基于模糊神經(jīng)網(wǎng)絡(luò)的斜軋建模研究

《基于模糊神經(jīng)網(wǎng)絡(luò)的斜軋建模研究》由會員上傳分享，免費(fèi)在線閱讀，更多相關(guān)內(nèi)容在學(xué)術(shù)論文-天天文庫。

1、的模糊神經(jīng)網(wǎng)絡(luò)，及其輸入?yún)?shù)的隸屬函數(shù)。訓(xùn)練數(shù)據(jù)通過有限元建模和實(shí)驗數(shù)據(jù)得到。因此，主要工作包括以下幾個方面：(1)在深入了解斜軋無縫鋼管生產(chǎn)工藝和理解斜軋理論的基礎(chǔ)上，運(yùn)用ANSYS有限元軟件對其進(jìn)行了建模仿真，采集到了模糊神經(jīng)網(wǎng)絡(luò)所需的樣本數(shù)據(jù)；(2)建立無縫鋼管斜軋過程的模糊神經(jīng)網(wǎng)絡(luò)，選取了隸屬函數(shù)，將樣本數(shù)據(jù)分成兩組：訓(xùn)練數(shù)據(jù)和檢驗數(shù)據(jù)。(3)N用有限元模擬得到的數(shù)據(jù)對模糊神經(jīng)網(wǎng)絡(luò)進(jìn)行訓(xùn)練。(4)通過用鋁管代替鋼管進(jìn)行斜軋規(guī)律研究，搭建了實(shí)驗平臺，驗證了有限元模擬斜軋的準(zhǔn)確性和可靠性。通過基于模糊神經(jīng)網(wǎng)絡(luò)斜軋建模研究，建立了工藝參數(shù)與質(zhì)量參數(shù)的關(guān)系，對斜軋工藝具有指導(dǎo)意義。

2、同時，得到的模糊神經(jīng)網(wǎng)絡(luò)的數(shù)學(xué)模型，為實(shí)現(xiàn)斜軋過程自動控制奠定基礎(chǔ)。關(guān)鍵詞：斜軋；模糊控制；模糊神經(jīng)網(wǎng)絡(luò)；建模ABSTRACTCross—rollingisanimportantproductiontechnologyintheproductionofseamlesspipe．Ithasbecomeabasictechnologyandakeymeansforallkindsofseamlesspipeproductionafter100years’development．Butitisoftendifficultyforpeopletobuildtheexactmathmodel

3、byusingclassicmechanicsapproach．ByusingFEM，thecalculatingprecisionincross—rollingprocessishigherthanusingothermethods．Butthesimulationtimeistoolongtobeon-linecontrolmathematicsmodels．ItiSenforcedtocarryonthedevelopmentandresearchonmathematicsmodelsofartificialneuralnet．Thisprojectisastudyofcro

4、ss—rollingprocesstopipe—elongatingoperationintheproductionofseamlesspipe．Themappingrelationshipbetweentechnicalparametersandqualityparametersisobtainedbycalculatingparametersincross．rollingprocess。suchasdeformationstate，mechanicalstate，pipequalityect．Andthefuzzyneuralnetworkandmembershipfuncti

5、onofinputparametersincross．rollingprocessareestablished．TrainingdataaregainedthroughFEMmodelingandexperimentaldata．Therefore，themainworkincludesthefollowingaspects：(1)Basedonadeepknowledgeofproductiontechnologyofcross‘rollingseamlesspipeandcross．rollingtheory,sampledatafortheneedoffuzzyneuraln

6、etworkarecollectedbyANSYS／LS—DYNAformodelingandSimulation．(2)Establishfuzzyneuralnetworkaboutcross—rollingprocessofseamlesspipe；selectmembershipfunction；anddividesampledateintotwogroups：trainingdateandtestingdate．(3)TrainfuzzyneuralnetworkbyusingexperimentaldataandFEMsimulationdata．(4)Buildane

7、xperimentalplatformandverifytheveracityandreliabilityoffinite．elementmodelingofcross．rollingthroughstudyingtheprincipleofcross—rollingbyusingaluminumpipeinsteadofsteelpipe．Basedonthemodelingresearchonfuzzyneuralnetworkofcross—rolling，th