Sliding Mode Control Of Boost Converter : Application to energy storage

Authorsystem via
manuscript, HIJAZI
published in "EPE, Barcelone Alaa (2009)"
: Spain
supercapacitors

Sliding Mode Control of Boost Converter: Application to energy storage

system via supercapacitors

Hijazi Alaa*, Di Loreto Michael**, Bideaux Eric**, Venet Pascal*, Clerc Guy*, Rojat
Gerard*
*AMPERE UMR CNRS 5005 **AMPERE UMR CNRS 5005
Bât. Omega, Université Lyon I, Villeurbanne Bât. Saint Exupéry, Insa de Lyon, Villeurbanne
Université de Lyon, F.69622, Lyon, France Université de Lyon, F.69622, Lyon, France
Tel.: 0033 4 72 43 11 92. Tel.: 0033 4 72 43 81 98.
Fax: 0033 4 72 43 11 93. Fax: 0033 4 72 43 85 35.
E-Mail: alaa.hijazi@univ-lyon1.fr E-Mail: michael.di-loreto@insa-lyon.fr
URL: http:// www.ampere-lab.fr

Acknowledgements
This work was co-supported by “Direction Générale des Entreprises” and “Région Rhone-Alpes”
under the contract Hysbus through the competitive pole (Lyon Urban Truck and Bus 2015).
hal-00475770, version 1 - 23 Apr 2010

Keywords
«Energy storage», «Robust control», «Supercapacitor».

Abstract
Sliding mode control of boost converter is studied. In order to improve dynamical performances with
static and dynamic specifications, we propose a systematic procedure to compute the gains of the
controller based on an optimization scheme. This method is applied to the control of an energy storage
system based on supercapacitors technology in order to regulate the output voltage. Given a system
with large variations of input voltage and load, it is necessary to guarantee good performance of the
controller for large variations of operating point. Our study shows the great influence of the controller
gains on the static and dynamic performances of the system. Hence, we point out a methodology for
choosing the gains.

Introduction
In order to supply voltage, current and frequency needed for the load, and to guarantee the desired
dynamics, electronics power converters must be suitably controlled.
Conventionally, classical PID controllers are used for the control of power converters [1], [2], [3].
Simple models of converters are usually obtained from signals averaging and linearization techniques,
these models may then be used for control design [4], [5].
On the other hand, the PID family of controllers failed to satisfactorily perform constrained
specifications under large parameter variations and load disturbances [6]. Another choice for
controlling power converter is to use the sliding control techniques. Sliding mode control (SMC) of
variable structure systems such as power converters is particularly interesting because of its inherent
robustness, its capability of system order reduction, and appropriateness for the nonlinearity aspect of
power converters [5], [7], [8].
However, despite being a popular research subject, SMC is still rarely applied in practical DC-DC
converters. It is mainly due to the fact that no systematic procedure is available for the design of SMC
in practical applications [9]. For example, the influence of the controller gains on the closed loop
system performances for a given application is not properly clarified, and most of the previous works
are limited to the study of the influence of these parameters only on the existence and stability of
sliding mode [10], [11]. In other cases an empirical approach is adopted for selecting these gains of

EPE 2009 - Barcelona P.1

 Sliding Mode Control Of Boost Converter : Application to energy storage system via HIJAZI Alaa
supercapacitors

SMC, computer simulation and experiments were performed to study the effect of the various control
gains on the response of the output voltage [12].
Therefore in this paper, we study the design of SMC for boost converters. After studying and
analyzing different existing solutions for sliding mode control of boost converter, we propose a control
mode that allows a direct control of the voltage of boost converter. The performances of the controller
in terms of robustness and dynamic response will be improved. Most of literature works are concerned
with the study of hitting, existence and stability conditions of the SMC. Our contribution goes beyond
this direction by involving the study of the influence of control parameters on system performances. In
this context, we develop an optimization algorithm in order to choose the controller parameters based
on a predefined specification for a given real application.

