The Sliding Mode Control Of PMSM Based On Identification And Compensation

With the advantages of small size, high reliability and high power density, the permanent magnet synchronous motor(PMSM) has been widely used in the fields of spaceflight, national defense, industrial automation and so on. However, as a nonlinear, strong coupling and time-varying controlled object, the PMSM is easily affected by many internal and external disturbances. Therefore, it is of great significance to improve the dynamic and static characteristics of the control system on the basis of strong anti-interference ability. This paper focuses on the sliding mode variable structure control theory, designs the identification algorithm of the inertia and the disturbance observer, then proposes a kind of terminal sliding mode(TSM) control method for the PMSM.Firstly, based on the mathematical model of PMSM, a terminal sliding mode speed controller has been designed. It effectively solves the problem of the traditional linear sliding mode controller that the system state variables can not converge to the equilibrium point in finite time. But the TSM algorithm will cause the problem of control singularity and poor dynamic response characteristics. In order to improve above defects, the paper proposes the nonsingular fast terminal sliding mode(NFTSM) control method, which not only eliminates the singularity region of state space, but aslo increases the convergence speed greatly when the system states are far away from the equilibrium point. By the way, the dynamic characteristics of the control system has been improved effectively.Secondly, to obtain the accurate system moment of inertia, two on-line identification schemes, the recursive least square method and the gradient correction method, are designed in the paper. Through the theoretical analysis and simulation, the gradient correction identification method shows the superiorities that the more rapid convergence speed in entire period, higher stable precision and stronger recovery capability from disturbances, and also solves the problem of the recursive least square medthod that identification value can not reach the target in finite time. So the gradient correction method is more applicable to the PMSM system.Thirdly, the extended sliding mode disturbance observer(ESMDO) is designed for the internal parameter perturbation and external load disturbance of the PMSM control system. By means of reasonable parameter configuration, ESMDO realizes the accurate observation of all kinds of system disturbances. Subsequently, the controller is compensated by the disturbance estimation value so that to improve the anti-interference ability.Finally, the PMSM experimental platform based on the DSP has been set up to verify the effectiveness of the presented algorithms. The results show a good practical effect of the algorithm very well.