https://doi.org/10.6113/JPE.2018.18.6.1805
ISSN(Print): 1598-2092 / ISSN(Online): 2093-4718
Humpback Whale Assisted Hybrid Maximum Power Point Tracking Algorithm for Partially Shaded Solar Photovoltaic Systems
Manoharan Premkumar† and Rameshkumar Sumithira*
†Department of Electrical and Electronics Engineering, GMR Institute of Technology, Rajam, India
*Department of Electrical and Electronics Engineering, Government College of Engineering, Salem, India
Abstract
This paper proposes a novel hybrid maximum power point tracking (MPPT) algorithm combining a Whale Optimization Algorithm (WOA) and the conventional Perturb & Observation (P&O) to track/extract the highest amount of power from a solar photovoltaic (SPV) system working under partial shading conditions (PSCs). The proposed hybrid algorithm is based on a WOA which predicts the initial global peak (GP) and is followed by P&O in the final stage to achieve a quicker convergence to a GP. Thus, this hybrid algorithm overcomes the computational burden encountered in a standalone WOA, grey wolf optimization (GWO) and hybrid GWO reported in the literature. The conventional algorithm searches for the maximum power point (MPP) in the predicted region by the WOA. The proposed MPPT technique is modelled and simulated using MATLAB/Simulink for simulating an environment to check its effectiveness in accurately tracking the MPP during the GP region. This hybrid algorithm is compared with a standalone WOA, GWO and hybrid GWO. From the simulating results, it is shown that the proposed algorithm offers high tracking performance and that it increases the output power level of a SPV system under partial shading. The algorithm also verified experimentally on various PSCs.
Key words: GWO, Hybrid GWO, Hybrid WOA, P&O, Partial shading, WOA
Manuscript received May 2, 2018; accepted Jul. 15, 2018
Recommended for publication by Associate Editor Xiaoqiang Guo.
†Corresponding Author: mprem.me@gmail.com Tel: +91-9500390495, GMR Institute of Technology
*Dept. of Electr. & Electron. Eng., Government College of Eng., India
Ⅰ. INTRODUCTION
The use of renewable energy has increased rapidly due to a reduction in conventional resources, the high cost of fossil fuels, and environmental issues. SPV is one of the most important renewable resources. However, it faces many challenges when compared with conventional resources due to its higher cost and lower efficiency. The important thing is that PV strings are non-linear [1].
The power output of a SPV string depends on the irradiation, temperature and load on the SPV module. When solar irradiation is uniform at a constant temperature, the PV power output is proportional to the voltage. The techniques used to find the optimum voltage that results in the maximum power are called MPPT. Various MPPT algorithms have been proposed by a number of researchers to increase the performance of SPV systems. These algorithms include but are not limited to P&O, incremental conductance (IC), hill climbing (HC), fractional open circuit voltage (FOCV) and fractional short-circuit current (FSCC) [2]. PV characteristics under a constant irradiance are shown in Fig. 1. When a SPV system is subjected to partial shading, the conventional algorithms fail to track the MPP. In a PV system, partial shading on a panel that reduces the performance of the SPV system results in multiple MPPs with many local peaks and one GP. Thus, locating the GP presents a serious challenge for selecting a proper technique for a SPV system.
|
(a) |
|
(b) |
Basically, a SPV system includes PV modules, a DC-DC converter and a load. The main purpose of the converter is to extract the maximum power from the PV modules using a proper MPPT technique and to deliver this power to a load. To match the load demand, a number of panels are connected in parallel/series combinations. A few of the panels are shaded due to cloud passage or adjacent trees as shown in Fig. 2. A certain amount of power is consumed by the shaded panels, and this power has to be generated by the unshaded panels. This consumption of power generates a hot spot that can harm the shaded panel.
Fig. 2. PV modules under partial shading.
To improve the efficiency of SPV systems, numerous MPPT techniques have been proposed as discussed earlier. In the HC method, a perturbation is used as the duty cycle of the converter [3]. In the P&O method, a perturbation is used as the operating voltage of the PV module. Since the perturbation is continuously changed, the above algorithm yields an oscillation at the MPP which results in power loss. To rectify the problems of the HC and P&O methods, the IC algorithm was proposed. However, the oscillations are not completely reduced. These algorithms fail to harvest power from panels when subjected to rapid changes in irradiation and partial shading. The focus of this research is to locate the GPs during partial shading. In order to improve the MPPT tracking by modulating the duty ratio, an improved IC algorithm is proposed in [4], [5].
