摘要
提出基于序列响应面模型的方法结合遗传算法的优化思路,对某全承载式大客车的振动频率进行优化。在有限元模型的基础上,借助灵敏度分析,结合零件分组,选取一部分重要结构件的厚度作为设计变量,以质量不超过原设计质量作为约束,在Matlab环境下对指定的振动频率分别进行优化。优化过程中将遗传算法每轮迭代得到的最优解作为新一轮有限元分析的输入,并用分析结果更新响应面模型。优化结果表明,该客车被优化的振动频率有明显的提高,所采用的方法有效减少了计算的相对误差,改善了优化效果。
With an idea proposed of adopting sequential response surface method(SRSM) combined with genetic algorithm(GA),an optimization is performed on the vibration frequencies of a bus with unitized body structure.Based on its finite element model,by means of sensitivity analysis and part grouping,with the thickness of some key structural parts as design variables and the requirement of 'mass not exceeding original one' as constraint,the designated vibration frequencies are optimized respectively in Matlab environment.In the process of optimization,the optimal solution obtained in each GA loop is taken as the input of the new run of FEA with the result used to update the response surface model.The optimization results show that the optimized vibration frequencies of the bus are raised remarkably,the relative error is reduced effectively,and the optimization effect is improved with the method adopted.
出处
《汽车工程》
EI
CSCD
北大核心
2010年第10期833-836,887,共5页
Automotive Engineering
基金
国家“863”计划项目(2007AA04Z133)资助