摘要
目前列车纵向动力学仿真的积分算法多以Newmark-β法为基础的固定步长算法为主,但随着列车编组数量增加以及工况变化,固定步长算法难以获得最佳的计算效率。基于此,文章采用显式预测-隐式迭代校正的方法,在Newmark-β法的基础上,通过限制迭代次数的方法变化步长,将其演变为一种变步长的积分算法。通过算例验证了变步长算法在动力学方程解变化剧烈的时候具有更高的计算精度,将该算法应用至1万t单编列车的全制动停车工况的纵向动力学仿真,结果表明,相较于其他算法,变步长的积分算法在保证精度的同时计算效率更高。
At present,the integration algorithm commonly used in train longitudinal dynamics simulation is fixed step algorithm mostly based on Newmark-βmethod,but with the increase of the number of train formation and the change of working conditions,the fixed step algorithm is difficult to obtain the best calculation efficiency.In this paper,the explicit prediction-implicit iterative correction method is adopted.Based on the Newmark-βmethod,the step size is changed by limiting the number of iterations,which is evolved into a variable step integration algorithm.It is verified by arithmetic examples that the variable step size algorithm has higher computational accuracy when the solution of the dynamical equations changes drastically.The algorithm is applied to the longitudinal dynamic simulation of 10000 t single train under full braking parking condition.The results show that compared with other algorithms,the integration in this paper not only ensures the accuracy,but also has higher calculation efficiency.
作者
郭炎冰
杨诗卫
杨璨
倪文波
GUO Yanbing;YANG Shiwei;YANG Can;NI Wenbo(School of Mechanical Engineering of Southwest Jiaotong University,Chengdu 610031,China;CRRC Meishan Rolling Stock Co.,Ltd.,Meishan 620032,China)
出处
《铁道车辆》
2023年第3期147-152,共6页
Rolling Stock
基金
中国中车股份有限公司重点科技项目(2019CHB087)。
