Identification And Verification Of High-speed Train Dynamics Model

High-speed railway is one of the important tools to promote the rapid development of China's national economy,accelerate the rapid circulation of resources,and shorten the space-time distance.However,as the operating speed of high-speed trains continues to increase,the interaction forces between trains,contact nets,wheel rails,and air have significantly increased,resulting in a serious deterioration of the dynamic operating environment of train operating systems.Therefore,in order to ensure the healthy development of high-speed railways and the safe operation of high-speed trains,effective modeling of high-speed train operating systems is required.Firstly,a high-speed train dynamics model is established using a mechanism modeling method,and then a discretization process is performed to obtain a single-particle difference equation model for describing a high-speed train operating system.Due to the time-varying characteristics of the parameters in the model,combined with the system identification theory,a forgetting factor recursive least squares method and a recursive maximum likelihood method were proposed to identify the time-varying parameters in the model.Under normal train operating conditions,both identification methods have achieved good identification results,and the discrepancy between the identification value and the empirical value in the model is small,meeting the needs of the actual system,indicating that the proposed identification method is effective.Under abnormal train operating conditions,the parameters in the train model are abruptly changed.The forgetting factor recursive least squares method is used to identify the abrupt changes in the model.The identification results show that the forgetting factor can not only track the slow time-varying parameters in the model in real time.The parameters can also be well tracked when the model parameters are abruptly changed to prevent data saturation and reduce data convergence errors.It can be seen that the combination method of mechanism modeling and system identification greatly improves the accuracy of model parameter identification,makes the established model more accurate,and better describes the actual running status of high-speed trains.Secondly,because the actual running status of high-speed trains will change with the changes in the train's operating environment,it is a complex process that changes over time.Therefore,in order to more accurately describe the actual operating system of high-speed trains,the data of traction and velocity is collected from the actual running process of the train as inputs and outputs for identification modeling.An ARX-RLS identification method and an ARMAX-RLS identification method based on the recursive least squares principle are proposed.Using these two identification methods,a fourth-order ARX model and a fourth-order ARMAX model are established by collecting partial data of the traction andvelocity.These two models only consider the input and output of the system and do not consider its internal mechanism,which effectively reduces the modeling complexity of the actual train operation system.Finally,the fourth-order ARX model and the fourth-order ARMAX model were verified by using another partial data of the traction and velocity collected.The simulation results verify the accuracy of the identification model through the fitting of the actual output of the system and the output of the identification model,and identification residuals.The verification results show that both models can reflect the actual output response of the high-speed train running system,but the accuracy of the fourth-order ARMAX model is higher.This lays the foundation for the next step to study the dynamic characteristics of high-speed trains and to achieve control and optimal design of high-speed train operating systems.