Available online at www.sciencedirect.

com

ScienceDirect
Solar Energy 132 (2016) 538546
www.elsevier.com/locate/solener

Designing a new robust sliding mode controller for maximum

power point tracking of photovoltaic cells
Mohammad Rasool Mojallizadeh, Mohammadali Badamchizadeh ,
Sohrab Khanmohammadi, Mehran Sabahi
Faculty of Electrical and Computer Engineering, University of Tabriz, Tabriz, Iran
Received 9 December 2015; received in revised form 9 February 2016; accepted 17 March 2016
Available online 8 April 2016
Communicated by: Associate Editor Bibek Bandyopadhyay

Abstract
This paper proposes a new sliding mode controller for maximum power point tracking of photovoltaic cells. By dening a suitable
sliding surface, the proposed control law does not require reference voltage/current. The proposed system is based on the one-loop control schemes which makes its implementation easy. Stability of the proposed system is ensured using Lyapunov stability theorem. It is
proved that the proposed control system is robust to system uncertainties. Moreover, a new state-dependent control magnitude is
designed which suppresses chattering of the system. A traditional sliding mode controller is considered in order to compare the results
of the proposed system. Simulation and experimental results are used to evaluate the robustness of the proposed system. It is shown that
the proposed system is robust to system uncertainties, environment changes and load variations.
2016 Elsevier Ltd. All rights reserved.

Keywords: Photovoltaics; Sliding mode control; Robustness; Maximum power point tracking

1. Introduction
Environmental problems and energy issues such as
rising oil price are motivating research on the development
of renewable energy sources (Ajanovic and Haas, 2015).
Among renewable energies, photovoltaic (PV) has
attracted a lot of interest with many important applications
because it is free, abundant and environmental friendly
(Mavromatidis et al., 2015; Fattori et al., 2014). Moreover,
Corresponding author at: Faculty of Electrical and Computer
Engineering, University of Tabriz, 29 Bahman Blvd., Tabriz, Iran.
Tel.: +98 41 33393729; fax: +98 41 33300819.
E-mail addresses: mojalli@tabrizu.ac.ir (M.R. Mojallizadeh),
mbadamchi@tabrizu.ac.ir (M. Badamchizadeh), khan@tabrizu.ac.ir
(S. Khanmohammadi), sabahi@tabrizu.ac.ir (M. Sabahi).

http://dx.doi.org/10.1016/j.solener.2016.03.038
0038-092X/ 2016 Elsevier Ltd. All rights reserved.

the cost of electricity from PV systems is close to that from

conventional sources (Benda, 2015).
In order to extract the maximum electrical power from
the PV cells under changing loads and environmental conditions, it is essential to provide PV systems with maximum
power point tracking (MPPT) controllers. In this context,
dierent MPPT controllers has been addressed in many literatures (Rezk and Eltamaly, 2015). In many PV systems a
switching converter is usually connected between the PV
modules and the load (Gonzalez Montoya et al., 2016).
Due to nonlinear characteristics of PV modules and switching converters, nonlinear controllers have attracted considerable attention (Kim, 2007; Chu and Chen, 2009; Chiu
et al., 2012; Zhang et al., 2015; Hamrouni et al., 2009;
Lauria and Coppola, 2014; Liu et al., 2014; Valencia and
Ramos-Paja, 2015; Mojallizadeh and Badamchizadeh,

