damage of the converter, and as it shown in (Sira-Ramirez, 1987), the currant must increase
continuously so that the surface v  v* tend to zero, this will include other objectives which are:
 the input current x1 must track, as closely as possible, a given reference signal x1* .

 And the current x3 in inductor L2 also must track, as closely as possible, a given reference signal x*3 .

1352
Proceedings of the International Conference on Industrial Engineering and Operations Management
Rabat, Morocco, April 11-13, 2017

x1* and x3 denote the corresponding currant’s references signals.

*

To solve the problem causes by the non-minimum phase nature, two nested controllers must be used for each
stage of the convertor: an inner current control loop and an outer voltage control loop, but also a single control loop
can be used thanks to the sliding surface proposed in(Hijazi et al., 2009) this surface allow to regulate the current and
the voltage in the same time. In this section, the both solutions will be introduced.

3.2.a Controller design with two loops

as mentioned before; two nested loops will be used, the first loop which is the voltage controller is usually realized
with standard linear control techniques, whereas the current control is implemented using the sliding mode approach
for the control of inductors current. The following figure show the sys tem structure which comprise four blocs:
 the PI controller which is a linear regulator used to generate the current references x1* and x*3 this bloc is
controller’s outer loop.
 the second bloc which is SMC with S1 is the inner controller synthetized with the sliding mode approach
using the first proposed sliding surface (4).
 the third bloc the PWM is the famous pulse-with modulation generator.
 the last bloc is the DBC which is the double boost converter illustrate on figure (1). Where the currant’s error
z  z 
z1 , z1   11  , and voltage’s error z 2 , z 2   21  .
 12 
z  z 22 

_
+
_
+
SMC
PI With PWM DBC
_ _
S1
+ +

Figure 2. structure of the system with a regulator designed with two loops

a.1 The choice of the sliding surface

To start an appropriate sliding function s must be chosen to guarantee the system stability in the sliding
modes s  0 (Slotine et al., 1991):
n 1
 
sx     k x  z x  (4)
 t 
z  x  the deviation of the controlled variable, k x a positive constant and n the system order.

The Double Boost Converter comprise two stages where each stage can be considered as a subsystem, each subsystem
is a second order system n  2 then we get two sliding surfaces:
 s   z  k z 
s   1    11 1 11  (5)
 s2   z12  k2 z12 

1353
Proceedings of the International Conference on Industrial Engineering and Operations Management
Rabat, Morocco, April 11-13, 2017

Where the currant’s error z1 :

 z   L x  x1*
z1   11    1 1
  
 
(6)

 z12   L2 x3  x3
*

a.2 Existence condition

To ensure the existence of the sliding mode; a dynamic of defined Lyapunov function must be negative V  0
. The Lyapunov function is defined as:
1
V  s2 (7)
2
its dynamic is:
V  ss (8)
And the dynamic of the sliding surface s is:
 s   u1 x 2  ( 1  u1 )x 2  L1 x1  k1 ve  k1 ( 1  u1 )x 2  k1 L1 x1 
* *
s   1     (9)
 s 2   x 2  u 2 x 4  ( 1  u 2 )x 4  L2 x3  k 2 x 2  k 2 ( 1  u 2 )x 4  k 2 L2 x 3 
  * *

In sliding mod control, we usually determine u as following (Hu, 2016):

1 if s0
u (10)
0 if s0

Using (10) to ensure the existent condition, s must be defined as flowing:

s  0 if s0
 (11)
s  0 if s0

Thanks to (11) there is no need to calculate the product ss and the sliding region can be easily deduced, this region
is limited by the following inequalities:
  L1x1*  k1 ve  k1 L1 x1*  0
s1  


 x2  L1 x1  k1 ve  k1 x2  k1 L1 x1  0
   * * (12)
s  
 
 x 2  L2 x3  k 2 x2  k 2 L2 x3  0
* *
s2  
 x2  x4  L2 x3  k 2 x2  k 2 x4  k 2 L2 x3  0
*
     *


At last the condition will be satisfied if:

x2  L1x1*
k1  (13.a)
ve  L1x1*  x2

x4  x2  L2 x*3

k2  (13.b)
x2  x*3  x4
a.3 Current loop design

where ueq t  and u n t  are the equivalent control and the switching control, respectively. The equivalent

control, ueq t  , proposed by (Utkin, 2013) is based on the nominal (estimated) plant parameters and provides the main
control action, while the switching control, u n t  , ensures the discontinuity of the control law across sliding surface.
If the initial error is not on the sliding surface s t  or there is a deviation of the representative point from s t  due to
parameter variations and disturbances, the controller must be designed such that it can drive the error to the sliding

