首页|面向脱轨后被动安全防护的转向架结构优化

面向脱轨后被动安全防护的转向架结构优化

扫码查看
原文链接
  • 国家科技期刊平台
  • NETL
  • NSTL
  • 万方数据
为了验证和优化列车转向架部件的脱轨后被动安全防护效果,提出基于多边形接触模型的列车脱轨后接触建模方法,并建立列车脱轨后动力学模型,模拟不同结构参数组合下的列车脱轨后运行行为.将所得横移量和结构参数组合作为反向传播神经网络模型的训练样本,拟合两者之间的非线性关系,采用基于锦标赛选择的自适应遗传算法搜索横移量最小解及其对应的参数组合.以拖车制动盘优化问题为例,运用该优化模型进行结构参数优化.研究结果表明:在100 km/h脱轨速度下,制动盘的最优结构参数组合的半径为335 mm、厚度为80 mm、安装位置为320 mm;与原设计相比,车辆脱轨后横移量减少80.2%,大幅度提高了制动盘被动安全防护能力.本文所提出的建模方法和优化模型可推广应用于其他列车结构的脱轨后被动安全防护能力优化和验证,具有较大的工程实用价值.
Structural optimization for post-derailment passive safety protection of bogie
In order to further verify and optimize the effect of passive safety protection of bogie components,a train dynamics modelling method was proposed based on a polygonal contact model(PCM)and train post-derailment dynamics model was established to simulate the running behaviour with different combinations of structural parameters.The obtained lateral displacement and parameter combinations were used as training samples for a Backpropagation(BP)model to fit the nonlinear relationship between them.The TAGA model was employed to search for the minimum solution of lateral displacement and its corresponding parameter combination.Taking the optimization of the brake disc as an example,the proposed optimization model was applied for structural parameter optimization.The results show that at a derailment speed of 100 km/h,the optimal structural parameter combination for the brake disc is a radius of 335 mm,a thickness of 80 mm,and an installation position of 320 mm.Compared to the original design,the lateral displacement of the vehicle after derailment decreases by 80.2%,significantly improving the passive safety protection capability of the brake disc.The modelling method and optimization model proposed in this paper can be widely applied to the validation and optimization of the passive safety protection capability of other train structures after derailment,demonstrating promising engineering practical value.

post-derailment behaviourpolygonal contact model(PCM)structural parameter optimisationbackpropagation neural networkgenetic algorithms

胡玉炜、唐兆、陈涛、彭子豪、庄达源

展开 >

西南交通大学轨道交通运载系统全国重点实验室,四川成都,610031

脱轨后行为 多边形接触模型(PCM) 结构参数优化 反向传播神经网络 遗传算法

国家自然科学基金四川省自然科学基金西南交通大学轨道交通运载系统全国重点实验室自主研究课题

521724072022NSFSC04152024RVL-T14

2024

10.11817/j.issn.1672-7207.2024.05.029

中南大学学报(自然科学版)
中南大学

中南大学学报(自然科学版)

CSTPCD北大核心
影响因子：0.938
ISSN：1672-7207
年,卷(期)：2024.55(5)
  • 4