Study On Control Strategy Of Permanent Magnet Synchronous Motor Based On BP Neural Network

Although the traditional oil vehicle is convenient for personal travel and improves people’s quality of life,it also brings problems such as energy consumption and environmental pollution.Due to the advantages of high driving efficiency,good driving experience,low operating cost and reducing urban emission pollution,governments all over the world are vigorously developing electric vehicles at present.Different from the traditional oil vehicle,which is driven by the engine,the electric vehicle relies on the motor to output power,so the performance of the motor is very important for the electric vehicle.Permanent magnet synchronous motor(PMSM)has been widely used in electric vehicles due to its advantages of small size,high working efficiency,large power density and low controller cost.At present,most of the permanent magnet synchronous motor control uses the classical PID controller,because its principle is relatively simple,the technology is mature and has a certain control precision,but the permanent magnet synchronous motor as a nonlinear system,possesses the characteristics of strong coupling and lead to the PID control algorithm is far cannot satisfy the requirement of the high precision fast response,aiming at this problem,this paper designed the AFSA-BP-PID controller.The main research contents are as follows:Firstly,Two types of permanent magnet synchronous motor are introduced in detail,and the three coordinate systems of permanent magnet synchronous motor,namely A-B-C coordinate system,α-β coordinate system and d-q coordinate system,and the coordinate conversion principle between them are deduced and analyzed in detail,and the mathematical model of permanent magnet synchronous motor is established respectively based on these three coordinate systems.Secondly,the vector control strategy of PMSM and the principle of space vector pulse width technology are analyzed.Based on the double closed loop vector control system of PMSM,in view of the disadvantage that the PID control algorithm used in PMSM can not adjust the parameters adaptively,a BP neural network is proposed to adjust the PID parameters automatically online.In view of the difficulty in determining the initial weights and thresholds of BP neural network,the artificial fish swarm algorithm was proposed to optimize the selection of initial weights of BP neural network,and the AFSA-BP-PID controller was established.Then,in order to verify the feasibility of AFSA-BP-PID control algorithm,a permanent magnet synchronous motor vector control system was built on Matlab/Simulink platform,and the BP-PID control strategy and AFSA-BP-PID control strategy were simulated.The experimental results show that the AFSA-BP-PID control algorithm can improve the dynamic performance and anti-interference ability of permanent magnet synchronous motor.Finally,the d SPACE real-time simulation system with higher efficiency and lower cost than the traditional development mode is introduced from the aspect of software and hardware.The real-time simulation model of PMSM vector control is built on d SPACE platform,and the hardware in the loop simulation experiment is carried out.The results show that afsa-bp-pid algorithm has better performance in the speed response ability of PMSM,which is different from the theory in this paper The results are consistent.