Comparison of Artificial Intelligence Based Maximum Power Point Techniques for Photovoltaic systems

Abstract

Maximum Power Point Tracking technologies are being used in traditional PV system charge controllers to enhance the power conversion efficiency. An MPPT controller will ensure power extracted from the PV panels during varying climatic conditions is always maximum. This will ensure that maximum power is flowing between the panel and load. As both Temperature and Irradiation levels vary during the day, maximum power point trackers are an inevitable component in a PV system. As solar energy holds a major share in renewable energy in the world market, an improvement in MPPT technique makes the efficiency of the PV system increase and in turn cost reduction possible. However, the efficiency of conventional MPPT Techniques suffers from failing in tracking MPPT at fast varying climatic conditions and falling in local maxima of maximum power point than global maxima. The issues of stability, tracking speed, and accuracy can be solved using intelligent MPPT techniques methods based on soft computing tools: Therefore, this paper aims to study and provide a comparative analysis of two AI-based MPPT techniques such as ANN and ANFIS. The MPPT techniques considered in this study are ANN and ANFIS. Performance evaluation is carried out using MATLAB Simulation. Experimental results indicate that the two methods ANN and ANFIS are more efficient than conventional MPPT techniques due to its capability to avoid local MPP and partially shaded conditions.