Research On Fault Diagnosis Method Of Suspension System Of High Speed Maglev Train

The maglev train is a new type of rail transit system,which can operate without contact with the track by electromagnetic force,and the suspension system is the key system to support the maglev vehicle.The suspension system fault not only includes the complete failure of each component,but also includes the tiny fault caused by performance degradation caused by long-time operation.No matter what kind of fault occurs,it may affect the normal operation of the maglev train,so it is necessary to stud y the fault diagnosis of the maglev train suspension system.Supported by the National Key R&D program of China "Research on modeling,fault diagnosis and fault tolerant control of permanent magnet electromagnetic hybrid suspension system",this paper takes the suspension system of high-speed maglev train as the research object.The fault model of joint structure suspension system is established.Robust observer based fault diagnosis is uses to diagnose the actuator and sensor of suspension system.Finally,by combining the fault model and data-driven,the problem of fault diagnosis under the condition of inaccurate fault model is solved using the existing experimental data.The main contents of this paper are as follows:1.Based on the analysis of the structure,composition and working principle of the suspension system,the fault model of the suspension system of the high-speed maglev train is established.The full state observer is designed to diagnose the fault of the model,and the effect of fault diagnosis is verified by simulation.The limitations of the full state observer method are analyzed.2.For the actuator fault of suspension system,the fault is decoupled from the original model by state and output transformation,so it is not affected by interference.The sliding mode observer is designed to realize the fault diagnosis of the actuator by deducing the stability condition of suspension system.Solving the Lyapunov stability theory transformed solving the parameters of the linear matrix inequality.The simulation shows the effectiveness of the method.3.For the sensor fault of suspension system,the sensor fault is integrated from the output equation and then introduced into the state equation,which can be compared to the actuator fault,and then the fault is completely decoupled from the original model to achieve fault detection and isolation.Two fault estimation schemes,adaptive observer and sliding mode observer,are designed.The simulation results show that the fault diagnosis based on sliding mode observer and adaptive observer can be realized effectively,and the effect of fault diagnosis based on adaptive observer is better.4.Because the actual system model is difficult to obtain,and the model parameters change with the operating conditions,and the input and output data of the system can be easily obtained online and reflect the system characteristics,therefore,the combination of model and data-driven method can effectively avoid the impact of model uncertainty and model parameter changes.In this paper,the fault diagnosis scheme designed for the suspension system can improve the reliability and safety of the suspension system,and provide relevant reference and engineering application for the realization of localization of high-speed maglev train.