M.R. Mojallizadeh et al. / Solar Energy 132 (2016) 538546

2016). Among nonlinear controllers, sliding mode control

(SMC) has attracted a lot of interest due to the system
order reduction, large signal stability, simple implementation, robustness to model uncertainty and disturbances
(Young et al., 1999; Utkin, 2013). Sliding mode control is
a nonlinear control strategy based on the variable structure
theory. In a variable structure control system, the structure
of the controller changes from one form to another (Utkin,
1993). Conventional sliding mode MPPT controllers are
based on the following methods:
(1) The rst method is based on the two-loop MPPT
control approach (Kim, 2007; Chiu et al., 2012;
Hamrouni et al., 2009; Lauria and Coppola, 2014;
Liu et al., 2014; Lalili et al., 2013; Valencia and
Ramos-Paja, 2015). In this method, the rst loop is
to nd the reference voltage/current of the PV module and the second loop is to regulate the PV module
voltage/current to the reference point. A terminal
sliding mode MPPT controller was proposed in
Chiu et al. (2012), which ensures the nite time convergence of the system. However, the chattering
problem is not addressed in this controller. The main
drawback of the rst method is that the interaction
between two loops needs to be addressed appropriately (dan Zhong et al., 2008).
(2) The second approach is based on the one-loop
control method (Chu and Chen, 2009; Zhang et al.,
2015; Yau and Chen, 2012; Valencia and
Ramos-Paja, 2015). In this scheme, the controller
does not need reference voltage/current. Therefore,
its implementation complexity and calculation burden is relatively low. When the value of the sliding
surface converges to zero, the PV power will be maximized. In Chu and Chen (2009), the sliding surface is
dened as the derivative of the PV power with respect
to PV current. In some other references Yau and
Chen (2012) and Zhang et al. (2015), the sliding surface is dened as the derivative of the PV power with
respect to PV voltage. In Zhang et al. (2015), a new
strategy is proposed which reduces the requirement
of voltage sensors. A new sliding mode controller
for grid-connected PV systems is proposed in
Valencia and Ramos-Paja (2015). This control system
is based on the one-loop method. However, its chattering problem increases the voltage ripple of the PV
module. Although these approaches are robust to
environment changes and load variations, these
controllers are not robust to system uncertainties.
The main challenges in designing MPPT controllers can
be summarized as follows: The controllers should be able
to track the maximum power point to maximize the
extracted PV power. Many MPPT controllers suer from
chattering problem (Valencia and Ramos-Paja, 2015). This
problem decreases the eciency of system. Moreover, chattering may excite the high frequency dynamics which may

539

lead the system into instability. Another challenge related

to PV systems is the ease of implementation. One-loop
schemes are usually easier to implement compared to
two-loop schemes (Zhang et al., 2015).
The main motivation for the proposed controller presented in this article comes from the fact that robustness
of the controller is important in PV systems where input
voltage variations and parasitic elements are always present. The contribution of this paper is designing a new sliding mode controller for MPPT of PV modules. The
proposed scheme is based on the one-loop control strategy.
The robust stability of the proposed controller is assured
by Lyapunov theory. Moreover, chattering amplitude of
the controller is suppressed using a new state-dependent
control magnitude (Levant, 2010). This paper is structured
as follows. Modeling of the PV system is presented in Section 2. The proposed sliding mode controller and its robust
stability are addressed in Section 3. Results of the simulations and experiments are given in Sections 4 and 5, respectively. Finally, conclusions are presented in the last section.
Table 1 represents the nomenclature of quantities that are
used in this paper.
2. Modeling of the photovoltaic system
A circuit schematic of the PV system is shown in Fig. 1.
The system is composed of a PV module and a boost converter. The equations of the PV module and the boost converter are given in Sections 2.1, 2.2, respectively.
2.1. Model of the PV module
The voltage-current characteristic of the PV module can
be written as Park and Choi (2015):



qV P

I P N p I ph  I rs eN s AK 0 T  1
1
where I P and V P are the output current and voltage of the
PV module, respectively. N s and N p are the number of the
series and parallel cells, respectively. T denotes the PV
module temperature, q is the charge of an electron,
A 2 1; 5 denotes the ideality factor, k 0 is the Boltzmanns
constant, I rs is the reverse saturation current and I ph
denotes the light generated current. Moreover, I ph and I rs
depend on solar irradiance and PV module temperature
and can be described by Eqs. (2), (3) (Park and Choi, 2015).
go
1=T 1=T 
3
I rs ir T =T ref e K 0 A ref


I ph is K l T  T ref k=1000
qE

2
3

where, ir denotes the reverse saturation current at the reference temperature (T ref 298K), is denotes the shortcircuit cell current at the standard condition, Ego denotes
the semiconductor band-gap energy, K l A=K is the temperature coecient and k denotes the irradiance in W/m2.
The PV module parameters are shown in Table 1. Powercurrent curve of the PV module is shown in Fig. 2. It can

