Sliding Mode Control of Maglev Train Suspension System with Neural Network Acceleration Feedback

Abstract:

In order to ensure the suspension stability of maglev train， the active control of suspension system is studied. Firstly， based on the minimum suspension unit of single electromagnet of maglev train， the corresponding control mathematical model of current is established. Combined with the simulation， it is shown that the proportion-integration-differentiation（PID）control algorithm is very sensitive to time-varying disturbances such as nonlinear load. Then， a sliding mode control method based on the stability proof of bifurcation theory is proposed. Combined with the parameter self-adjusting function of radial basis function （RBF） neural network， a suspension control module with vibration suppression is constructed to effectively suppress the vibration of electromagnet. Finally， the Simulink control model is constructed and the single electromagnet suspension experimental platform is built for relevant simulation and experiments. The results show that the effect of electromagnet vibration on the suspension performance is particularly obvious. The proposed control algorithm can effectively suppress the electromagnet vibration in the presence of complex disturbances and improve the dynamic performance of the suspension system.