|
热力发电 2014
doublehiddenlayerrbfprocessneuralnetworkbasedonlinepredictionofsteamturbineexhaustenthalpy, PP. 36-40 Keywords: steamturbine,exhaustenthalpy,rbfprocessneuralnetworkwithtwohiddenlayers,onlineforecasting Abstract: inordertodiagnosetheuniteconomicperformanceonline,theradialbasisfunction(rbf)processneuralnetworkwithtwohiddenlayerswasintroducedtoonlinepredictionofsteamturbineexhaustenthalpy.thus,themodelreflectingcomplicatedrelationshipbetweenthesteamturbineexhaustenthalpyandtherelativeoperationparameterswasestablished.moreover,theenthalpyoffinalstageextractionsteamandexhaustfroma300mwunitturbinewastakenastheexampletoperformtheonlinecalculation.theresultsshowthat,theaveragerelativeerrorofthismethodislessthan1%,sotheaccuracyofthisalgorithmishigherthanthatofthebpneutralnetwork.furthermore,thismethodhasadvantagesofhighconvergencerate,simplestructureandhighaccuracy.
|