Maximum Power Point Estimation Based on Operating Conditions Classification for Photovoltaic Systems: A Case Study for Partial Shading

Abstract

As a kind of operating condition, partial shading fault is an unavoidable occurrence in photovoltaic power generation, causing variations in the photovoltaic system's output power, panel current-voltage data, and power-voltage data, thus changing the maximum power point. The maximum power point information and the operating condition have to be known in order to make an accurate analysis and also increase the system safety, production, efficiency, and availability due to these variations during the analysis. In this paper, a novel customized maximum power point estimation method is proposed to detect the maximum voltage using a different operating condition classification system based on the common vector approach. Operating conditions consist of a standard test and three different types of partial shading, and 100% accuracy in classification is achieved. Then, the support vector regression is employed for the maximum voltage estimation in the classified operation condition. The data set used was obtained from the PSIM package simulation of a 250 W photovoltaic system under different studying conditions. The experimental results show that the proposed estimation method significantly reduces estimation errors and outperforms conventional voltage estimation at the maximum power point.