湖南大学学报（自然科学版）2025，Vol.52Issue(10)：1-11,11.DOI:10.16339/j.cnki.hdxbzkb.2025200
一种轮轨接触力解耦识别方法
A Decoupling Identification Method for Wheel-rail Contact Forces
摘要
Abstract
To achieve precise identification of the wheel-rail contact force and investigate the influence patterns of heavy haul train operations,this work establishes the functional relationships governing the influence rules of train axle weight,train traction force,and running speed on the contact forces.A decoupling identification methodology for the contact force is originally proposed.Initially,the numerical simulations are employed,with varying levels of each variable tested by using the control variable method within specified intervals.Furthermore,the functional relationships for the contact force at the wheel-rail interface in three-dimensional directions are derived relative to the train axle weight,train traction force,and running speed.Subsequently,a decoupling analysis is conducted to examine the non-contact point strain to the contact force.The data acquired from the simulation tests are then utilized in training a neural network model based on backpropagation(BP)neuron network.The study finally obtains the expression of the contact force as a function of train axle weight,train traction and running speed,and realizes the real-time prediction of the decoupling identification of wheel-rail contact force through the trained neural network,and the R2 of the fitting and training results reaches more than 0.9.关键词
重载列车/轮轨系统/神经网络/接触力/解耦Key words
heavy haul trains/wheel-rail systems/neural networks/contact forces/decoupling分类
交通工程引用本文复制引用
韦凯,张镇晖,肖浩田,尹盛文,崔晓璐..一种轮轨接触力解耦识别方法[J].湖南大学学报（自然科学版）,2025,52(10):1-11,11.基金项目
国家重点研发计划资助项目(2022YFF0609400),National Key Research and Development Program of China(2022YFF0609400) （2022YFF0609400）