The scanning based MPPT was proposed in [6], [7] to introduce a dynamic controller under PSCs and rapid changes in insolation to harvest the maximum power. An alternative approach called the evolutionary algorithm (EA) technique was proposed in [8], [9]. It is capable of handling non-linear functions. Metaheuristics methodologies such as particle swarm optimization (PSO) [10], [11], GWO [12], Firefly [13], ant colony optimization (ACO) [14], artificial bee colony (ABC) [15], [16], etc. have been used for numerous applications and achieve faster convergence towards a GP under PSCs. A sensorless MPPT algorithm was proposed in [17], which shows less oscillation in the output power around the MPP. The ACO algorithm-based technique is found to track the global peak with a reduced tracking time and a lower computation burden. A GWO based tracking method was proposed in [12], which is able to track the maximum power under the PSC. The convergence time of the GWO is higher than the GWO combined with P&O proposed by [18].
In all of the recent MPPT techniques, the variables such as the duty cycle, voltage, and current are considered as a population to find a better solution (maximized power) under the PSC, where the objective function is considered as the output power of PV modules. In the literature, the PSO method is often used to track the GP under the PSC. However, PSO based techniques exhibit a high settling time, power oscillations and a high exploration region. To improve the convergence time and extract the maximum power from PV modules, it is necessary to find an alternate method. Recently, the authors of [19] proposed WOA, which is inspired by humpback whales attacking prey for hunting. In [20], a WOA based MPPT technique was proposed. However, it exhibited more power oscillations when the algorithm tracks the MPP around the GP.
After a thorough convergence analysis, this paper combines the WOA and P&O MPPT techniques. This combination is developed to achieve the maximum power with less power oscillation and a fast convergence rate to handle rapid variations of solar irradiation and PSC. Hence, this research attempt to exploit the WOA in combination with the conventional P&O for developing a MPPT technique to obtain effective tracking abilities under PSCs.
The paper is structured as follows. A mathematical model of a PV module and the characteristics of a PV under PSCs are discussed in section II. Section III illustrates the WOA and its application in developing MPPT, and an introduction to the hybrid technique where it is combined with P&O. Simulation and hardware results and comparisons with WOA, GWO and hybrid GWO are discussed in section IV. Finally, the paper is concluded in section V.
Ⅱ. MATHEMATICAL MODELLING AND CHARACTERISTICS OF A PV MODULE UNDER PARTIAL SHADING
A. Modelling of a PV Module
A PV cell/module can be represented as equivalent diode model as shown in Fig. 3. A number of panels is connected in series/parallel combination and the PV current is represented by Equ. (1).
where Ipv is the PV current, Io is the diode saturation current, a is the ideal factor, Rp is the shunt resistance, Rs is the series resistance, Vt is the thermal voltage, and Np and Ns represent the number of parallel and series PV cells, respectively.
Fig. 3. Equivalent circuit of a PV cell.
The PV module thermal voltage is represented in Equ. (2).
where q is the electron charge which is equal to 1.6 x 10-19 C, k is the Boltzmann’s constant = 1.38 x 10-23 J/K, and T is the temperature of the p-n junction in Kelvin.
The current in a SPV cell is given in Equ. (3).
where Ipv,n is the PV current under nominal conditions i.e. 1000W/m2 and 25°C, G is the panel surface irradiation, Gn is the irradiation under normal conditions, and KI is the temperature coefficient. The saturation current is given by Equ. (4).
where, Io,n is saturation current under nominal conditions, and Eg is the band gap energy. A reverse saturation current under nominal conditions is given by Equ. (5).
Eqns. (1)-(5) are modeled and simulated using the MATLAB/Simulink simulating environment. The simulation results for a uniform irradiation are shown Fig. 1, and PV array characteristics under PSC are shown in Fig. 4. Fig. 4 illustrates the I-V and P-V characteristics of the PV array under PSCs.
|
(a) |
|
(b) |
B. PV Array under Partial Shading Conditions
A PV string consisting of several modules is connected in parallel/series combinations. If there are few cells shaded due to the factors discussed earlier, the shaded cells absorb more energy due to the reverse voltage across them. This energy is converted to heat, which results in a thermal breakdown and leads to cell breakdown. This hotspot can be reduced by a bypass diode. The function of the diode is to avoid negative voltage. The diode starts to function when Equ. (6) is satisfied.
