Research On Signal Filtering Algorithms In Maglev Suspension Control

The suspension system of the low-speed maglev train is a typical non-linear open-loop unstable system, and research on filtering and noise-reducing algorithms is a key technology in suspension control because of the complicacy of the electromagnetic environment and the uncertainties of the operational state parameters. This dissertation concentrates on the comparison of the application of different filtering algorithms in the suspension control system, using theoretical analysis as well as simulations and experiments. The main jobs are listed as follows:Firstly, a dynamic model of a single electro-magnet suspension system is established, and a current loop and a position control loop are designed. Analyses of the impacts to the stability of the whole system due to control parameters are undertaken in both the time domain and the frequency domain, and the noise model of the suspension system is introduced. Moreover, a model is estabilished in the experimental way to discribe the nosie of the suspension system, and the Power Spectral Density (PSD) is employed to describe the frequency characteristics of the noise.Secondly, the anti-interference performance of several low-pass filters in common use is explored, and the performance of the notch filter in suppressing the nerrow-band alternating-current noise is dicussed. In addition, the influences of the low-pass filter and the notch filter on the stability of the suspension control system are also analyzed.Thirdly, considerring the characteristics of the noise in the suspension control system, the Kalman filter is used. Theoretical analysis and simulation validation have been undertaken on the application of the Kalman Filter in the suspension control system. The simulation result indicates that using the Kalman filter, the suspension control system can significantly reduce the impact of the noise and its robustness is therefore improved.Finally, the Least Squares (LS) adaptive filter is applied to the suspension control system to suppress the noise, and the Recursive Least Squares (RLS) algorithm is applied to reduce the computational complicacy. Simulation shows that the adaptive filter performs well in suppressing the noise.According to the analysis and comparison of the algorithms listed above, filtering algorithm which meets the requirements of the maglev train is proposed, which provides a significant referrence for the suspension control algorithm.