American Journal of Applied Sciences 7 (8): 1168-1173, 2010

ISSN 1546-9239
2010 Science Publications

Voltage Oriented Control Applied to a Grid Connected Photovoltaic

System with Maximum Power Point Tracking Technique

Rym Marouani and Abdelkader Mami

Laboratory of Analyse and Control of Systems, Department of Electric Engineering,
National School of Engineering of Tunis, PG. 37, Le Belvedere 1002, Tunis, Tunisia

Abstract: Problem statement: This study presented a grid connected photovoltaic system with voltage
oriented control. This system contains the problem of degradation of photovoltaic power with climatic
factors and the problem of synchronization of currents to be injected to the grid. Approach: In order to
optimize the photovoltaic energy generation, we developed a new maximum power point tracking
algorithm and we use a phase locked loop to solve the problem of synchronization of grid voltage and
current. Results: The model of the grid connected photovoltaic system had been implemented in the
MATLAB/SIMULINK software and simulation studies have been presented. Conclusion: Results had
proved the effectiveness of the proposed system, the utility of the maximum power point tracker and
the robustness of the phase locked loop.

Key words: Grid connected photovoltaic system, voltage oriented control, maximum power point
tracking, phase locked loop

INTRODUCTION MATERIALS AND METHODS

The installation of renewable energy power Modeling of the grid connected photovoltaic system:
generation has been one of the new trends in power The configuration of the GCPS presented in this study
electronic technology. can be described by Fig. 1. The system consists of a
Among various renewable energy types, PVG, a DC-bus, an inverter, a grid filter and a grid
photovoltaic power generation systems are expected to utility, as cited in (Kim, 2007; Kim et al., 2009).
play an important role as a clean electricity power
source in meeting future electricity demands.
Modeling of the PVG: The I-V characteristic of the
One of the most important applications is the Grid
PVG is represented by (1), as in (Marouani and Bacha,
Connected Photovoltaic System GCPS whose makes
2009).
the object of this study, where its typical structure is
composed by a photovoltaic generator PVG, a DC-bus,
a PWM inverter and a grid utility. Ip = I ph Iss exp ( Vp VT ) 1 (1)
However, the photovoltaic grid application
presents the problem of achieving the maximum power
With:
point tracking aiming at maximizing the extracted
Ip and Vp = The PV current and voltage respectively
energy irrespective of the irradiation conditions.
Iph = The light-generated current
In this study, we present a new maximum power
point tracking algorithm which calculates the reference
DC voltage versus the solar illumination in order to
force the photovoltaic generator to operate at its
maximum power point.
In our case, efficient simulation before practical
control implantation is required. This study investigates
the utilization of the MATLAB/SIMULINK software
whose enables an efficient control development based Fig. 1: Typical configuration of a grid connected
on specific system libraries or toolboxes. photovoltaic system
Corresponding Author: Rym Marouani, Laboratory of Analyse and Control of Systems, Department of Electric Engineering,
National School of Engineering of Tunis, PG. 37, Le Belvedere 1002, Tunis, Tunisia Tel: 21697693722
1168
 Am. J. Applied Sci., 7 (8): 1168-1173, 2010

Fig. 2: The entire schematic of the studied grid connected photovoltaic system with the voltage oriented control

K1, K2 and K3 are the controller signals applied to

the switches.
The inverters current Iinv is given by (4):

Iinv = K1Ia + K 2 I b + K 3Ic (4)

where, (Ia, Ib, Ic) are the currents to be injected to the

Fig. 3: The real power control grid.

Iss = The reverse saturation current Modelling of the grid utility: Considering the grid
VT = n.KB.T/q = The thermodynamic potential voltages (Ed, Eq) as control inputs, the grid filter
n = The ideality factor of the PN junction currents (Id, Iq) as state variables, the electrical model of
KB = The Boltzmans constant (1.381023 the grid in the d-q referential axis is given by (5), as in
J/K) (Weslati et al., 2008):
T = The temperature of a photovoltaic
array (K) dId R K 1
= Id + Iq + d Vp E d
q = The electron charge (1.61019 C) dt L L L
(5)
dIq = I R I + K q V 1 E
Modeling of the DC bus: The DC bus current Ic is dt d
L q L p L q
given by (2), as in (Marouani and Bacha, 2009):
Where:
Ic = CdVpv / dt (2) R and L = The resistance and the inductance of the
grid filter respectively
Modeling of the inverter: The grid connection inverter = The angular frequency of the grid
consists on a three-phase voltage source inverter with Kd and Kq = The controller signals of the inverter
three independent arms. Each one includes two
switches which are complementary and controlled by Voltage oriented control: The entire schematic of the
the Pulse Width Modulation PWM, as in (Lee et al., Grid Connected Photovoltaic System GCPS with the
2008; Weslati et al., 2008; Yu et al., 2006). Voltage Oriented Control VOC is described by Fig. 2
The inverters voltages (Va, Vb, Vc) are expressed and detailed in (Kim et al., 2009; Ostrem et al., 2006).
in terms of the upper switches as follows:
The power control: This subsystem controls the real
Va 2 1 1 K1
and reactive power of the inverter, (Kim et al., 2009):
V = Vp 1 2 1 K (3)
b 3 2 The real power is controlled by the q-axis current Iq
Vc 1 1 2 K 3
as shown in Fig. 3. A maximum power point
1169
 Am. J. Applied Sci., 7 (8): 1168-1173, 2010

