企业电能耗的BP神经网络预测模型  

Forecasting model of electric energy consumption with BP neural network

作　　者：张有山[1] 李光远 邢军[1] 郭良栋[1] 

机构地区：[1]辽宁科技大学理学院,辽宁鞍山114051 [2]国家电网鞍山供电公司,辽宁鞍山114001

出　　处：《辽宁科技大学学报》2014年第4期343-349,共7页Journal of University of Science and Technology Liaoning

基　　金：辽宁科技大学大学生创新创业训练计划项目

摘　　要：在我国节能减排的大环境下,企业的电能耗预测问题一直都备受关注。准确预测电能耗对企业决策有重要的意义。本文以鞍山某矿山公司电能耗历史数据为研究对象,通过Grubbs检验进行数据的离群检测,以剔除某些失真数据,并用主成分分析法对电能耗的影响因素进行降维处理,建立了BP神经网络的电能耗预测模型。利用Matlab对数据分析,给出了该公司未来一年内各月份的电能耗情况。In the environment of energy saving and emission reduction in China,the forecasting of energy consumption has always been a concerning problem,therefore,it has an important significance for enterprises. In this paper,the data is provided by Anshan mineral company,Grubbs test is applied in the outlier detection of data,the principal component analysis is used to reduce the dimensions of the electrical energy consumption factors,using theory of neural network,a nonlinear multivariable forecasting model is established based on BP neural network. By actual data testing,the electric energy consumption of every month in next year can be predicted.

关 键 词：BP神经网络 电能耗 离群检测 主成分分析 预测模型 

分 类 号：TM761[电气工程—电力系统及自动化]