THD reduction with reactive power compensation for fuzzy logic DVR based solar PV grid connected system

Abstract

Dynamic voltage restorer (DVR) is used to protect sensitive loads from voltage disturbances of the distribution generation (DG) system. In this paper, a new control approach for the 200 kW solar photovoltaic grid connected system with perturb and observe maximum power point tracking (MPPT) technique is implemented. Power quality improvement with comparison is conducted during fault with proportional integral (PI) and artificial intelligence-based fuzzy logic controlled DVR. MPPT tracks the actual variable DC link voltage while deriving the maximum power from a photovoltaic array and maintains DC link voltage constant by changing modulation index of the converter. Simulation results during fault show that the fuzzy logic based DVR scheme demonstrates simultaneous exchange of active and reactive power with less total harmonic distortion (THD) present in voltage source converter (VSC) current and grid current with fast tracking of optimum operating point at unity power factor. Standards (IEEE-519/1547), stipulates that the current with THD greater than 5% cannot be injected into the grid by any distributed generation source. Simulation results and validations of MPPT technique and operation of fuzzy logic controlled DVR demonstrate the effectiveness of the proposed control schemes.

Graphical abstract

Keywords

fuzzy logic / maximum power point tracking (MPPT) / proportional integral (PI) / control / voltage restorer

1 Introduction

Distributed generation (DG) is considered as a promising alternative in reduced cost and improved power quality reliability. Large grid connected solar photovoltaic (PV) power plants are an important developing direction of power generation using solar PV technology. The DG system gets power from utility grid where it is lack of power and supports extra energy to grid when the system runs at light loads. Power quality at the time of transmitting active power can be improved by reactive power compensation and harmonics control function of grid connected inverter. Solar PV array operation is dependent on environmental conditions, and when it is connected to utility grid it can exercise double function of active power generator, and reactive power compensator. The reason for this is that, when there is little or no solar radiation the system can supply reactive power to utility grid, maintaining total apparent power constant on its rated value. This local reactive power supply helps to decrease voltage drops and losses along the distribution supply (grid), and avoids unnecessary overcharges in cables and transformers. Custom power devices like dynamic voltage restorer (DVR), a series active power filter is the latest development of interfacing devices between distribution supply (grid), and consumer. These devices improve power quality by compensating reactive power, and reduce harmonics generated or absorbed by load. This paper is to introduce a model based analytical approach to design a suitable three-phase grid connected inverter for a 200 kW solar PV system with its maximum power point tracking (MPPT) control. The integrated role of DVR with its proportional integral (PI) and dynamic fuzzy logic control schemes are compared, and highlighted with simulation results.

Voltage regulation of a solar PV array (KC200GT model) was discussed by reducing the settling time, losses, and avoids oscillations with overshoot in Ref. [1]. Using average state equations in converter modeling, small-signal models for voltage and current control have been derived, and validated using PSIM simulator. Analysis, and control of a DC-DC buck converter with constant output voltage were explained in Ref. [2]. A linear feedback voltage loop was used to achieve control of variable input voltage. Closed-loop input voltage of experimental, and simulated converter with v_ref= 20 V were compared and analyzed. In Ref. [3], an analysis of PV array modeling was done using gradient-based MPPT for averaged model of power converter. A behavioral model of a grid connected PV system was presented and discussed. Using anti-islanding test it was proved that gradient-based MPPT inverter shuts down when utility was lost at time t = 12 s. In Ref. [4], a comparison was made for various topologies of DC-DC converters connected to DC-AC inverter which created high voltage string based on converter-per-panel approach. Experimental validation of all simulations conducted at 25°C was tested with a boost converter model with a 30 V output voltage, a 75 W output power and a 92%−94% efficiency from 10 W to 60 W. Buck converter efficiency was tested at 96 % over the range of 20−70 W. Simulation of improved incremental conductance MPPT control was highlighted in Ref. [5], through which variable effect of photocurrent was removed at changing solar radiation. Distance between operating point and maximum power operating point was determined on dynamic characteristics by introducing variable gain. In Refs. [6,7], mitigation of voltage sags and swells was discussed; PI controlled DVR provided the most effective solution by establishing the proper voltage quality level. Under unbalanced fault conditions, the behavior of PI and proportional resonant (PR) controllers was addressed in Ref. [8], along with the variation of active and reactive power during single-phase to ground fault. Five control strategies to manage voltage fluctuations (±5%) from nominal voltage arising from a double load were presented in Ref. [9]. A new scheme for total harmonic distortion (THD) reduction using a step variable inductor system at inverter output in conjunction with adjustable hysteresis band control was proposed in Ref. [10]. But it resulted in very high switching frequency of the inverter which might not be feasible from switching devices point of view. All the aforementioned studies show that no artificial intelligence based technique like fuzzy control and neural network, was discussed in surveys which could enhance power quality for grid connected DG systems.