tracking algorithm MPPT to be addressed later,

generates the reference value of the DC-bus
voltage Vdc* to be compared to the actual value Vdc
The error between Vdc* and Vdc is processed as Iq*
through a proportional-integral PI controller
The reactive power is controlled by the d-axis
current Id as shown in Fig. 4. For a unity power Fig. 4: The reactive power control
factor, the desired value of the reactive power Qinv*
is compared to the reactive power Qinv. The error Table 1: Reference DC-bus voltage versus the illumination value
between Qinv* and Qinv is processed as Id* through a E (w m2) Vdc* (V)
100 162.6
PI controller 200 186.2
300 203.0
The maximum power point tracking technique: The 400 216.0
PVG operates at a certain voltage which corresponds to 500 226.5
the maximum power point under different climatic 600 235.3
conditions, i.e., the solar illumination E. In our case, the 700 242.8
800 249.3
MPPT generates the reference DC-bus voltage versus 900 255.0
the solar illumination given by Table 1. This method 1000 260.0
has been detailed in (Marouani and Bacha, 2009; Tarik
Duru, 2006). The current controller compares the reference and
Using the curve fitting technique, the MPPT can be the actual values of the current and generates the
a polynomial function of Vdc* versus E given by (6): desired value of the inverter voltage to be injected to
the grid utility and which are given by (10):
Vdc* = 3.4393 + 0.1328E + 2.2352.104 E 2
(6) Vd* = V cos
1.5460.107 E 3 + 4.2288.1011 E 4 * (10)
Vq = V sin
The current control: The AC-side circuit equations in
the synchronously rotating d-q reference frame are Where:
given by (7), (Kim et al., 2009):

V = Vd + Vq
2 2

Vd = E d LdId dt LIq (11)

= tg1 ( Vd Vq )
(7)
Vq = E q LdIq dt + LId
Where:
Where: Vd and Vq = The active and reactive reference
Ed and Eq = The d- and q-axis components of the grid voltages respectively
side voltage respectively V = The absolute value of the reference
Vd and Vq = Those of the inverter voltage voltage
= The system angular frequency = The reference angle correction given by
To provide direct use of the sensed currents which the VOC system
present intrinsic error, (7) may be transformed to (8) Based on these equations, a current controller may
using PI controllers: be implemented as shown in Fig. 5.
Vd = E d LIq d The phase locked loop: The PLL used in this study is
(8)
Vq = E q + LId q based on zero crossing voltage detection and the
utilization of the quadrate of the input signals, as it is
Where: shown in Fig. 6, as in (Chayawatto et al., 2009; Chung,
2000; Han et al., 2009):
d = ( K p + K i s )( I*d Id )
(9) Ed*= 0: The synchronization signals based on this
q = ( K p + K i s )( I*q Iq ) technique has a simple structure. By keeping Ed*=

0, the synchronous reference voltage component Ed
where, Kp and Ki are the proportional and integral gains is minimized and thereby the PLL will remain
of the PI controllers. locked to the input voltages E and E
1170
 Am. J. Applied Sci., 7 (8): 1168-1173, 2010

Fig. 5: The current controller

H VCO ( s ) = 1 s (14)

RESULTS

A complete Simulink-MATLAB simulation of the

Grid Connected Photovoltaic System with the Voltage
Oriented Control and the Maximum Power Point
Tracking algorithm has been carried out with the
following parameters:

The PVG is composed of 4 series modules. Those

modules are amorphous. Each one presents the
Fig. 6: The phase locked loop following characteristics: Nominal peak power: 60
wp, Nominal voltage: 67 V, Nominal current: 0.9
Using of the quadrate of the input signals: The A, Open-circuit voltage: 92 V and Short-circuit
input signals are multiplied with internal signals as current: 1.19 A
in (12): The DC-bus capacitance: C = 200 103 F
The grid filter: R = 5, L = 0.02 H
E d = E sin + E cos (12) The grid voltage: 220V/50 Hz

