基于机器学习的自适应光伏超短期出力预测模型.pdfVIP

第 39 卷 第 2 期 电 网 技 术 Vol. 39 No. 2 2015 年 2 月 Power System Technology Feb. 2015 文章编号：1000-3673 （2015 ）02-0307-05 中图分类号：TM 721 文献标志码：A 学科代码：470·4051 基于机器学习的自适应光伏超短期出力预测模型 高阳，张碧玲，毛京丽，刘勇 （北京邮电大学网络教育学院，北京市 海淀区 100876 ） Machine Learning-Based Adaptive Very-Short-Term Forecast Model for Photovoltaic Power GAO Yang, ZHANG Biling, MAO Jingli, LIU Yong (School of Network Education, Beijing University of Post Telecommunications, Haidian District, Beijing 100876, China) ABSTRACT: For the reason that forecast performance of solar 果表明：所提的自适应预测模型在均方根误差(root mean radiation and cloud cover information is still very poor in square error，RMSE) 、平均绝对百分比误差(mean absolute China, the accuracy improvement of direct prediction for percentage error ，MAPE) 和希尔不等系数(Theil inequality photovoltaic (PV) power by introducing meteorological data is coefficient，TIC)上性能最好。 limited. To further improve the prediction accuracy, a 关键词：自适应预测；自回归和滑动平均模型；神经网络； self-adapting very-short-term forecast model for PV power is 小波分析；超短期光伏出力预测 proposed based on the feature mining of historical PV output data. Firstly, the parameters of support vector machine (SVM) DOI ：10.13335/j.1000-3673.pst.2015.02.002 classifier are calculated using wavelet analysis and 0 引言 characteristics of the historical PV output data. Then, based on the established SVM classifier, the power curve type of the 光伏发电作为一种重要的绿色能源，近年来在 sequential 15 minutes is determined through historical PV 国内外得到了快速的发展。但是其出力会跟随太阳 output data of the previous 30 minutes. An appropriate f