540

M.R. Mojallizadeh et al. / Solar Energy 132 (2016) 538546

Table 1
Nomenclature.
Quantity

Comment

PV module
k
Solar irradiance level (W/m2)
T
PV module temperature (K)
ir
Reverse saturation current of the PV cell (5:98  108 A)
iph
Light-generated current of the PV cell (A)
IP
Current of the PV module (A)
Voltage of the PV module (V)
VP
DV P
Uncertainty in the PV voltage (which is unknown in practice)
(V)
irs
Reverse saturation current at the standard condition (A)
Short-circuit current at the standard condition (3.45 A)
is
A
Ideality factor of the cell (1.12)
Ego
Semiconductor band-gap energy (1.2 ev)
Kl
Temperature coecient of the short-circuit current
(12  104 A/K)
k0
Boltzmanns constant (1:3805  1023 J/K)
T ref
Reference temperature of the PV module (298 K)
Np
Number of parallel cells (1)
Number of series cells (36)
Ns
Boost converter
L
Nominal inductor (1.5 mH)
DL
Uncertainty in the inductor (which is unknown in practice) (H)
C
Nominal capacitor (500 lF)
DC
Uncertainty in the capacitor (which is unknown in practice)
(F)
R
Nominal resistance of the load (X)
DR
Uncertainty in the load (which is unknown in practice) (X)
Vo
Output voltage of the converter (V)
Controllers
Parameters of the proposed SMC
g0 ; d; k
(g0 0:1; d 0:01; k 0:005)
kc
Parameter of the conventional SMC (0.01)

module and therefore can maximize the output power. By

utilizing state-space averaging method (Ngamkong et al.,
2012), the practical state-space equations of system can
be written as follows (Wai and Shih, 2011):
(
DV P
1u
V o V PLDL
I_ P  LDL
4
Vo
V_ o 1u I P 
CDC

CDCRDR

where, R is nominal resistance of the load, C is nominal

capacitance, L is nominal inductance, u is averaged control
signal, I P is the PV current, V P is the PV voltage and V o is
the output voltage. DR, DC; DL and DV P are uncertainties
of R; C; L and V P , respectively which maybe time dependent. The state-space equation of PV system can be written
as follows:
(
I_ P  1u
V o VLP W 1
L
5
V_ O 1u I P  V o W 2
C

CR

where, the lumped uncertainties can be written as Eq. (6). It

can be seen from Eqs. (5), (6) that the practical state-space
model of PV system is uncertain. It should be noted that
many existing methods do not take into account the system
uncertainties. Therefore, many existing MPPT controllers
are not robust to system uncertainty.
8
< W 1 LDV P V P uV o V o DL
LLDL
6
: W 2 uI P I P DC RDC CDRDRDCV o
CCDC
RCRCRDC CDRDRDC
3. Design of the sliding mode controller

be observed that the maximum power changes along dierent irradiance levels. Moreover, there is a PV current (I P )
which maximizes the PV power.
2.2. Model of the boost converter
A boost converter is utilized in the PV system for regulation of output current of the PV module (I P ). The input
impedance of boost converter can be tuned by duty cycle.
Thus, the boost converter can be considered as an adjustable loss free resistor (Haroun et al., 2015). As a result, the
boost converter can change the operating point of PV

The design of SMC systems involves two steps: design of

the sliding surface, and design of the control signal. The
sliding surface is selected based on the control objectives.
Afterwards, the control signal is designed to ensure the
existence condition of sliding mode. Thus, the system can
converge to its sliding surface in a nite time and remains
on it. From Fig. 2, the sliding surface is selected based
on the Eq. (7).


@P P @RP I 2P
@RP

I P 2RP I P
0
7
@I P
@I P
@I P
where RP V P =I P . Therefore, when the state variables converge to the sliding surface, the PV power will be maximized. The sliding surface is selected as follows:

Fig. 1. Circuit diagram of the PV system.

M.R. Mojallizadeh et al. / Solar Energy 132 (2016) 538546

541

Fig. 2. Characteristic of the PV module under dierent irradiance levels (T 25 (C)).

s , 2RP I P

@RP
@I P

In this study, the proposed control signal is dened by:

u ueq un

where, ueq is the equivalent control and un is the curbing

control. The aim of the curbing control is to drive the system states to the sliding surface. Moreover, the curbing
control eliminates the perturbation eect of the system
uncertainties in practical applications. The equivalent control ensures that the system trajectory stays on the surface
thereafter. By considering W 1 0, the equivalent control
can be calculated from the following condition:
s_ @s=@I P I_ P 0 ! 1  ueq V o V P =L 0
! ueq 1  V P =V o

