Analyzing the performance of photovoltaic systems using support vector machine classifier

Abstract

The performance of a photovoltaic system depends on several parameters such as temperature, clouds, season, etc. This makes the study of Photovoltaic (PV) performance from monitoring databases very complex given the size of the information and the complexity of the phenomena involved. This article proposes an efficient classification and analysis approach based on the Support Vector Machine (SVM) classification in order to simplify and optimize the processing of this data and the study of the performance of PV systems. More precisely, we are applying the SVM; the data changes from variables to multiclass for simplifying the analysis. We have detailed all the calculation steps. Based on the application of artificial intelligence (Classification), the recorded data (output power, temperature of the PV module, solar irradiation, ambient temperature, etc.) can be analyzed and processed easily. The results of the performance analysis of a 270 W mono-crystalline PV module after seven (07) months of monitoring were studied. • Application of classification in the analysis of monitoring database. • SVM method for analyzing the performance of photovoltaic systems. • Simplifier the big data monitoring for analyzing the performance of photovoltaic systems. • Using SVM classification for treated the electrical parameters (P max ).