Research And Implementation Of PID- Neural Networks Control For Permanent Magnet Synchronous Motor By Model-Based Design

Permanent Magnet Synchronous Motor(PMSM) has characteristics like small size, high torque density and high efficiency, and it is widely used in the field of servo control. PMSM is a complex system which is very sensitive to motor parameter variations and external disturbance, therefore, how to design an advanced performance PMSM drive system has been a research emphasis. Artificial Neural Networks has advantage in solving problems of nonlinear and uncertain systems. In this research, we will combine the Artificial Neural Networks and PID controller to build a new PID- Neural Networks(PIDNN) controller, this kind of controller will replace PID controllers in Space Vector Pulse Width Modulation(SVPWM). In this way, we hope to get a better result in PMSM control system. In order to improve the efficiency of the system design and focus on the algorithms design, the model-based design(MBD) method is used to serve as the main design method. The main contents are summarized as follows:1. The principle and algorithm of PID- neural network controller is analyzed, and the method of improving the control performance is proposed. A test case is given to verify the feasibility and reliability of the algorithm, and the results show that the PID- neural network controller can achieve the desired results by adjusting the connection weights of neural networks in real time.2. The principle of SVPWM is analyzed, with the help of Embedded Coder Toolbox, the model of PMSM neural network PID control system is build. By comparing simulation results of PID- neural network controller and PID controller, it can be seen that PID- neural network controller is fully able to achieve the requirements of PMSM speed requirement and has faster response capability.3. The model-based design method is introduced, and according to its process, the model checking, fixed-point model building, SIL and PIL test are completed and after completion of these tests, a coder model is build to generate coder. The hardware platform tests show that the PID- neural network controller is fully able to achieve the requirements, the motor speed can track the given and show the faster response capability.