DISCUSSION
The PLL generates a rotating reference angle
synchronized to the voltage signals E and E.
When the grid voltage is restarted, the PLL The simulation is based on the variation of the
will effectively synchronize and the system may be solar illumination E as shown in Fig. 7.
reconnected to the grid. So, the reference DC voltage calculated by the
The loop filter LF is a PI controller used in order to MPPT and used by the VOC has the following
give a good trade-off between stability and waveform (Fig. 8).
dynamic performance. Its transfer function is given Following this variation of the solar illumination,
by (13): the current injected to the grid can be described by
Fig. 9 and 10.
In the transient of change of E, we can see the
H LF ( s ) = k p + k i is (13) variation of the grid current in Fig. 10 and 11.
So, we can demonstrate the importance of the
kp and ki are respectively the proportional and the Maximum Power Point Tracking algorithm MPPT in
integral gains of the PI controller. resolution of the problem of the degradation of the
The Voltage Controlled Oscillator VCO is an climatic factors and the utility of the Voltage Oriented
integrator described by (14): Control in the control of the GCPS.
1171
 Am. J. Applied Sci., 7 (8): 1168-1173, 2010

Fig. 7: The solar illumination waveform Fig. 12: The angle calculated by the PLL

Fig. 13: The grid voltage waveform

Fig. 8: The reference DC voltage waveform CONCLUSION

This study is focused on the modelling and

simulation of a grid connected photovoltaic system
controlled by the voltage oriented control.
The MPPT has proved its utility in tracking of the
maximum power point and the optimization of the
photovoltaic power generation.
The phase locked loop used in this study has
Fig. 9: The grid current waveform proved its robustness and effectiveness in
synchronization of the grid voltage and current.

REFERENCES

Chayawatto, N., K. Kirtikara, V. Monyakul, C. Jivacate

and D. Chenvidhya, 2009. DC-AC switching
converter modeling of a PV grid-connected system
under islanding phenomena. Renew. Energy,
34: 2536-2544. DOI: 10.1016/j.renene.2009.05.002
Fig. 10: Zoom of the grid current waveform Chung, S.K., 2000. Phase-locked loop for grid-
connected three-phase power conversion systems.
IEE Proc.-Elect. Power Appli., 147: 213-219. DOI:
10.1049/ip-epa:20000328
Han, Y., L. Xu, M.M. Khan, G. Yao and L.D. Zhou et al.,
2009. A novel synchronization scheme for grid-
connected converters by using adaptative linear
optimal filter based PLL (ALOF-PLL). Simulat.
Model. Pract. Theory, 17: 1299-1345. DOI:
Fig. 11: Variation of the grid current at the transient of 10.1016/j.simpat.2009.05.004
change of E Kim, I.S., 2007. Robust maximum power point tracker
using sliding mode controller for the three-phase
We can see also the angle calculated by the Phase grid-connected photovoltaic system. Solar Energy,
Locked Loop in Fig. 12. 81: 405-414. DOI: 10.1016/j.solener.2006.04.005
We can see also the waveform of the grid voltage Kim, S.K., J.H. Jeon, C.H. Cho, E.S. Kim and J.B. Ahn,
in Fig. 13. 2009. Modeling and simulation of a grid-connected
We can prove here the robustness of the Phase PV generation system for electromagnetic transient
Locked Loop PLL in the synchronization of the grid analysis. Solar Energy, 83: 664-678. DOI:
voltage. 10.1016/j.solener.2008.10.020
1172
 Am. J. Applied Sci., 7 (8): 1168-1173, 2010

Lee, S.H., S.G. Song, S.J. Park, C.J. Moon and Tarik Duru, H., 2006. A maximum power tracking
M.H. Lee, 2008. Grid-connected photovoltaic algorithm based on Impp = f(Pmax) function for
system using current-source inverter. Solar Energy, matching passive and active loads to a photovoltaic
82: 411-419. DOI: 10.1016/j.solener.2007.10.006 generator. Solar Energy, 80: 812-822. DOI:
Marouani, R. and F. Bacha, 2009. A maximum-power- 10.1016/j.solener.2005.05.016
point tracking algorithm applied to a photovoltaic Weslati, Y., A. Sellami, F. Bacha and R. Andoulsi,
water-pumping system. Proceeding of the 8th 2008. Sliding mode control of a photovoltaic grid
International Symposium on Advanced connected system. J. Elect. Syst., 4: 24-38.
Electromechanical Motion Systems and Electric http://www.esrgroups.org/journal/jes/papers/4_3_6.
Drives Joint Symposium, July. 1-3, IEEE Xplore pdf
Press, Lille, pp: 1-6. DOI: Yu, B., Y. Jung, J. So, H. Hwang and G. Yu, 2006. A
10.1109/ELECTROMOTION.2009.5259078 robust anti-islanding method for grid-connected
Ostrem, T., W. Sulkowski, L.E. Norum and C. Wang, photovoltaic inverter. Proceeding of the
2006. Grid connected Photovoltaic (PV) inverter Conference Record of the 2006 IEEE 4th World
with robust Phase-Locked Loop (PLL). Proceeding Conference on Photovoltaic Energy Conversion,
of the IEEE/PES Transmission and Distribution May 2006, IEEE Xplore Press, Waikoloa, HI.,
Conference and Exposition, Aug. 15-18, IEEE pp: 2242-2245. DOI:
Xplore Press, Caracas, pp: 1-7. DOI: 10.1109/WCPEC.2006.279956
10.1109/TDCLA.2006.311434

1173