組合式多分力計及神經(jīng)網(wǎng)絡的應用.pdf

1、浙江大學碩士學位論文組合式多分力計及神經(jīng)網(wǎng)絡的應用姓名：王川申請學位級別：碩士專業(yè)：機械制造及自動化指導教師：齊津李培玉20000101AbstractBeingcombinedwiththeresearchanddevelopworkoftlleNinthFiveyearproject——“SixcomponentDynamometeranditsCalibrationSystem”thisdissertationmainlyacco

2、mplishestwopartsofwork：themechanicalemulateforAssembledMulticomponentDynamometerandtheresearchofmathematicalmodalityoftheMulticomponentDynamometerThedissertationincludessevenchaptersandtllecontentsareasfollow：Inchapter1，

3、theconceptionofmulti—componentmeasuringtechnologyandthesurveyofitsdevelopmentareelaborated，thentheresearchbackgroundandmaincontentsoftllisdissertationiSintroducedInchapter2，theframeworkofAssembledMulticomponentDynamomete

4、risintroduced，thenthemethodsareusedtogettheforcetransfermatrix：theoreticallydeductionandcomputeofFiniteElementMethod(FEM)Inchapter3，thestabilityoftheelement，whichisusedtoeliminatethecouplingcausedbyadjectivecomponen4isan

5、alyzedSotheoptimalvalueofthereed’Sthicknessiseduced，whichisthepivotaldimensionoftheelementInchapter4，thecommonmathematicalmodalityofMulticomponentDynamometerisexpafimedTwomethodsareintroducedtoobtaintheapplicationmodal：1

6、establishingthecalibrationequationsandresolvingthem2Stepby—stepRegressionAnalyzeInchapter5，NeuralNetworkTechnologyanditsapplicationsintestingdatatreatmentisbrieflyintroducedinchapter6，commonproblemsofMulti—componentDynam

7、ometerNeuralNetworkisexpatiateda“SixinSixOut”Multi—componentDynamometerNeuralNetworkisconstructed，emulationexperimentistakentochecktheeffectInchapter7，improvedmodalofMulti—componentDynamometerNeuralNetworkisbroughtforwar

8、dInordertotrainingandtestingthisnewmathematicalmodalitysometestingdataareinputtedintothemodal，andtheeffectisappraisedAtlast，someconclusionsofthewholedissertationwasdrawnandtheprospectoftheresearchisbroughtforwardsKeyword