• 查询稿件
  • 获取最新论文
  • 知晓行业信息

轮轨几何接触点的人工神经网络快速计算方法

Fast calculation method for wheel-rail geometric contact points based on artificial neural network

  • 摘要: 现有的轮轨几何接触点计算方法的求解过程较为繁琐,计算效率较低。结合迹线法构造钢轨廓形NURBS曲线权因子与轮轨接触点的几何关系,提出一种基于BP神经网络的轮轨几何接触点的快速计算方法,实现不同钢轨廓形条件下轮轨几何接触点的快速计算。实例分析表明:训练的人工神经网络能够高效准确地实现轮轨几何接触点的计算。

     

    Abstract: The current calculation methods for wheel-rail contact points are complicated, and the calculation efficiency is low. In this paper, the contact trace method was introduced to establish the geometry relationship between weight factors of rail NURBS curve and the wheel-rail contact points. A fast calculation algorithm for wheel-rail geometric contact points based on BP neural network was proposed to implement the rapid calculation under different rail profiles. The real case analysis shows that the trained artificial neural network can calculate the wheel-rail geometric contact points efficiently and accurately.

     

/

返回文章
返回