10

In this study, the curbing control is dened as follows:


un g  sgns k  s
11
g g0 jsj d
where, g0 > 0, k > 0 and d > 0 are designing constants. It
can be seen from Eq. (11) that the control amplitude g
depends on the value of the sliding surface (s). Therefore,
chattering amplitude of the proposed system will be
suppressed.
From Eq. (8), the value of @s=@I P can be calculated as
follows:
@s
@RP
@ 2 RP
3
IP
@I P
@I P
@I 2P

12

By Considering RP V P =I P , the following equations are

achieved:


@RP
P
@I@P VI PP I1P @V
 VI 2P
@I P
@I P
P
13
2
2
@ RP
VP
1 @ VP
2 @V P


2
3
2
2
I P @I
@I
I
I @I P
P

From Eq. (1), V P can be written as a function of I P :



N s  A  k0  T
N P  I ph N P  I rs  I P
ln
VP
q
N P  I rs
Derivatives of the PV voltage are as follows:
@V P
N s  A  k0  T
N P  I rs

<0
q
@I P
N P  I ph N P  I rs  I P
@V 2P
N s  A  k0  T
N P  I rs

<0
2
q
@2I P
N P  I ph N P  I rs  I P

14

15
16

From Eqs. (13)(16), Eq. (12) can be rewritten as follows:

@s
1 @V P @ 2 V P V P

 2 <0
17
@I P I P @I P
@I 2P
IP
From Eqs. (5), (9), (10):
1  ueq  un
VP
Vo
W1
I_ P 
L
L
1  1  V P =V o  un
VP
Vo
W1

L
L
V o =Lun W 1
18
Considering Eqs. (10), (11), (18) and the Lyapunov function V 1=2s2 , time derivative of Lyapunov function is
as follows:


_V s_s s @s V o un W 1
@I P L


@s V o
s
g jsj d  sgns k  s W 1
@I P L 0


@s V o 2 V o
Vo

g0 s g0 djsj ks2 sW 1
@I P L
L
L


@s V o 2 V o
Vo 2
6
g0 s g0 djsj
ks  jsjjW 1 j
@I P L
L
L


@s V o
Vo

g0 s2 ks2 g0 djsj  jsjjW 1 j < 0; s 0

@I P L
L
19
Table 2
Simulation conditions.
Parameter

t = 0:25 (ms)

t = 25:50 (ms)

R (X)
k (w/m2)
T (C)

150
1000
50

100
900
20

542

M.R. Mojallizadeh et al. / Solar Energy 132 (2016) 538546

From Eq. (19), the stability condition of the proposed system is obtained as follows:
g0 d > L=V o jW 1 j

20

In order to suppress the chattering magnitude, the design

parameters are selected such that g0  d. Algorithm of
the proposed control scheme is shown in Algorithm 1.
Algorithm 1. Algorithm of the proposed MPPT control
system.

where, t represents the time. Simulation conditions are

shown in Table 2. It can be seen that the solar irradiance,
module temperature and load are variable over the time.
Simulation results of the proposed and conventional
controllers are shown in Fig. 3. The plot shows that the
proposed controller tracks optimal PV current with zero
steady-state error. Therefore, it can be concluded that the
proposed system is robust towards the system uncertainties. It is worthwhile to note that conventional controller
is not robust to system uncertainties and is unable to
converge to the steady-state. The output voltages of boost
converter, control signals and the values of sliding surface
are shown in Fig. 3. In order to compare the controllers,
R 0:05
extracted PV energy is dened as 0 I P  V P dt. The
extracted PV energies related to each controller are shown
in Table 3. In this scenario, the maximum theoretical PV
energy in the time interval t 0; 50 (ms) is 0.7428 (J).
Table 3 shows that the proposed controller can increase
the eciency up to 10%. Moreover, the proposed system
exhibits fast transient that correspond to step changes in
parameters.
5. Experimental results
In order to verify the results derived in previous sections, the experiment conguration is used as shown in
Fig. 4. The eectiveness of the proposed system is evaluated by two experiments. In the rst experiment, robustness of the system is veried through irradiance and in
the second experiment, it is examined through load
variation.
5.1. Experimental instrumentation

4. Simulation results
A conventional sliding mode controller (Chu and Chen,
2009) is considered in order to compare the results with the
proposed controllers responses. In the design procedure of
the conventional method, the uncertainties of the system
are not taken into account (Chu and Chen, 2009). The conventional control signal (uc ) is calculated as follows:
uc ueq k c  s

