Research On Parameter Identification Of Permanent Magnet AC Servo System

Due to the superior control performance, permanent magnet ac servo system is more and more widely applied in various fields. In the actual operation, electrical and mechanical parameters of the motor changes because of the change of windings temperature, magnetic circuit saturation, and the mutation of load. If we don’t adjust the controller parameters correspondingly, control performance of the servo system will decline. Therefore, research on parameter identification algorithm and self-tuning of controller parameters based on identification value is of great practical significance.This paper is mainly about identification algorithm of the electrical parameters(resistance, inductance) and moment of inertia for permanent magnet synchronous motor, the main research contents are as follows:Firstly, the vector control theory with rotor field oriented is illustrated, and the design principle of current and speed loop parameters is mainly analyzed. Then the simulation model of PMSM AC servo system is set up in the MATLAB/Simulink environment to analyze system dynamic and steady-state performance.Secondly, the theory of model reference adaptive algorithm to identify parameters is introduced in this paper, in which reduced order current observer is treated as the adjustable model. Adapted law for parameter identification is also derived based on PoPov’s hyperstability theory. Then simulation model for the parameter identification module is established to analyze the influence of different adaptive adjustment parameters on the identification results.Then the moment of inertia identification method is studied. The paper illustrates three methods of identification: the off-line identification method based on arbitrary trajectory planning、discrete-time model reference adaptive algorithm and forgetting factor recursive least square method. Three simulation models is set up respectively in MATLAB/Simulink environment. In the same simulation conditions, the constant inertia, variable inertia simulation is carried out respectively for the latter two methods in order to compare the identification performance.Finally, this paper studies realization of parameter identification in the hardware platform. The related hardware circuit and software programming ideas are introduced for the experiment, which uses model reference adaptive algorithm to identify electrical parameters and uses offline method to identify the moment of inertia. The experimental results show that the identification methods adopted in this paper is practical.