基于RSM模型及NSGA-Ⅱ算法的低地板车辆曲线通过性能优化  

Optimization of Curving Performance for Low Floor Rail Vehicles Based on RSM Model and NSGA-Ⅱ Genetic Algorithm

作　　者：杨阳[1] 丁军君[1] 李芾[1] 戚壮[1] 蒋宽 

机构地区：[1]西南交通大学机械工程学院,四川成都610031 [2]广州电力机车有限公司技术中心,广东广州510850

出　　处：《铁道学报》2017年第3期25-30,共6页Journal of the China Railway Society

基　　金：国家自然科学基金(51305359);中央高校基本科研业务费(2682016CX029)

摘　　要：为了提高优化效率并改变通过单一参数变化对车辆动力学性能进行优化的方式,提出一种基于RSM近似模型及NSGA-Ⅱ遗传算法的车辆曲线通过性能优化方法,以获得更加精确的结果并缩短设计周期。通过建立70%低地板有轨电车动力学模型,以悬挂参数为设计变量,以曲线通过性能指标为响应拟合其RSM近似模型,在近似模型的基础上对曲线通过性能进行多参数、多目标优化。优化结果表明,与原始设计优化方式相比,轮轨横向力、轮轴横向力、脱轨系数及轮重减载率均有不同程度的降低,且优化效率得到大幅提高。In order to improve optimization efficiency and change the way of optimizing vehicle dynamic performance by variation of a single parameter,this paper proposed an approach which could optimize vehicle performance on the curve track based on approximate RSM model and genetic algorithm NSGA-Ⅱ,aiming to obtain more accurate results and shorten design cycle.By establishing the dynamic model of 70%low-floor trams,with suspension parameters as design variables,and the curving performance index as response,to fit,its RSM approximation model,the multi-parameter and multi-objective optimization was conducted on the curving performance.The optimization results showed that this method,compared with the original design optimization method,delivered reduced wheel-rail lateral force,wheel axle lateral force,derailment coefficient and wheel load reduction rate,as well as greatly improved optimization efficiency.

关 键 词：低地板车辆 RSM模型 NSGA-Ⅱ遗传算法 曲线通过性能 性能优化 

分 类 号：U270.2[机械工程—车辆工程]