一种电力系统短期负荷预测的RBF优化算法  

Optimal Algorithm of Electric Power System's Shortdated Burden Forecasting Based on Radial Basis Function Neural Network

作　　者：胡迎松[1] 陈明[1] 范志明 

机构地区：[1]华中科技大学计算机科学与技术学院,湖北武汉430074 [2]武汉铁路建设集团有限公司,湖北武汉430023

出　　处：《华中科技大学学报（城市科学版）》2003年第3期8-10,共3页Journal of Huazhong University of Science and Technology

摘　　要：针对径向基函数网络在电力系统短期负荷预测中的应用,将神经网络训练的冲量改进算法运用到径向基函数网络的训练中,提高了训练收敛速度和精度.详细介绍了短期负荷预测的建模过程,实验表明该优化算法能有效改善预测性能.Contraposing the application of radial basis function(RBF) neural network on forecasting electric power system's shortdated burden the impulse algorithm of neural network is applied to RBF's training process. This optimizing algorithm can improve the speed and precision of convergence. Furthermore the modeling steps of forecasting electric power system's shortdated burden is introduced. The experimental results show that the optimizing algorithm is effective.

关 键 词：电力系统 短期负荷预测 径向基函数 神经网络 冲量 

分 类 号：TM715[电气工程—电力系统及自动化]