During PSCs, multiple MPPs, i.e. one GP and multiple local peaks (LP), are observed in the PV characteristics as shown in Fig. 4 due to the existence of the bypass diode. Fig. 5(a) shows the 3S configuration i.e. three panels connected in series, and Fig. 5(b) shows the 2S2P configuration in which there are two parallel paths and each path has two series connected panels. The P-V curve is simulated for different shading patterns with a distinct GP for both of the configurations as shown in Fig. 6.
|
(a) |
|
(b) |
|
|
|
|
| (a) |
||
|
|
|
|
| (b) |
||
The proposed power extraction method can be formulated as an optimized problem and it is given in Equ. (7).
where P(d) is the PV output power, d is the duty cycle of the dc-dc converter. dmin is the lower bound and dmax is the upper bound of the duty cycle and they are taken as 50% and 75%, respectively. The upper bound is kept at 75% to reduce the reverse recovery problem on the power electronic switch.
Ⅲ. OVERVIEW OF THE PROPOSED HYBRID WOA MPPT TECHNIQUE
A. Whale Optimization Algorithm (WOA) Application to MPPT
The WOA is a metaheuristic algorithm inspired by the hunting approach of a humpback whale. Since humpback whales have a superior hunting technique, they are considered to have solved the non-linear issue. The hunting behavior of a humpback whale is called the bubble-net feeding method [19]. Humpback whales prefer to hunt small fish near the surface. The hunting is done by producing bubbles along a circular path as shown in Fig. 7(a). The important steps to follow in WOA are encircling the target, bubble-net feeding movement, and the search for targets in the exploration phase. Figs. 7(b) and 7(c) show the shrinking method of encircling the target and the search for targets in exploration phase, respectively.
|
(a) |
|
(b) |
|
(c) |
Humpback whales identify the position of and encircle prey. The initial position of the search space is unknown and the algorithm assumes the current best solution as the target prey. With the initial assumption, the algorithm defines the best search agent while other agents try to update its position towards the best. Humpback whales swim about the prey within a shrinking circle and a spiral-shaped path at the same time, and update their position. This behavior is mathematically modeled as follows:
where i represents the current iteration, 
and 
are the coefficient vectors, 
is the position vector for the current best solution, which is updated when a better solution is arrived at for each iteration, and 
is whale’s position vector. 
, which indicates the distance and the shape of logarithmic spiral, p is a random number in [0,1], and l is a random number, which is selected as [-1,1]. If p<0.5, the shrinking encircling hunting mechanism will be selected; and if p>0.5, spiral hunting mechanism is selected. The coefficient vectors 
and 
are computed as shown in Eqns. (10)-(11).
where 
decreases from 2 to 0 linearly over an iteration in both the exploitation and exploration, and the values 
and 
are the random vectors in [0,1].
B. Conventional P&O MPPT Algorithm
Before and after perturbation, the P&O algorithm tracks the maximum operating voltage by observing the change in the PV power. P&O is the best and most popular method of MPPT among researchers and it is the reference to develop any new MPPT method. By sensing the PV voltage and current, the PV power is calculated by the algorithm. Based on changes in the power, the algorithm provides the perturbation in the converter duty cycle. The rule is represented in Equ. (12).
where dnew is the new duty cycle, dold is the old duty cycle and 
is the perturbed duty ratio. If 
is considered to be large, this results in a steady state oscillation and fast convergence.
C. Proposed Hybrid WOA-P&O MPPT Technique
The combination of the WOA and P&O is called hybrid WOA-PO MPPT algorithm and it is a smart computation algorithm. During uniform irradiance, the P&O algorithm tracks the MPP, and during non-uniform irradiance, the hybrid WOA-PO tracks the GP. First, the WOA comes into action followed by P&O. When the whales are close to each other, the P&O is in progress at the location of best whale in the WOA process.
A flowchart of the proposed hybrid WOA-PO algorithm is shown in Fig. 8. The steps to be followed to implement the hybrid MPPT are as follows.