An optimal utilization of a solar PV farm which could operate as a static-compensator (STATCOM) and regulate in the distribution voltage within ± 3% at point of common coupling (PCC) at nighttime was proposed in Ref. [11]. A control approach based on hysteresis voltage and fuzzy logic was presented in Ref. [12], which compensated network faults and mitigated their effects under different loading conditions. The harmonic spectrum performances of open loop (26.05%) converter, proportional-integral-derivative (PID) (7.98%) controlled converter and fuzzy (6.86%) controlled converter were compared in Ref. [13]. The steady-state and transient performances were compared in terms of settling time, overshoot and steady-state error. These parameters were found to be 0.062 s, 0.38 % and 0.002, respectively, with the fuzzy logic controller (FLC). Thus, the performance of the converter was found to be better when the fuzzy controller was considered over the PID controller.

This paper describes the design and simulation of a novel high performance power conditioning system (PCS) for a three-phase grid-connected solar PV system, and its associated control schemes. The perturb and observe MPPT control scheme is applied to control the three-phase voltage source converter (VSC) average model which tracks maximum power from solar PV grid-connected system. The whole simulation model includes two solar PV arrays (100 kW × 2 = 200 kW) whereby each array delivers 100 kW at 1000 W/m² and 25°C standard temperature conditions (STC) using theoretical and empirical equations at variable solar radiation among other variables. Feed-forward current control in synchronous rotating d-q frame is implemented along with the VSC-based control scheme. With application of fault on utility grid side, power quality performance comparison is made between the PI and the fuzzy logic controlled DVR. Both types of controllers with MPPT control technique provide less distortion in maximum active and reactive power from the solar PV array, less fluctuation of DC bus voltage to VSC, and fast tracking of optimum operating point at unity power factor. The simulation of both controllers operating DVR is found quite satisfactory to eliminate unbalanced voltage sag. The fuzzy logic [14] controlled DVR scheme also shows simultaneous exchange of active and reactive power with the distribution system during fault. Finally, the total harmonic distortion (THD) values of grid injected current, load current and VSC current have been calculated using fast fourier transform (FFT) analysis as per IEEE-519/1547 standard, which proves the correctness of implemented control schemes. Validation of the model and algorithms of the control system have been conducted through computer simulations using SimPowersystems [15] of Matlab/Simulink software.

2 Model of solar PV energy conversion system

The model of proposed solar PV energy conversion system shown in Fig. 1, consists of a solar PV array and its PCS for connecting to the electric utility grid. An ideal solar PV cell is modeled electrically with a network containing a light generated current source, an anti-parallel diode representing a nonlinear impedance of P-N junction with series and parallel intrinsic resistances. Fig. 2 depicts the P-V and V-I curves obtained from simulations by using one of the two 100 kW solar PV array (specifications are taken from manufacturer datasheet of PV model Sunpower SPR-305-WHT and are given in Table 1 [15]). As shown, the red dots on blue curves indicate module manufacturer specifications (V_oc, I_sc, V_mp, I_mp) under standard test conditions (25°C, 1000 W/m²). These results show a good agreement for different weather conditions such as at 25°C, and 250/500/750/1000 W/m². This solar PV array exhibits highly nonlinear P-V characteristics, which strongly depends on variable environmental conditions as demonstrated in Fig. 3. Since maximum power point (MPP) changes with variations in solar radiation and operating temperature of a solar cell, the solar PV array has to be continuously operated within the region of MPP for an optimized utilization of the solar PV system.

2.1 Solar PV power conditioning system

PCS is an electronics device which uses electronic interface with MPPT capabilities and connects it effectively to the utility grid. In this way, the solar PV array is capable of simultaneously, and independently performing both instantaneous active and reactive power flow control. Figure 4 illustrates a schematic diagram consisting of two solar PV arrays (100 kW each) connected to the utility grid system (specifications listed in Table 2). Two separate DC-DC boost converters with perturb and observe MPPT control, as given in Ref. [16], track the MPP of each solar PV array, and are connected between the solar PV array and a DC link capacitor C_dc.

