基于神经网络的盾构液压推进系统的参数辨识（英文）-豆柴文库

您所在位置：网站首页 / 基于神经网络的盾构液压推进系统的参数辨识（英文）.docx / 文档详情

在线预览结束，喜欢就下载吧，查找使用更方便

5 金币

下载文档

/ 2

下载提示

如果您无法下载资料，请参考说明：

1、部分资料下载需要金币，请确保您的账户上有足够的金币

2、已购买过的文档，再次下载不重复扣费

3、资料包下载后请先用软件解压，在使用对应软件打开

文本预览

基于神经网络的盾构液压推进系统的参数辨识（英文）
Parameteridentificationofshieldhydraulicpropulsionsystembasedonneuralnetwork
Abstract:
Shieldhydraulicpropulsionsystemisacomplexnonlinearsystemwithmanyparametersthataredifficulttomeasure.Inordertoobtainaccurateandreliablecontrolofthesystem,itisnecessarytoidentifyitsparameters.Thispaperproposesaparameteridentificationmethodbasedonneuralnetworkforshieldhydraulicpropulsionsystem.Themethodusesbackpropagationalgorithmtotraintheneuralnetwork,andthenusesthetrainedneuralnetworktoidentifytheparametersofthesystem.Theeffectivenessandaccuracyoftheproposedmethodareverifiedbysimulationexperiments.Theresultsshowthatthemethodhasgoodaccuracyandrobustness,andcaneffectivelyidentifytheparametersoftheshieldhydraulicpropulsionsystem.
1.Introduction
Theshieldhydraulicpropulsionsystemisanimportantcomponentofthemoderntunnelboringmachine(TBM).ItsfunctionistopushtheTBMforwardbyinjectinghydraulicfluidintothepiston-cylindersystem.Theaccuracyandreliabilityoftheshieldhydraulicpropulsionsystemaredirectlyrelatedtotheboringspeed,efficiency,andsafetyofthetunnelconstruction.However,theshieldhydraulicpropulsionsystemisacomplexnonlinearsystemwithmanyparametersthataredifficulttomeasure.Inordertoobtainaccurateandreliablecontrolofthesystem,itisnecessarytoidentifyitsparameters.
2.Parameteridentificationmethodbasedonneuralnetwork
Neuralnetworkisawidelyusedmethodinsystemidentification.Inthispaper,aparameteridentificationmethodbasedonneuralnetworkisproposedfortheshieldhydraulicpropulsionsystem.Themethodconsistsoftwosteps:neuralnetworktrainingandparameteridentification.
2.1Neuralnetworktraining
Theneuralnetworkusedinthismethodisathree-layerfeedforwardnetwork,includinganinputlayer,ahiddenlayer,andanoutputlayer.Theinputlayerconsistsofthevaluesofinputvariables,suchaspistondiameter,cylinderlength,fluidviscosity,etc.Theoutputlayerconsistsofthevaluesofoutputvariables,suchaspistonforce,hydraulicpressure,etc.Thenumberofneuronsinthehiddenlayerisdeterminedbytrialanderror.Inthispaper,wechoose10neu