Investigation of the behavior of a three phase gridconnected photovoltaic system to control active and reactive power with DPC

Transcription

1 Available online at Energy Procedia 6 (2011) MEDGREEN 2011-LB Investigation of the behavior of a three phase gridconnected photovoltaic system to control active and reactive power with DPC I.Hamzaoui a, F.Bouchafaa a*, A.Hadjammar a a Laboratory of Instrumentation, Faculty of Electronics and Computer, University of Sciences and Technology Houari Boumediene, BP 32 El-Alia Bab-Ezzouar Algiers, Algeria. Abstract This paper proposes an extended direct power control (DPC) for a three-phase PWM multilevel inverter fed by photovoltaic generation system (PVGS) and connected of a grid. The model contains a detailed representation of the main components of the system that are the solar array, and the grid side inverter multilevel inverter NPC VSI. In order to extract the maximum amount of from the photovoltaic generator, we propose an intelligent control method (fuzzy logic controller) for the maximum power point tracking (MPPT) of a PVGS. The DPC approach for multilevel inverter NPC which makes it possible to achieve unity power factor (UPF) operation by directly controlling its instantaneous active and reactive power. The other is a fuzzy logic controller, in the multi-dc-bus voltage control loop, developed to provide active power command. To achieve UPF operation, the reactive power command is set to zero. It is shown via simulation results that the proposed DPC has high performance. Moreover, the controller multi DC bus link voltages based on fuzzy logic has excellent performance in transient and steady states, a good robustness, a good dynamic behaviour response, and a good rejection of impact of load disturbance Published by Elsevier Ltd. Open access under CC BY-NC-ND license. Selection and/or peer-review under responsibility of [name organizer] Keyword: Direct power control, Fuzzy logic control, MPPT, Photovoltaic, Multilevel inverter NPC, multi DC bus voltage. * Corresponding author. Tel.: (213) ; fax: 213 (0) address: fbouchafa@gmail.com Published by Elsevier Ltd. Open access under CC BY-NC-ND license. doi: /j.egypro

2 494 I.Hamzaoui et al. / Energy Procedia 6 (2011) Introduction In recent years, the efforts to spread the use of renewable energy resources instead of pollutant fossil fuels and other forms have increased. PVGS are increasing in size as they become more affordable and supporting schemes start to include larger installations. In a near future, PVGS are going to be very common, and it is expected that they will contribute with a significant share to power generation. One of the most common control strategies structures applied to decentralized power generator is based on Power direct control employing a controller for the multi DC link voltage and a controller to regulate the injected current to the utility network. Control method is the key issue for the PWM inverters development. FLC regulator is adopted in the outer voltage control loop for maintain the DC output voltage. The currents drawn from the power system should be sinusoidal and in phase with respective phase voltages to satisfy required power factor. Many strategies are proposed to make currents sinusoidal [1]. As part of our work, we will focus on voltage inverter at three levels to NPC structure. The latter can increase the voltage supplied to the load through their topology. Thus, they can generate more voltage sinusoidal possible and improve the total harmonic distortion through the high voltage levels provided by the structure of this new converter. The system components and power control scheme were modelled in terms of dynamic behaviours. The proposed models were implemented in Matlab/Simulink. This paper describes the dynamic performance of the PVGS connected through inverters to distribution network. In the last, authors propose a DPC approach based to calculate active and reactive powers. The multi DC capacitor voltage is regulated and permits to obtain the reference of active power. Fuzzy logic control (FLC) is studied in order to give better performances in time response and system steadiness. This solution is simple and owns dynamics and robustness performances. 2. System configuration Figure 1 show the configuration of the grid-connected PV system, which consists of two solar cell arrays and the three phase multilevel inverter NPC VSI. The control structure of the grid-connected PV system is composed of two structure control: 1. The MPPT Control, 2. The DPC Control the active and regulate the reactive power injected into the grid; - Control the multi DC bus voltage; - Ensure high quality of the injected power. i PV1 U PV1 ic 1 C1 U c1 i d1 S11 S 21 i d0 S 12 S 22 S 13 S 23 R L i net1 O U PV2 ic 2 C2 U c2 S 31 S 32 S 33 V A V B V C vnet 1 i net2 vnet 2 vnet 3 N i PV2 i d2 S 41 S 42 S 43 S 43 Fig.1. General diagram of grid connected photovoltaic system

