应用小波-人工神经网络组合模型研究电力负荷预报  

Hybrid Model of Wavelet and Artificial Neural Network and Its Application on Forecasting Power System Load Time Series

作　　者：王文圣[1] 朱聪[1] 丁晶[1] 

机构地区：[1]四川大学水利水电学院

出　　处：《水电能源科学》2004年第2期68-70,共3页Water Resources and Power

基　　金：国家自然科学基金资助项目(50279023;40271024)。

摘　　要：针对负荷时间序列的非线性和多时间尺度特性,提出了将小波分析与人工神经网络相结合进行负荷预报的方法——小波-人工神经网络组合模型。该模型吸取了小波分析的多分辨功能和人工神经网络的非线性逼近能力。以月、日平均负荷预报为例对模型进行验证,结果表明:该模型的拟合、检验精度较高。Based on the multi-time scale and the nonlinear character of electric load time series, a hybrid model of wavelet and artificial neural network (ANN) is presented. The suggested model has super advantage because of drawing some merits of wavelet and artificial neural network. The forecast accuracy has been increased and the length time of forecast also has been increased. Two cases study, the short and long term forecasting about daily and monthly electricity load of Sichuan province in China, have been researched. The results show that the presented model based upon wavelet analysis and ANN is satisfied.

关 键 词：小坡分析 人工神经网络 组合模型 负荷预报 

分 类 号：TM715[电气工程—电力系统及自动化]