1354
Proceedings of the International Conference on Industrial Engineering and Operations Management
Rabat, Morocco, April 11-13, 2017

surface. The error under the condition that will move toward and reach the sliding surface is said to be on the reaching
phase. The control’s law defined by:

ut   ueq t   un t  (14)

 the equivalent control

the equivalent control can be calculated from s  0 , and the obtained control’s low is:

ueq1 
1  ueq1 x2  k1 x2   L1x1*  k1 x1*  k1 ve (15.a)
x2

ueq 2 
1  ueq2 x4  k2 x4   L2 x*3  k2 x*3  k2 x2  x2 (15.b)
x4
 the switching control

The switching control law, is introduced as (Slotine et al., 1991):

un1  1signs1 
un ( t )   (16)
un 2   2 signs2 
Where  1 and  2 are positive constants and sign function is defined as:
1  s  0

signs   0  s  0
 1  s  0

Then the find controller law is:
 ueq1  un1   ueq1  1signs1  
u    (17)
 ueq 2  un 2   ueq 2   2 signs2 
   
a.4 Voltage loop design

The aim now is to design a tuning law for x1* , x*3 . And usually, the voltage control is achieved with standard
linear control techniques, that’s why the PI controller is used to generate the currant references. A low pass filter is
adopted, then the PI controller structure is de fined by:
 t  a
x1*   k p1 z 21  ki1 z21 
 0 sa  (18.a)

 t  b
x*2   k p 2 z 22  ki 2 z 22 
 0  sb  (18.b)

Where ( k p1 , k p 2 ) the proportional gains, ( k i1 , ki 2 ) integral gains, and (a, b) positive constants.
 z   y*  y1  , y1  x2  , y2  x4  , y1*  x*2  and y*2  x*4  .
2 2 2 2
With z 2   21    *1 
 
 z 22   y2  y2 

3.2.b Controller design with single loop

In this part, another sliding surface will be used for the controller design, this surface proposed by (Hijazi et
al., 2009) allows to ensure the regulation of the currant and the voltage in the same time, which make the controller
design less complicated and more performer.
The figure bellow illustrates the new general system structure using the second surface, which comprise this time three
blocs:
 the first bloc is SMC with S2 which is the inner controller synthetized with the sliding mode approach using
the sliding surface (4) proposed by (Hijazi et al., 2009).
 PWM: pulse-with modulation generator.

1355
Proceedings of the International Conference on Industrial Engineering and Operations Management
Rabat, Morocco, April 11-13, 2017

 the last bloc is the DBC which is the double boost converter illustrate on figure (1). Where the voltage’s error
z11 and z12 .

_
+
SMC
With DBC
PWM
S2
_
+

Figure 3. structure of the system with a regulator designed with single loop

b.1 The choice of the sliding surface

The surface used in this controller is proposed by (Hijazi et al.+, 2009), and can be introduced like:

s1  k3 z21  k4 z11

s (19)
  k6 z12
s2  k5 z22

Where k3 , k 4 , k5 and k6  R  , z 2  
 z 21
  

   C1 x2  x*2  and
 
 z 22   C2 x4  x4
*
 z1 correspond to (6)

As reference current depends on the operating point, will be extracted from (3):
x1* 
x 
* 2
2 (20.a)
Rve

x*3 
x 
* 2
4 (20.b)
Rx*2
Sliding surface coefficients k3 , k 4 , k5 and k 6 must be chosen to ensure that the sliding mode exists at least around the
desired equilibrium point, and the dynamics of the system will reach the surface and lead toward the equilibrium point.

b.2 Existence condition

As mentioned before, to ensure the existence of the sliding mode; a dynamic of Lyapunov function must be negative
1
V  0 . the same Lyapunov function defined before V  s 2 and its dynamic V  ss will be used.
2
the dynamic of the sliding surface s (20) is:
s1  k3ve  k3 1  u1 x2  k3 L1 x1*  k 4 1  u1 x1  k 4 x3
 (21)
s  
s2  k5 x2  k5 1  u 2 x4  k5 L2 x3  k6 1  u 2 x3  k6
* x4
 R
Using (11) and (20), the sliding region which is limited by the following inequalities can easily deduced

1356
Proceedings of the International Conference on Industrial Engineering and Operations Management
Rabat, Morocco, April 11-13, 2017

 k ve  k z   k x*  k L x*  k z  k x*  k z  k x*  0
s1  3 3 21 3 2 3 1 1 4 11 4 1 4 12 4 3
 k 3 ve  k 3 L1 x1*  k 4 z12  k 4 x*3  0
 (22)

s    z 22 x*4
k z   k x  k z   k x  k L x  k z  k x  k
* * * *
 k6
  5 21 5 2 5 22 5 4 5 2 3 6 12 6 3 6 R
0
R
s2 
 k z   k x*  k L x*  k z 22  k x 4  0
*