21

where, k c > 0 is designed based on Chu and Chen (2009). It

will be shown that this controller is not robust to system
uncertainties. In the simulation, robustness of the system
is examined towards system uncertainties. In this context,
the following sinusoidal disturbance is considered which
is commonly used in system analysis (Wai and Shih, 2011):
W 1 t 5sin1000t
W2 0

22

Specications of the PV module and boost converter

and parameters of the proposed controller are shown in
Table 1. The proposed controller is implemented in an
ATmega2560 microcontroller. The proposed algorithm
which is shown in Algorithm 1 has been written in C language. The Arduino integrated development environment
(IDE) is used to compile the program and download it to
the microcontroller. The PV current (I P ) is sensed using
an ACS712 Hall eect current sensor module. PV voltage
(V P ) and output voltage of PV system (V o ) are sent to
A/D pins of microcontroller through resistive voltage
dividers. The resolution and sampling rate of the A/D
converter are set to 10 bits and 125 kHz, respectively.
Low-pass RC lters are used in the feedback to reduce
the measurement noise. Afterwards, the proposed control
signal (u) is calculated and a PWM signal in 62.5 kHz is
generated to control the switch of the DC converter. The
voltages and currents are recorded through a digital oscilloscope. The PV module is emulated by a programmable
DC source (Lu and Nguyen, 2012). The PV emulator is
composed of a buck-boost converter and a microcontroller. The emulator utilizes a piecewise linear approach

M.R. Mojallizadeh et al. / Solar Energy 132 (2016) 538546

543

I P (A)

2
1
I P (The proposed controller)

0
0

0.005

0.01

0.015

0.02

0.025

0.03

0.035

I P (The conventional controller)

05

Optimal current

Vo (V)

40
V o (The proposed controller)

20

V o (The conventional controller)

0
0

0.005

0.01

0.015

0.02

0.025

0.03

0.035

0.04

0.045

0.05

1
0.5
u (The proposed controller)
(The conventional controller)

0
0

0.005

0.01

0.015

0.02

0.025

0.03

0.035

0.04

0.045

0.05

1013
X: 0.02209
Y: -0.07583

X: 0.03462
Y: 0.1152

0
S (The proposed controller)
S (The conventional controller)

-5
0

0.005

0.01

0.015

0.02

0.025

0.03

0.035

0.04

0.045

0.05

Time (s)
Fig. 3. Simulation response.

Table 3
Performance of the controllers.
Controller

Extracted PV
energy (J)

Eciency
(%)

The proposed controller

The conventional controller (Chu and
Chen, 2009)

0.7428
0.6475

97.3
87.2

to emulate the voltage-current curve of the PV module.

Block diagram of the experimental system is shown in
Fig. 5.
5.2. First experiment: Robustness to irradiance
In this case, the irradiance (k) varies between 700 and
1000 (w/m2), and vice versa while the load and temperature
are constant (R 50 (X) and T 25 (C)). Fig. 6 shows
transient response of experimental system. The rst transient corresponds to irradiance variation from 1000 to
700 (w/m2) and the last transient corresponds to irradiance
variation from 700 to 1000 (w/m2). PV current (I P ), PV
voltage (V P ), PV power (I P  V P ) and the output voltage
of boost converter (V o ) are shown in Fig. 6. In this experiment, during the rst time interval t 0; 200 (ms), the
irradiance level is 1000 (w/m2) and the optimal PV current
and voltage are I P 1:08 (A) and V P 14:33 (V), respectively. Accordingly, maximum PV power is about 15.48
(W). It can be seen from Fig. 6 that the tracking

performance of the proposed controller is good enough.

Since the resolution of the A/D converter and PWM generator are limited, the extracted PV power is about 15.35
(W), which is lower than the maximum PV power. During
the second time interval t 200; 600 (ms), irradiance level
is 700 (w/m2). In this condition, optimal PV current and
voltage are about 0.84 (A) and 14.25 (V), respectively.
Therefore, the maximum PV power is 11.97 (W). Compared to the previous time interval, the maximum PV
power is decreased about 3.4 (W). It can be seen that the
proposed system successfully tracks maximum PV power.
The tracking performance in the third time interval
t 600; 1000 (ms) is the same as the performance of the
rst time interval. It can be seen that PV voltage is nearly
constant. Based on the rst experiment, the proposed control system is robust to irradiance variation.
5.3. Second experiment: Robustness to load
In the last experiment, the proposed controller is tested
by applying changes in load resistance (R) while the irradiance and temperature are constant (k 1000 (w/m2) and
T 25 (C)). Fig. 7 shows transient response of the experimental system. The rst transient corresponds to load
variation from 100 to 50 (X) and the last transient corresponds to load variation from 50 to 100 (X). In this case,
during the time intervals t 0; 200 (ms) and
t 600; 1000 (ms), load resistance is 100 (X). Moreover,

