相空间重构和溷沌神经网络融合的短期负荷预测研究.pdfVIP

第24 卷 第 1 期 中 国 电 机 工 程 学 报 Vol.24 No. 1 Jan 2004 2004 年 1 月 Proceedings of the CSEE ©2004 Chin.Soc.for Elec.Eng. 文章编号：0258-8013 (2004) 01-0044-05 中图分类号：TM715 文献标识码：A 学科分类号：4704051 相空间重构和混沌神经网络融合的 短期负荷预测研究 孙雅明, 张智晟 （天津大学电气自动化工程学院，天津300072） A NEW MODEL OF STLF BASED ON THE FUSION OF PSRT AND CHAOTIC NEURAL NETWORKS SUN Ya-ming, ZHANG Zhi-sheng (School of Electrical Engineering Automation, Tianjin University, Tianjin 300072, China) ABSTRACT: A new model of Short-Term Load Forecasting 融合的电力系统STLF 方法用于实际运行系统在理论上取得 (STLF) based on the fusion of Phase Space Reconstruction 了有效的进展。 Theory(PSRT) and Improved Chaotic Neural Networks(ICNN) 关键词：短期负荷预测；混沌神经网络；相空间重构理论； is presented in this paper. The ICNN model has strong 最近邻点；遗传算法；神经网络 sensitivity to the initial load value and to the walking of the whole chaotic track. And it can characterize the complicated 1 引言 dynamics behaviors and has global searching optimal ability. 电力系统STLF 是电力系统正常、安全调度操 The input dimension of ICNN is decided by PSRT, and the 作、电力市场运营和经济运行的重要依据。传统多 training samples are formed by means of the stepping dynamic 层前馈神经网络(Neural Networks NN) 已在国内外 space track and nearest neighbor point set of the forecasting [1,2]