  5 21 5 2 5 2 3 6 R 6
R
From those inequalities and in order to ensure that the sliding mode exists at least around the equilibrium point (3)
with z1 , z 2 tend to zero at (3), the following condition must be satisfied:
k3 x1*  x*3 (23.a)

k4 ve  x1*  x*2
x*
x*3  4
k5 R (23.b)

k6  x*2  x*4  L2 x*3
b.3 Controller design

the same control law (14) defined before will be used:

 ueq1  un1 
u 
 ueq 2  un 2 
 
 the equivalent control
the equivalent control can be calculated from s  0 , and the obtained control’s low is:

k4 x3  k3 L1x1*  k3ve (24.a)

ueq1  1 
k4 x1  k3 x2
x4
k6  k5 L2 x*3  k5 x2
ueq 2  1  R (24.b)
k6 x3  k5 x4

 the switched control

The switching control low is defined to ensure the negativity of V .

Using (8), (21) and (24):

V1  u1n k3 x2  k4 x1 .s1 (25.a)
V  u k x  k x .s
2 2n 4 4 6 3 2 (25.b)
We look for u1n and u 2 n as:
1
u1n  u~
k3 x2  k4 x1  1n (26.a)

1
u2 n  u~
k4 x4  k6 x3  2n (26.b)

We have
V1  s1u~1n (28.a)
V  s u~
2 2 2n (28.b)
~ ~
The suitable choice of u1n and u 2 n
u~1n  1 sin g s1  (29.a)
u~2 n   2 sin g s2  (29.b)

1357
Proceedings of the International Conference on Industrial Engineering and Operations Management
Rabat, Morocco, April 11-13, 2017

To ensure the stability’s condition V  0 :

V1  1 s1 (30.a)
V2   2 s2 (30.b)
Then the final control’s law:
k4 x3  k3 L1x1*  k3ve  1 sgns1  (31.a)
u1  1 
k4 x1  k3 x2

 k5 L2 x*3  k5 x2   2 sgns2 
x4
k6
u2  1  R (31.b)
k 6 x3  k5 x4
4. Simulation results

In this section, the simulation’s results of the cascade boost converter with both controllers designed
previously is illustrated by using SIMPOWER of MatlabSimulink, in order to verify the theoretical results using the
parameters value shown in tables 1.

Table 1. system parameters

Double boost Value Controller 1 Value Controller 2 Value
parameters parameters parameters
Inductors L1  35 H , k1 , k 2 200000 k3 , k5 0 ,5  0 ,15
L2  100H
Capacitors C1  6 mF , 1 , 2 0.03 k 4 , k6 0 ,2  0 ,15
C2  600 F
Frequency 10 kHz ki1 , ki 2 10  3 , 4 0,5
Input voltage 12 .8 k p1, k p 2 40-2
x*2 30  60
Load 10  30  50 50 100
resistors
x*2 x*4
Switchers Ideal
x*4 50  70

80
80

70
70

60
60

50
50
Vc2,Vr2
Vc1,Vr1

40 40

30 30

20 20

10 10
Vc1 Vc2
Vr1 Vr2
0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
temps(s) temps(s)

Figure 4. The intermediate voltage vc1 in C1 Figure 5. The intermediate voltage vc 2 in C2

Figure 4 and 5 shown respectively the intermediate and the output regulated voltages ( vc1 , vc 2 ) presented with the
blue color, those signals are obtained thanks to the first controller which is synthetized by two nested loops, the
regulated voltages pursuit the given references ( the doted red signals) but the system is clearly instable specially the
intermediate voltage ( vc1 ), then the simulation results show the weakness of the first regulator and how it need to be
improved.

1358
 Proceedings of the International Conference on Industrial Engineering and Operations Management
Rabat, Morocco, April 11-13, 2017

110
60
100

90
50
80

40 70
Vc1,Vr1

Vc2,Vr2
60
30
50

40
20
30

20
10
Vc1 10 Vc2
Vr1 Vr2
0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.1 0.2 0.3 0.4 0.5 0.6
temps(s) temps(s)

Figure 6. the intermediate voltage vc1 in C1 Figure 7. the intermediate

voltage vc 2 in C2

Figure 6 and 7 shown respectively the intermediate and the output regulated voltages ( vc1 , vc 2 ) obtained thanks to the
second controller which is synthetized with a single loop using the (Hijazi et al., 2009) sliding surface, each figure
illustrate two signals, the first one which is the doted red is the reference signal and the second signal which is bleu
represent the voltage signal. After 0.3 s the reference signal changed and the voltage signal pursuit that change with
a transition period that lasts 18 ms. Also, the two figures shown that the regulated voltages pursuit the given references
as good as expected, the system is stable, then the simulation results show the efficiency of the second controller.