544

M.R. Mojallizadeh et al. / Solar Energy 132 (2016) 538546

Fig. 4. Photograph of experimental equipment.

is shown in Table 4. It can be seen from Table 4 that the

system can nearly maximize the extracted PV power
regardless of its condition. Therefore, the proposed control
system is robust to irradiance level, load variation and system uncertainties.

6. Conclusions
Fig. 5. Block diagram of the experimental system.

in the second time interval t 200; 600 (ms), the load

resistance is 50 (X).
In this experiment, since the irradiance and temperature
are constant, optimal PV current and voltage are I P 1:08
(A) and V P 14:33 (V), respectively. Therefore, maximum
PV power is about 15.48 (W). The response of the proposed control system is shown in Fig. 7. Due to hardware
limitations, the extracted PV power is not exactly maximum. However, the performance of the proposed system
is acceptable.
By performing a simple calculation, it can be seen that
the prototype boost converter is not ideal and the average
value of its eciency is about 88%. Therefore, it can be
concluded that the experimental boost converter has some
uncertainties. Hence, it should be noted that the proposed
system is robust to system uncertainties.
Based on the experimental results, The steady-state performance of the proposed system under dierent conditions

This paper has presented the design of a new sliding

mode controller for maximum power point tracking of
PV cells. Comparing to the two-loop MPPT control
approach, the proposed controller does not required reference voltage/current for control law synthesis. Therefore,
its implementation complexity and calculation burden is
relatively low. Chattering amplitude of the controller is
suppressed using state-dependent control magnitude.
Robust stability of the proposed control system is ensured
using the Lyapunov theory. Simulation results show that
while the conventional controllers are not robust to system
uncertainties, the proposed controller rejects the disturbances. Therefore, the transient and the steady-state
responses of the proposed system are good enough in the
presence of system uncertainties. As a result, the proposed
controller can increase the eciency up to 10%. Moreover,
the experiments show that the proposed system is robust to
irradiance and load voltage variations. Moreover, the proposed SMC is shown to be eective for handling the plant
uncertainties. In the proposed system, the PV module

M.R. Mojallizadeh et al. / Solar Energy 132 (2016) 538546

=1000(W/m2)

=700(W/m2)

545

=1000(W/m2)

Fig. 6. First experiment: Response with the solar irradiance changing periodically stepwise between 700 and 1000 (w/m2).

R=100( )

R=50( )

R=100( )

Fig. 7. Second experiment: Response with the load changing periodically stepwise between 50 and 100 (X).
Table 4
Experimental results: steady-state performance of the proposed system
under dierent conditions.
Condition number

Condition 1

Condition 2

Condition 3

R (X)
k (W/m2)
Maximum extractable
power (W)
Extracted power (W)

50
700
11.97

50
1000
15.48

100
1000
15.48

11.97

15.35

Eciency (%)

100

99.16

15.48
100

reaches its desired operating point. Thus, the eciency of

the system can be improved.
References
Ajanovic, A., Haas, R., 2015. Driving with the sun: why environmentally
benign electric vehicles must plug in at renewables. Solar Energy 121,
169180.
Benda, V., 2015. Photovoltaics towards terawatts-progress in photovoltaic
cells and modules. IET Power Electron. 8 (12), 23432351.

