云模型优化LSSVR的短期电力负荷预测.pdfVIP

云模型优化LSSVR的短期电力负荷预测张捍东。等

云模型优化LSSVR的短期电力负荷预测

ShortTermPowerLoadBasedontheLSSVR

Forecasting

OptimizedbyCloudModel

绻挥东1旅荔1场敌2

摘要：准确的短期电力负荷预测是电力系统安全经济运行的重要依据。针对短期负荷影响因素的非线性特性，研究了基于最tJ,一-

真结果表明，与其他进化算法相比。云进化算法优化LSSVR模型具有更高的预测精度和鲁棒性。

关键词：最小二乘支持向量机云模型人工智能技术遗传算法进化算法负荷预测

中图分类号：TPl81文献标志码：A

shorttennloadisbasisofsafeandeconomicoftheaccordance

Abstract：Accurate

powerforecastingsignificantoperationpowersystem．In

termmethodonleast

withthenonlinearcharacteristicsofthefactorsshorttermloadbased

affectingload，thelongforecastingsquaresupport

vectorresearched．Inviewoftheditficultiesofselectionforshorttermloadmodel

regression(LSSVR)isparameterLSSVR，theforecasting

isLSSVRcloudmodelbasedmodel

basedonthewith

(CBEA—LSSVR)whichoptimizedevolutionaryalgorithm(CBEA)isproposed．This

LSSVRcloudloadatnextisforecasted．Theresultofsimulation

for

implementsparameteroptimizationbyusingmodel，thepowerperiodrolling

LSSVRanditoffer

showsthatwithothercloudmodelbasedwellmakes

comparingevolutionaryalgorithms；theevolutionaryalgorithmoptimizes

androbustness．

higherforecastingaccuracy

vectorCloudmodelArtificialGenetic

Keywords：Leastsquaresupportmachine(LSSVM)intelligencetechniquealgorithmEvolutionary

Load

algorithmforecasting

0引言l短期电力负荷预测概述

短期负荷预测是指在负荷的历史数据基础上．结国内外学者提出了多种短期电力负荷预测方法。

合