A pulse width modulation (PWM) three-phase controlled average model based VSC is synchronized, and connected to the utility grid through a three-phase two-winding transformer. The controlled VSC current I_inv compensates the reactive current of the utility grid depending on the load connected and produces active current corresponding to the solar PV output power. The AC-side terminals of the VSC are connected to the PCC [17] through an interfacing RC-RL filter, comprising of a series resistance-capacitance, and a parallel resistance-inductance (specifications laid down in Table 3), respectively. In this paper, the main grid supplies distribution network. A three-phase series resistance-inductance-capacitance (RLC) load is connected between the main grid and a VSC, where C_pv(F) is the link capacitance between the solar PV array and the DC-DC boost converter; S_b is the power switch for the DC-DC boost converter; S₁−S₄ are the power switches for the VSC; L_f(H) and C_f(F) are the inductance and capacitance of the inductance-capacitance (LC) filter, respectively; and I_inv(A), I_g (A) and I_load (A) is the VSC output current, grid current, and load current, respectively. A single-phase to ground fault is created at the PCC on the grid side during switching period 1 s to 2 s.

2.2 Rotating reference frame transformation

For a three-phase sinusoidal signal, the following equations are obtained after abc_to_dq0 transformation [17],

I_d={\displaystyle \frac{\mathrm{2}}{\mathrm{3}}}\left(I_a\sin (\omega t)+I_b\sin \left(\omega t-{\displaystyle \frac{\mathrm{2}\pi }{\mathrm{3}}}\right)+I_c\sin \left(\omega t+{\displaystyle \frac{\mathrm{2}\pi }{\mathrm{3}}}\right)\right),

I_q={\displaystyle \frac{\mathrm{2}}{\mathrm{3}}}\left(I_a\cos (\omega t)+I_b\cos \left(\omega t-{\displaystyle \frac{\mathrm{2}\pi }{\mathrm{3}}}\right)+I_c\cos \left(\omega t+{\displaystyle \frac{\mathrm{2}\pi }{\mathrm{3}}}\right)\right),

I_{\mathrm{0}}={\displaystyle \frac{\mathrm{2}}{\mathrm{3}}}(I_a+I_b+I_c).

The THD simulation block measures the THD of a periodic distorted signal. The THD has a null value for a pure sinusoidal voltage or current, which is defined as the root mean square (rms) value of the total harmonics of signal, divided by the rms value of its fundamental signal. It is defined in terms of current as

THD={\displaystyle \frac{I_h}{I_f}},

where

I_h=\sqrt{I_{\mathrm{2}}^{\mathrm{2}}+I_{\mathrm{3}}^{\mathrm{2}}+I_{\mathrm{4}}^{\mathrm{2}}+\cdots +I_n^{\mathrm{2}}}(\mathrm{A}),

I_{\mathrm{f}}=\mathrm{r}\mathrm{m}\mathrm{s}\mathrm{v}\mathrm{a}\mathrm{l}\mathrm{u}\mathrm{e}\mathrm{o}\mathrm{f}\mathrm{f}\mathrm{u}\mathrm{n}\mathrm{d}\mathrm{a}\mathrm{m}\mathrm{ent}\mathrm{a}\mathrm{l}\mathrm{c}\mathrm{u}\mathrm{r}\mathrm{r}\mathrm{ent}(\mathrm{A})\mathrm{.}

2.3 Boost converter analysis

The DC-DC boost converter (5 kHz) allows the solar PV voltage to extend from 272 V to 900 V at the maximum power. It produces a chopped output voltage, and therefore controls the average DC voltage relation between its input and output aiming at continuously matching characteristics of a solar PV system to equivalent impedance presented by the DC bus of the VSC. Operating in continuous current mode, the relations for the steady-state voltage and current of the DC-DC boost converter are given by Eqs. (1)−(2),

(1)V_{dc}={\displaystyle \frac{V_a}{\mathrm{1}-D}},

(2)I_{dc}={\eta }_b(\mathrm{1}-D)I_a,

where V_{dc} (V) is the DC link voltage to the VSC, V_a (V) is the solar PV array voltage, D (=0.5) is the DC-DC boost converter duty ratio, I_{dc} (A) is the DC bus current at the VSC side, {\eta }_b is the efficiency of the boost converter, and I_a (A) is the solar PV array output current. The incremental value used to increase and decrease the value of duty ratio is 0.0003. In steady-state continuous conduction mode, the state space Eq. (3), describing the dynamics of the DC-DC converter is given in Ref. [18].

