International Conference on Renewable Energies and Power

Quality (ICREPQ’09)
European Association for the
Development of Renewable Energies,
Environment and Power Quality Valencia (Spain), 15th to 17th April, 2009

Maximum power point tracker of a photovoltaic system using sliding mode control
M. I. Arteaga Orozco, J. R. Vázquez, P. Salmerón, S. P. Litrán, F. J. Alcántara

Departamento de Ingeniería Eléctrica y Térmica

Escuela Politécnica Superior, Universidad de Huelva
Ctra. de Palos de la Frontera s/n, 21819, Palos de la Frontera, Huelva, Spain
Phone: +34-959217590, Fax: +34-959217304

e-mail: maria.arteaga@die.uhu.es, vazquez@uhu.es, salmeron@uhu.es, salvador@uhu.es, benju@uhu.es

This target of control system is that the PV array will

Abstract. In this paper, a maximum power point tracker maximize the electrical power with a given irradiance.
method using sliding mode control for a photovoltaic system is The control should guarantee that the dc power will be
presented. The system includes a photovoltaic array, a DC/DC transformed with high efficiency to the load.
converter and a DC/AC inverter connected to a load. The
designed control regulates the converter output voltage and it In order to archive the maximum power point (MPP) of
maximizes the power generated by the photovoltaic array. To the PV array, it is necessary maintain it at their optimum
obtain it, the sliding surface used to control the DC/DC point operating. The MPP varies with the solar radiation
converter is adjusted according to PV array output power. The and the temperature. The characteristic curves specify a
control law designed and the results in a simulation platform unique operating point at which maximum possible
will be presented.
power is delivered. At the MPP, the PV system operates
at its highest efficiency.
Key words
Many methods have been developed to determine the
Photovoltaic system boost converter, sliding mode MPP. For example, Ibrahimm and Houssing employed
control, maximum power point. the look-up table on a microcomputer to track MPP,
Midya et al. applied a dynamic MPP tracker to PV
1. Introduction appliances, and Kuo and Liang proposed a single-stage
MPP controller using the slope the power versus voltage,
Nowadays, photovoltaic (PV) systems are used as energy [1-3]. Dual boost converter based MPP tracking using
source in many cases. Most commonly applied PV fuzzy logic has been reported, [4]. Another approach is
systems can be found in remote and rural areas where no based on variable structure control and is applied to a
public grid is available. A typical small photovoltaic buck converter, [5]. Too, there is an approach based on a
power system (off-grid) can contain the following perturbation and observation method, where the reference
components: solar PV array, with a number of voltage varies periodically when the MPP is reached, [6].
series/parallel interconnected solar modules and
protection elements, a DC/DC converter, a DC/AC In this work, a control that regulates the voltage of the
inverter and a control system (Figure 1). DC/DC converter and maximizes the power generated by
the photovoltaic array will be presented. The system
CONTROL
consists of three closed loops, two inner loops and one
SYSTEM outer loop. The DC/DC converter has a control loop
which regulates the output voltage irrespective of the
input voltage variations, and a sliding mode control that
+ + + DC
DC searches the maximum power point of the PV array. The
proposed system has an independent control loop for the
LOAD

PV
Array
DC AC
PWM inverter.
- - -
CONVERTER INVERTER In this control system, it is necessary to measure the PV
array output power and to change the duty cycle of the
DC/DC converter control signal. So, the PV array output
Fig. 1. Basic structure photovoltaic system power is measured and compared to the previous PV
array output power. Depending on the result of the
comparison, the optimal reference current of sliding
mode control is changed and the process is repeated until
the maximum power is reached.

The rest of this paper is organized as follows. Section II

introduces the basic principle of the PV system and a
system description. Section III describes the proposed
sliding controller. Simulation results and conclusions are
finally discussed in the last section.

2. System Description
A. Photovoltaic system
(a)
A real PV array has been modeled. It consists of 30 PV
modules with 36x2 monocrystalline silicon solar cells
each one, connected in series and parallel. Each module
can produce 106 W of DC electrical power with an area
of 126.5 square centimeters. The array is configured as
follows: fifteen modules are connected in series, resulting
a nominal operating voltage of 325 V. Then, 2 of these
series strings are connected in parallel, resulting in a
current of 6A. The rated power of the PV array is 2.6 kW
(DC).

