復(fù)雜殼體類塑件氣輔注射成型工藝優(yōu)化研究.pdf

1、南昌大學(xué)碩士學(xué)位論文復(fù)雜殼體類塑件氣輔注射成型工藝優(yōu)化研究姓名：羅章申請學(xué)位級別：碩士專業(yè)：機(jī)械設(shè)計及理論指導(dǎo)教師：江雄心20110520AbstractABSTRACTHowtoeffectivelyandquicklydetermineprocessparametersofgas—assistedinjectionmolding(GAIM)isaquiteimportantandcomplexprocess，whichdirectl

2、yaffectonqualityofproductsUptOnowitisstillaproblemhowtochoosethekeyprocessparametersandoptimizethemtoobtaingoodproductperformance21inchescolorfrontpanelistakenasstudyobjectOnthebasisoforthogonalexperimentaldesign，GAIMpro

3、cessparametersoptimizationsystemisestablishedbyusingahybridsystemthatcombinesradicalbasisfunction(RBF)neuralnetworkwithgeneticalgorithm(GA)。Itcallbeusedtodetermineoptimumprocessparametercombinationrapidly，andalsoprovidea

4、newsolutionofprocessparameteroptimizationforGAIMThemaincontentsofthisthesisarelistedasfollows：(1)neinfluencefactorsofseveralquiteimportantprocessparametersforthequalityofGAIMpartproductsarecomprehensivestudiednecommondef

5、ectsofGAIMaresummarizedandtherelevantsolutionsaregiven(2)Thepreliminaryoptimizationisdonebasedonorthogonalexperimentaldesignandmold—flow2010simulationmethods(3)theGAIMprocessparametersoptimizationmathematicalmodelisestab

6、lishedbyusingmulti—objectiveoptimizationmethod(4)TheRBFneuralnetworkissetuptoshowtherelationshipofGAIMprocessparametersandthequalityofpartproductsAccordingtothepreliminaryoptimumGAIMprocessparametersgottenbyusingorthogon

7、alexperimentaldesign，trainingsampleisgeneratedtotraintheRBFneuralnetworkandthecorrectnessoftheRBFneuralnetworkiSverified(5)TheGAIMmultiobjectiveprocessparametersoptimizationsystemisbuiltwithRBFneuralnetworkmodelandGATheb