Real-time Determine The Temperature Of Stator And Rotor Of PMSM

The permanent magnet synchronous motor(PMSM)has been widely used in traction drive fields such as industrial robots,CNC machine tools,rail transit and electric vehicles thanks to its high power density,high instantaneous power,high power factor and high control accuracy.Due to its high power density,permanent magnet synchronous motors are often employed in environments with sm all installation space and complicated working conditions.The heat dissipation environment is always harsh,and results in an increase of the risk of excessive temperature,and the probability of permanent magnet demagnetization and stator winding insulation falling off etc.The control performance and service life of the motor will also be consequently affected.In order to avoid the above potential issues,it is very important to online observe the temperature of stato r winding and permanent magnets.In this paper,temperature observers based on two different estimation schemes,such as the Position Offset-Based Parameter Estimation(POPE)injection and the Model Reference Adaptive System(MRAS),are investigated,which are based on the temperature sensitivity of stator winding resistance and permanent magnet flux linkage.The effectiveness of the two temperature observers is then tested on PMSMs.The main contents of this thesis are as detailed below:1)Design the PMSM stator and rotor temperature observer based on the MRAS algorithm.The adaptive law of parameter identification based on the Popov hyper-stability theory for PMSMs is investigated.The stator and rotor temperature observers are designed by using the approximate linear relationship between stator winding resistance/permanent magnet flux linkage and temperature.The simulation model of the PMSM parameter observer is built in Simulink and the test results show that the parameter observers can quickly track the changes of stator winding resistance and permanent magnet flux linkage.The parameter observer is programmed in C language,and tested on a surface-mount permanent magnet synchronous motor(SPMSM),being fed bya NXP MKV31F256VLL12 chip based mach ine drive.It is found that at i_d=0 control,the stator winding resistance and permanent magnet flux linkage cannot be solved simultaneously from one q-axis voltage equation because the estimation model is inherently rank-deficient.2)Design the PMSM rotor temperature observer based on the POPE scheme.According to the resulting state of the d-axis voltage after continuous injection of the positive and negative offset signals of the rotor position,the full-rank identification equation of the permanent magnet flux linkage can be derived.The observation of temperature is then experimentally investigated on the SPMSM platform,of which the accuracy of temperature rise in PMSM is ensured by the high and low temperature test chamber.Besides,the permanent magnet flux linkage calculated from the no-load back EMF via.FFT is used as a reference value.The experimental results show that when the PMSM temperature changes continuously,the POPE algorithm can quickly and effectively track the change of the permanent magnet flux linkage and the identification accuracy is high enough for temperature tracking.The on-line monitoring of PMSM rotor temperature within the error tolerance range can be realized by the temperature observer,which is of certain engineering significance.