Demagnetization Analysis And On-line Detection Of Permanent Magnet Traction Machine For Elevator

Permanent magnet synchronous motor(PMSM)with high efficiency and high power density has been widely used in elevator traction machine.However,its internal permanent magnets(PMs)generally have shortcomings such as high temperature coefficient and low mechanical strength.Under abnormal working conditions,irreversible demagnetization is prone to occur,resulting in elevator jitter and noise,and even dangerous stalling and topping accidents.At present,there are still gaps in the equipment used for the demagnetization detection of the PMSM of the elevator,so the research on the demagnetization characteristics and demagnetization detection of the PMSM is of great significance in elevators and other fields.This paper takes the multi-pole low-speed PMSM used in the elevator as the research object,analyzes the demagnetization distribution law of the PMSM under the vector control,and proposes an online demagnetization detection method based on the variable parameter sliding mode algorithm,which is used to quantitatively analyze the state of the PM inside the motor without collecting the encoder data.Finally,a motor loading experimental platform is built to verify the theory of the demagnetization characteristics of the PMSM and the proposed demagnetization detection method.In terms of demagnetization characteristics analysis,Simulink and Ansys EM are used to analyze the armature reaction effect and PM demagnetization law of the motor at different speeds under vector control.The alternating armature reaction has different effects.Based on the motor under id=0 control,no matter what speed it is at,the armature reaction always acts near the q-axis,resulting in the demagnetization effect of the PM is always distributed at both ends.It is pointed out that the focus of anti-demagnetization should be placed at the end of the PM,and the overall working point of the PM decreases with the increase of load.The simulation result of different types of demagnetization motors shows that the no-load BEMF(back electromotive force)value of the motor can accurately represent the proportion of local and uniform demagnetization faults,which can be used for quantitative evaluation of demagnetization faults.In addition,through the motor loading experiment,the effect of different degrees of demagnetization of PMSM is analyzed by power analyzer,which verifies the theory that the demagnetization fault of the PMSM will cause the power factor and efficiency to decrease.The greater the demagnetization degree is,the higher the load level is,the more obvious the effect is.In terms of demagnetization detection,based on the traditional sliding mode algorithm(SMO)to calculate the extended BEMF of the motor,the changes of the PM flux linkage,stator resistance and inductance parameters with the load are taken into account,which is used to construct the error equation of the motor no-load BEMF estimation.By collecting the three-phase voltage and three-phase current of the motor under different loads,combined with the influence law of the resistance and inductance parameters in the equation on the estimation results,the variable parameter strategy is proposed.By adjusting the resistance and inductance calculation parameters in the sliding mode algorithm,the no-load BEMF is obtained which can be used to evaluate the demagnetization of the PMSM.In this paper,six groups of motors with different degrees of local and uniform demagnetization faults are tested.The data processing shows that this method is applicable to both local demagnetization and uniform demagnetization faults,and the error can reach within ± 1.5%.Finally,the paper explores the influence of selecting operation data under different loads on the estimation results by simulation,and comes to the conclusion that at least three groups of load conditions should be selected,and the moderate load area is the best.