铁道科学与工程学报2016，Vol.13Issue(10)：2056-2063,8.
基于混合代理模型的高速轨道车辆悬挂参数多目标优化
Multi-objective optimization of the suspension parameters for high speed rail vehicle based ona hybrid surrogate model
摘要
Abstract
Both the riding quality and traveling stability of the rail vehicle are affected by the suspension parame-ters , and the comprehensive dynamics performance of the rail vehicle can be improved by the multi-objective op-timization of the suspension parameters. The Kriging, second order response surface and RBF surrogate models of the lateral Sperling index, derailment coefficient and reduction ratio of wheel load are constructed. In these mod-els, design variables are regarded as suspension parameters that affect both the riding quality and traveling stabil-ity. Three kinds of surrogate models with the highest fitting accuracy on the corresponding dynamics index are se-lected to establish the hybrid surrogate model based on the coefficients of determination ( R2 ) . On this basis, the multi-objective optimization model of suspension parameters is established based on the objective function con-structing with hybrid surrogate model. Then, the optimal values of suspension parameters are obtained by using the NSGA-II algorithm. The multi-body dynamics simulation results indicate that both of the riding quality and traveling stability will improve significantly after the optimization of the suspension parameters.关键词
混合代理模型/Kriging/RBF/RSM/悬挂参数/NSGA-IIKey words
hybrid surrogate model/Kriging/RBF/RSM/suspension parameters/NSGA-II分类
交通工程引用本文复制引用
解欢,杨岳,童林军,曾威,谢素超..基于混合代理模型的高速轨道车辆悬挂参数多目标优化[J].铁道科学与工程学报,2016,13(10):2056-2063,8.基金项目
西京学院科研基金资助项目(XJ130238) （XJ130238）
国家自然科学基金资助项目(51405516) （51405516）
