基于多特征分析提取的随机森林超短期光伏功率预测  

Ultra-short-term photovoltaic power prediction for random forests based on multiple feature analysis and extraction

作　　者：张程珂 刘会灯 朱渝宁 贾凡 郭恒青 张金良[3] ZHANG Chengke;LIU Huideng;ZHU Yuning;JIA Fan;GUO Hengqing;ZHANG Jinliang(State Grid Chongqing Power Supply Company,Chongqing 400014,China;State Grid Chongqing Urban Power Supply Company,Chongqing 400015,China;North China Electric Power University,Beijing 102206,China)

机构地区：[1]国网重庆市电力公司,重庆400014 [2]国网重庆市区供电公司,重庆400015 [3]华北电力大学,北京102206

出　　处：《电力需求侧管理》2023年第6期50-56,共7页Power Demand Side Management

基　　金：国家自然科学基金项目(71774054)。

摘　　要：随着新能源的大规模利用,光伏渗透率稳步增长,准确的光伏功率预测能为电网企业带来较多效益。基于此提出了一种多特征分析提取的随机森林预测模型,用于超短期光伏功率预测。首先,对收集到的光伏数据进行预处理,清理缺失值和重复值;再次,对影响因素进行相关性分析,选取相关性强的因子;然后,对筛选后的因子进行输入特征量选择,将处理后的特征向量作为预测模型的输入;最后,建立随机森林预测模型,并与BP、RBF、MLP模型对比。实证结果表明,所提模型具有较好的拟合度和更高的预测精度,对光伏预测工作有一定的指导意义。PV penetration is steadily increasing with the large-scale utilization of new energy sources.Accurate PV power prediction can bring more benefits to grid enterprises.Based on this,a random forest prediction model with multi-feature analysis extraction is proposed for ultra-short-term PV power prediction.Firstly,the collected PV data is pre-processed to clean up the miss-ing and duplicate values.Then,correlation analysis is performed on the influencing factors and factors with strong correlation are se-lected.Next,feature engineering is performed on the screened fac-tors and the processed feature vector is used as input of the predic-tion model.Finally,the random forest prediction model is built and compared with BP,RBF and MLP models.Empirical results show that the model proposed has better fit and higher prediction accura-cy,which is of certain guidance for PV prediction work.

关 键 词：光伏发电 功率预测 超短期负荷预测 随机森林 特征值分析 

分 类 号：TM714[电气工程—电力系统及自动化] TK018[动力工程及工程热物理]