Research On Sensorless Control System Of Permanent Magnet Synchronous Motor Based On Cerebellar Model

Permanent Magnet Synchronous Motor(PMSM)has been widely used in the fields of drive traction,position servo and high-efficiency power output due to its high power density,flexible and diverse structure,and strong reliability.These have also promoted the design and control of permanent magnet synchronous motors to become an important research content and research hotspot.The use of traditional mechanical sensors increases the system hardware cost,motor volume and failure rate,which limits the application occasions.How to obtain the position and speed information of the rotor of a permanent magnet synchronous motor without using a mechanical sensor to achieve high-performance control of the permanent magnet synchronous motor has become a research hotspot at home and abroad.Therefore,the text proposes a neural network controller based on the integral separation type cerebellar model to realize the speed control of the sensorless system of the permanent magnet synchronous motor.The research contents of this article are as follows:(1)The basic structure and principle of PMSM are introduced,and the mathematical model of the motor before and after coordinate transformation is derived.Then,the basic principles and methods of vector control are summarized,and the space vector pulse width modulation technique is analyzed in detail.The PMSM vector control system using the control strategy is completed in Matlab / Simulink.(2)The basic principle of the high-frequency injection method is introduced,and the two main sensorless vector control methods of the high-frequency injection method are analyzed and derived.According to the characteristics and application scope of the two methods,the pulse high-frequency injection method with little influence on the torque current is selected as the focus of the research.The position sensor system of surface mount permanent magnet synchronous motor using pulsed high frequency injection method was simulated by Matlab / Simulink.The results show that the method can effectively extract the rotor position angle of permanent magnet synchronous motor.(3)Introduced the structure and learning algorithm of Cerebellar Model Articulation Controller(CMAC).Aiming at the situation that the parameters of the speed loop PID controller in the PMSM sensorless system are difficult to set,an integral separation type PID controller based on the cerebellar model is proposed.The CMAC-PID controller uses the gradient descent method to modify the weight of the CMAC neural network,and the integral part uses an integral separation algorithm.The integral separation type CMAC-PID controller can set PID parameters in real time according to the operation of the motor after learning.The simulation results show that the integral-separated CMAC-PID controller has a smaller overshoot than the PID controller,a shorter start-up stabilization time,faster and more stable after a sudden load,and better dynamic response of the system.(4)A PMSM control platform with TI's DSP28335 as the control chip was built,the software and hardware design of the system was completed,and the feasibility of the CMAC-PID controller was verified.