混粉電火花加工工藝及神經(jīng)網(wǎng)絡(luò)仿真研究.pdf

1、大連理工大學碩士學位論文混粉電火花加工工藝及神經(jīng)網(wǎng)絡(luò)仿真研究姓名：鄧宗慶申請學位級別：碩士專業(yè)：機械制造及其自動化指導教師：趙福令20070601混粉電火花加工工藝及神經(jīng)網(wǎng)絡(luò)仿真研究StudyonthemachiningtechnologyandneuralnetworksimulationofpowdermixedEDMAbstractElectricaldischargemachining(EDM)hasbeenwidelyappl

2、iedinthefieldsofaerospace，moldmakingandmanufacturingofsomepartswitllcomplexsnucturesTheappearanceofpowdermixedEDM(PMEDM)extremelyenrichestheeffectofEDMtechnologyTheconditionofdischargingischangingbyusingspecialdielectric

3、mixedwithconductivefinepowdersofsomeconcentrationThusitCalleffectivelyimprovethesurfacequality，decreasethesnrfaeonghness，realizethesurfacefinishingoflargeareaandwidentheapplicationofEDMButthephysicalmechanismofPMEDMisunc

4、learandthemachiningprocessisstillnecessarytobeimproved，furthermore，thisnewtechnologyislackofcriterionsoPMEDMhasn’tbeenwidelyusedtillnowTheeffectruleofpmcessparameterstothetechnologytargetshasbeenanalyzedinthispaperthroug

5、htheorthogonalexperimentofPMEDMMeanwhilethepaperutilizesMATLAB70softwaretosetupthebackpropagation凹)networkmodel，andbyusingthemodelbuiltabovetopredictthemachiningresultsofPMEDMn圮paperhasdonesomeworkasfollows：(1)mmachining

6、processhasbeensystemicstudiedthroughorthogonalexperimentofPMEDMn把effectruleandsignificanceofprocessparametersthatselectedinorthogonalexperimentdesigntomachined$urfacerougimess(SR)andmaterialremovalrate(MRR)havebeenobtain

7、edThedifferenceofSRandIvIRRwithdifferentmachiningpolari鑼isalsostudiedTheresultsarehelpfultomake5111“ethereasonablemachiningparameters(2)AccordingtothechactcrLsticofPMEDMtechnologyanditscomplexitycombiningwiththeadvantage

8、ofartificialneuralnetwork(ANN)，asimulationmodelbasedonbackpropagationalgorithmforforecastingthetechnologyresultsofPMEDMhasbeenestablished(3)Theorthogonalexperimentsampledata啪fullyrepresentthesystemcharacteristicofPMEDMdu