Artificial intelligence-based maximum power point tracking controllers for Photovoltaic systems: Comparative study

详细信息    查看全文

• 作者：Mostefa Kermadi ; ^{mostefa.kermadi@g.enp.edu.dz} ; El Madjid Berkouk ^{emberkouk@yahoo.fr el_madjid.berkouk@g.enp.edu.dz}
• 关键词：Maximum power point tracking (MPPT) ; Photovoltaic (PV) ; PID control ; Fuzzy Logic (FL) ; Artificial Neural Network (ANN) ; Genetic Algorithms (GA) ; Particle Swarm Optimization (PSO) ; Buck-Boost Converter
• 刊名：Renewable and Sustainable Energy Reviews
• 出版年：2017
• 出版时间：March 2017
• 年：2017
• 卷：69
• 期：Complete
• 页码：369-386
• 全文大小：2656 K
• 卷排序：69

文摘

In Photovoltaic (PV) systems, maximum power point tracking (MPPT) is an indispensable task. To date, various MPPT techniques have been proposed in the literature using classical and artificial intelligence methods. However, those techniques are tested on different PV systems and under different environmental conditions. In this work, we attempt to summarize and to give a comprehensive comparative study of the most adopted Artificial Intelligence (AI)-based MPPT techniques. The MPPT techniques which will be described are based on: Proportional-Integral-Derivative (PID), Fuzzy Logic (FL), Artificial Neural Network (ANN), Genetic Algorithms (GA) and Particle Swarm Optimization (PSO). The developed MPPT controllers are tested under the same weather profile in the same photovoltaic system which is composed of a PV module, a DC-DC Buck-Boost converter and a DC load. Initially, Modelling and simulation of the system is performed using the MATLAB/Simulink environment. Thereafter, the sliding mode control is applied to the converter in order to improve its performance. In a further stage, the different steps of development for each MPPT technique are presented. Simulation is performed to confirm the validity of the proposed controllers under the same variable temperature and solar irradiance conditions. Finally, a comparative study is carried out in order to evaluate the developed techniques regarding two principal criteria: the performance and the implementation cost. The performance is evaluated using comparative analysis of the tracking speed, the average tracking error, the variance and the efficiency. To estimate the implementation cost, a classification is carried out according to the type of the used sensors, the type of circuitry and the software level complexity. Recommendations that expected to be useful for researchers in the MPPT area about the validity of each MPPT technique are given in the last section.