Fig. 8. Flowchart of the proposed hybrid WOA-PO MPPT.
| Step 1: The position of the whale is initialized with an equal search space between 50% and 75% of the duty cycle. Step 2: To maximize the PV output power at each whale position, trigger the converter and recalculate the output power. Step 3: The position of the whale is adjusted as follows: The objective function is given in equation (13): where P represents the PV power, d is the duty ratio, i is the number of iterations, and k is the number of whales. The duty cycle is defined in equation (14) by modifying equation (9) with the assumption of p<0.5 (shrinking encircling mechanism). where D is the cycle, i is the number of iterations, and k is the number of attacking whales. Step 4: Repeat steps 3 and 4 until the convergence of all the whales. Step 5: When the algorithm locates the MPP, P&O begins to track the GPs. The step size should be small to reduce the PV output power oscillation. This also offers more efficiency in tracking. |
The hybrid algorithm is proposed for PV panels under PSCs. The position of the whale represents the duty cycle for the dc-dc converter, which eliminates the use of a PI loop. This results in a reduction of the controller gain tuning burden and a simple structure design. A block diagram of the proposed MPPT technique is shown in Fig. 9. Depending on climatic conditions i.e. solar irradiation and temperature, the PV output power keeps changing. When the PV power changes, the proposed hybrid MPPT is reinitialized by evaluating the PV power. To improve the tracking accuracy, the number of whales should be more. However, this also increases the computation burden. Therefore, in this paper, 10 whales are considered to reduce the computation time.
Ⅳ. SIMULATION AND HARDWARE RESULTS
To test the performance of the proposed hybrid WOA-PO based MPPT algorithm, its performance was compared with that of GWO, WOA and hybrid GWO-PO algorithms. The above four algorithms were implemented for a solar PV based dc-dc converter under partial shading conditions and changes in the insolation level. The solar panel configurations for testing are taken as 3S and 2S2P under PSCs. The structure of the PV system is shown in Fig. 9, and it consists of a PV panel, a controller with MPPT, a dc-dc converter, and a load. The parameters of the PV system for the simulation studies are takes as follows: Pmax = 250 W, Voc = 32.9 V, Isc = 10.62 A, Vmpp = 26.3 V, Impp = 9.51 A, number of cell/panel = 54, temperature coefficient of Voc = -0.123 %/°C, and temperature coefficient of Isc = 0.00318 %/°C. A boost converter is selected to test the efficacy of the proposed algorithm. The values of various components for the designed converter are as follows: L = 20 mH, Cout = 330 µF, Cin = 100 µF, Vin = (40-100) V, Vout = 385 V, and the output ripple voltage is < 1%.
Fig. 9. Block diagram of the proposed MPPT technique.
A. Rapid Change in Insolation
Fig. 10 shows the PV output power, voltage and current for the 3S configuration under a rapid change in insolation for the four algorithms: WOA-PO, GWO-PO, WOA and GWO. For simulating the PV system, pattern 1 is created at t = 0.4 sec, pattern 2 is created at t = 0.6 sec, and pattern 3 is created at t = 0.8 sec. During pattern 1, the hybrid WOA-PO tracks the GP of about 158.6 W, the hybrid GWO-PO tracks the GP of 147.6 W, the WOA tracks the GP of 146.9 W, and the GWO tracks the GP of 163.2 W. At t = 0.6 sec the pattern is changed to pattern 2. During pattern 2, the proposed algorithm locates the GP at 208.3 W, the hybrid GWO-PO locates the GP at 188.2 W, the WOA locates the GP at 184.2 W, and the GWO locates the GP at 170.6 W. During pattern 3, the hybrid WOA-PO tracks the GP at 255.5 W, the hybrid GWO-PO tracks the GP at 239.2 W, the WOA tracks the GP at 235.2 W, and the GWO tracks the GP at 187.8 W. From the simulation results shown in Fig. 10, the GWO fails to track the GP accurately and it settles at a LP. It is concluded that the proposed hybrid algorithm harvests the high tracking speed and it exhibits less oscillation when compared to the other three metaheuristics algorithms.
