Application of support vector machines in photovoltaic power prediction

Abstract

PV power generation is affected by environmental factors such as solar radiation intensity, temperature and humidity, and PV power generation is characterized by volatility and instability, and the prediction accuracy of traditional prediction algorithms is low. In this paper, support vector machine is used as the PV power generation prediction algorithm, and the training and testing samples for the experiment are selected from the historical data in the laboratory. The support vector machine model trained in MATLAB simulation environment is used to predict and analyze the laboratory PV power generation. The simulation experiment results show that the stability of PV prediction based on support vector machine is high and the prediction error is small, which overcomes the error brought by the traditional prediction algorithm pursuing empirical risk minimization and improves the accuracy of the prediction system.