Research Of PMSM Field Oriented Control System Based On STM32

With the rapid expansion of electric vehicle industry,the research on motor control system aroused people's wide concern and becomes a hot spot.On account of the low-carbon and environmentally-friendly character,Permanent Magnet Synchronous Motor(PMSM)is attracted more attention among many kinds of motors,meanwhile Field Oriented Control(FOC)is preferred because of the excellent dynamic speed and torque response in the area of motor control.Therefore,the reach on improving the performance of FOC of PMSM is extremely important and necessary.So far,control parameters of FOC cannot voluntarily adapt to changes and the position sensor is indispensable in the control process,the paper is focus on how to improve those problems.First of all,this paper analyses the basic structure and mathematic mode of PMSM.Through transformation of coordinates this paper illuminates how FOC works.Then,this paper discusses the characteristic of different Vector control strategies meanwhile elaborates the realization principle of FOC.By MATLAB/Simulink platform,the author establishs the simulation system of FOC of PMSM which confirms the theory referred.And focus on the problem of speed deviation caused by the fixed parameter in PI controller,this paper using Neural Networks Control to improve the performance of traditional PI controller which makes the system possesses parameter adaptive ability and shows a better speed response.Then,the paper discuss the accuracy of speed response,response time and so on using different kind of neural network in the simulation system which lays a certain theoretical foundation for the further improvement of the FOC on PMSM.At last,the author studies on how to run the hardware system when there is no position sensor and establishs the PMSM FOC hardware system by using STM32F405RGT6 as MCU.The result shows a good and steady performance whether there is a position sensor or not.