maximum power point tracking technique for partially

SEEE DIGIBOOK ON ENGINEERING & TECHNOLOGY, VOL. 01, MAY 2018 ALTERNATE ENERGY TECHNOLOGIES

978-81-933187-0-6 © 2018 SEEEPEDIA.ORG Society for Engineering Education Enrichment

Yagini. k, [email protected]; Dr. E. Lathamercy , [email protected];

Maximum Power Point Tracking Technique for

Partially Shaded PV System Using Hybrid DEPSO Method

YAGINI.K, Dr. E. LATHAMERCY

Government College of Technology, Coimbatore, India. [email protected]

Abstract— In photovoltaic (PV) power generation, partial shading is an unavoidable complication that significantly reduces the efficiency of the overall system. Under this condition, the PV system produces a multiple-peak function in its output power characteristic. Thus, a reliable technique is required to track the global maximum power point (GMPP) with in an appropriate time. This study aims to employ a hybrid evolutionary algorithm called the DEPSO technique, a combination of the differential evolutionary (DE) algorithm and particle swarm optimization (PSO), to detect the maximum power point under partial shading conditions. The paper starts with a brief description about the behavior of PV systems under partial shading conditions. Then, the DEPSO technique along with its implementation in maximum power point tracking (MPPT) is explained in detail. Finally, Simulation and experimental results are presented to verify the performance of the proposed technique under different partial shading conditions. Index Terms— Differential evolution (DE) algorithm, maximum power point tracking (MPPT), partial shading, particle swarm optimization (PSO), photovoltaic (PV) system. I.INTRODUCTION The world is facing many problems and environmental pollution. The research and development of photovoltaic (PV) technologies have become a hot topic in the world. Solar PV is now used around the world as an important technology for the conversion of solar energy because of its cleanliness and security. The solar PV industry is one of the fastest growing high-tech industries. The major problem solar PV is Maximum Power Point Tracking (MPPT).Various MPPT algorithms were discussed and decrease in PV output power, and hot-spot generated damages the PV cells. Since the dynamics of the PV power system under partial shading is time varying, (MPPT) is equipped with features such as tracking Global Maximum Power Point (GMPP) at different conditions ,e.g., shading ,degradation of PV cell, and adaptability of P-V characteristics change in PV array, smooth, and steady tracking behavior. A number of MPPT techniques such as Hill Climbing (HC), Perturb and Observe (P&O) and Incremental conductance (IC) and have been proposed for improving the efficiency of the PV system. The HC method uses a perturbation in the operating voltage of the PV system. Both these methods yield oscillation at (MPP) owing to the fact that the perturbation continuously changes in both directions to maintain the MPP resulting in power loss. The module efficiency, the IC method was proposed which reduced the oscillations but not completely. Both P&O and IC methods fail during those time intervals characterized by changing atmospheric conditions. A few improved IC algorithms were also proposed to improve

the MPP tracking capability during fast-changing irradiance level and load. To achieve a fast MPP response, a simple trigonometric rule has been presented to establish relationship between the load line and I-V curve. A dynamic MPPT controller for PV system under fast-varying insolation and partial shading conditions is proposed, which uses a scanning technique to determine the maximum power –delivering capacity of the panel at a given operating condition. The focus of the research here is to determine the global peak(GP) during PSCs; in order to alleviate some of the issues like lower tracking efficiency and oscillations generated in the PV output power, an alternative approach is to employ evolutionary algorithm (EA) techniques, which has the capability to handle nonlinear objective functions. Meta-heuristic optimization methodologies such as particle swarm optimization (PSO) and firefly have been extensively used for various engineering applications. In the proposed method, the P&O identifies the nearest local MPPs in the first step and then the PSO starts searching the GMPP in the second step. 1)As a result of using this combination, the search space exploration is reduced in early iterations. 2) Given the random values in the algorithm architecture, the meta heuristic approach of evolutionary algorithms is not avoided. 3) In contrast to the other modified forms of PSO, the algorithm continues to search the GMPP until a justifiable stopping condition.

164

http://seeepedia.org/

mailto:[email protected]



SEEE DIGIBOOK ON NATIONAL CONFERENCE ON ALTERNATE ENERGY TECHNOLOGIES 2018


4) Given the simple approach of the proposed hybrid technique, the computational burden of the algorithm is reduced.

Fig.1 Block diagram of MPPT technique II. CHARACTERISTICS OF A PV SYSTEMS UNDER PSCs A. Basic characteristics of a PV cell A PV cell can be represented by an equivalent single diode model. The symbol used in the model are defined as follows: Ipv PV current source; D a diode connected in parallel to the current source; Rs the sum of resistance due to all the components the come in path of current which is desirable to be as low as possible; Rp to represent the leakage across the P-N junction; I difference between the photocurrent Ipv and the diode current Id,which is given bys I = Ipv –Io[exp (qV+qRsI/Ns Ks Ta -1) –V+RsI/Rps…(1) Where Io is the saturation current, α is diode ideality factorElectron, T is the temperature in kelvin .Ns is the number of series.

