基于随机森林算法的中短期用电量预测  

Medium and short-term electricity demand prediction based on random forests algorithm

作　　者：乔黎伟 王静怡 郭炜 李国文 韩俊杰 QIAO Liwei;WANG Jingyi;GUO Wei;LI Guowen;HAN Junjie(Institute of Economic and Technology, State Grid Jiangsu Electric Power Co., Ltd., Nanjing 210008,China;Beijing TsingsoftInnovation Technology Co., Ltd., Beijing 100085, China)

机构地区：[1]国网江苏省电力有限公司经济技术研究院,江苏南京210008 [2]北京清软创新科技股份有限公司,北京100085

出　　处：《电力科学与技术学报》2020年第2期150-156,共7页Journal of Electric Power Science And Technology

摘　　要：传统电力预测模型不能很好地将多种影响因素纳入考虑,也无法对关联因素进行筛选。针对该类问题,该文将信息论中的互信息及人工智能随机森林算法引入中短期用电量预测中。互信息可以根据多种变量与用电量间的平均互信息值大小辨识出关联性高的因素,不同产业可能高关联于不同变量。考虑不同关联因素,采用随机森林算法对不同产业进行针对性预测建模。以江苏省的用电量数据作为实际算例,并将上述方法与未采用互信息的方法以及未针对性分产业建模进行对比。仿真结果表明,上述方法具有科学性和有效性,且有较高的预测精度。Traditional power forecasting models can not efficiently take various factors into account,neither to identify the relation factors.In this paper,mutual information in information theory and the artificial intelligence random forests algorithm are introduced into the medium and short-term electricity demand prediction.Mutual information can identify the high relation factors based on the value of average mutual information between a variety of variables and electricity demand.Different industries may be highly associated with different variables.The random forests algorithm is used to build the different industries forecasting models according to the different correlation factors.The data of electricity consumption in Jiangsu Province is taken as a practical example.In the example,the above methods are compared with the methods without mutual information and the industries.The simulation results show that the above method is scientific,effective,and can provide higher prediction accuracy.

关 键 词：电力系统 中短期负荷预测 互信息 随机森林算法 

分 类 号：TM863[电气工程—高电压与绝缘技术]