|
|
|
|
| (a) |
||
|
|
|
|
| (b) |
||
|
|
|
|
| (c) |
||
|
|
|
|
| (d) |
||
For the 2S2P configuration simulation, three patterns are created. Pattern 4 is created at t = 0.4 sec, pattern 5 is created at t = 0.6 sec, and pattern 6 is created at t = 0.8 sec. During pattern 4, the hybrid WOA-PO tracks the GP of about 231 W, the hybrid GWO-PO tracks the GP of 222 W, the WOA tracks the GP of 174.9 W, and the GWO tracks the GP of 164 W. At t = 0.6 sec the pattern is changed to pattern 5. During pattern 5, the proposed algorithm locates the GP at 233.5 W, the hybrid GWO-PO locates the GP at 223.3 W, the WOA locates the GP at 176.7 W, and the GWO locates the GP at 168 W. For pattern 6, the hybrid WOA-PO tracks the GP at 243.4 W, the hybrid GWO-PO tracks the GP at 232.8 W, the WOA tracks the GP at 183.8 W, and the GWO tracks the GP at 171.4 W. From the simulation results shown in Fig. 11, both the WOA and the GWO fail to track the GP accurately and it settles at a LP. It is concluded that the proposed hybrid algorithm harvests the highest tracking speed and exhibits less oscillation when compared to the other three metaheuristics algorithm.
|
|
|
|
| (a) |
||
|
|
|
|
| (b) |
||
|
|
|
|
| (c) |
||
|
|
|
|
| (d) |
||
B. Partial Shading Conditions
The proposed algorithm is also applied to the 3S configuration for pattern 7 under the partial shading condition and it is compared with the other three MPPT algorithms as shown in Fig. 12. The simulation is also carried out on the 2S2P configuration for pattern 8 and the results are shown in Fig. 13. From Fig. 12 and Fig. 13, it is clear that the proposed hybrid WOA-PO MPPT algorithm locates the GP with a quicker convergence rate when compared with the other MPPT algorithms discussed in this paper. Even though the WOA offers a slightly (very little) faster convergence than the WOA-PO as shown in Fig. 13, the oscillation introduced by the WOA is higher than that of the WOA-PO. It is concluded that the proposed hybrid algorithm shows the quickest convergence rate among the WOA, hybrid GWO-PO and GWO MPPT algorithms.
|
(a) |
|
(b) |
|
(a) |
|
(b) |
C. Extreme Change in Insolation
To demonstrate the strength of the proposed hybrid algorithm, a PV system with the 3S configuration is exposed to an extremely rapid change in insolation. The resulting waveforms are shown in Fig. 14. The insolation is changed at 0.2 sec interval each. The time period up to 0.4 sec for the hybrid GWO-PO exhibits less oscillation. However, the hybrid WOA-PO locates the GP quicker than the others with a small oscillation. The period from 0.4 sec to 0.6 sec for both of the hybrid algorithms converges quickly. However, the hybrid WOA-PO offers a slight oscillation. At 0.6 sec, the hybrid GWO-PO tracks less power than the hybrid WOA-PO and WOA MPPTs. The proposed algorithm converges the GP accurately and tracks more power from the panel. At 0.8 sec, the proposed algorithm exhibits its robustness. It tracks more power than the other three algorithms with a very high fast convergence rate. It is clear that the hybrid GWO-PO is able to locate the GP during extremely rapid changing insolation but with less power tracking. It is concluded that the proposed hybrid WOA-PO MPPT can effectively handle conditions such as PSCs, rapidly changes in insolation and extremely rapid changing insolation.
|
|
|
|
| (a) |
||
|
|
|
|
| (b) |
||
|
|
|
|
| (c) |
||
A performance comparison of the proposed hybrid algorithm, GWO-PO, WOA and GWO in terms of the efficacy of the tracking and the tracking speed for both configurations in terms of various shading patterns and simulation results are shown in Table I. From Table I, it can be seen that proposed hybrid algorithm has greater tracking speed and efficiency than the other three MPPTs. Fig. 15 and Fig. 16 show tracking comparisons of all the algorithms for both patterns in terms of tracking efficiency and tracking speed. The characteristics of the proposed algorithm and the other three algorithms are presented in Table II.
Fig. 15. Power extraction level of all MPPTs.
