改进粒子群算法在中长期电力负荷组合预测模型中的应用  

Application of improved particle swarm optimization to mid-long-term power load combination forecasting model

作　　者：李小龙[1] 李郁侠 师彪[1] 牛艳利[1] 王煜[1] 

机构地区：[1]西安理工大学水利水电学院,陕西西安710048

出　　处：《武汉大学学报（工学版）》2011年第3期380-382,387,共4页Engineering Journal of Wuhan University

摘　　要：为提高电力负荷预测的精度,提出了基于改进粒子群算法的电力负荷组合预测模型求解方法.该方法以回归分析、比例系数、灰色模型为基础建立负荷组合预测模型,利用改进粒子群算法优化组合预测模型的权值,并与单个预测模型进行比较.预测结果表明,基于改进粒子群算法的电力负荷组合预测模型运算速度快,预测精度高,相对误差小.To forecast power load accurately,the power load combination forecasting model method based on improved particle swarm optimization algorithm is proposed.This method takes regression analysis,proportionality coefficient and grey model forecasting as the basic forecasting model.It optimizes the weight of combination forecasting through the improved particle swarm optimization algorithm.Comparing the forecasting data with the single model＇s data,the results show that the operation speed of combination forecasting model based on improved particle swarm optimization algorithm is fast;at the same time,it has higher forecasting accuracy and little relative error.

关 键 词：改进粒子群算法 电力负荷预测 组合预测 相对误差 

分 类 号：TM715[电气工程—电力系统及自动化]