基于改进BP神经网络光伏发电量预测研究  

Research on Improve the Prediction of BP Neural Net Photovoltaic Power Generation

作　　者：戴锦[1] 

机构地区：[1]南昌航空大学测试与光电工程学院,南昌330063

出　　处：《南昌航空大学学报（自然科学版）》2015年第3期91-97,共7页Journal of Nanchang Hangkong University(Natural Sciences)

基　　金：集成光电子学国家重点联合实验室开放课题(IOSKL2012KF14)

摘　　要：分析了影响光伏电池输出的主要因素,建立基于改进BP神经网络的光伏发电量预测模型。由光伏输出影响因素的分析,利用光照强度及环境温度对改进BP神经网络进行训练,对比了传统数学模型、传统BP模型与改进的BP模型的预测结果,结果表明该模型有较准确的预测能力。The main influence factors of the photovoltaic cell output has been analyzed and the prediction model on improved the photovoltaic power generation of BP neural net was established. Based on the analysis of influence factors,using the light intensity and environment temperature to train improving BP neural network,and the predictive results of the traditional mathematical model,the traditional BP model and the improved BP model were compared. The results show that the model has relatively accurate prediction ability.

关 键 词：改进BP神经网络 传统数学模型 预测能力 

分 类 号：TM914.4[电气工程—电力电子与电力传动]