VAV空調(diào)的自適應(yīng)模糊控制與通透以太網(wǎng)BAS研究.pdf

1、北京建筑工程學(xué)院碩士學(xué)位論文VAV空調(diào)的自適應(yīng)模糊控制與通透以太網(wǎng)BAS研究姓名：戴正偉申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：控制理論與控制工程指導(dǎo)教師：張少軍20100301北京建筑T程學(xué)院碩fj學(xué)位論文AbstractSinceVariableAirVolume(VAV)systemsislargenon—linearandhysteretic，itisdifficultevenimpossibletoobtaintheaccuratemodel

2、SoinVAVairconditioningsystem，classicalcontrolschemecannotgeteffectivecontrolandusingintelligentcontrolcanovercometheshortcomingsofclassicalcontrolschemeIntheintelligentcontrolmethod，fuzzycontrolsystemhasaneasywaytounders

3、tandtheexpression。butdifficulttoautomaticallygenerateandadjustthemembershipfunctionandfuzzycontrolrules；neuralnetworkhasastrongselfadaptivelearningabilitybutcannotbederivedfrominputoutputrelationshipcouldeasilybeexpresse

4、dinamanneracceptableFuzzylogicandneuralnetworkscombinedcanabsorbtheadvantagesofbothtoovercomethedrawbacksofeachisanovelintelligentcontrolschemeinVAVsystemscontr01Meanwhile，inordertoexplorespecificworkstoachieveintelligen

5、tapproach，itisnecessarytesearchtheBuildingAutomationSystem(BAS)fortheprojectapplicationinVAVsystemscontr01Thispaperstudiedonthestrengthsandweaknessesoffuzzycontrolandneuralnetworkcontrol，combinedbothanddevelopedtheapplic

6、ationoffuzzyneuralnetworkcontrolmethodsintheⅥWsystemsAccordingtothestructureofthefuzzysystemusedneuralnetworklearningalgorithmforfuzzycontrolsystem，anddecidedanequivalentstructurewithneuralnetwork，eachlayerandnodeofneura

7、lnetworkcorrespondstoapartoffuzzysystemThroughneuralnetworklearning。adjustedthecenterpointandwidthofthemembershipoffuzzyset，alsoadjustedmembershipfunctionshapeon—line，andusedfuzzyadaptivecontrolonroomtemperatureinvariabl

8、eairvolumeairconditioningsystemsThefuzzyneuralnetworklearningalgorithmhasbeenimprovedinthepaperbasedontheanalysisoftheactualsituationofthenumberofnodesinthecomplexstructurenetworkWithadditionalmomentummethodandadaptivele

9、arningratemethod，toacceleratethelearningspeedofthenetwork，avoidingthenetworkintoalocalminimumvalueUsedneuralnetworkidentifier(NNI)structureinthefeedbackchannelinordertoprovideonlinelearningforteacher’Scontrollersignals，i