基于遗传算法的实测轮轨蠕滑曲线模拟  

Simulation Research of Measured Wheel-rail Creep Curve Based on Genetic Algorithm

作　　者：张洪吉 安博洋[1,2] 王小韬[3] 陶功权[4] 丁昊昊[5] 王平[1,2] ZHANG Hongji;AN Boyang;WANG Xiaotao;TAO Gongquan;DING Haohao;WANG Ping(MOE Key Laboratory of High-speed Railway Engineering,Southwest Jiaotong University,Chengdu 610031,China;School of Civil Engineering,Southwest Jiaotong University,Chengdu 610031,China;China Railway Eryuan Engineering Group Co.,Ltd.,Chengdu 610031,China;State Key Laboratory of Traction Power,Southwest Jiaotong University,Chengdu 610031,China;School of Mechanical Engineering,Southwest Jiaotong University,Chengdu 610031,China)

机构地区：[1]西南交通大学高速铁路线路工程教育部重点实验室,成都610031 [2]西南交通大学土木工程学院,成都610031 [3]中铁二院工程集团有限责任公司,成都610031 [4]西南交通大学牵引动力国家重点实验室,成都610031 [5]西南交通大学机械工程学院,成都610031

出　　处：《铁道标准设计》2024年第8期71-77,共7页Railway Standard Design

基　　金：国家自然科学基金项目(52108418,U1934214);中央高校基本科研业务费理工科科技创新项目(2682022CG003)。

摘　　要：轮轨黏着-蠕滑特性直接影响着轮轨系统的工作性能和列车的运行安全。为在轮轨接触数值模型中合理地表征实测黏着-蠕滑曲线的影响,在KP模型和改进FASTSIM模型的基础上,提出了黏着-蠕滑特征参数初值计算方法,结合遗传算法设计参数拟合过程并选取合适的适应度函数,形成了一种获取描述轮轨黏着-蠕滑特性特征参数的通用方法,随后以160 km/h速度的实测蠕滑数据为例,分析对比了本文算法拟合蠕滑曲线的计算效率及与理想工况下局部接触结果的差异。研究表明:(1)本文提出的算法在拟合不同速度工况对应的实测黏着-蠕滑曲线时均具有较高的精度,提出的拟合特征参数取值范围提高了遗传算法收敛速度,与遍历法相比计算速度提高18.5倍;(2)考虑潮湿环境实测轮轨黏着-蠕滑曲线时,轮轨黏滑分布、切向接触应力较干态理想工况有显著差异,主要体现在接触斑最大切向接触应力减小、滑动区面积随蠕滑率增大的速度减缓等。本文算法满足车辆-轨道耦合动力学数值仿真要求,可更真实地模拟实际环境条件下的轮轨相互作用。Wheel-rail stick-slip characteristics directly affect the performance of wheel-rail system and the safety of train operation.In order to reasonably characterize the influence of the actual creep curve in the wheal-rail contact model,this paper proposes the initial calculation method of the stick-slip characteristic parameters on the basis of KP model and modified FASTSIM model,and combines with the genetic algorithm to design the parameter fitting process and select the appropriate fitness function.A general method is developed to obtain the characteristic parameters describing wheel-rail stick-slip characteristics.Then,the measured creep data at 160 km/h is taken as an example to analyze and compare the computational efficiency of the proposed algorithm in fitting the creep curve and the difference between the local contact results and ideal conditions.The results show that:(1)The algorithm proposed in this paper has high accuracy in fitting the measured creep curves corresponding to different velocity conditions.The range of fitting characteristic parameters proposed improves the convergence speed of the genetic algorithm,which is 18.5 times faster than that of the traversal method.(2)Considering the measured wheel-rail creep curve in wet environment,the wheel-rail stick-slip distribution and tangential contact stress are significantly different from those in the dry ideal condition,which is mainly reflected in the decrease of the maximum tangential contact stress of the contact area and the deceleration of the sliding area with the increase of creepage.The proposed algorithm meets the requirements of vehicle-track dynamics numerical simulation and can simulate the wheel-rail interaction in real environment conditions more realistically.

关 键 词：轮轨蠕滑曲线 轮轨黏着-蠕滑特性 遗传算法 轮轨接触 蠕滑率 

分 类 号：U211.5[交通运输工程—道路与铁道工程]