Fig.2 (a) Patterns. (b) pattern 2 (c) P-V curves under PSC III. COMPARING PSO & DEPSO A.Basic overview of PS0 PSO is a swarm-based evolutionary algorithm that investigates the search space and determines the components and settings required to optimize a special objective function

Fig. 3 Simple vector diagram of PSO The PSO algorithm hires a certain number of particles (N) to explore the D-dimensional search space of the problem. At each iteration, every single particle represents a solution to the problem on the basis of the particle’s location in the search xi. The particles move stochastically using a velocity vector of Vi, a resultant of three vectors: 1) the best location experienced by the particle (Pbi); 2) the best location experienced by the entire swarm (Gb); and 3) a portion of itself in the last iteration. Fig. 3 shows a simple representation of the movement of a particle in the search space. Pbi and Gb must be updated at each iteration throughout the optimization process. To do so, a fitness function should be defined to evaluate the location of each particle at each step. The mathematical form of obtaining the velocity and updating the location of each particle. The mathematical equation is

.......(1) Here, the subscript I represents the particle number; k denotes the iteration number;

are the velocity vector and location of the ith particle, respectively, at the kth iteration; r1 and r2 are random values chosen from a uniform distribution from 0 to 1; c1 and c2 are the cognitive and social coefficients, respectively; and w is the inertia weight, which decreases continuously throughout the optimization and controls the scale of particle.

Fig 4 Generation of trial vector using target and mutant vectors. B. Overview of DEPSO

165




The DE algorithm is a particle-based global optimization algorithm consisting of three main operation sections, namely mutation, crossover, and selection eq .2 shows the mutation unit of the DE algorithm, which works as a search engine .

..(2) Here, vi,Gis the mutant vector and xi,Gis the target vector ofthe ith particle at the Gth step. r1, r2, and r3 are random numbers in the range of 1, 2, . . .nop, where nop denotes the number of particles. K and F are the scaling and combination factors, respectively. The crossover or recombination in the DE technique is a non uniform operation that generates trial vectors on the basis of components from the population. The DE algorithm uses a non-uniform crossover operation to generate trial vectors on the basis of the components of the population. The crossover enables the algorithm to propose better solutions by shuffling the data of the successful combinations. ui, Gi s the trial vector which is generated based on the two other vectors, namely target vector (xi,G) and mutant vector (vi,G). The target vector is indeed used for the selection step, which is the last step in the algorithm. To be exact, the new target vector (xi,G+1) takes its members based on comparing the fitness values for the old target vector (xi,G) and the trial vector (ui,G). Equation (12) shows the trial vector of the crossover and Fig. 4 describes how the trial vector is generated using target and mutant vectors.

..(4) Here, j ∈1, 2, . . .,D, CR is the crossover constant between 0 and 1, rn djis a random value between 0 and 1, and rn iis a random number chosen from the domain of j. The maximum dimension of the search space is denoted by D. Fig. 9 shows the process of obtaining a new trial vector in the DE algorithm for a two-dimensional (2-D) search space.

Fig.5 Process of obtaining new trial vector for DE algorithm. The final operation is the selection operation that directs the movement toward prospective areas in the search. The selection of the parents is independent of the fitness values; however, the produced children (offspring) in the mutation and crossover must be evaluated by the fitness function and compared with the parents. The parents stay in the population if the fitness values have no effect on the

selection of parents. The children produced in the mutation and crossover are then evaluated and compared with the parents. The parents remain in the population if they have better fitness values than their children. The selection procedure can be expressed as follows:

..(5) In the PSO algorithm, as the iteration goes on, the diversity of particles decreases significantly. This phenomenon increases the probability of being trapped in the local optima of the solution space. By contrast, the DE algorithm successfully explores the local optima of the search space by utilizing the differential information; however, this algorithm degrades its search quality in finding the global optima. The DEPSO algorithm keeps individuals from being trapped in the local optima by combining the DE operator with the PSO algorithm, which diversifies the PSO technique. Several applications of training, clustering, and optimization have been presented in literature, which shows that DEPSO outperforms both the PSO and DE algorithms in terms of solution quality and convergence speed. Gband Pbi are important components in the searching process of the PSO algorithm, in which every individual tries to improve its position toward reaching Gb as the best position found by the swarm. However, the decremental weighted velocity of the individuals decreases the ability of the swarm to diversify after a certain iteration. are kept from being stuck to the local optima.