Supercapacitors
& DC-DC
converters

Electric supply
Trolleybus DC- Motors &
DC Converter auxiliaries
hal-00475770, version 1 - 23 Apr 2010

Fig. 1: Architecture of the trolleybus supply system

The application concerned by our study is the supply of an electrical bus via supercapacitors in case of
electrical microcuts. The supply system of trolleybus consists of two parallel DC-DC converters (Fig.
1). The first converter (Trolleybus converter) regulates the voltage of the bus Vb to 330 Volts, the
other converter (Supercapacitors converter) manages the energy transfer between supercapacitors and
trolleybus. When the voltage of trolleybus is more than 350 Volts, the supercapacitors will be charged.
When the voltage decreases under 310 Volts, the supercapacitors discharges to insure the continuity of
the supply of the auxiliary of trolleybus (pump, air compressor, fans). The supercapacitors voltage
must be always between 120 and 300 Volts. In this paper, we will focus on the discharge phase of
supercapacitors. As the voltage of trolleybus is greater than the voltage of supercapacitors, our study is
limited to the study of the control of boost converter.
In section II, we will present the mathematical model of boost converter. After that we will analyze the
design of SMC for boost converter and then we study the influence of the choice of controller
parameters on the system performances. This analysis requires an optimization algorithm to compute
the parameters.
In section III, the importance of application of SMC for the alimentation of electrical bus is illustrated.

Sliding mode control for boost converter

Mathematical Model of Boost Converter

Fig. 2: Boost converter circuit

EPE 2009 - Barcelona P.2

 Sliding Mode Control Of Boost Converter : Application to energy storage system via HIJAZI Alaa
supercapacitors

Fig. 2 shows the circuit of boost converter. It consists of a DC input voltage E which represents the
voltage of the supercapacitors, a smoothing inductor L, a controlled switch S, a freewheeling diode D,
a filter capacitor C and finally the load which is modeled by a resistor R. Assuming that the circuit is
in continuous conduction mode, which means that the inductors current never falls to zero, the
mathematical model of boost converter can now be easily deduced by applying Kirchhoff’s laws. The
model of boost converter in continuous conduction mode is:
dV
C 1 u iL i ,
dt (1)
diL
L E 1 u V ,
dt
where u is the switch state or the switch duty cycle in the case of average model, and are
respectively the output voltage and the inductor current of the boost converter. In SMC, we usually
determine u as following:
1 if S 0,
u (2)
0 if S 0,
hal-00475770, version 1 - 23 Apr 2010

Where is the sliding surface.

Design of sliding mode controller
The objective of boost converter control is to regulate the output voltage to a reference voltage
. The design of sliding mode controller for boost converter starts with the choice of sliding
surface. As it is shown in [13], it is clear that direct surface can be tend to zero only if
the current increases continuously. Usually, a cascade control structure is applied to control boost
converter, which leads to solve the control problem using two control loops [14], [15], [16] : The
output voltage loop generates the reference current from voltage error and the inner current loop
controls inductor current via sliding mode (Fig. 3).
+ Linear
controller Sliding Converter
+ controller u
- -

Fig. 3: Cascade control structure

This control of the output voltage of DC-DC converter meets the criteria of stability and existence of
sliding mode. However, it has been shown in [17] that it is difficult to determine the gains of the
voltage loop since sliding mode is a highly nonlinear method. Furthermore, since SMC is only applied
to current regulation, the voltage loop will be more sensitive to high frequencies phenomena and to
uncertainties on the reference current. In order to improve the performances of the controller, we
propose to study a control mode based on a sliding surface which involves output voltage. Let
( , ) be the desired equilibrium point, we take the following surface:

S K V V K iL I , (3)
Where and .
As reference current depends on the operating point, we will extract it, that is:
V i
I , (4)
E
where .

EPE 2009 - Barcelona P.3

 Sliding Mode Control Of Boost Converter : Application to energy storage system via HIJAZI Alaa
supercapacitors

Sliding surface coefficients ( , ) 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.

Existence condition:
The existence condition of sliding mode implies that both and will tend to zero when t tend to
infinity, which means that the dynamic of the system will stay into the sliding surface. The existence
condition of the sliding mode is 0 (when 0), the fulfillment of this inequality ensures the
existence of sliding mode around the commutation surface.
Let us write the model of boost converter (1) in a state space where the equilibrium point is the origin.
We obtain:
dx x V
C 1 u x I ,
dt R
(5)
dx
L E 1 u x V ,
dt
where and .