In order to allow the interaction between a DC/DC

converter and the PV array, a simulation model for a PV
array has been developed, including the dependence of (b)
the PV array output with the irradiance and temperature.
Fig. 3. Characteristic of a PV array with constant irradiation
The model was implemented in Simulink, helped by the and temperature of 25ºC: (a) current-voltage graph and (b)
SimPowerSystem blockset (figure 2). power-voltage graph.

In general, PV modules have nonlinear voltage-current

characteristics, and there is only one unique operating
point for a PV with a maximum output power under
particular conditions.

B. Boost converter
Fig. 2. Equivalent circuit of a PV cell
In this section, the state-space model of the boost
This equivalent circuit models the general form the converter that will be used to evaluate the sliding mode
equation that relates current and voltage [7] in a control is presented.
photovoltaic cell:
A boost converter is simply is a particular type of power
⎛ q (V ) converter with an output DC. This type of circuit is used
+ Rs I ⎞ V +R I
⎜ ηKTK ⎟ s to ‘step-up’ a source voltage to a higher, regulated
I = Il − Io ⎜ e − 1⎟ − (1) voltage, allowing one power supply to provide different
⎜ ⎟ Rsh
⎝ ⎠ driving voltages. The basic boost converter circuit
consists of only a switch (typically a transistor), an
where I and V are cell output current and voltage, Io is the inductor, and a capacitor. The specific connections are
cell reverse saturation current; Il is the light-generated shown in figure 4.
current; Rs and Rsh are series and shunt resistance, q is
electronic charge, K is Boltzmann’s constant, and TK is
cell temperature in ºK.

Figs. 3(a) and 3(b) present the current-voltage and

power-voltage output characteristics of a photovoltaic
array for different solar isolation with temperature of
25ºC.
Fig. 4. Boost converter
Applying Kirchhoff’s rules around the loops, it can 3. Control design
obtain the ideal mathematical model of this circuit:
Sliding mode control is one of the effective nonlinear
di robust control approaches since it provides system
L L = V PV − v o (1 − u ) (2)
dt dynamics with an invariance property to uncertainties
dv o v once the system are controlled in the sliding mode [8].
C = i L (1 − u ) − o
dt R (3)
Considering the model of DC/DC converter (2)-(3) the
where iL is the current across the inductor, v o is the following sliding surface S (.) is proposed:
voltage in the capacitor. Parameters R , L and C are
supposed to be known constants. u ∈ {0, 1} defines the u=⎨
⎧1 if S (vo , u , ξ ) > 0
(5)
switch position and V PV is voltage supplied by ⎩0 if S (vo , u , ξ ) < 0
photovoltaic array. S (vo , u, ξ ) = iL − iLREF + K 0ξ (6)
ξ& = v − V , ξ (0) = 0
o e (7)
The gain from the boost converter is directly proportional
to the duty cycle ( D ), or the time the switch is ‘on’ each
cycle. where K 0 is a positive constant and i LREF is reference
vo 1 value of inductor current and Ve is desired output
= (4)
V PV 1 − D voltage (400V).

When boost converter is used in PV applications, the The sliding surface assures that the sliding motion is
voltage input change continuously with atmospheric reached and regulates output voltage boost converter. By
conditions. Therefore, the duty cycle should change to measuring the array voltage and current, the array power
track the maximum power point of photovoltaic array. is calculated and compared to the previous array output
power. Depending this comparison i LREF is re-
This converter should support input voltages in a wide calculated:
range from 100 to 325 V. Under such conditions, the duty
ratio D is adjusted to regulate the output voltage at 400 i LREF [n + 1] = i LREF [n] + Δi LREF (8)
V. For the given range, D is in a range of [0.76-0.20]
and the output current is maximum when D = 0.33 . where i LREF [ n] and Δi LREF are reference current and its
perturbation, respectively. Defining to
The 400 Volts obtained at boost converter are applied to
an IGBT two-level inverter to generate a sinusoidal
ΔPPV
output voltage of 50 Hz. The IGBT inverter uses Pulse Δi LREF = α (9)
Width Modulation (PWM) at a 1050 Hz carrier ΔV PV
frequency. The circuit is discretized at a sample time of ΔPPV = P[n] − P[n − 1] (10)
1 μs . The IGBT inverter is controlled with a PI regulator
ΔV PV = V PV [n] − V PV [n − 1] (11)
in order to maintain to 230 Vrms, 50 Hz at the load
terminals.
where V PV [ n] , I PV [n] and PPV [ n] represent the
Considering this model as starting point, our aim is to voltage, current and power of PV array, respectively, in
find a control law that provides voltage regulation at the which [n] denotes the iteration number; and α is a
converter output, and guarantees the maximum power positive constant.
point of the PV array.

