基于GANO算法的电力系统中长期负荷预测研究  

Medium and long-term load forecasting based on multifactor-influenced gray artificial neural network optimal combination algorithm in power system

作　　者：顾曦华[1] 邢棉[2] 牛东晓[1] 程利敏[1] 

机构地区：[1]华北电力大学工商管理学院,河北保定071003 [2]华北电力大学数理学院,河北保定071003

出　　处：《华北电力大学学报（自然科学版）》2007年第6期32-36,共5页Journal of North China Electric Power University：Natural Science Edition

基　　金：国家自然科学基金资助项目(50077007);高等学校博士点专项基金(20040079008)

摘　　要：进行负荷预测时,由于中长期负荷历史数据较少而制约因素较多,因此预测难度较大。在分析了灰色预测和神经网络预测优缺点的基础上,提出了多因素灰色神经网络组合预测模型(GANO)。该模型首先采用灰色GM(1,n)模型处理多因素的影响,进而利用BP神经网络训练电力历史负荷数据,最后利用统计方差的倒数建立较为理想的优化组合预测模型。该优化模型结合了各模型优点且综合考虑了电力负荷的多种制约因素。经算例验证,优于单一历史负荷预测模型,有效地提高了中长期负荷预测精度。Because of less history load data and more restricting factors, it is more difficult to predict medium and long-term load. The paper analyzes the merits as well as defects of grey forecasting method and artificial neural net- work （ANN） method and proposes a forecasting method named multifactor-influenced gray artificial neural network optimal combination algorithm （GANO）. The author firstly utilized grey GM （ 1, n ） to deal with the multifactor-influenced, then used BP neural network to train the history load data. Finally, reciprocal of statistic variance was used to create the optimal combination forecasting model. The new model synthesizes various restricting factors of power load and combines the merits of every model. It is verified that the new model is superior to single history load fore.sting model and can effectively improves the accuracy of power load forecasting.

关 键 词：电力负荷预测 人工神经网络 灰色 优化组合 精度 

分 类 号：TM715[电气工程—电力系统及自动化]