(3)S\left(\begin{array}{c}{I}_a\\ V_{dc}\end{array}\right)=\left(\begin{array}{cc}0& {\displaystyle \frac{-\left(1-D\right)}{L_b}}\\ {\displaystyle \frac{-\left(1-D\right)}{C_{dc}}}& 0\end{array}\right)\left(\begin{array}{c}{I}_a\\ V_{dc}\end{array}\right)+\left(\begin{array}{cc}{\displaystyle \frac{1}{L}}& 0\\ 0& {\displaystyle \frac{1}{C}}\end{array}\right)\left(\begin{array}{c}{V}_a\\ I_{dc}\end{array}\right).

3 Control system techniques

In this paper, each solar PV cell of the two 100 kW solar PV arrays is based on single-diode equivalent electrical circuit. In DG systems, all the available electric power is delivered to the grid. To achieve this, the PV system needs a control system that senses variations in the PV array operating conditions, and leads the system to a new operating point (V_mp and I_mp), called MPPs where the maximum power can be extracted. The MPPT technique is an essential part of a solar PV energy conversion system. The MPPT algorithms described in Ref. [19] are necessary because the PV arrays exhibit nonlinear voltage-current characteristics with a unique point at which the maximum power can be extracted. In Ref. [20], a comparison of different MPPT techniques is presented. The main techniques have been found to be perturb and observe or dithering [21], incremental conductance, constant voltage, fuzzy level control, data-based look up Table [17,21−23], neural network, and ripple correlation factor.

In this category, one of the most efficient and commonly used methods is perturb and observe. This technique is very popular because of its simplicity and ease of implementation. It is implemented in this paper in order that the PV power can be generated efficiently even under changing weather conditions. The proposed technique is a simple approach which can actually calculate direction in which the operating point of the solar PV array is perturbed to reach the MPP. The principle of the perturb and observe controller is to provoke perturbation by increasing or decreasing the PWM duty cycle D value and observe the resulting change in PV output power. If the instantaneous PV output power P_a(k) is greater than the previously computed PV output power P_a(k− 1), the direction of duty cycle perturbation is maintained, otherwise it is reversed.

In Fig. 2 (a) Region I, if {\displaystyle \frac{d{P}_a}{d{V}_a}}>\mathrm{0} and D (k) = D (k− 1) +C, there is an increase in voltage; in Region II, if and D(k) =D (k− 1) −C, there is a decrease in voltage, and C is the incrimination step.

It is expected that using this MPPT technique, maximum output power can be obtained and controlled. Section 3.1 envisages the controlling action for the average model of the VSC using discrete voltage and current controllers. Section 3.2 and Section 3.3 explain operation of the fuzzy logic and the PI controlled dynamic voltage restorer, respectively.

3.1 Controlling action of VSC using voltage and current controllers

In the control scheme shown in Fig. 5 and the research made in Ref. [17], it has been proved that the inverter output voltage can be changed by changing its modulation index [24]. When the output voltage of the inverter is higher than the grid voltage, reactive power is supplied by the PV system to the grid. However, the same gets absorbed when inverse action takes place. It has been proved using the power transfer theory [25], and the further research made in Ref. [17]. Sinusoidal PWM is one of most popular modulation techniques among others applied in power switching inverters. In this technique output voltage (line to line) is obtained in the linear modulation range with the modulation index value varying from 0 to 1, where V_g (V) is the grid voltage; V_dc-m (V) is the measured value of DC voltage; V_dc-r(V) is the reference value of DC voltage; I_g(A) is the grid current; V_dm, V_qm(V) are the measured direct axis and quadrature axis components of the grid voltage, respectively; V_d , V_q(V) are the direct axis and quadrature axis components of the grid voltage, respectively; I_dm, I_qm(A) are the measured direct axis and quadrature axis components of the grid current respectively; I_d-r(A) is the regulated direct axis current from the DC voltage regulator, and m_a is the modulation index [17].