Replacing by its value from (4) in the expression of the commutation surface (3), we obtain:
hal-00475770, version 1 - 23 Apr 2010

K V V
S K V V K iL ,
RE
(6)
K V
K V K iL K V .
RE
Writing this last equation in the new coordinate system ( , ), we have:
K V K V
K x K V K x K I K V ,
RE RE (7)
K x K x ,
K V
Where K RE
.
By calculating to be negative for u=0 and 1, we can deduce that the sliding region is limited by the
following inequalities:
K K x K E V V I
x K K 0,
RC L C L L RC C
(8)
K x K E V K
0.
RC L RC

From these two inequalities and in order to ensure that the sliding mode exists at least around the
equilibrium point ( 0, 0), the following condition must be satisfied:
K RCE
. (9)
K V L

Stability condition:
The stability of the system is guaranteed if the dynamic of the system in sliding regime is directed
toward the desired equilibrium point. Our aim is to determine the dynamic of the new variable state x
and x when the sliding regime is reached. Given the state space model (5), the commutation surface
(6), and, from =0, the equivalent average control that must be applied to the system in order that the
system state slides along the surface is given by:

EPE 2009 - Barcelona P.4

 Sliding Mode Control Of Boost Converter : Application to energy storage system via HIJAZI Alaa
supercapacitors

E K x V
L K RC
u 1 . (10)
x V K x I
L K C
By replacing the equivalent control (10) in the state space model (5), and from 0, we deduce the
dynamic of at the sliding regime:
x K x V
dx E I
K R
. (11)
dt K x K
C V x L I
K K

Let be a lyapunov candidate function, we have so that:

K V V
EK R R
V x . (12)
K K
C Veref x L Ieref
K x K
The condition for to be negative is:
hal-00475770, version 1 - 23 Apr 2010

Veref
K K
C Veref x L Ieref x 0 . (13)
K K

Taking into account the sliding region defined by (8) and the existence condition (9), we can
demonstrate that the condition (13) is always satisfied along the sliding region of the commutation
surface. Based on the stability theorem in the sense of Lyapunov we can say that the system is globally
asymptotically stable.
Influence of controller parameters on system performance
Clearly, inequality (9) provides only the general information about the existence of sliding mode. On
the other hand, choosing controller parameters has a significant influence on the performances of
closed-loop system especially when the system presents large variations around nominal value like our
system. The first criteria for choosing these parameters must be the size of sliding part of commutation
surface. In fact, sliding condition is only satisfied on a subpart of the surface and not on the entire
surface as shown in (8). In consequence, the controller parameters must be carefully chosen to ensure
that the system dynamic will intercept the commutation surface in the sliding part. In this way we
prevent undesirable behavior which is reflected by a response overshoot (Fig. 4).

O O

(a) (b)

Fig. 4: Sliding mode with (a) and without overshoot (b).

EPE 2009 - Barcelona P.5

 Sliding Mode Control Of Boost Converter : Application to energy storage system via HIJAZI Alaa
supercapacitors

Let A( , ), be the intersection of the system dynamics with the commutation surface, and let
B( , ) be the limit of sliding part as shown in (Fig. 4). Supposing that at 0 the output
voltage is equal and the current in the inductor is zero. As at 0 the surface will be negative,
then 1 so that the state space vector will be:
dx x V
C ,
dt R
(14)
dx
L E.
dt
By solving these two equalities taking into account initial conditions, we obtain:
L I
x V e RCE V . (15)

This equation represents the dynamic of the system before intercepting the surface at the point A.
From 0 and from (14) we deduce the coordinate of point A:
L A I
K
x A V e RCE V . (16)
K

Solving this last equation we obtain:

hal-00475770, version 1 - 23 Apr 2010

K
LK V
K
LV
K RE V e RCE
K
lambertw CE R L K REV LV (17)
R E C

x A
REL

is the lambert function defined for a complex number Z by the inverse function of
.
From the equation of the commutation surface (7) and the limit of sliding zone defined by (8), we
deduce the coordinate of point B, we have:

V I
K RC C E V
K K L L (18)
x B .
K K K 1
K K RC C
We can now deduce that the system intercepts the commutation surface in the right part if the
controller parameters are selected in such a way that:
. (19)
On the other hand, theoretical approach of sliding mode supposes that the hysteresis band is null,
therefore frequency tends to infinity. It is clear that we cannot keep this assumption due to frequency
limitation caused by the characteristic of circuit components and losses. Usually, an hysteresis band
∆ is added around the surface in order to fix the operating frequency, looking at Fig. 5 we can
calculate the rise time " " and fall time " ":
2∆ 2∆ 1
t t f (20)
S S t t

EPE 2009 - Barcelona P.6

 Sliding Mode Control Of Boost Converter : Application to energy storage system via HIJAZI Alaa
supercapacitors

u=1 u=0

2∆

1 2

Fig. 5: Sliding Regime with Hysteresis Band

From equations (7) and (20), we deduce the expression of the hysteresis band function of the surface
parameters, the operating point and the circuit parameters, we obtain:
hal-00475770, version 1 - 23 Apr 2010