3 I.Hamzaoui et al. / Energy Procedia 6 (2011) Electrical model of photovoltaic cell Many mathematical models have been developed to represent their highly nonlinear resulting from of semiconductor junctions. We will present our work the model with two diodes (Fig.2) [1,2,3]. i d1 i d2 I Rs V I ph Rsh V I Fig.2. Model of a photovoltaic cell with two diodes The current generated by the module is given by the following equation: I IphId1Id2Ish q s I I ph IS1 exp A1 KT VR I qvr I 1 I S2 s exp A2 KT VRs I 1 Rsh Where V and I represent the output voltage and current of the PV, respectively; R s the series parasitic resistance of a solar array and R sh its shunt parasitic resistance; q is the electronic charge; I ph corresponds to the light-generated current of the solar array. I S1, 2 represent the current saturation of the two diodes; A 1, 2 is ideality factor of the junction of D 1 and D 2, K the Boltzmann s constant, T the cell temperature. The characteristics of a PV cell of changes in current and power based on the voltage of the PV cell is shown in figure 3. (1) Fig.3. PV array I V and P V characteristics. 4. Fuzzy Logic MPP Tracking controller The maximum power that can be delivered by a PV panel depends greatly on the insulation level and the operating temperature. Therefore, it is necessary to track the maximum power point all the time. Recently fuzzy logic controllers have been introduced in the tracking of the MPP in PV systems; they have the advantage to be robust and relatively simple to design as they do not require the knowledge of the exact model.

4 496 I.Hamzaoui et al. / Energy Procedia 6 (2011) They do require in the other hand the complete knowledge of the operation of the PV system by the designer. The five linguistic variables used are: NB (Negative Big), NS (Negative Small), ZE (Zero Approximately), PS (Positive Small), PB (Positive Big) [4]. The two FLC input variables are the error E and change of error CE at sampled times k defined by [3]: P(k) P(k 1) E(k) V(k) V(k 1) CE(k) E(k) E(k 1) Where P(k) is the instantaneous power of the photovoltaic generator. The input E(k) shows if the load operation point at the instant k is located on the left or on the right of the maximum power point on the PV characteristic, while the input CE(k) expresses the moving direction of this point. The fuzzy inference is carried out by using Madani s method, (Tab.1), and the defuzzification uses the centre of gravity to compute the output of this FLC which is the duty cycle [3,4]: (2) n j1 n j1 j j j (3) The control rules are indicated in Table 1 with E and CE as inputs and d as the output. Table1. Fuzzy rule table CE E NG NP ZE PP PG NG ZE ZE PG PG PG NP ZE ZE PP PP PP ZE PP ZE ZE ZE NP PP NP NP NP ZE ZE PG NG NG NG ZE ZE Obviously, it can be deduced that the fuzzy controller is fast controller in the transitional state and presents also a much smoother signal with less fluctuations in steady state. A fast and steady fuzzy logic MPPT controller was obtained. It makes it possible indeed to find the point of maximum power in a shorter time runs (Fig.4) [3].

