Abstract:The dynamic track stabilizer can effectively improve the quality of ballast track bed, however, its current operation mode will lead to the deterioration of ballast track cross level. To solve this problem, combined with the theory of cumulative subsidence of the ballasted track bed, the functional relationship between the amount of track sinkage and the speed, the left and right side downforce and frequency of the dynamic track stabilizer is investigated by experimental method. And the transfer function model between the horizontal change of the track and the downforce difference between the left and right side of the stabilizer is established. Considering the track level detect mode of the dynamic track stabilizer, feed-forward-feedback control and Smith's prediction compensation are introduced to pre-adjust and ensure the closed-loop stability of the system after detection lag. A grey wolf hybrid particle swarm algorithm (GWO-PSO) is introduced for PID parameter optimisation. In the Simulink simulation environment, a track level anti-deterioration control system was built. The simulation results show that the feedforward feedback smith prediction control strategy and the PID parameter optimization using the GWO-PSO algorithm can make the system a 19.66% improvement in the system's ITAE metrics and an increase in dynamic performance. In the simulated line simulation, the control system reduces the magnitude of track level irregularity by 52.1%, which effectively improves the track level degradation problem under the action of dynamic track stabilizer.