1 1
∆ . (21)
2f 1 1
K V K V V E V
K K
L E RC REC RC L L
Hence, from the expression of ∆ function of and , a restriction on the choice of the controller
parameters holds. In fact, values of these parameters must guarantee that the hysteresis band is greater
than the perturbation generated by the converter, and in the same time, the value of the hysteresis band
should be limited to guarantee the robustness.
Furthermore, as our objective is to control the output voltage and as we can see in (3) the expression
of commutation surface depends on the current errors, so a major study for optimization of controller
gains should be the analysis of the sensitivity of the controller facing a measurement or an estimation
error of current reference. In fact, since the current reference is unknown, we are constrained to extract
it from the load current (4). However, this latter can be measured or observed through an extended
Luenberger observer. In both cases, an error can be occurred which affects the response of the system
in terms of steady state error. So the choice of controller parameters must take into account to make
the closed loop system less sensitive to an error in the current part of the surface.
Supposing that the measured reference current ( ) can be expressed as the sum of the expression
of reference current given by (4) and an error term ( ), where e is defined as the error percentage,
so that:
I I eI (22)
By replacing this expression of in the expression of the commutation surface (3), we have:
K eV i
S K V V K iL I ,
E
K eV V
K V V K iL I ,
ER
2 (23)
K eV V V K eVeref
K V V K iL I ,
ER RE
2
K eV K eVeref
K V V K iL I ,
ER RE

EPE 2009 - Barcelona P.7

 Sliding Mode Control Of Boost Converter : Application to energy storage system via HIJAZI Alaa
supercapacitors

2
K Veref
The term RE
is constant so we can say that:

K eV
S K V V ∆V K iL I ∆iL .
ER
Where ∆V and ∆iL represents the voltage and the current steady state errors.
We have:
K eV K eV (24)
K ∆V K ∆iL .
ER RE
The control given by (2) will tend S given by (23) to zero, so that Veref ∆Vb and I
∆iL . From the equality of power input-output we can say that:
V ∆V i E I ∆iL . (25)
From equation (24) and (25), we determine the expression of ∆ :

V V V V 4
∆ . (26)
2
hal-00475770, version 1 - 23 Apr 2010

Simulations:
Simulations were performed on a typical boost converter circuit. The input voltage of the boost E is
given by 120 supercapacitors (3000 F) and can varies between 120 and 300 volts. The charge power is
between 20 and 50 kW. The computer aided design software “Simplorer V7.0” is used for the
simulation. The parameters of controllers are given by a genetic algorithm. The criteria are based on
industrial specifications. Thus the objective of the genetic algorithm is to determine the values of
parameters that ensure, regardless of the operating point, that the system will intercept the sliding part
of commutation surface while respecting the following conditions:
K RCE V
stability and existence condition ,
K V L RE
∆ max ∆ , ∆ ,
max ∆ ∆ for .

Figure 6 shows that the change in supercapacitor's voltage ( ) has no influence on output voltage ( )
so the controller is absolutely robust with respect to supercapacitor’s voltage variation.
On the other hand, Figure 7 shows the recovering features of the proposed controller to the imposed
load. The load has been changed by various steps in the range 2Ω , 4Ω . As expected, the sliding
mode controller is robust when the load resistance is subject to a sudden variation.
Figure (7.a) and (7.b) shows that dynamic response of output voltage and inductor current has an
excellent behavior with non oscillation and that is independent of the operating point. The proposed
controller is therefore very appropriate for systems which present large variations around its nominal
operating point as our concerned system. The shape of the output voltage shows fast recovery time,
which is a very important factor in application where the supercapacitors are used as an auxiliary
supply system.
Given that the duration of electrical failure can be very small, it is very important to test the startup
behavior of the controller. Figure 8 presents the startup phase plane trajectories of sliding mode
control for different operating points. As mentioned before, the boost converter will regulate the
output voltage to 330 volts when the voltage of the bus decreases below 310 volts in the case of
electrical microcuts. As it is shown, the behavior of phase plane trajectories is similar. Looking to
figure 7 and 8, it is important to note that the system is directed toward the sliding part of the
commutation surface, which validates the proposed algorithm for choosing the parameters of the

EPE 2009 - Barcelona P.8

 Sliding Mode Control Of Boost Converter : Application to energy storage system via HIJAZI Alaa
supercapacitors

controller. Unlike classical control via PID, values of parameters are not determined to satisfy desired
performance only around nominal values but to cover larges variations in the operating condition.

Fig. 6: Output voltage with change of supercapacitor’s voltage.

