基于遗传算法的铝合金转向节差压铸造工艺参数优化.pdfVIP

HotWorkingTechnology2020,Vol.49,No.l9

基于遗传算法的错合金转向节差压铸造

工艺参数优化

陈川川，苏小平，周大双，李智

南京工业大学机械与动力工程学院，江苏南京211800)

摘要：以铝合金转向节为研究对象，以铸造模拟软件预测的结果为质量指标，应用正交试验法进行试验设计，基

于BP神经网络建立差压铸造工艺参数与质量指标之间的非线性映射关系。采用遗传算法对工艺参数进行优化，获得

近似最优解。这从一定程度上有助于改变合理的铸造工艺参数依赖于工程师经验获得的现状，为汽车铝合金转向节差

压铸造工艺参数的优化提供了重要参考。

关键词：转向节；差压铸造；遗传算法；BP神经网络；工艺优化

DOI:10.14158/j.cnki.1001-3814.

中图分类号：TG249.2文献标识码:A文章编号：1001-3814(2020)19-0066-04

OptimizationofDifferentialPressureCastingProcessParameters

ofAluminumAlloySteeringKnuckleBasedonGA

CHENChuanchuan,SUXiaoping,ZHOUDashuang,LIZhi

(SchoolofMechanicalandPowerEngineering,NanjingTechUniversity,Nanjing211800,China)

AbstractAluminum：alloysteeringknucklewastakenastheresearchobject,castingsimulationsoftwareprediction

resultsweretakenasthequalityindex,theorthogonaltestmethodwasappliedfortheexperimentdesign,andthenon-linear

mappingrelationshipbetweencastingprocessparametersandqualityindexwasestablishedbasedonbackpropagation(BP)

neuralnetwork.Thegeneticalgorithmwasusedtooptimizetheprocessparameters,andtheapproximateoptimalsolutionwas

obtained.Inacertainextent,itishelpfultochangethecurrentsituationthatreasonablecastingprocessparametersare

obtainedbasedonengineersexperience,andprovidesanimportantreferencefortheoptimizationofdifferentialpressure

castingprocessparametersofautomotivealuminumalloysteeringknuckle.

Keywords：steeringknuckle;differentialpressuredie-casting;geneticalgorithm;backpropagatio