3.2 Controlling action with fuzzy logic controlled dynamic voltage restorer

The basic function of medium voltage DVR is to inject a voltage component of desired amplitude, frequency and phase [8] between the PCC and the grid in series with the utility or load voltage as discussed in the scheme in Ref. [26]. The voltage sag described in Ref. [27] must be detected fast, and corrected with a minimum of false operations. A forced commutated VSC, as shown in Fig. 6, is considered in DVR along with its energy storage device to maintain the voltage across the capacitor. From Eq. (4), three-phase supply voltage is transformed from abc to positive, negative, and zero sequence components [7].

(4)\left(\begin{array}{c}{V}_1\\ V_2\\ V_0\end{array}\right)={\displaystyle \frac{1}{3}}\left(\begin{array}{ccc}1& a& a^2\\ 1& a^2& a\\ 1& 1& 1\end{array}\right)\left(\begin{array}{c}{V}_a\\ V_b\\ V_c\end{array}\right),

The aim of the second control scheme is to compensate for the voltage disturbance in the grid voltage magnitude on occurrence of fault. This scheme also regulates and maintains the DC link voltage to the VSC (with MPPT) constant at a point where a sensitive load is connected with the solar PV grid system under system fault disturbances. Series voltage is injected through a three-phase linear transformer by a VSC connected to the DC power source. FLCs are an active choice when precise mathematical formulations are not accurately possible. Unlike other controllers, FLC is able to tolerate uncertainty and imprecision to a great extent [26,28]. It has two crisp inputs, the first being the difference between the grid voltage and the reference voltage (is taken as 1 per unit) e, and the second being the derivative of error de.

FLC, which is based on the mamdani’s system, consists of three stages [12], as depicted in Fig. 7, the fuzzification, rule execution, and defuzzification. In the first stage, the crisp variables e and de are converted into fuzzy variables E and dE using triangular and trapezoidal membership functions shown in Fig. 8 (a) and (b). Both E and dE are divided into seven fuzzy sets each: NL (negative large), NM (negative medium), NS (negative small), Z (zero), PS (positive small), PM (positive medium) and PL (positive large).

In the second stage of FLC, the fuzzy variables E and dE are processed by an inference engine that executes a set of control rules contained in 49 rule bases. These control rules are formulated using the knowledge of DVR behavior. In this paper, the max-min inference algorithm is used in which final membership degree is equal to maximum of product of membership degree of both E and dE. The output variables from inference engine are converted into crisp values in the defuzzification stage. The centroid defuzzification algorithm is used through which the crisp value is calculated, as exhibited in Fig. 8 (c), using the center of gravity of membership function. The reference voltages for the PWM generator (with carrier frequency f_c = 15 kHz) are FLC crisp output commands.

3.3 Controlling action with PI controlled dynamic voltage restorer

The PI controller generates an angle β which controls the PWM signal generator (with carrier frequency f_c = 15 kHz). Simultaneous exchange of active and reactive power also takes place. The discrete PI (K_p = 5.5, K_i = 1200) controller processes the error signal to generate the required angle to drive this error to zero so that, the grid rms voltage V_g is brought back to reference voltage after application of fault. At fault, grid voltage is measured by a three-phase sequence analyzer which measures the rms voltage while the compensating voltage is injected on the grid side through a three-phase linear transformer. This action also controls the load and VSC voltage current values. The three-phase sequence analyzer outputs the positive, negative, and zero sequence components (magnitude and phase) of a set of three balanced signals. An inductance-capacitance (L = 1 mH, C = 750 µF) filter is used which prevents the current harmonics produced by inverter-based DG system from infiltrating into the distribution network side.

