摘要
【目的】针对高速车辆悬挂系统优化存在参数多、计算耗时等问题,提出基于参数敏感度分层的高速车辆悬挂系统优化设计方案。【方法】首先,构建高速车辆单车动力学仿真模型并验证模型是否合理,运用最优拉丁超立方抽样法均匀抽取样本点代入动力学模型,计算动力学响应;随后,采用代理模型替代计算耗时的动力学模型以提高优化效率;然后,借助敏感度分析确定优化变量后,对该变量进行分层,对于分层后的两层变量,分别采用近邻培养移植算法、下山单纯形法推进优化流程;最后,对比优化解、原始解和使用非支配排序遗传算法Ⅱ(Non-dominated Sorting Genetic AlgorithmⅡ, NSGA-Ⅱ)得出的结果。【结果】结果表明,在最优解下对非线性临界速度和脱轨系数的优化率分别为14.584%和9.615%,综合优化率高于NSGA-Ⅱ所得结果,并减少了设计迭代次数,改善了高速车辆动力学性能,验证了优化方法的可行性。
[Objective]To address the issues of numerous parameters and time-consuming calculations in the optimization of high-speed vehicle suspension systems,a layered optimization design based on the parameter sensitivity stratification was proposed.[Methods]Firstly,a dynamic simulation model of a single high-speed vehicle was constructed and validated for pragmatic.The optimal Latin hypercube sampling method was utilized to evenly extract sample points for calculating dynamic responses in the model,and a surrogate model was employed to replace the time-consuming dynamic model in order to enhance optimization efficiency.Secondly,after determining the optimization variable through sensitivity analysis,the variable was stratified.For the two stratified variables,the nearest neighbor cultivation transplantation algorithm and the downhill simplex method were used to advance the optimization process.Finally,the optimization results were compared with the original solution and those obtained from the non-dominated sorting genetic algorithmⅡ(NSGA-Ⅱ).[Results]The results demonstrate that the optimization respectively reduces the nonlinear critical speed and derailment coefficient by 14.584%and 9.615%,surpassing the NSGA-Ⅱin comprehensive optimization rate and reducing the design iterations,thereby improving the dynamic performance of high-speed vehicles and validating the feasibility of the optimization method.
作者
武福
杜泽阳
李忠学
杨喜娟
蒋鹏民
WU Fu;DU Zeyang;LI Zhongxue;YANG Xijuan;JIANG Pengmin(College of Mechanical Engineering,Lanzhou Jiaotong University,Lanzhou 730070,China;College of Electronic and Information Engineering,Lanzhou Jiaotong University,Lanzhou 730070,China)
出处
《机械强度》
北大核心
2026年第3期87-95,共9页
Journal of Mechanical Strength
基金
国家自然科学基金项目(56062028)
甘肃省教育厅产业支撑计划项目(2021CYZC-11)
甘肃省教育厅创新基金项目(2022A-036)。
关键词
动力学性能
多目标优化
悬挂参数
参数敏感度分层
代理模型
Dynamic performance
Multi-objective optimization
Suspension parameter
Parameter sensitivity stratification
Surrogate model