5 I.Hamzaoui et al. / Energy Procedia 6 (2011) Fig.4. PV characteristic MPPT under standard climatic conditions. 5. Mathematic model of the Three-level NPCVSI The basic topology of neutral-point-clamped three-level PWM inverter is shown in figure 1. There are four switches S 11 to S 41 and two clamping diodes D D11 and D D10 through each bridge. V A, V B and V C are phase voltages; i net1, i net2 and i net3 are the line currents of three phase grid. L net and R net represent inductance and resistance of interconnecting reactors respectively. The value of capacitors connected in series in DC side is C, the voltages of which are U c1 and U c2. The voltage, V MN, represents the potential between neutral point of the output voltage and the midline of the grid In order to deduce the knowledge model of the inverter, we introduce the connection function S is of the switch which describes the state of every switch (1=closed, 0=opened). In this function i is the number of the of the commutation cell, i{1,2,3} s: is the number of the semi-conductor [5]. Different switching states of the three-level NPC VSI are shown in Table 2. As seen in this table, in order to prevent a capacitor leg short circuit, all switches in a leg are never turned on simultaneously. According to table 2, the equivalent main topology is obtained by replacing every leg with a one-knife three-state switch, just as shown in figure 5 [5,6]. We can define three switching functions C a1~3, C b1~3, Cc 1~3 farther more, as to develop the mathematic model of total system. Table 2. Different switching sates in three level NPC VSI C K S i1 S i2 S i3 S i Each phase in the NPC inverter can produce three distinct levels by connecting the output either to the positive (U c1 ), negative (U c2 ) or null (0) potential. In a three-phase system it results in 3 3 =27 output voltage states, some of them apply the same voltage vector [7]. There are two possible configurations for each small vector and three for the zero vectors. Therefore, 19 different vectors V 0 to V 18 are available as shown in figure 6. The switching space vectors can be classified into four categories according to its magnitude, namely: zero space vector V 0 which corresponds to three configurations that produces null output voltage; small vectors (V 1, V 2, V 3, V 4, V 5 and V 6 ) that create a space vector with amplitude equal to U PV /3; medium vectors (V 7, V 8, V 9, V 10, V 11 and V 12 ) with an amplitude of U PV / 3 and the large voltage vectors (V 13, V 14, V 15, V 16, V 17 and V 18 ) that generate a space vector with amplitude equal to 2.U PV /3.

6 498 I.Hamzaoui et al. / Energy Procedia 6 (2011) U c1 +1 C a i net1 (R-L) net V net1 V V V 14 U PV M U c2 0-1 C b i net2 i net3 C c V net2 V net3 N V V 10 V V 3 V 2 V 4 V V V6 V V V V 7 V V Fig.5. Equivalent switching structure for the NPC inverter Fig.6. Space vector diagram of three-level inverter 6. Extended direct power control 6.1. Fuzzy logic controls of multi DC bus link voltages To remedy to the problem of the instability of the output multi DC voltage of the PWM inverter [8,9], we propose to enslave it by using a PI type fuzzy logic regulator [10,11]. In order to maintain multi DC bus link voltage U pv, respective of reference U cref and load variations. The inputs to the controller are the DC voltage error, e(k) and the change of voltage error, e(k). The output of the fuzzy controller is change reference current ic ref (k) [12] The principle of direct power control The basic principle of the Direct Power Control (DPC) was proposed by Noguchi [13] and is derived from the well-known Direct Torque Control (DTC) for induction machines. The DPC is based on the instantaneous power theory, which was developed for 3-phase PWM rectifiers. Unlike two loops system of voltage and current, DPC regulates active and reactive part of the instantaneous apparent power directly, using relay control without current controller [14]. The relay control can be performed by selecting an optimum switching state of the converter, so that the active and reactive power errors can be restricted in appropriate hysteresis bands, which is possible by using a switching table and several hysteresis comparators. Two important aspects must be considered to guarantee the correct operation of the system: Correct selection of the switching states; Measurement and calculate of the active and reactive power. Utilizing the line voltages and currents measurement, active and reactive power component can be estimated. After researching the impact of every switch vector on the inverter s instantaneous power, the switching table, the most important part of DPC, is carried out. The DPC strategies result in the switching frequency varying with the variations of active and reactive power, network conditions, and hysteresis band width of the power controllers [15]. As shown in figure 7, the active power command, P ref, is provided from a DC-bus voltage controller block. The unity power factor operation is achieved by controlling the reactive power q ref to be zero [16]. Errors between the commands and the estimated feedback power are input to the hysteresis comparators and digitized to the signals S p and S q. Also, the phase of the power-source voltage vector is converted to n.

