Research On Speed Control System Of Permanent Magnet Synchronous Motor Based On Improved PID

With the rapid development of electric vehicle industry,motor speed control has become a research hotspot in recent years and attracted extensive attention.Among many motors,PMSM has attracted more and more attention because of its high efficiency,low loss,energy saving and environmental protection.In the field of motor control,the proportional integral derivative（PID）control strategy is still used.PID control strategy has simple structure and reliable operation,but in the case of large load fluctuation and high precision,the traditional PID control strategy is difficult to meet the demand.Firstly,this paper introduces the physical model and mathematical model of permanent magnet synchronous motor,and expounds the principle of vector control technology and space vector control.Then the traditional PID control strategy is introduced and analyzed.Based on the shortcomings of traditional PID,based on vector control,this paper combines radial basis function（RBF）neural network with traditional PID control to control the parameters of permanent magnet synchronous motor.Using the self-learning ability of neural network,the PID parameters can be adjusted online and in real time according to the actual operation of the motor.Then,according to the structure and learning RBF neural network algorithm,the gradient descent method modifies the parameters of RBF neural network,and the integral term adopts variable speed integral to improve the control accuracy.Aiming at the above theory,this paper establishes the simulation model in Matlab /Simulink,and compares the speed,torque and current parameter waveforms of permanent magnet synchronous motor under traditional PID control and RBF-PID control strategy.In order to further verify the feasibility of RBF-PID control strategy,the algorithm is written in C language in keil5 software development environment,and the relevant software and hardware experimental platform is built with stm32f407igt6 as the main control chip.Experiments verify the performance of RBF-PID control algorithm,such as forward and reverse,acceleration and deceleration.The simulation and experimental results prove the feasibility and effectiveness of RBF neural network in motor control system,which has important reference significance for speed control in practical engineering.