Research On Drive Control System For Electric Vehicle Based On Neural Network Self-turning

Environmental problems and energy problems become more and more serious with the rapid increase in the number of cars in the world.Electric vehicles with no noise and zero pollution advantages,will become the main trend of the automotive industry.How to develop safe,reliable and high-performance electric vehicles is the focus of the development of car.The advantages of electric vehicle drive system are an important factor in the rapid development of electric vehicles.Permanent magnet synchronous motor(PMSM)has the advantages of good driving performance,simple structure,small volume,light weight,low loss and high efficiency.Therefore,permanent magnet synchronous motor(PMSM)is widely used in vehicle drive system.Because PMSM is one of strong coupling and nonlinear time-varying systems,the performance of speed control is very sensitive to the change of motor's parameters and load fluctuation during operation,the research on the key technology and advanced control theory of electric vehicle PMSM has become hot.On the basis of expounding the superiority of intelligent control,the speed control problem of the permanent magnet synchronous motor drive control system for electric vehicle is studied deeply.Because of the nonlinear of PMSM,traditional PID is difficult to adapt to the parameter change and load fluctuation.It is difficult that electric vehicle is driving safely and reliably under the complex road conditions.In order to solve the above problems,improve the efficiency of electric vehicle drive system,speed up the response and enhance the anti-jamming and robustness of the system,this paper proposes a self-tuning speed control method based on recurrent chebyshev neural network.The control method includes recurrent chebyshev neural network control and estimated compensation control.Recurrent chebyshev neural network on-line parameter adjustment is based on lyapunov stability law.Based on this law,we calculate the best learning rate to track the error more quickly.Compared with the traditional PID speed controller,recurrent chebyshev neural network speed controller has fast response,small overshoot,short adjustment time,less subject to external changes and changes in motor parameters.In order to verify the feasibility and correctness,a simulation model based on recurrent chebyshev neural network and traditional PID vector control system is built on MATLAB/Simulink platform.In the case of load disturbance and parameter change,the simulation experiment is carried out.Simulation results show that the recurrent chebyshev neural network controller can improve the dynamic and enhance the anti-perturbation ability and robustness of the system.Finally,the PMSM vector control system with TI TMS320F28335 chip as the core is designed,and the related experimental research is carried out to further verify that the electric vehicle drive control system based on the recurrent chebyshev neural network self-tuning is little affected by the load fluctuations and changes in motor parameters,can automatically adjust the controller parameters to adapt to complex situations.