Hybrid Intelligent Machine Learning based Ultra-short Term Generation Prediction of Photovoltaic Systems

Abstract

This work developed an ultra short-term photovoltaic power prediction model based on hybrid intelligent technology. The proposed model adopts a series of data processing technologies, including input variable selection based on statistical analysis, attribute reduction based on principal component analysis (PCA) and feature subset division based on the K-means clustering algorithm, to obtain a more relevant and effective data as input information for prediction. The model uses an adaptive neural fuzzy inference system (ANFIS) to train and learn the input information to obtain the output prediction results. The particle swarm optimization (PSO) algorithm is adopted in the training process to optimize the ANFIS parameters to reduce the prediction error. The proposed solution is evaluated through simulation experiments and the numerical results demonstrate that it can achieve effective prediction accuracy and has good adaptability.