%0 Journal Article
%T fruitflyoptimizationalgorithmbasedhighefficiencyandlownoxcombustionmodelingforaboiler
%A zhangzhenxing
%A sunbaomin
%A xinjing
%J ÈÈÁ¦·¢µç
%P 25-30
%D 2014
%X inordertocontrolnoxemissionsandenhanceboilerefficiencyincoal£¿firedboilers,thethermaloperatingdatafromanultra£¿supercritical1000mwunitboilerwereanalyzed.onthebasisofthesupportvectorregressionmachine(svm),thefruitflyoptimizationalgorithm(foa)wasappliedtooptimizethepenaltyparameterc,kernelparametergandinsensitivelosscoefficientofthemodel.then,thefoa£¿svmmodelwasestablishedtopredictthenoxemissionsandboilerefficiency,andtheperformanceofthismodelwascomparedwiththatofthega£¿svmmodeloptimizedbygeneticalgorithm(ga).theresultsshowthefoa£¿svmmodelhasbetterpredictionaccuracyandgeneralizationcapability,ofwhichthemaximumaveragerelativeerroroftestingsetliesinthenoxemissionsmodel,whichisonly3.59%.theabovemodelscanpredictthenoxemissionsandboilerefficiencyaccurately,sotheyareverysuitableforon£¿linemodelingprediction,whichprovidesagoodmodelfoundationforfurtheroptimizationoperationoflargecapacityboilers.
%K ultra£¿supercritical
%K 1000mwunit
%K boiler
%K efficiency
%K noxemissions
%K supportvectormachine
%K fruitflyoptimizationalgorithm
%U http://rlfd.paperopen.com//oa/darticle.aspx?type=view&id=201412005