2Ω 4Ω 3Ω 2Ω
hal-00475770, version 1 - 23 Apr 2010

Fig. 7: Output voltage and inductor current waveforms for step load variations with different initial
voltage of supercapacitors ((a) => E=300 V, (b) => E=150 V).

EPE 2009 - Barcelona P.9

 Sliding Mode Control Of Boost Converter : Application to energy storage system via HIJAZI Alaa
supercapacitors

Fig. 8: SLMC phase plane trajectories for startup according to different levels of load and super
capacitor voltages.

Conclusion
A new sliding mode control design of boost converter was developed in this paper. A control mode
that improves the performances of the system was proposed. The influence of the control parameters
on the performances of the system was studied. An optimization algorithm was developed in order to
calculate the optimal values of parameters based on a predefined specification. The importance of the
application of this type of controllers for the supply of electrical bus via supercapacitors is highlighted.
Results show excellent dynamic response of controller and robustness to load and input voltage with
large variations around nominal values. Further research may focus to the study of the observation of
load current, and on the stability of the controller-observer closed loop.

References
[1] S. Arulselvi, G. Uma and M. Chidambaram, "Design of PID controller for boost converter with RHS zero",
4th International Power Electronics and Motion Control Conference (IPEMC 2004), 14-16 august 2004, China.
[2] L. Guo, J.Y. Hung and R. M. Nelms, "Design and implemenation of a digital PID controller for a buck
converter," Proceedings of the 36th Intersociety Energy Conversion Engineering Conference, July/August 2001,
hal-00475770, version 1 - 23 Apr 2010

vol. 1, pp. 187-192.

[3] H. Mingzhi and X. Jianping, "Nonlinear PID in Digital Controlled Buck Converters", Applied power
electronics conference APEC 2007, 25 Feb-1 March 2007, Anahein CA USA, pp. 1461-1465
[4] A.J. Foreyth and S.V. Mollov, "Modeling and control of Dc-Dc converters", IEEE Power Engineering
Journal , vol. 12 no. 5, 1998, pp 229-236.
[5] M. Rashid, Power Electronics Handbook, Academic Press, 2001.
[6] V.S.C Racirah and P.C Sen, "Comparative Study Of Proportional- Integral Sliding Mode, and Fuzzy Logic
Controllers for Power Converters", IEEE Transactions On Industry Applications, vol. 33 no. 2, 1997, pp 518-
524.
[7] V.I Utkin, "Sliding Modes and Their Applications on variable structure system", MIR publisher Moscow,
1978.
[8] V.I Utkin, "Sliding Modes in control optimization", springerverlag, 1981.
[9] Siw-Chong Tan, Y.M Lai, Chi K.Tse, "General Design Issues of Sliding-Mode Controllers in DC-DC
Converters", IEEE transactions On Industrial Electronics, vol. 55, no. 3, March 2008
[10] G. Spiazzi, P. Mattavelli and L. Rossetto, "Sliding control of DC-DC converters", 4 Congresso Brasileiro de
Elettronica de Potencia (COBEP), Belo horizonte, December 1997, pp 59-68.
[11] D. Cortes and J. Alvarez, "Robust sliding mode control for the boost converter", Proceedings of VIII IEEE
International Power Electronics Congress, (CIEP 2002), 20-24 October 2002, pp. 208 - 212.
[12] S-C Tan, Y.M Lai, Chi K Tse, Luis Martniez Slamero, Chi-Kin Wu, "A Fast response sliding mode
controller for boost type converter with a wide rang of operating conditions", IEEE transactions On Industrial
Electronics, vol. 54, no. 6, March 2007
[13] H. Sira-Ramirez, "Sliding motions in bilinear switched networks", IEEE Trans. on Circuits and Systems,
Vol. 34,No. 8, August, pp. 919-933, 1987
[14] V.I. Utkin, "Sliding mode control in electromechanical systems", press,1999
[15] M. Ahmed, M. Kuisma, P. Silventoinen, "Implementing Simple Procedure for controlling Switch Power
supllying using SMC as a control techniques", Proceeding of the XIII International Symposium, SIELA, 29-30
May 2003, Plovdive Bulgarie
[16] H. Gueldemir, "Sliding Mode Control of Boost Converter", Journal of applied science, vol. 5, no. 3, p.588-
592, 2005.
[17] M. Ahmed, M. Kuisma, P. Silventoinen, "Comparison between PID Control and Sliding Mode Control for
Buck Converter", Proceedings of the Symposium on Power Electronics, Electrical Drives, Automation and
Motion. SPEEDAM 2004. CAPRI - ITALY, 16-18 June 2004.

EPE 2009 - Barcelona P.10