Fig. 5. Photovoltaic array interconnected to AC load

4. Simulations results 5. Conclusions
A Simulink-Matlab simulation of closed-loop PV (figure In this paper, a sliding mode integral control of the boost
5) has been carried out when the load is 0.5 kW. converter has been analyzed. The reported controller uses
the output power of the PV array, output voltage of the
250 converter and an input signal, which in this case is the
Output voltage
3000 switching signal. The control law provides voltage

Output Power (W) and Irradiance solar (W/m2)

regulation at the converter output, and guarantees the
Current and voltage photovoltaic array

200 2500
maximum power point of the PV array.
Output Power
2000
150 A practical case developed in a Matlab/Simulink
1500
simulation platform has been presented, and the results
100 confirm the adequate performance of designed control.
1000 So, the dynamical response of the control is around one
Irradiance
period of the voltage supply.
50
500

Output current Besides, the control law can be easily implemented by

0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
0
0.16 means of standard operational amplifiers, analog
Time (seg) multipliers and digital devices in an experimental
Fig. 6. Evolution of V PV , I PV and PPV to difference platform.
irradiances solar
References
Figure 6 shows the evolution of the current, voltage and
power PV array for 1000 W/m2 and for successive [1] Ibrahim, H. E.-S. A. and Houssiny, F. F., “Microcomputer
irradiance steps applied at t=0.6 seg (800W/m2) and t=1.2 Controlled Buck Regulator for Maximum Power Point
seg (600 W/m2). It can be seen than the system reaches Tracker for DC Pumping System Operates from
Photovoltaic System,” Proceedings of the IEEE
the maximum power point in each step as it can be
International Fuzzy Systems Conference, August 22-25,
verified from figure 3. Vol. 1, pp. 406-411 (1999).
[2] Midya, P., Kerin, P. T., Turnbull, R. J., Reppa, R. And
Figure 7 shows the load voltage for the irradiance step Kimball, J., “Dynamic Maximum Power Point Tracker for
and figure 8 shows the evolution of the modulation index. Photovoltaic Applications,” Proceedings of the IEEE
The time response is nearly 10 ms for a final output of Power Electronics Specialists Conference, PESC, Vol. 2,
230 Vrms. pp. 1710-1716 (1996).
[3] Kuo, Y. C., Liang, T. J. and Chen, F. C., “Novel
Maximum-Power-Point-Tracking Controller for
400 Photovoltaic Energy Conversion System”, IEEE
Transactions on Industrial Electronics, Vol. 48, pp. 594-
1000 601 (2001).
[4] Veerachary, M., Senjyu, T., and Uezato, K. “Feedforward
Irradiance step variations (W/m2)

200
900 maximum power point tracking of PV systems using fuzzy
controller”. IEEE Transactions on Aerospace and
Voltage Load (V)

800 Electronic Systems,Vol. 38, 3 (July 2002), 969-981.

0 [5] Miao, Z., W. Jie, et al., “The application of slide
700 technology in PV maximum power point tracking system”,
Fifth World Congress on Intelligent Control and
600
-200 Automation, WCICA, 2004.
[6] Koutroulis, E. And Voulgaris N. C., “Development of a
Microcontroller-based photovoltaic Maximum power point
tracking control system”, IEEE Transactions on Power
-400
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 Electronic, Vol. 16, No. 1, January 2001.
Time (s) [7] Gow, J. A. and C. D. Manning, “ Development of a model
Fig. 7. Evolution of load voltage to difference irradiances solar for photovoltaic arrays suitable for use in simulation
studies of solar energy conversion systems”, Sixth
International Conference on Power Electronics and
Variable Speed Drives, 1996.
[8] V. Utkin, J. Guldner, and J. Shi, “Sliding Modes in
Electromechanical Systems”, London, U.K.: Taylor &
Francis, 1999.
[9] R.D. Middlebrook and Slobodan Cuk, “A General Unified
Approach to Modelling Switching-Converter Power
Stages”, IEEE PESC’76 Rec., pp. 18-34, Cleveland, OH,
June 8-10.

Fig. 8. Evolution of index modulation