基于族群进化计算的多项式回归电力负荷预测  

Polynomial Regression of Power Load Forecasting Based on Ethnic Group Evolutionary Computation

作　　者：沙宇恒[1] 潘晓英[2] 张小庆[1] 宋晓林[1] 

机构地区：[1]陕西电力科学研究院,陕西西安710054 [2]西安邮电学院,陕西西安710121

出　　处：《陕西电力》2012年第2期27-31,共5页Shanxi Electric Power

基　　金：国家自然科学基金(61105064);陕西省自然科学基金(2011JM8007);陕西省教育厅专项科研计划项目(2010JK837)

摘　　要：提出了1种基于族群进化计算的多项式回归电力负荷预测方法,该方法利用族群进化理论来有效优化多项式回归分析模型中的参数,取得了较好的效果。同时,选择某地区用电量历史数据进行预测分析,并与季节指数模型预测的结果进行了对比,从数据对比分析结果看,该文提出的预测值与实际值的平均误差和相对误差均小于季节指数模型,证明了该方法的可行性和有效性。The polynomial regression of power hind forecasting method based on ethnic group evolutionary computation is proposed. The ethnic theot.＇y of evolution is used to optimize tile parameters in the polynomial regression analysis model, and achieve good results. In experiments, a regional electricity consumption historical data is analyzed, and the predictable results are compared with the seasonal index model. From the comparison of data analysis, the average e, rror and relative error of prcdictive value anti actual value of proposed method in this paper are less than those of seasonal index model. The feasibility and ettectiveness of the method are prow＇d.

关 键 词：多项式回归分析 族群进化算法 参数优化 电力负荷预测 

分 类 号：TM715[电气工程—电力系统及自动化]