摘要
为快速获得改善车辆横向平稳性的最优悬挂参数,提出基于自适应模拟退火算法和非线性序列二次规划算法的组合优化策略对动车组悬挂参数进行优化设计。建立动车组动力学模型,利用最优拉丁超立方抽样方法选取对横向平稳性影响较大的悬挂参数作为设计变量;以横向平稳性为目标函数构建Kriging代理模型,并利用可决系数检验代理模型精度;采用自适应模拟退火算法对代理模型进行全局范围内初步寻优,在初步最优解的基础上采用非线性序列二次规划算法进行局部空间精确求解。研究结果表明,基于Kriging代理模型和组合优化策略的优化效率明显提高,车辆横向平稳性得到显著改善,并且优化前后运行稳定性均满足要求。
In order to obtain the optimal suspension parameters for improving the lateral stability of the vehicle,a combinatorial optimization strategy based on adaptive simulated annealing algorithm and non-linear programming by quadratic lagrangian algorithm is proposed to optimize the suspension parameters of EMU.The dynamic model of the EMU is established,and the design variables are regarded as the suspension parameters that affect the lateral stability by using the optimal Latin hypercube method.Kriging surrogate model is constructed with the objective function of the lateral stability,and the fitting accuracy of the surrogate model is tested by the coefficient of determination.The surrogate model is initially optimized by using the adaptive simulated annealing algorithm in the global scope,and the non-linear programming by quadratic lagrangian algorithm is used to solve the problem accurately in the local scope on the basis of the initial optimal solution.Research results show that the optimization efficiency is obviously improved based on Kriging surrogate model and the combinatorial optimization strategy.The lateral stability of the vehicle is significantly improved,the running stability meets the requirements before and after optimization.
作者
张昕磊
李永华
王衍
李继康
ZHANG Xinlei;LI Yonghua;WANG Yan;LI Jikang(School of Locomotive and Rolling Stock Engineering,Dalian Jiaotong University,Dalian 116028,China;School of Computer and Communication Engineering,Dalian Jiaotong University,Dalian 116028,China)
出处
《机械与电子》
2023年第1期75-80,共6页
Machinery & Electronics
基金
国家自然科学基金资助项目(51875073)。
关键词
悬挂参数
横向平稳性
组合优化策略
最优拉丁超立方
Kriging代理模型
suspension parameters
lateral stability
combinatorial optimization strategy
optimal Latin hypercube
Kriging surrogate model
作者简介
张昕磊(1997-),男,河北唐山人,硕士研究生,研究方向为车辆结构分析与现代设计方法;李永华(1971-),女,黑龙江青冈人,博士,教授,研究方向为车辆结构分析与现代设计方法、车辆结构疲劳可靠性分析、机械产品数字仿真与优化设计、质量与RAMS工程。
