Sensorless Control Of Permanent Magnet Synchronous Motors Based On Online Parameter Identification

Permanent magnet synchronous motors are widely used in electric vehicles,new energy power generation,robots and other fields because of their high power density,high torque / rotational inertia ratio,and high efficiency.The implementation of high-performance control strategies for permanent magnet synchronous motors often requires real-time rotor position acquisition.However,due to the high cost,large volume,and many leads of the position sensor,the reliability of the system will be reduced.Therefore,the current research focus at home and abroad is to achieve high-performance control of permanent magnet synchronous motors without using mechanical position sensors.The method based on the fundamental wave model in position sensorless control is widely used due to its convenient implementation,strong versatility,and excellent dynamic and static performance.However,because the back electromotive force is too small at low speed,it is not easy to accurately estimate,so how to enhance the low speed performance of this method has become an important and difficult point in the study of sensorless control.At the same time,the position sensorless control is affected by the nonlinear factors of the inverter and the electrical parameter perturbation,etc.,which will reduce the accuracy of position estimation and weaken the control performance.In response to these problems,this thesis carried out in-depth analysis and research,and proposed a variety of solutions,summarized as follows:In view of the problem that the command voltages are not equal to the actual output voltages due to the nonlinear factors of the inverter,a phase voltage measurement circuit is proposed,the voltage sampling range and accuracy are analyzed,and the phase and amplitude error caused by the hardware low-pass filter is compensated.On this basis,a method of directly measuring voltage combining phase and amplitude compensation is proposed to improve the voltage accuracy.Experimental results show the effectiveness of this method.Aiming at the problem that the inaccuracy of motor parameters affects the accuracy of position estimation,a multi-parameter online identification method for permanent magnet synchronous motors using voltage measurement is proposed.Firstly,the problems of under-ranking and mutual coupling in multi-parameter online identification are analyzed.The accuracy of parameter identification is improved through the steps of identification and excitation signal injection.Then the influence of injected high-frequency sinusoidal voltage and low-frequency square wave current on the control performance of the system is analyzed.Then,the influence of the amplitude and frequency of the injected signal on the accuracy of parameter identification is further analyzed.Finally,it is verified by experiments that the method can accurately identify the direct and quadrature-axis inductances,stator resistance and permanent magnet flux linkage at different temperatures.The identification accuracy is high,and the method is simple and easy to implement in engineering.In order to enhance the low-speed performance based on the fundamental wave model method,an adaptive position sensorless control strategy using phase voltage measurement is proposed.First,use the measured voltages instead of the command voltages.Then,by injecting high-frequency sinusoidal and direct current voltages into the stationary reference frame,a robust online identification method for the direct and quadrature-axis inductances and the stator resistance is proposed.Next,the identification parameters are updated in real time based on the active flux model position observer.Finally,the experimental results confirm that the method effectively improves the performance of sensorless control.Aiming at the adverse effect of the injected signal on the operation of the system,a reduced-order position observer with voltage measurement,integrated stator resistance and inductance adaptation is further proposed.First,the nonlinear dynamic estimation error of the observer,the stability of the observer gain,and the selection of resistance adaptive parameters are analyzed.Then,through offline measurement,the function of the direct and quadrature-axis inductances with respect to the current is established and used to update the observer parameters in real time.Finally,experiments show that the proposed strategy can effectively improve the accuracy of position and speed estimation in the low speed range,achieve reliable loading,speed regulation and speed reversal at low speed,and enhance the ability of the sensorless control of low speed operation.