4 Simulation of control schemes and their results

To prove the capabilities of the above mentioned control schemes, a solar PV grid connected test system with its implemented P&O MPPT, DVR with a separate PI and a FLC is modeled with the Matlab/Simulink and SimPowersystems block set [15]. A three-phase series RLC load is connected to the output side. Figures 9 and 10 show the simulation results without the DVR, and a single-phase to ground fault (on phase A) is created at PCC on the grid side during switching period 1 s to 2 s out of the total simulation period of 3 s. From Fig. 9, it is observed evidently that no exchange of active reactive power takes place. Fig. 10 demonstrates the uncompensated 3-Ф discrete output voltage from the VSC and the grid, with the MPPT control, respectively. Fig. 11 depicts the uncompensated 3-Ф discrete output current from the VSC and the grid injected current, with the MPPT control, respectively. The fault and ground resistance values are 0.066 Ω and 0.001 Ω, respectively. To show the effectiveness of the developed perturb and observe MPPT, a comparison between the fuzzy logic and the PI controlled dynamic voltage restorer has been made during fault. The same solar radiation intensity levels shown in Fig. 3 (non-uniform type) have been used for both controllers. The active and reactive output power from the 3-Ф VSC, the connected load, and the grid have been compared in Figs. 12 and Fig. 13, respectively. The compensated 3-Ф discrete output voltage from the VSC, and the 3-Ф grid voltage are shown in Fig. 14 (a), (b), respectively whereas the compensated 3-Ф discrete output current from the VSC and the 3-Фgrid injected current are displayed in Fig. 15 (a), (b), respectively. Figure 16 analyzes the comparison in the actual DC link voltage to the VSC and reference voltage of the VSC control during fault. Table 4 gives the THD values of the VSC current, load current and grid injected current using FFT analysis.

4.1 MPPT technique control and dynamic voltage restorer validation

The solar PV array has been tested through digital simulation and laboratory implementation, with a PC-based emulation model in an IBM AT computer in Ref. [29]. It is interfaced with electrical load system using a data translation DT2821 data acquisition board and the MATLAB software drivers. The dynamic performance of the developed system has been studied using the PI and the fuzzy logic based controllers. Using the PI and the FLCs, it has been proved that the PV array can be operated at its maximum available solar power point at different solar radiations and temperatures values.

The goal is achieved by controlling the array output voltage to the permanent magnet DC motor load using the chopper converter. A current loop and a voltage loop controller have been designed for the grid-connected inverter in Ref. [30]. A feed-forward controller (as per block diagram of Fig. 5) has been tested by an experimental setup which cancels the effect of grid voltage on inverter output current. This setup consists of a 300 V voltage source connected to the input of interface. The chosen gain values for the voltage loop controller are (K_p = 0.0162, K_i = 0.0254) and the current loop controller are (K_p = 0.0094, K_i = 2.9609). It is proved that the current loop controller shapes the inverter output current whereas the voltage loop controller maintains the capacitor voltage and provides a reference inverter output current for the solar PV inverter. The investigation of the application of a fuzzy controller for controlling the DC capacitor voltage and its comparison with the PI controller is proved in Ref. [28]. The results show that the proposed FLC has a faster dynamic response with a high accuracy of tracking the DC voltage even at a sudden change in loads. The proposed perturb and observe MPPT algorithm using a boost converter has been implemented and tested using a digital controller based on dSPACE DSP unit in Ref. [31]. A comparison is made between experimental and theoretical optimal voltages with dynamic performance of the solar PV system. As per Fig. 2 for different weather conditions such as at 25°C, and 250/500/750/1000 W/m², maximum power is achieved when PV voltage is approximately 270 V. A shunt harmonics voltage compensation approach for the DC current harmonic component injected into the grid by the VSC has been presented in Ref. [32]. An active filter is employed to perform the DC voltage compensation to reduce the DC current component flowing in the grid power lines. This component causes the magnetic core saturation of the transformer, increased magnetizing currents with power losses and, consequently, overheating. The DC current component value for the grid current and the VSC current is found to be 6.2 mA and 9.3 mA using the fuzzy-logic control whereas the same value is 6.8 mA and 9.8 mA using the PI controller, respectively (Table 4). This simulation results show the effectiveness of the FLC over the PI controller for proposed compensation approach with reduced value of DC component for grid and controller current.

