基于遗传算法优化BP神经网络的光伏阵列短期功率预测  

Short Term Power Prediction of Photovoltaic Array Based on BP Neural Network Optimized by Genetic Algorithm

作　　者：付宗见[1] 梁明亮[1] 王艳萍[1] FU Zongjian;LIANG Mingliang;WANG Yanping(Zhengzhou Railway Vocational&Technical College,Zhengzhou 451460,China)

机构地区：[1]郑州铁路职业技术学院,郑州451460

出　　处：《电子器件》2020年第3期516-521,共6页Chinese Journal of Electron Devices

基　　金：河南省科技攻关项目(182102210141);郑州铁路职业技术学院科技创新团队项目(17KJCXTD02)。

摘　　要：提高光伏系统短期预测准确率对光伏系统平稳运行、协调电力系统资源具有重要意义。由于短时间的光伏阵列功率具有随机非平稳特征,现有小波预测、神经网络预测方法受到训练初始值局限,不能准确预测短期光伏系统功率。就非平稳随机特征的光伏阵列功率初始值引入遗传算法,优化BP神经网络,提出基于遗传算法的改进BP神经网络方法,对光伏阵列短期功率进行预测。实验表明该方法能适应于不同天气状况下的短期光伏阵列功率预测,并具有较高的准确度。The accurate short-term prediction of photovoltaic(PV)system is of great significance to the photovoltaic system. Due to the stochastic and non-stationary characteristics of short-term photovoltaic array power,the existing wavelet prediction and neural network prediction methods are limited by the initial value of training. It cannot accurately predict the short-term photovoltaic system power. In this paper,genetic algorithm is introduced to optimize BP neural network. An improved BP neural network method based on genetic algorithm is proposed to predict the short-term power of photovoltaic array. Experiments show that this method can be applied to short-term photovoltaic array power prediction under different weather conditions,and has high accuracy.

关 键 词：光伏阵列 功率预测 遗传算法 神经网络 

分 类 号：TM615[电气工程—电力系统及自动化]