[1]刘 爽,黄 宇.基于深度学习的车钩碰撞等效代理模型研究[J].机械与电子,2026,44(08):10-17.
　LIU Shuang,HUANG Yu.Research on Deep Learning-based Equivalent Surrogate Model for Coupler Collision[J].Machinery & Electronics,2026,44(08):10-17.
点击复制

基于深度学习的车钩碰撞等效代理模型研究()
分享到：

《机械与电子》[ISSN:1001-2257/CN:52-1052/TH]

卷:
44
期数:
2026年08期
页码:
10-17
栏目:
研究与设计
出版日期:
2026-08-25

文章信息/Info

Title:
Research on Deep Learning-based Equivalent Surrogate Model for Coupler Collision
文章编号:
1001-2257(2026)08-0010-08
作者:
刘 爽1; 2; 黄 宇2
1.大连交通大学轨道交通装备设计与制造技术国家地方联合工程研究中心,辽宁 大连 116028; 2.大连交通大学交通工程学院,辽宁 大连 116028
Author(s):
LIU Shuang1; 2; HUANG Yu2
(1.National Local Joint Engineering Research Center for Rail Transit Equipment Design and Manufacturing Technology,Dalian Jiaotong University,Dalian 116028,China; 2.School of Transportation Engineering,Dalian Jiaotong University,Dalian 116028,China)
关键词:
车钩; 碰撞; LS-DYNA; 神经网络
Keywords:
coupler; collision; LS-DYNA; neural networks
分类号:
U270.3
文献标志码:
A
摘要:
针对车钩碰撞响应有限元仿真计算耗时、多工况分析效率低的问题,提出一种基于LS-DYNA 仿真数据与深度学习的车钩碰撞响应快速预测方法。首先建立车钩碰撞有限元模型,获取不同初始速度、碰撞角度和位移距离下的节点力响应数据;然后构建以工况参数为输入、节点力响应为输出的神经网络预 测模型,并与有限元结果进行对比验证。结果表明:车钩节点合力变化趋势与x 方向力Fx 高度一致,Fx 在整体受力中占主导作用;所建模型能够较好预测不同工况下的碰撞响应变化规律,可为车钩碰撞安全分析及多工况快速评估提供参考。
Abstract:
To address the high computational cost of finite element simulation for coupler impact response and the low efficiency of multi-condition analysis,a rapid prediction method for coupler impact response based on LS-DYNA simulation data and deep learning is proposed.First,a finite element model of coupler collision is established to acquire nodal force response data under different initial velocities,collision angles,and displacement distances.Then,a prediction model based on neural networks is constructed with the operating condition parameters as inputs and the nodal force responses as outputs,and its prediction results are validated against the finite element results.The results indicate that the variation trend of the resultant nodal force on the coupler is highly consistent with that of the x-direction force Fx ,and Fx plays a dominant role in the overall force response.The proposed model can effectively predict the variation patterns of collision responses under different operating conditions,providing a reference for coupler impact safety analysis and rapid multi-condition evaluation.

参考文献/References:

[1] 朱涛,张敬科,吴启凡,等.车钩缓冲装置对轨道列车碰撞安全性的影响综述[J].交通运输工程学报,2021,21(1):233-249. [2] 姚远.铁路货车缓冲装置性能优化研究[D].大连:大连交通大学,2020. [3] Durali M,Shadmehri B.Nonlinear analysis of train derailment in severe braking[J].Journal of Dynamic Systems, Measurement,and Control,2003,125(1):48-53. [4] Duncan I B,Webb P A.The longitudinal behaviour of heavy haul trains using remote locomotives[C]∥ Fourth International Heavy Haul Railway Conference, 1989:587-590. [5] 刘爽.重载列车非稳态冲动及安全运行的建模与仿真研究[D].大连:大连交通大学,2017. [6] 苗伟明,孙丽萍.基于装配体的13号车钩有限元分析[J].大连交通大学学报,2007,28(4):37-41. [8] 赵士忠,王晋乐,田爱琴,等.动车组中间车钩动态力学性能研究[J].铁道科学与工程学报,2018,15(5):1103-1110. [9] Bolandi H,Li Xuyang,Salem T,et al.Bridging finite element and deep learning:high-resolution stress distribution prediction in structural components[J].Frontiers of Structural and Civil Engineering,2022,16(11):1365-1377. [10] Baltic S,Asadzadeh M Z,Hammer P,et al.Machine learning assisted calibration of a ductile fracture locus model[J].Materials and Design,2021,203:109604. [11] Kohar C P,Greve L,Eller T K,et al.A machine learning framework for accelerating the design process using CAE simulations:an application to finite element analysis in structural crashworthiness[J].Computer Methods in Applied Mechanics and Engineering,2021,385:114008. [12] Xie Jian,Zhang Junyuan,Li Zihang,et al.Spatiotemporal modeling based on manifold learning for collision dynamic prediction of thin-walled structures under oblique load[J].Computer Methods in Applied Mechanics and Engineering,2025,440:117926. [13] Sakaridis E,Karathanasopoulos N,Mohr D.Machine-learning based prediction of crash response of tubular structures[J].International Journal of Impact Engineering,2022,166:104240. [14] 林翔,严波,牟哲岳,等.基于数值模拟和机器学习的汽车碰撞代理模型[J].重庆大学学报,2021,44(2):86-93. [15] 许平,黄启,许拓,等.基于数据驱动的地铁车辆等效缩比模型钩缓结构设计[J].交通运输工程学报,2024,24(5):217-233 [16] He Jiaxing,Xu Ping,Xing Jie,et al.Optimization of energy-absorbing parameters for train crash energy management using a variable-fidelity neural network surrogate model[J].Engineering Structures,2025,334:120242. [17] Guo Weinian,Xu Ping,Yang Chengxing,et al.Machine learning-based crashworthiness optimization for the square cone energy-absorbing structure of the subway vehicle[J].Structural and Multidisciplinary Optimization,2023,66(8):182. [18] 戴志睿.基于有限元法的13号车钩结构强度分析[D].大连:大连交通大学,2021. [7] 许平,何家兴,邢杰,等.不同初始姿态下城际列车车钩碰撞特性及脱轨风险分析[J].中南大学学报(自然科学版),2024,55(6):2370-2381.

备注/Memo

备注/Memo:
收稿日期:2026-05-06 基金项目:中国铁路沈阳局集团有限公司科技研发计划重点课题(2024Y03) 作者简介:刘 爽 (1979-),男,辽宁岫岩人,博士,副教授,研究方向为列车系统动力学及关键技术、列车自动驾驶、列车安全防护动力学,通信作者,E-mail:Liushuang_Dalian@163.com;黄 宇 (2001-),男,四川达州人,硕士研究生,研究方向为列车系统动力学。
更新日期/Last Update: 2026-08-28