5 Conclusions

This paper has emphasized an approach of design, and simulation for a three-phase grid-connected 200 kW solar PV system, and its associated control schemes. The control scheme includes an average model based three-phase VSC with its validated and effective perturb and observe MPPT for the solar PV grid connected system. The simulation results from Figs. 12 to 16 show that active power transfer takes place from solar PV arrays through VSC, and reactive power is compensated. In addition, better tracking of the actual DC voltage to the VSC around the reference value of 100 V with perturb and observe MPPT through the FLC and the PI controller is demonstrated. Initially, the whole active power is being supplied by the grid to the load (as shown by in negative direction of active power flow). But, as solar radiation level increases, the PV system partially supplies 3-Ф load through the VSC and the grid. Similarly, initially, the grid supplies the reactive power requirement of the inductive load, however, at around 0.12 s, the reactive power requirement is completed by the PV array through the inverter and continues till 0.2 s. However, after 0.2 s the reactive power requirement of the load is completed by the grid. From Figs. 12 and 13, it is observed evidently that the fuzzy controller and the PI controller have better performance with regard to time in reaching steady values of the active power and reactive power (with compensation), respectively from the VSC, the load connected and the grid. Figure 14 shows the compensated 3-Ф discrete output voltage from the VSC and the 3-Ф grid voltage, using the fuzzy logic controlled DVR whereas Fig. 15 shows the compensated 3-Ф discrete output current from the VSC and the 3-Ф grid injected current, using the fuzzy logic controlled DVR. A closer look at the VSC output current and the grid voltage shows that the output current is of the same shape as the grid voltage and it is in phase with the grid voltage. It clearly reflects the effectiveness of the grid interfacing in tracking the input current, controlling the VSC output current and conveying the power from the solar PV arrays to the grid. Figure 16 (b) demonstrates fuzzy logic controlling tracking of the actual DC link voltage around the reference voltage. It is apparent that the maximum undershoot from the reference voltage is around 18 V as compared to the response using the PI controller given in Fig. 16 (a).

In this work, the main aim was to improve the power quality using the feed-forward current control for an average model based VSC solar PV grid connected with the PI and the fuzzy logic based DVR control system. The controllers based on the PI and the fuzzy logic can provide a more effective and flexible solution for nonlinear systems. As per Table 4, the THD for the VSC current, the load current and the grid injected current with the PI controller is 3.43%, 1.33% and 0.61% whereas the same values are 3.44%, 1.34% and 0.58%, with the fuzzy logic implemented system. The DC current component for the VSC current and the grid current is reduced with the FLC as compared to the PI controller. It is apparent that the proposed system has proved its utility in the optimization of PV power generation and performs power quality control to reduce the THD current as per standards IEEE-519/1547, which stipulates that a current with a THD greater than 5% cannot be injected into the grid by any DG source.

References

[1] Villalva M G, de Siqueira T G, Ruppert E. Voltage regulation of photovoltaic arrays: small-signal analysis and control design. IET Power Electronics, 2010, 3(6): 869−880

[2] Villalva M G, Ruppert E. Dynamic analysis of the input-controlled buck converter fed by a photovoltaic array. Brazilian Journal of Control and Automation, 2008, 19(4): 463−474

[3] Ropp M E, Gonzalez S. Development of a MATLAB/Simulink model of a single-phase grid-connected photovoltaic system. IEEE Transactions on Energy Conversion, 2009, 24(1): 195−202

[4] Walker G R, Sernia P C. Cascaded dc-dc converter connection of photovoltaic modules. IEEE Transactions on Power Electronics, 2004, 19(4): 1130−1139

[5] Kim T Y, Ahn H G, Park S K, Lee Y K. A novel maximum power point tracking control for photovoltaic power system under rapidly changing solar radiation. In: Proceedings of IEEE International Symposium on Industrial Electronics. Pusan, Korea, 2001, 1011−1014

[6] Srisailam C, Sreenivas A. Mitigation of voltage sags/swells by dynamic voltage restorer using PI and fuzzy logic controller. International Journal of Engineering Research and Applications, 2002, 2(4): 1733−1737

[7] Hingorani N G. Introducing custom power. IEEE Spectrum, 1995, 32(6): 41−48

[8] Dash A R, Babu B C, Mohanty K B, Dubey R. Analysis of PI and PR controllers for distributed power generation system under unbalanced grid faults. In: Proceedings of 2011 International Conference on Power and Energy Systems. Chennai, India, 2011, 1−6

[9] Niemi R, Lund P D. Alternative ways for voltage control in smart grids with distributed electricity generation. International Journal of Energy Research, 2011, 36(10): 1032−1043

[10] Wandhare R G, Agarwa V. A novel technique for THD control in grid connected photovoltaic systems using step variable inductor approach. In: Proceedings of 35th IEEE Photovoltaic Specialists Conference. Honolulu, USA, 2010, 844−848

[11] Varma R K, Khadkikar V, Seethapathy R. Nighttime application of PV solar farm as STATCOM to regulate grid voltage. IEEE Transactions on Energy Conversion, 2009, 24(4): 983−985

