Research On Fault Diagnosis And Fault-Tolerant Control Technology For Suspension System Of Low Speed Maglev Train

Suspension system is one of the key parts of maglev train. Its security and reliability is directly responsible to the safety of maglev train. Therefore the research on fault diagnosis (FD) and fault tolerant control (FTC) methods for suspension system is of great significance.This paper does some groping study of FD and FTC for suspension system of low-speed maglev train. Many methods are adopted to design FD and FTC strategies, and then thorough analysis, compare and improvement are done to them. The effect of these FD and FTC strategies is proved by simulation.First, an active FTC strategy is designed for sensor faults of suspension system which is composed of two subsystem- FD subsystem and control law redistribution subsystem. Two strategies are adopted for the design of the FD subsystem. The first one is base on tracking differentiator (TD) and can only diagnose accelerometer fault. Simulation shows that the FD strategy based on "fast" (a kind of TD) is the most effective one. After that, the strategy is improved considering the fluctuation of the rail. The second one is based on principal component analysis combined with TD which is more adaptive to each kind of sensor faults of suspension system. Simulation shows that this strategy can reduce the time consuming and misinformation of FD system. The simulation also shows the time consuming of each FD strategy.Based on the FD subsystem, an active FTC strategy is designed based on control law redistribution for different sensor faults. Simulation shows that when sensor faults occur, new control law can drive the system stable again. The simulation also shows that the FD subsystem satisfies the requirement of active FTC.At last, a passive fault tolerant controller is designed based on linear matrix inequation (LMI) considering actuator invalidation. Simulation shows that when a single actuator is cut off, the passive fault tolerant controller can still make the system stable and maintain certain dynamic and static performance, thus serious accidents such as carriage falling are avoid.