Diagnosis Of Inter-Turn Short-Circuit In Direct-Drive Permanent Magnet Synchronous Motor Based On Electromagnetic Signals

Direct-drive permanent magnet synchronous motor(DDPMSM),which has various advantages such as light weight,high efficiency and no-mechanical transmission and so on,is widely used in transportation industry,aviation area and national defenses.In high-power applications,such as electric vehicles and industrial robot,DDPMSM with multiple branches parallel(MB-DDPMSM)is widely applied.However,MB-DDPMSM suffers from mechanical,thermal,electrical or environmental stress,which leads to inter-turn short-circuit fault(ISF).In the early stage of ISF,MB-DDPMSM is hardly affected and still allowed to keep working.The ISF quickly triggers other types of stator winding faults,such as phase-to-phase,which may lead to high repair or replacement cost,or even catastrophic failure.ISF location technique improve the reliability,availability and safety of DDPMSM,and is the basis of fault tolerance and maintenance.Therefore,it is necessary for theoretical research and practical application to carry out the ISF diagnosis research.ISF causes the variation of the magnetic flux density.Thus,magnetic flux density contains direct and critical information of fault coil location.According to the above analysis,an electromagnetic signal-based ISF diagnosis method is proposed in this paper to diagnose the MB-DDPMSM.First,the working principle of the search coil(SC)is analyzed,and a new type of search coil array is proposed to measure the fluxes of stator teeth.Second,search coil differential back EMF and search coil residual back-EMF as the fault indicators are used to detect ISF at the early stage and identify the fault coil location.Third,an ISF diagnosis algorithm is proposed,and an ISF intelligent diagnosis system of MB-DDPMSM is established using knowledge graph.Finally,the experimental platform is built and the corresponding experimental research work is carried out.The main contents of this paper are as follows.1.The working mechanism and arrangement principle of search coil array are studied.Firstly,an analytical model of the SC back EMF for DDPMSM with ISF is presented,and the accuracy of the analytical model is verified by finite element simulation.Then,according to the established model,a new type of search coil array is proposed.For DDPMSMs with different slot-pole combinations and winding ways,the arrangement of search coil array is various.Therefore,the winding distribution of DDPMSM is analyzed by using the slot potential star diagram,which provides a basis for the formulation of the arrangement principle of search coil array in DDPMSM.2.The fault indicators of ISF in MB-DDPMSM is studied.Using the proposed mathematical model,the influence of ISF on the back-EMF of each search coil in the search coil group is analyzed.According to the relationship between the back EMF of the detection coil and the position of the short-circuit coil group,the early fault detection and the positioning feature quantity of the short-circuit coil group are selected.According to the change law of the back EMF of the detection coil in the fault coil group,the positioning fault indicator of the fault coil is selected.The effectiveness of the selected fault indicators is verified by the finite element simulation results.Compared with the existing fault location methods,the sensitivity of the selected feature quantity is verified.3.The ISF diagnosis method of MB-DDPMSM is proposed.According to the relationship between fault indicators and fault coil location,an ISF diagnosis algorithm is proposed,and according to the proposed diagnosis algorithm,an ISF intelligent diagnosis system of MB-DDPMSM is established using knowledge graph is established.The simulation results show that the diagnosis results are consistent with the preset results,and the system can realize early inter turn short circuit fault detection and early fault coil location.4.The ISF experiment of MB-DDPMSM is carried out.The MB-DDPMSM is made and the experiment platform is built,which verify the accuracy of the established mathematical model and the effectiveness of the proposed fault indicators.