Study On The Influence Of Tunnel Piston Wind On The Failure Of The Platform Screen Door Based On The Fully Automatic Operation Mode

In the fully automatic operation system,the number of trains on existing routes is constantly increasing to meet passengers’travel demand.The alternating load caused by tunnel piston wind ramps up when the train speed level improves.The most direct effect is that the efficiency and robustness of opening and closing platform screen doors would be depressed.With the rapid development of the fully automatic operation system and the increasing demand for passengers’ travel,it is of great significance to study the piston wind pressure applied to the door and to predict the output force of the motor.In this paper,the formation principle of piston wind and the theoretical calculation method was firstly introduced,and the force applied to the platform screen door was analysed.Then,the derivation which solves the motor output force calculation formula of the platform screen door system was proposed.The piston wind simulation model in a dual-train tunnel was constructed based on the turbulence unsteady flow model.The piston wind in the tunnel caused by train movement are simulated by using the dynamic mesh technology.Moreover,the pressure field and velocity field of the dual-train model were analyzed,and the flow field distribution characteristics in the tunnel were obtained.The finite volume method and SIMPLE algorithm were used to calculate the pressure field and velocity field in the tunnel.Based on this model,the influence of train speed and the tunnel interval length between two trains on the wind pressure load of the screen door was studied.It is concluded that the maximum wind pressure of the screen door is affected by both the tunnel interval length and train speed.Finally,based on genetic algorithm improved BP neural network,a model for predicting the force of screen door motor according to the train running speed and the tunnel length between two trains was proposed.The piston wind numerical simulation sample was used to train the neural network,and this model was verified by the test sample.It has been proved that this model can predict the output force of the door motor effectively,which could make the platform screen door close normally without the influence of piston wind pressure.