Fig. 16. Tracking performance level of all MPPTs.
| PV Configuration |
Shading Patterns |
MPPTs |
PV Output Power (W) |
Convergence Time (Seconds) |
Pmax from P-V Curve (W) |
Tracking Efficiency (%) |
| 3S |
Pattern 1 |
WOA-PO |
158.6 |
0.03879 |
164 |
96.7 |
| GWO-PO |
147.6 |
0.03947 |
90 |
|||
| WOA |
146.9 |
0.04269 |
89.57 |
|||
| GWO |
163.2 |
0.04487 |
99.51 |
|||
| Pattern 2 |
WOA-PO |
208.3 |
0.03043 |
209 |
99.66 |
|
| GWO-PO |
188.2 |
0.03748 |
90.04 |
|||
| WOA |
184.2 |
0.03874 |
88.13 |
|||
| GWO |
170.6 |
0.04019 |
81.62 |
|||
| Pattern 3 |
WOA-PO |
255.5 |
0.01268 |
256 |
99.84 |
|
| GWO-PO |
239.2 |
0.02987 |
93.43 |
|||
| WOA |
235.2 |
0.03241 |
91.87 |
|||
| GWO |
187.8 |
0.03493 |
73.33 |
|||
| 2S2P |
Pattern 4 |
WOA-PO |
231 |
0.03543 |
232 |
99.56 |
| GWO-PO |
222 |
0.03642 |
95.61 |
|||
| WOA |
174.9 |
0.04124 |
75.43 |
|||
| GWO |
164 |
0.04361 |
70.61 |
|||
| Pattern 5 |
WOA-PO |
233.5 |
0.03121 |
234 |
99.78 |
|
| GWO-PO |
223.3 |
0.03478 |
95.42 |
|||
| WOA |
176.7 |
0.03958 |
75.64 |
|||
| GWO |
168 |
0.04125 |
71.79 |
|||
| Pattern 6 |
WOA-PO |
243.4 |
0.02144 |
244 |
99.75 |
|
| GWO-PO |
232.8 |
0.03147 |
95.40 |
|||
| WOA |
183.8 |
0.03987 |
75.40 |
|||
| GWO |
171.4 |
0.04054 |
70.49 |
| MPPTs |
Periodic Tuning |
Response |
Power Oscillation |
GP Convergence |
Complexity |
| Proposed WOA-PO |
No |
Very fast and accurate |
Very less |
Guaranteed |
Medium |
| GWO-PO |
No |
Fast and moderate accuracy |
Less |
Guaranteed |
Medium |
| WOA |
No |
Slow and less accuracy |
Slightly high |
Guaranteed |
Medium |
| GWO |
No |
Slow and very less accuracy |
High |
Guaranteed |
Medium |
D. Hardware Results
The proposed WOA-PO hybrid algorithm is experimentally verified for the both the 3S and 2S2P PV array configurations. The experimental setup is shown in Fig. 17. The ratings for each of the PV modules are Vmpp=17.58 V, Impp=5.69 A, Voc=22.1 V and Isc=5.88 A. The shading on the panel is created by transparent sheets. In this paper, a dSPACE1103 is selected to generate the PWM signals and it has a lot of ADC and DAC channels. A Hall effect sensor (ACS712) is used sense the PV current and a LV25-P/SP5 is used to sense the PV voltage. The sensed PV voltage and current are processed by a signal conditioning unit before being sent to the dSPACE1103.
Fig. 17. Experimental setup for the proposed algorithm.
Two shading patterns are created to validate the proposed algorithm. The waveforms are captured using the dSPACE control desk NG, and processed using MATLAB/Simulink software. The experimental waveforms of PV parameter for 3S configuration are shown in Fig. 18. During the first shading pattern at 20 sec, the hybrid WOA converges to the GP of 224 W, the hybrid GWO converges to the GP of 200 W, the WOA converges to the GP of 185 W, and the GWO converges to the GP of 168W. The tracking speed of the hybrid WOA is higher than that of the other MPPT algorithms. At 40 sec, the shading pattern is changed, and the algorithms restart the search to find the new MPP. The hybrid WOA converges to the GP of 259 W, the hybrid GWO converges to the GP of 235 W, the WOA converges to the GP of 220 W, and the GWO converges to the GP of 183W. At 80 sec, the shading pattern is changed again to the pattern at 20 sec, and the MPPT algorithms display the same converges to GPs.
|
|
|
|
| (a) |
||
|
|
|
|
| (b) |
||
|
|
|
|
| (c) |
||
|
|
|
|
| (d) |
||
Experimental waveforms for the 2S2P configuration are shown in Fig. 19. During the first shading pattern at 20 sec, the hybrid WOA converges to the GP of 210 W, the hybrid GWO converges to the GP of 180 W, the WOA converges to the GP of 172 W, and the GWO converges to the GP of 163W. The tracking speed of the hybrid WOA is higher than that of the other MPPT algorithms. At 40 sec, the shading pattern is changed, and the algorithms restart the search to find a new MPP. The hybrid WOA converges to the GP of 248 W, the hybrid GWO converges to the GP of 219 W, the WOA converges to the GP of 204 W, and the GWO converges to the GP of 175W. At 80 sec, the shading pattern is changed again to the pattern at 20 sec, and the MPPT algorithms display the same converges to the GPs.
