基于灰色理论和BP神经网络串联组合的能源需求预测  

Energy Demand Forecasting Based on the Series Combination of Gray Theory and BP Neural Network

作　　者：徐志芬[1] 王书俊[2] 

机构地区：[1]华北电力大学经济管理学院,河北保定071000 [2]淄博供电公司,山东淄博255000

出　　处：《山东电力高等专科学校学报》2013年第4期7-12,共6页Journal of Shandong Electric Power College

基　　金：河北省社科基金项目(HB11YJ011)

摘　　要：本文对某省能源现状进行了分析,进而提出了能源需求预测的必要性,在利用统计数据分析某省能源消耗趋势后,利用灰色GM(1,1)模型进行预测,然后把预测结果和影响因素同时作为BP神经网络的输入进行预测,结果表明,基于GM-BP串联组合模型的预测精度要高于单纯地GM(1,1)预测。最后分析了某省新能源的发展优势。The energy situation in a province was analyzed in this article,and proposing the need for energy de-mand forecasting.After analyzing energy consump-tion trends in a province according to the statistical data, and to predict the energy demand base on gray GM （1,1）,then we use the predicting results and impact factors at the same time as the input of BP neural network to predict ,the results show that the prediction accuracy of GM-BP series model combi- nation is higher than GM （1,1） model.Finally,we analyze the development superiority of new energy in a province.

关 键 词：GM(1  1)模型 BP神经网络 串联组合预测 

分 类 号：TM715[电气工程—电力系统及自动化]