7 V 18 V 23 V 24 V 4 V 11 V 17 V 19 V V 3 10 V 22 V 5 V 12 V 13 V 6 V 25 V 16 V 1 V 8 V20 V 21 V 26 V 15 I.Hamzaoui et al. / Energy Procedia 6 (2011) i PV1 R-L i i net1 V net1 UC1 i d1 S 11 S 12 S 13 M S 21 i d0 S 22 S 23 V A V B i net2 V net2 N UC2 S 31 S 32 S 33 N V C i net3 Vnet3 i d2 S 41 S 42 S 43 i PV2 ipv1 ipv2 Uc2 Uc1 S i2 S i1 S i3 S i4 i net1 i net2 Uc ref + - MPPT FLC Uc2 Uc1 + + UPV FLC icref voltage corrector DC Bus control P PV - -+ P ref Switching State Power and Voltage Selector Estimator bloc S P S q n Sector V 2 V9 V V P est Arctang q est q ref Fig.7. Diagram of Extended direct power control system It is known that the calculation of the active power is a scalar product between the voltages and the currents, whereas the reactive power can be calculated by a vector product between them [17]. We can rewrite in (,)as (4): q P V 1 V 3 net2.i net1 net1 V V i.i net2 net2 V net3 net1 V net3.i net3 net3 V i net1 net2 V net1 V i net2 net3 P V.i V.i q V.i V.i (4) With (4), the instantaneous value of active and reactive power can be calculated easily by measuring the phase voltages and line currents of the grid. The relay controllers are introduced to regular the error of power inverter drawn from the network. As to utilize the complexity of the switching vectors in three-level inverter, multi-level hysteresis comparators are proposed in DPC controller. Active and reactive power error has its own hysteresis comparator respectively. S p and S q are the output of hysteresis comparators, which are defined by judgment in (5): P H H P H P2 P1 SP Sq P1 P1 P2 H P H P H P q H q q H q q H Where, H p1, H p2 and H q the error between the commands and the estimated feedback power. q (5)

8 500 I.Hamzaoui et al. / Energy Procedia 6 (2011) The phase of grid voltage vector should be identified before selecting the appropriate switching state. For digitizing the phase angular, the stationary coordinates are divided into 12 sectors, as shown in figure n 6.3. Simulation result In order to verify the effectiveness of the direct power control (DPC) for a three-phase PWM multilevel inverter fed by photovoltaic generation system (PVGS) and connected of a grid. Simulations have been performed using Matlab/Simulink software environment. We use the algorithm enslavement elaborated previously (Fig.7) to control the active and reactive power of the cascade we applied load variation between two instants t=1s and t=3s for active power and t=2s to t=4s for reactive power. Figures 8, shows the simulation results when we use a feedback control with PI fuzzy logic regulator to stabilize multi DC link voltage, fuzzy logic MPPT for photovoltaic systems and direct power control for active and reactive power. We show the performances of the FLC of the input voltage of the three-level PWM inverter (Fig.8). We note that, the input voltage of PWM inverter follows perfectly its reference which is constant and are no effect for the load variation (Fig.8). We observe that the differences between the input voltages of the three-level NPC inverter are decreased to have a value null in steady state, so the different input DC voltages of the three-level NPC VSI are constant and practically equal too. (U C1 =U C2 ) after a transient state and insensible to any perturbation (Fig.8). In consequence the output voltage of the three-level NPC VSI is symmetrical. We show the active power and the reactive power follows perfectly theses references respectively and the current i d and i q are practically proportional then P and Q respectively. This enabled us to obtain high dynamic performances of the controllers and decoupling. The currents i d1 is the opposite of the current. The inverter current i d0 has a mean value practically null. The line currents are very close to sinusoidal waveforms, a good regulation of the multi DC link voltage is obtained, and UPF operation is achieved. We remark that the network voltage and current are in phases then the power factor of network is unit. 7. Conclusion In this paper, we have studied an approach of modelling and control of a grid connected photovoltaic system with using a conventional direct power control for NPC three-level inverter. In this work, the aim was to control the voltage of the solar panel in order to obtain the maximum power possible from a PV generator, whatever the solar insulation and temperature conditions. Since quite a few control scheme had already been used and had shown some defects, it was necessary to find and try some other methods to optimize the output, fuzzy logic controller seemed to be a good idea. The controllers by fuzzy logic can provide an order more effective than the traditional controllers for the nonlinear systems, because there is more flexibility. The steady-state and dynamic results illustrating the operation and performance of the proposed control scheme are presented. As a result, it was confirmed that the fuzzy logic controller gives excellent performance to regulate of multi DC bus link voltage. Moreover in transient state, a good rejection of impact load disturbance, and a good robustness. Exploiting the proposed DPC, the ride-through capability of the three-level PWM inverter is improved significantly. The simulation results are consistent with the theoretical analysis and verify the excellent performances of the proposed scheme.