546

M.R. Mojallizadeh et al. / Solar Energy 132 (2016) 538546

Chiu, C.-S., Ouyang, Y.-L., Ku, C.-Y., 2012. Terminal sliding mode
control for maximum power point tracking of photovoltaic power
generation systems. Solar Energy 86 (10), 29862995.
Chu, C.-C., Chen, C.-L., 2009. Robust maximum power point tracking
method for photovoltaic cells: a sliding mode control approach. Solar
Energy 83 (8), 13701378.
dan Zhong, Z., bo Huo, H., jian Zhu, X., yi Cao, G., Ren, Y., 2008.
Adaptive maximum power point tracking control of fuel cell power
plants. J. Power Sources 176 (1), 259269.
Fattori, F., Anglani, N., Muliere, G., 2014. Combining photovoltaic
energy with electric vehicles, smart charging and vehicle-to-grid. Solar
Energy 110, 438451.
Gonzalez Montoya, D., Ramos-Paja, C., Giral, R., 2016. Improved design
of sliding-mode controllers based on the requirements of mppt
techniques. IEEE Trans. Power Electron. 31 (1), 235247.
Hamrouni, N., Jraidi, M., Chrif, A., 2009. Theoretical and experimental
analysis of the behaviour of a photovoltaic pumping system. Solar
Energy 83 (8), 13351344.
Haroun, R., El Aroudi, A., Cid-Pastor, A., Garica, G., Olalla, C.,
Martinez-Salamero, L., 2015. Impedance matching in photovoltaic
systems using cascaded boost converters and sliding-mode control.
IEEE Trans. Power Electron. 30 (6), 31853199.
Kim, I.-S., 2007. Robust maximum power point tracker using sliding
mode controller for the three-phase grid-connected photovoltaic
system. Solar Energy 81 (3), 405414.
Lalili, D., Mellit, A., Lourci, N., Medjahed, B., Boubakir, C., 2013. State
feedback control and variable step size mppt algorithm of three-level
grid-connected photovoltaic inverter. Solar Energy 98 (Part C),
561571.
Lauria, D., Coppola, M., 2014. Design and control of an advanced pv
inverter. Solar Energy 110, 533542.
Levant, A., 2010. Chattering analysis. IEEE Trans. Autom. Control 55
(6), 13801389.
Liu, Y.-H., Chen, J.-H., Huang, J.-W., 2014. Global maximum power
point tracking algorithm for pv systems operating under partially
shaded conditions using the segmentation search method. Solar
Energy 103, 350363.

Lu, D.D., Nguyen, Q.N., 2012. A photovoltaic panel emulator using a

buck-boost dc/dc converter and a low cost micro-controller. Solar
Energy 86 (5), 14771484.
Mavromatidis, G., Orehounig, K., Carmeliet, J., 2015. Evaluation of
photovoltaic integration potential in a village. Solar Energy 121, 152
168.
Mojallizadeh, M., Badamchizadeh, M., 2016. Adaptive passivity-based
control of a photovoltaic/battery hybrid power source via algebraic
parameter identication. IEEE J. Photovoltaics PP (99), 18.
Ngamkong, P., Kochcha, P., Areerak, K., Sujitjorn, S., Areerak, K., 2012.
Applications of the generalized state-space averaging method to
modelling of dc-dc power converters. Math. Comput. Modell. Dynam.
Syst. 18 (3), 243260.
Park, J.-Y., Choi, S.-J., 2015. A novel datasheet-based parameter
extraction method for a single-diode photovoltaic array model. Solar
Energy 122, 12351244.
Rezk, H., Eltamaly, A.M., 2015. A comprehensive comparison of dierent
mppt techniques for photovoltaic systems. Solar Energy 112, 111.
Utkin, V., 1993. Sliding mode control design principles and applications to
electric drives. IEEE Trans. Indust. Electron. 40 (1), 2336.
Utkin, V., 2013. Sliding mode control of dc/dc converters. J. Franklin
Inst. 350 (8), 21462165.
Valencia, P.A.O., Ramos-Paja, C.A., 2015. Sliding-mode controller for
maximum power point tracking in grid-connected photovoltaic
systems. Energies 8 (11), 1236312387.
Wai, R.-J., Shih, L.-C., 2011. Design of voltage tracking control for dc-dc
boost converter via total sliding-mode technique. IEEE Trans. Indust.
Electron. 58 (6), 25022511.
Yau, H.-T., Chen, C.-L., 2012. Fuzzy sliding mode controller design for
maximum power point tracking control of a solar energy system.
Trans. Inst. Measure. Control 34 (5), 557565.
Young, K., Utkin, V., Ozguner, U., 1999. A control engineers guide to
sliding mode control. IEEE Trans. Control Syst. Technol. 7 (3).
Zhang, F., Maddy, J., Premier, G., Guwy, A., 2015. Novel current sensing
photovoltaic maximum power point tracking based on sliding mode
control strategy. Solar Energy 118, 8086.