基于响应面方法的车钩缓冲器特性曲线优化分析  

Optimal Analysis of the Hysteretic Characteristic Curve for Draft Gear Based on Response Surface Method

作　　者：常崇义[1] 王成国[1] 王永菲[1] 赵鑫[1] 

机构地区：[1]铁道科学研究院铁道科学技术研究发展中心,北京100081

出　　处：《中国铁道科学》2007年第6期84-90,共7页China Railway Science

基　　金：铁道部科技研究开发计划项目(2006YF13)

摘　　要：为了更加真实地反映车钩缓冲器实际工作状态和进行缓冲器特性曲线优化,提出采用列车纵向动力学系统对缓冲器的特性曲线进行分析。在优化方面,从响应面方法的基本原理出发,建立响应面的子区域构造方法,运用凝聚函数和惩罚技术处理约束的遗传算法求解加约束条件的二次响应面最优解。以大秦线实际线路、列车编组以及列车纵向动力学性能为优化计算的条件,结合缓冲器数值模型建立缓冲器特性曲线的优化模型,运用所提出基于遗传算法的响应面方法对重载列车缓冲器特性曲线进行优化求解。结果说明所提方法对求解缓冲器特性曲线的优化问题非常有效。In order to truly reflect the work state of the draft gear and optimize the hysteretic characteristic curve for draft gear, it is proposed that the Longitudinal Train Dynamics （LTD） should be the most powerful tool for the analysis and simulation of the draft gear model. In the aspect of numerical optimization technique, according to the basal principle of Response Surface Method （RSM）, the adjustment method for the size of RSM successive subregions is built, and the quadratic response surface model with constrains is solved by Genetic Algorithms （GA）, which adopts surrogate constraint handling method and penalizing factor. Computational condition of the optimization is the Daqin railway line, the heavy haul con- nected trains and the danger running case. Based on numerical model of draft gear, the optimal model of its hysteretic characteristic cure is provided. Then an attempt has been made to optimize the optimal model using RSM based on GA. The results demonstrate that the optimal method is efficient and effective in solving the optimal design problem of hysteretic characteristic curve of draft gear.

关 键 词：车钩缓冲器 特性曲线优化 响应面方法 遗传算法 中心复合设计 显著性检验 约束凝聚处理 子区域 纵向动力学 

分 类 号：U270.34[机械工程—车辆工程]