. Fig.6 Flowchart of DEPSO The search space of the problem consists of a vector of different values for the terminal voltage of the PV panel. The location vector of the problem, a 1 × N vector, where N denotes the number of particles hired. Each location represents a voltage value that is a potential solution to the MPPT problem. The particles are evaluated based on the output power of the PV panel with respect to the proposed terminal voltage

..(6) Equation (6) shows the location vector of the problem, a 1 × N vector, where N denotes the number of particles hired. In practice, during partial shading, instantaneous variations in

166




the insolation level cause sharp fluctuations in the generated power. Therefore, the condition presented .Must be satisfied to initialize the algorithm. The condition indicates the minimum allowed variation in the output power to run the algorithm and to find the new MPP, which is given by ΔP

Where J (Xi) returns the output power of the PV panel, respective to the location of ith particle in the search space. IV. EXPERIMENT RESULTS

Fig . 7 Simulation of PSO

Fig.8 Output of PSO A. specifications of PSO Power of array 30W Voltage 17.2V Current 1.27A Response time 0.5 sec

. Fig.9 Simulation model of DEPSO

Fig.10 Output of DEPSO The power, voltage and current for the configuration under PSCs employing DEPSO is shown above .In model of DEPSO based MPPT converges to the GP of 100w. D B. Specification of DEPSO

Power of array 100W Voltage 17.25V Current 3.25 A Response time 0.35 sec

The DEPSO–based MPPT can handle partial shading conditions efficiently and it out performs Ant-colony based technique with respect to fast convergence to GP, tracks more power. V. CONCLUSION This paper proposed a new evolutionary computing approach called DEPSO technique to design a maximum power extraction algorithm for PV systems to work under partial shading condition .In view of accessing the effectiveness of this new hybrid MPPT DEPSO its performance was compared with existing MPPT namely PSO optimization technique and from obtained results , it was found that the DEPSO based MPPT exhibits superior performance compare Ant-colony based MPPT. The proposed method is highly robust, reliable, system independent (DEPSO) can track the Global Maximum Power Point (GMPP) very accurately and quickly in comparison to state of art methods with a good dynamic as well as a steady state response in every type of environmental conditions. REFERENCES

[1]. H. Patel and V. Agarwal, ‘’Maximum power point tracking scheme for PV systems operating under partially shaded conditions, ’’IEEE Trans.Ind . electron. vol. 55, no. 4, pp.302-310,Apr. 2008.

[2]. B. Subudhi and R.Pradhan,’ ’A comparative study on maximum power point tracking techniques for photovoltaic power systems, ’’IEEE Trans. Sustain. Energy,vol.4, no.4, pp.88-98, jan. 2013.

[3]. Jeyakkannan, N., and B. Nagaraj. "Online Monitoring of Geological Methane Storage and Leakage Based on Wireless Sensor Networks." Asian Journal of Chemistry 26 (2014).

[4]. M.A.Elgendy and B.Zahawi ,’’Assesment of P&O MPPT algorithm implementation techniques for PV pumping applications,IEEE Trans.Sustain. Energy, vol. 3,no.1,pp.21-31, jan. 2012.

167




[5]. S.Kjaer,’’Evaluation of hill climbing and the incremental conductance maximum power point trackers for PV systems,ÍEEE.Trans ,Energy. Convers., vol. 5,no.4, pp.231-262 ,2013

[6]. M. A. G. Brito, L. Galotto,and C. A. Canesin,’’Evaluation of the main MPPT techniques for photovoltaic applications,’’IEEE.Trans.Ind. Electron ., vol.60,n0.3, pp.1156-1167, Mar.2013.

[7]. Vijayakumar. "Bio Inspired Algorithm for PID Controller Tuning and Application to the Pulp and Paper Industry." Sensors & Transducers 145.10 (2012): 149.

[8]. N.Femia, G.Petrone ,G.Spagunolo, and M.Vitelli, ‘’Optimization of Perturb and Observe maximum power point tracking method ,’’IEEE Trans. Power electon .,vol.20, no.4 pp 963-973, jul.2005.

[9]. R. Kotti and W.Shireen, ‘’Efficient MPPT control for PV systems adaptive to fast changing irradiation and partial shading conditions,’’ Solar Energy,vol.114, pp 397-342, Mar. 2015.

[10]. K. Ishaque , Z. Salam, M.Amjad, and S.Mekhief, ‘’An improved particle swarm optimization (PSO)-based MPPT for PV with reduced steady state oscillations,’’ IEEE Trans. Power Electron., vol.29 .no. 8, pp.3627-3638, Aug.2012.

[11]. Agalya, A., "CONCENTRATION CONTROL OF CONTINUOUS STIRRED TANK REACTOR USING PARTICLE SWARM OPTIMIZATION ALGORITHM."

[12]. K. Sundareshwaran, S. Peddapati, and ‘’MPPT of PV systems under partial shading conditions through a colony of flashing fireflies,’’IEEE Trans. Convers vol.29, no.2, pp. 463-472, Jun.2014.

[13]. S. Mirajalili, S .M. Mirjalili, and A. Lewis,’’Grey wolf. [14]. optimizer, Adv Eng. Software, vol.69, pSp.46-61. Jun

.2014S

168


maximum power point tracking technique for partially

Documents