Research On Fault Diagnosis Of Marine Permanent Magnet Synchronous Motor

Permanent magnet synchronous motor(PMSM)has the advantages of high efficiency,high power density and wide speed range.Therefore,PMSM is the best choice for the main motor in the podded marine electric propulsion system.If the motor fails to be handled in time,it will not only damage the motor itself and the whole electric propulsion system,but also threaten the safety of the whole ship and personal safety.At this time,the fault diagnosis of permanent magnet synchronous motor itself is very important.This paper mainly studies the fault diagnosis methods for the main electrical faults of permanent magnet synchronous motor(PMSM)in electric ships.In this paper,the mathematical model and vector control strategy of permanent magnet synchronous motor are introduced.Then,the characteristics,causes and failure effects of phase-to-phase short circuit,single-phase grounding and inter-turn short circuit fault and demagnetization fault are analyzed.The model of PMSM drive system is established by using MATLAB/Simulink software,and the common electrical faults such as inter-phase short circuit and single-phase grounding are simulated.The 2D finite element model of PMSM is established by Ansoft software,and the simulation of inter-turn short circuit fault and loss-of-excitation fault of PMSM is carried out.Based on the established model,the specific performance impact analysis of the inter-turn short-circuit fault motor is carried out.The simulation shows that the parameters of the phase current,back EMF,torque,core loss and copper loss of the permanent magnet synchronous motor occur with the increase of the fault degree.Corresponding change.Based on the improved neural network fault diagnosis method,this paper studies the fault diagnosis method which is based on wavelet decomposition and extracts the relevant features of fault diagnosis,and then identifies them by integrated neural network.Four groups of sample data are obtained by simulation,the first three groups are training samples of neural network,and the last group is testing samples of the fault diagnosis method.The simulation results show that the method has the characteristics of accurate diagnosis and good real-time performance,and meets the target requirements.