Variable-section multi-objective optimization of drive axle housing based on RSM and NSGA-Ⅱ
For the sake of the comprehensive performance of a commercial vehicle drive axle housing,the response surface method(RSM)combined with non-dominated sorting genetic algorithm(NSGA-Ⅱ)was used for optimizing the variable cross section of the axle housing.The parametric model of the axle housing was established based on Creo,and the static and modal finite element simulation was carried out with Workbench.The variable cross-section of the axle housing after parametric sensitivity analysis was taken as the design variable,and the maximum deformation,first-order frequency and weight of the axle housing were taken as the optimization objectives.The sample space was generated by the central composite design(CCD)method.Based on the sample data,a second-order response surface model reflecting the relationship between test factors and performance indexes was constructed,and the fitting accuracy was verified.On this basis,NSGA-Ⅱ was used to optimize the axle housing with variable section,and the Pareto frontier solution satisfying the conditions was obtained.The results showed that the optimized drive axle housing could reduce the weight by 6.61%,reduce the maximum deformation by 7.28%,and increase the first-order frequency by 2.23%when the performance index meets the design requirements.
idrive axle housingfinite element analysismulti-objective optimizationresponse surface modelNSGA-Ⅱ algorithm
NETL
NSTL
万方数据
维普
