Inductance Parameter Identification And Control Of Permanent Magnet Synchronous Traction Motor

As the rapid development of rail transit, it has attracted people’s extensive attention about how to improve the traction performance of train. Permanent magnet synchronous traction motor (PMSTM) has advantages such as high efficiency, less volume, and lighter, ect. Therefore, PMSM can improve the performance of the train significantly when it’s used in the train traction drive system.The inductance parameter of interior permanent magnet synchronous traction motor (IPMSTM) changes as the variation of motor’s current, which due to the asymmetric magnetic circuit between d-axis and q-axis of IPMSTM and cross saturation of motor’s magnetic field. And this feature of IPMSTM affects the control performance of the motor. This thesis analyzes the phenomenon of magnetic saturation when IPMSTM’s running. And this thesis also analyzes the reason of the inductance’s variation. The author used the finite element analysis software-Ansoft to calculate the inductance parameters of a600kW IPMSTM, and obtained the curve of inductance which varied with current. And then, this thesis discusses how the control of traction drive system is affected when inductance parameters varied with current. The author also compared the maximum torque per ampere control (MTPA) curves of fixed q-axis inductance and variable q-axis inductance which changes as q-axis current’s changing. We find out that when the q-axis inductance parameter changed as q-axis current’s variation, if the controller didn’t change correspondingly, the real output torque of IPMSM would not reach the command value, and the traction effect would be affected. This thesis used MATLAB building the control system of IPMSTM to simulate above issues.Based on the practical application of IPMSTM, the thesis identified the motor’s inductance parameters online with the recursive least squares algorithm with forgetting factor. This method obtained the current time estimated value of inductance from current time sample value and last time estimated value. And this method avoided a problem that the storage capacity was limited in ordinary least squares method. The simulation results indicated that this method could get the estimated value of q-axis inductance accurately and quickly.The thesis carried out the experiment about vector control of IPMSM with parameter identification, based on Myway, a kind of control platform of IPMSM. Experiment results showed that the response of IPMSM to the torque command was improved significantly by using this method.