浅谈基于改进回归法的电力负荷预测(pdf 6页).pdfVIP

30 卷 1 期 电 网 技 术 Vol. 30 No. 1 2006 年1 月 Power System Technology Jan. 2006 文章编号：1000-3673 （2006 ）01-0099-06 中图分类号：TM715 文献标识码：A 学科代码：4704051 基于改进回归法的电力负荷预测 钷， 敏，刘涤尘 （武汉大学 电气工程学院，湖北省 武汉市 430072 ） Power Load Forecasting Based on Improved Regression LI Po，LI Min，LIU Di-chen （School of Electrical Engineering，Wuhan University，Wuhan 430072 ，Hubei Province，China ） ABSTRACT: A load forecasting method based on improved 为长期、中期、短期和超短期预测。中期和长期预 regression is proposed. On the basis of historical data analysis 测主要用于厂网规划和水库调度等，预测周期的范 and parameter estimation, firstly the singular values are rej ected 围分别为1~ 12月和1~ 10年；短期负荷预测可预测未 by ridge regression, then the main factors impacting the load are 来1~7 日或稍长时间的负荷，用于发电机组的经济 extracted by principal component regression and the analytical 调度、制定功率交换计划等；超短期负荷预测预测 form of the model is obtained. Applying this method to actual power system, the results show that not only the proposed 1天之内的负荷，主要用于预防控制、紧急处理和 [1,2] method is suitable to short-term load forecasting, but also to ultra 频率控制等 。 short-term load forecasting. In addition, the fuzzy functions for 负荷与很多因素存在强相关的线性关系，本文 specified factors are built, and in ultra short-term load forecasting 提出了一种基于改进回归法的负荷预测方法，在对 the similar load days are extracted by clustering analysis. 历史数据进行分析和参数估计的基础上，先用岭回 Through different simplification the proposed method ca