摘要
为进一步提高光伏发电功率预测精度,提出一种基于相似日和交叉熵理论的光伏发电短期功率组合预测方法.首先采用模糊C均值聚类方法对历史样本数据分类,并提出一种基于隶属度的指标来选取相似日.然后采用最小二乘支持向量机、时间序列法和BP神经网络法分别预测光伏发电功率,通过交叉熵算法动态设置各预测时刻下单一方法的权重值,建立光伏发电功率的组合预测模型.算例结果表明,所提方法能够动态识别单一预测方法包含的信息量,能确定更加合理的权重值,从而提高光伏发电功率的预测精度.
In order to further improve the photovoltaic（PV） power forecasting accuracy, a short-term combination forecasting model based on similar days and cross entropy theory is proposed. Firstly, the fuzzy C-means clustering method is used to classify the historical samples, and a selection index based on membership degree is proposed to select similar days. Then, the LSSVM, ARMA and BP neural network are used to predict the PV power. The weights of three single forecasting methods are dynamically set by the cross entropy algorithm, and the short-term combination forecasting model of PV power is established. The results show that this method can dynamieally identify the information of single methods and obtain appropriate weights. As a result,the forecasting accuracy of PV power can be improved.
作者
季顺祥
王琦
姚阳
陈佳浩
刘瑾
Ji Shunxiang;Wang Qi;Yao Yang;Chen Jiahao;Liu Jin(School of NARI Electrical and Automation,Nanjing Normal University,Nanjing 210042,China;Jiangsu Key Laboratory of Gas and Electricity Interconnection Integrated Energy,Nanjing Normal University,Nanjing 210023,China)
出处
《南京师范大学学报（工程技术版）》
CAS
2018年第2期19-28,共10页
Journal of Nanjing Normal University(Engineering and Technology Edition)
基金
江苏省研究生科研与实践创新计划项目(KYCX17_1078)
关键词
光伏发电
组合预测
相似日
隶属度
交叉熵
photovolatic （PV） power generation
combination forecasting
similar days
membership
cross entropy