|
|
|
|
| (a) |
||
|
|
|
|
| (b) |
||
|
|
|
|
| (c) |
||
|
|
|
|
| (d) |
||
Experimental waveforms for the boost converter load parameters with the proposed hybrid algorithm are captured in a 4-channel DSO and shown in Fig. 20 for both of the PV array configurations. From Fig. 18 and Fig. 19, can be concluded that the proposed hybrid algorithm successfully tracks the MPP during a change in shading pattern and that it restarts the tracking algorithm. The combination of the WOA and the conventional P&O results in a quicker convergence to the GP with less tracking time and a high tracking efficiency, which enables it to locate the MPP for the solar PV system. Finally, it is concluded that the proposed hybrid WOA-PO MPPT algorithm is able to adjust itself to changes in insolation and PSCs while enhancing the tracking efficiency.
|
(a) |
|
(b) |
Ⅴ. CONCLUSIONS
In this paper, a novel hybrid WOA-PO MPPT algorithm is proposed to locate the GP when a PV system is subjected to PSCs. Two PV configurations of 3S and 2S2P under PSCs are tested with the proposed MPPT to verify its effectiveness. When the PV system exhibits multiple peaks, the proposed MPPT tracks the GP with less tracking time, high accuracy and high efficiency under both changes in insolation and under any PSCs. From the obtained simulation results, the proposed algorithm is shown to be superior to the other algorithms in terms of accuracy, efficiency and tracking time. To reduce the computational burden, the simulation is carried out for 30 iterations with 10 search agents (number of whales) and the numerical results are presented. It is noticed that the standard deviation of the hybrid WOA-PO MPPT is comparatively less, which enables the proposed algorithm to track the GP effectively. Simulation and hardware results demonstrated the capability of the proposed hybrid WOA-PO MPPT in tracking the GP of a PV system under various PSC and changes in insolation level for both the dynamic and steady-state conditions.
REFERENCES
[1] F. Salem and M.A. Awadallah, “Detection and assessment of partial shading in photovoltaic arrays, J. Electr. Syst. Informat. Technol., Vol. 3, No. 1, pp. 23-32, 2016.
[2] M. Premkumar, N. Dhanasekar, R. Dhivakar, and P. Arunkumar, “Comparison of MPPT algorithms for PV system-based DC-DC converter,” Advances in Natural & Applied Science, Vol. 17, No. 9, pp. 212-221, Mar. 2016.
[3] X. Xiao, X. Huang, and Q. Kang, “A hill-climbing-method- based maximum power point tracking strategy for direct-drive wave energy converters,” IEEE Trans. Ind. Electron., Vol. 63, No. 1, pp. 257-267, Jan. 2016.
[4] K. Soon Tey and S. Mekhilef, “Modified incremental conductance MPPT algorithm to mitigate inaccurate responses under fast-changing solar irradiation level,” Solar Energy, Vol. 101, pp. 333-342, 2014.
[5] D. C. Huynh and M. W. Dunnigan, “Development and comparison of an improved incremental conductance algorithm for tracking the MPP of a solar PV panel,” IEEE Trans. Sustain. Energy, Vol. 7, No. 4, pp. 1421-1429, Oct. 2016.
[6] A. Tobon, J. P. Restrepo, J.P.V. Ceballos, S.I.S. Garces, J. Herrera, and A. Ibeas, “Maximum power point tracking of photovoltaic panels by using improved pattern search methods,” Energies, Vol. 10, pp. 1-15, 2017.
[7] Y. Mahmoud and E. F. Elsaadany, “A novel MPPT technique based on an image of PV modules,” IEEE Trans. Energy Convers., Vol. 32, No. 1, pp. 213-221, Mar. 2017.
[8] A. Mohapatra, B. Nayak, P. Das, and K.B. Mohanty, “A review on MPPT techniques of PV system under partial shading condition,” Renew. Sustain. Energy Rev., Vol. 80, pp. 854-867, 2017.