9 I.Hamzaoui et al. / Energy Procedia 6 (2011) f j gy ( ) The results obtained with this solution confirm the good performances of the proposed solution and are full of promises to use this systems in high voltage and great power applications as electrical power applied to decentralized power generator is based on Power direct control. 8. References [1] D.Y.Lee and al. "An Improved MPPT converter using current Compensation Method for Scaled PV- Applications", /03/$17.00(C) 2003, IEEE. PP [2] M. Azab, "A New Maximum Power Point Tracking for Photovoltaic Systems", Procedings of World Academy of Science, Engineering and Technology Volume 34 October 2008 ISSN [3] T. Esran, P.I. Chapman, "Comparaison of photovoltaic array Maximum Power Point Tracking Techniques", IEEE Transactions of Energy Conversion [4] N. Patcharaprakitia, and al. "Maximum power point tracking using adaptive fuzzy logic control for grid-connected photovoltaic system", in IEEE Power Eng. Society Winter Meeting, 2002, PP [5] Lie Xu, Dawei Zhi, Liang zhong Yao. "Direct Power Control of Grid Connected Voltage Source Converters", IEEE-PES Conf., Jun. 2007, pp.1-6. [6] T. Lu, Z. M. Zhao, L. Q. Yuan, and S. P. Wang, "Instantaneous energy balancing in three-level neutral point clamped converters", in Proc. IEEE VPPC 08, Harbin, China, 2008, pp [7] L. A. Serpa, J. W. Kolar. "Virtual-Flux Direct Power Control for Mains Connected Three-Level NPC Inverter Systems", Power Conversion Conf. Apr. 2007, pp [8] J.Rodriguez and al. "Calculation of the DC-bus Capacitors of the Back-to-back NPC Converters", EPE-EPE-PEMC 2006, Portoroz, Slovenia, PP [9] Zadeh L.A."Fuzzy sets. In" Inform. Contr.,Vol.8, PP , [10] Mazumder, S. K."A Novel Discrete Control Strategy for Independent Stabilization of Parallel Three- Phase Boost Converters by Combining Space-Vector Modulation with Variable-Structure Control", IEEE Trans. on Power Electronics, July 2003, Vol.18, No.4, pp [11] N.Hur, J.Jung, K.Nam, "A Fast Dynamic DC-Link Power-Balance Scheme for a PWM Converter- Inverter System", IEEE Tansactions on industrial electronics.vol.n.4, August Energy Conversion and Management 50 (2009) pp [13] T. Noguchi, H. Tomiki, and al. "Direct Power Control of PWM Converter without Power-Source Voltage Sensors", IEEE Transaction on Industry Applications, Vol.34, No.3, May/June 1998, pp [14] H. Akagi, Y. Kanazawa, A. Nabae, "Generalized Theory of The Instantaneous Reactive Power in Three-Phase Circuits", IPEC, Tokyo'83, pp [15] T. Lu, Z. M. Zhao, and al. "General-purpose control platform based on dual DSPs for power electronic converters, Journal of Tsinghua University (Science and Technology), Vol.48, N.10, pp , October [16] T. Lu, Z. M. Zhao, and al. "A novel direct power control strategy for three-level PWM rectifier based on fixed synthesizing vectors", in Proc. ICEMS 08, Wuhan, China, 2008, pp [17] Restrepo and al. "Algorithm evaluation for the optimal selection of the space vector voltage using DPC in power systems", Proceedings of IEEE European Conference on Power Electronics and Applications, pp.1-9, September [18] Wang Jiu-he, and al. "Direct Power Control System of Three Phase Boost Type PWM Rectifiers", Proceeding of the CSEE, Vol. 26, pp , Sep

10 502 I.Hamzaoui et al. / Energy Procedia 6 (2011) , f, j gy ( ) Fig.8. Performances an extended DPC for a three-phase multilevel inverter fed by PVGS connected of a grid