Research Of On-line Monitoring And Fault Diagnosis System For Magnetic Levitation Control System Of Medium-low Speed Maglev Train

Because of the advantages of low noise,strong climbing ability and small maintenance workload,the engineering application of medium and low speed maglev trains is vigorously promoted in China.Suspension control system is one of the core equipment of maglev train.Its reliability and stability affect the operation of maglev train and the safety of passengers.Therefore,the research of online monitoring and fault diagnosis for suspension control system is very urgent.Taking the levitation control system of maglev train as an example,this paper explores the online monitoring and fault diagnosis technology of maglev train.The main work of this paper is as follows:The mathematical model of the suspension system is established in this paper.The characteristics of the inductance of the electromagnet,the electrical characteristics of the load electromagnet and the mechanical characteristics of the suspension module are analyzed in detail,and the device parameters and the possible fault forms that the controller needs to be monitored online are summed up in combination with the hardware structure and function of the suspension controller.Aiming at the devices that need on-line monitoring in the suspension control system,the parameter identification algorithm is designed and verified by simulation.For the charging resistance,the resistance value is calculated by Ohm law.For the support capacitor,the capacitance value of the capacitor voltage is calculated by the capacitance voltage formula.The KVL law,the inductance definition formula and the DAC and the multistage operational amplifier are used to calculate the current ripple.The resistance value and inductance value of the electromagnet are calculated.For the electromagnet temperature,the temperature value of the electromagnet is calculated by the relationship between the conductor resistivity and its temperature.Aiming at the device that needs fault diagnosis in the suspension control system,the fault diagnosis algorithm is designed and verified by simulation.For the fuse and the contactor,the fault state is obtained by using the relationship between the driving state,the capacitance voltage and the input current.For the voltage sensor,the fault state is obtained by using the relationship between the input voltage,the capacitance voltage and the voltage waveform of the electromagnet in the switch tube.For the MOSFET,the electromagnet voltage is used.The fault state of the air gap and acceleration sensor is obtained by using the air gap and acceleration sensor,the relationship between the three signal of air gap or acceleration and the continuous time range of a certain signal is a constant value,and the current integral of the input current sensor is used in the charging stage.The fault state is obtained by the relationship with the capacitance voltage,and the fault state of the electromagnet current sensor is obtained by using the current relationship and the range of the slope of the ripple.MATLAB's GUI diagnostic interface design.In order to facilitate the observation of the operation of the suspension control system and the specific characteristics of various signals after the failure,the GUI of MATLAB has designed a clear,direct and flexible display interface.The two interfaces of the main and secondary interfaces have the on-line monitoring function of the charging resistance value,the support capacitance value,the output of the voltage and current sensor,the output of the air gap sensor,the inductance value of the electromagnet,the resistance value and the temperature value,as well as the fault diagnosis function of the fuse,the contactor,the MOSFET,the voltage current sensor and the air gap and the acceleration sensor.Through the modeling and analysis of the suspension system,the parameter identification algorithm which needs the on-line monitor,and the exploration of the diagnosis algorithm which needs the fault diagnosis device,the simulation verification is completed.The validity of the algorithm is also need to be verified in practical applications.