[12] Ezoji H, Sheikholeslami A, Rezanezhad M, Livani H. A new control method for dynamic voltage restorer with asymmetrical inverter legs based on fuzzy logic controller. Simulation Modelling Practice and Theory, 2010, 18(6): 806−819

[13] Nagarajan C, Madheswaran M. Performance analysis of LCL-T resonant converter with fuzzy/PID controller using state space analysis. Electrical Engineering, 2011, 93(3): 167−178

[14] Raj V, Sudhakaran M, Kumar S S, Roy S R, Palanivelu T G. Multi level inverter based dynamic voltage restorer with PI and fuzzy logic controller. In: 32nd National Systems Conference. IIT Roorkee, India, 2008, 65−70.

[15] The Mathworks. MATLAB/SIMULINK, Inc. 7.10.0.499 (R2010a). 2016-06

[16] Lee S W, Kim J H, Lee S R, Lee B K, Won C Y. A transformer-less grid-connected photovoltaic system with active and reactive power control. In: Proceedings of the 6th IEEE International Power Electronics and Motion Control Conference. Wuhan, China, 2009, 2178−2181

[17] Gupta A, Chanana S, Thakur T. Power quality improvement of solar pv transformer-less grid connected system with maximum power point tracking control. International Journal of Sustainable Energy, 2013, 33(4): 921−936

[18] Molina M G, Juanicó L E, Rinalde G F, Taglialavore E, Gortari S. Design of improved controller for thermoelectric generator used in distributed generation. International Journal of Hydrogen Energy, 2010, 35(11): 5968−5973

[19] Gomathy S, Saravanan S, Thangavel S. Design and implementation of maximum power point tracking (MPPT) algorithm for a standalone PV system. International Journal of Scientific & Engineering Research, 2012, 3(3): 1−7

[20] Esram T, Chapman P L. Comparison of PV array maximum power point tracking techniques. IEEE Transactions on Energy Conversion, 2007, 22(2): 439−449

[21] Mohan N, Undeland T M. Power Electronics: Converters, Applications, and Design. New York: John Wiley & Sons, 1995

[22] Ansari M F, Chatterji S, Iqbal A. A fuzzy logic control scheme for a solar photovoltaic system for a maximum power point tracker. International Journal of Sustainable Energy, 2010, 29(4): 245−255

[23] Coelho R F, Concer F M, Martins D C. A MPPT approach based on temperature measurements applied in PV systems. In: Proceedings of the 9th IEEE/IAS International Conference on Industry Applications. Sao Paulo, Brazil, 2010, 1−6

[24] Molina M G, Juanico L E. Dynamic modeling and control design of advanced PV solar system for distributed generation applications. Journal of Electrical Engineering: Theory and Applications, 2010, 1(3): 141−150

[25] Moin H. Investigation to improve the control and operation of a three-phase PV grid tie inverter. Dissertation for the Doctoral Degree. Dublin Institute of Technology, 2011

[26] Kumar S V R, Nagaraju S S. Simulation of D-STATCOM and DVR in power systems. ARPN Journal of Engineering and Applied Sciences, 2007, 2(3): 7−13

[27] Azim M R, Hoque M A. A fuzzy logic based dynamic voltage restorer for voltage sag and swell mitigation for industrial induction motor loads. International Journal of Computers and Applications, 2011, 30(8): 9−18

[28] Ferdi B, Benachaiba C, Berbaoui B, Dehini R. STATCOM DC-link fuzzy controller for power factor correction. Journal of Acta Electrotechnica, 2011, 52(4): 173−178

[29] Altas I H, Sharaf A M. A fuzzy logic power tracking controller for a photovoltaic energy conversion scheme. Electric Power Systems Research, 1992, 25(3): 227−238

[30] Tsang K M, Chan W L. Three-level grid-connected photovoltaic inverter with maximum power point tracking. Energy Conversion and Management, 2013, 65: 221−227

[31] Mahdi A J, Tang W H, Wu Q H. Improvement of a MPPT algorithm for PV systems and its experimental validation. In: Proceedings of International Conference on Renewable Energies and Power Quality (ICREPQ). Granada, Spain, 2010, 1−6

[32] Menniti D, Pinnarelli A. A novel compensation approach for DC current component in a grid-connected photovoltaic generation system. In: Proceedings of IEEE International Conference on Power and Energy Society General Meeting. San Diego, USA, 2012

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg