Short-term PV/T module temperature prediction based on PCA-RBF neural network

International Conference on Energy Engineering and Environmental Protection 2017

能源学;生态环境科学

Li, Jiyong^1,2 ; Zhao, Zhendong^1,2 ; Li, Yisheng^1,2 ; Xiao, Jing^1,2 ; Tang, Yunfeng^1,2

College of Electrical Engineering, Guangxi University, University Road 100, Xixiangtang District, Nanning

530004, China^1

Guangxi Key Laboratory of Power System Optimization and Energy Technology, Guangxi University, University Road 100, Xixiangtang District, Nanning

530004, China^2

关键词: Component extraction;    Generalization performance;    Meteorological factors;    Prediction accuracy;    RBF Neural Network;    Short-term forecasting;    System controllers;    Temperature prediction;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/121/5/052045/pdf
DOI  :  10.1088/1755-1315/121/5/052045

学科分类：环境科学（综合）

来源: IOP

【 摘 要 】

Aiming at the non-linearity and large inertia of temperature control in PV/T system, short-term temperature prediction of PV/T module is proposed, to make the PV/T system controller run forward according to the short-term forecasting situation to optimize control effect. Based on the analysis of the correlation between PV/T module temperature and meteorological factors, and the temperature of adjacent time series, the principal component analysis (PCA) method is used to pre-process the original input sample data. Combined with the RBF neural network theory, the simulation results show that the PCA method makes the prediction accuracy of the network model higher and the generalization performance stronger than that of the RBF neural network without the main component extraction.