2.5 0.5

2
0.4

1.5
0.3

1
0.2
s1

s2

0.5
0.1
0

0
-0.5

-1 -0.1

-1.5 -0.2
0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.1 0.2 0.3 0.4 0.5 0.6
t(s) t(s)

Figure 8. The sliding surface s1 (21.1) Figure 9. The sliding surface s2 (21.2)
-3
x 10
0.025 6

5
0.02

0.015
3
z11

z12

0.01 2

1
0.005

0
-1

-0.005 -2
0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.1 0.2 0.3 0.4 0.5 0.6
t(s) t(s)

Figure 10. The currant’s error z11 (6) Figure 11. The currant’s error z12 (6)

1359
Proceedings of the International Conference on Industrial Engineering and Operations Management
Rabat, Morocco, April 11-13, 2017

-3
x 10
0.08 8

6
0.06

4
0.04

z'22
z'21

0.02
0

0
-2

-0.02
-4

-0.04 -6
0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.1 0.2 0.3 0.4 0.5 0.6
t(s) t(s)

Figure 12. The voltage’s error z 21 (19) Figure 13. The voltage’s error z 22 (19)

Figure 8 and 9 shown the sliding surfaces where s1  0 and s2  0 , and show that the sliding mode is obtained, where
the figure 10, 11, 12 and 13 shown respectively the currant’s errors z11 and z12 , and the voltage’s errors z 21 and z 22
, all the errors of the system vanishes (errors=0) which ensures the pursuit of the references and shows the reliabilit y
of the synthesized regulator.

5. Conclusion
To get high conversion ratio which is needed in plenty of applications, there is different topologies to obtain
that aim. In this paper, we studied and controlled a double cascade boost converter, this converter has a satisfying high
conversion ratio. To control this circuit, we used a sliding mode approach. This controller is designed with two
different ways: firstly, the controller is designed using two nested loops utilizing the sliding mode approach for the
inner currant loop and PI controller for the outer voltage loop, secondly we designed a controller using the sliding
mode approach but with another sliding surface which allows us to regulate the currant and the voltage in the same
time, the simulation results show that the controllers meet their objectives besid es the second controller has a better
results in the term of stability and tracking and a good minimization of regulator’s blocs .

6. References
Byrnes, C.I., Isidori, A., 1988. Local stabilization of minimum-phase nonlinear systems. System & Control Lettres. 11, 9–17. doi
:10.1016/0167-6911(88)90105-3
FLOQUET, T., 2000. Contributions à la commande par modes glissants d’ordre supérieur. Université Lille1 - Sciences et
Technologies (Productique, automatique et informatique industrielle).URL http://ori.univ-lille1.fr/notice/view/univ-
lille1-ori-160192 (accessed 11.10.16).
Hijazi, A., Di Loreto, M ., Bideaux, E., Venet, P., Clerc, G., Rojat, G., 2009. Sliding mode control of boost converter: Application
to energy storage system via supercapacitors, in: EPE. Barcelone, Spain, p. électronique.
Hu, H., 2016. Sliding M ode Control: The Delta-Sigma M odulation Approach [Bookshelf]. IEEE Control Syst. 36, 139–140.
doi:10.1109/M CS.2016.2536079
Rensburg, J.F.J. van, Case, M .J., Nicolae, D.V., 2008. Double-Boost DC to DC Converter, in: 34th Annual Conference of IEEE
Industrial Electronics, 2008. IECON 2008. Presented at the 34th Annual Conference of IEEE Industrial Electronics,
2008. IECON 2008, pp . 707–711. doi:10.1109/IECON.2008.4758040
Sira-Ramirez, H., 1987. Sliding motions in bilinear switched networks. in:IEEE Transactions on Circuits and Systems ( Volume:
34, Issue: 8, Aug 1987 ) Page(s): 919 - 933 DOI: 10.1109/TCS.1987.1086242
Slotine, J.-J.E., Li, W., others, 1991. Ap plied nonlinear control. prentice-Hall Englewood Cliffs, NJ.
Utkin, V.I., 2013. Sliding M odes in Control and Optimization. Springer Science & Business M edia.
Zhao, Q., Lee, F.C., 2003. High-efficiency, high step-up DC-DC converters. IEEE Trans. Power Electron. 18, 65–73.
doi:10.1109/TPEL.2002.807188

Biography
EL OTMANI Fadwa is a doctoral student in LTI laboratory, Hassan II Casablanca University. He holds a Master’s
degree in information processing (2015) and a bachelor’s degree in Electronics and Ind ustrial data processing
(2013). Her research interests focus on the area of Nonlinear Control of dc-dc converters.

1360