支持向量机预测可变参数的永磁同步电机快速终端滑模控制

A novel non-singular fast terminal sliding mode (VNFTSM) control strategy of permanent magnet synchronous motor (PMSM) using support vector machine to intelligently predict its variable parameters is proposed to solve the problem that the speed regulation performance of the traditional sliding mode control of PMSM is affected by the change of the state variable of the motor system. Firstly, a new type of non-singular fast terminal sliding mode surface with fast convergent variable parameters is designed. The sliding mode surface achieves global fast convergence and eliminates the singularity phenomenon. The sliding mode parameters change with the system state, so that the permanent magnet synchronous motor can always maintain excellent speed regulation performance under different system states. Then, a support vector machine (SVM) intelligent algorithm is used to model the state variable parameters of the sliding mode surface, and the state variable parameters of the sliding mode surface can be adjusted online according to the state of the motor system. Finally, a variable parameter non-singular fast terminal sliding mode speed controller of PMSM is designed by using the second-order model of angular velocity and q-axis stator reference current to reduce the model error. Numerical simulation results show that the parameter prediction model of the sliding mode surface state variable parameters established by support vector machine predicts the optimal values of the state variable parameters of the sliding mode surface according to the changes of the motor system parameters. Compared with the traditional NFTSM control, the control performance of the proposed VNFTSM control scheme is not affected by the change of motor parameters, and has the advantages of fast convergence speed, high robustness, small speed overshoot, and high control accuracy.