[9] J. Ying Shi, F. Xue, Z. Jian Qin, W. Zhang, L. Tao Ling, and T. Yan, “Improved global maximum power point tracking for photovoltaic system via cuckoo search under partial shaded conditions”, J. Power Electron., Vol. 16, No. 1, pp. 287-296, Jan. 2016.
[10] K. Ishaque and Z. Salam, “A deterministic particle swarm optimization maximum power point tracker for photovoltaic system under partial shading condition”, IEEE Trans. Ind. Electron., Vol. 60, No. 8, pp. 3195-3206, Aug. 2013.
[11] Y. H. Liu, S. C. Huang, J. W. Huang, and W. C. Liang, “A particle swarm optimization-based maximum power point tracking algorithm for PV systems operating under partially shaded conditions,” IEEE Trans. Energy Convers., Vol. 27, No. 4, pp. 1027-1035, Dec. 2012.
[12] S. Mohanty, B. Subudhi, and P. K. Ray, “A new MPPT design Using grey wolf optimization technique for photovoltaic system under partial shading conditions,” IEEE Trans. Sustain. Energy, Vol. 7, No. 1, pp. 181-188, Jan. 2016.
[13] D. F. Teshome, C. H. Lee, Y. W. Lin, and K. L. Lian, “A modified firefly algorithm for photovoltaic maximum power point tracking control under partial shading,” IEEE J. Emerg. Sel. Topics Power Electron., Vol. 5, No. 2, pp. 661- 671, Jun. 2017.
[14] S.Titri, C. Larbes, K. Youcef Toumi, and K. Benatchba, “A new MPPT controller based on the ant colony optimization algorithm for photovoltaic systems under partial shading conditions,” Applied Soft Computing, Vol. 58, pp. 465-479, 2017.
[15] A. Soufyane Benyoucef, A. Chouder, K.Kara, S. Silvestre, and O. Ait sahed, “Artificial bee colony based algorithm for maximum power point tracking (MPPT) for PV systems operating under partial shaded conditions,” Applied Soft Computing, Vol. 32, pp. 38-48, Apr. 2015.
[16] K. Sundareswaran, P. Sankar, P. S. R. Nayak, S. P. Simon, and S. Palani, “Enhanced energy output from a PV system under partial shaded conditions through artificial bee colony,” IEEE Trans. Sustain. Energy, Vol. 6, No. 1, pp. 198-209, Jan. 2015.
[17] M. Metry, M. B. Shadmand, R. S. Balog and H. Abu-Rub, “MPPT of photovoltaic systems using sensorless current- based model predictive control,” IEEE Trans. Ind. Appl., Vol. 53, No. 2, pp. 1157-1167, Apr. 2017.
[18] S. Mohanty, B. Subudhi, and P. K. Ray, “A grey wolf-assisted perturb & observe MPPT algorithm for a PV system,” IEEE Trans. Energy Convers., Vol. 32, No. 1, pp. 340-347, Mar. 2017.
[19] S. Mirjalili and A. Lewis, “The whale optimization algorithm,” Advances in Engineering Software, Vol. 95, pp. 51-67, 2016.
[20] C. H. S. Kumar and R.S. Rao, “A novel global MPP tracking of photovoltaic system based on whale optimization algorithm,” Int. J. Renew. Energy Develop., Vol. 5, No. 3, pp. 225-232, Sep. 2016.
Manoharan Premkumar was born in Coimbatore, India. He received his B.E. degree in Electrical and Electronics Engineering from the Sri Ramakrishna Institute of Technology, Coimbatore, India, in 2004; and his M.E. degree in Applied Electronics from the Anna University of Technology, Coimbatore, India, in 2010, where is presently working towards his Ph.D. degree. He is presently working as an Assistant Professor at the GMR Institute of Technology, Rajam, India. His current research interests include microinverters, non-isolated and isolated dc-dc converters, and solar PV MPPT techniques.
Rameshkumar Sumithira received her B.E. degree in Electrical and Electronics Engineering from Institute of Road and Transport Technology, Erode, India, in 2002; her M.E. degree in Power Systems from the PSG College of Technology, Coimbatore, India, in 2004; and her Ph.D. degree in Electrical Engineering from Anna University, Chennai, India, in 2013. She is presently working as an Assistant Professor at the Government College of Engineering, Salem, India. Her current research interests include power systems, matrix converters, and single-phase off-grid inverters.