Research On Kinematic Parameter Identification Algorithm Of PMSM

Compared with ordinary AC variable frequency motor, permanent magnet synchronous motor has advantages including simple structure, high power factor, low operation loss, simple and convenient control, widely used in the high performance servo field which demands accurate positioning, fast response, wide speed range. However, with the improvement of the actual demand and the deepening of the research, the requirement to PMSM servo system is rising. PMSM’s mechanical performance is affected by the moment of inertia and load torque. If the system can identify the inertia and load torque in real-time and adjust its control system on the basis of identification results quickly, the performance of servo system can greatly improve. In this paper, the moment of inertia and load torque identification of permanent magnet synchronous motor are studied as follow:In this paper, we establish the PMSM vector control mathematical model and its simulation model, the inertia identification is based on the model reference adaptive algorithm(MRAC) and recursive least squares algorithm(RLS). The identification results of the two algorithms under different given conditions are analyzed by MATLAB simulation. The influence of the gain coefficient on MRAC and the influence of the forgetting factor on RLS are studied. In view of the algorithms’ contradiction in rapidity and stability, the paper puts forward the corresponding improvement method.To load identification, this paper uses state reconstruction creating two kinds reduced-order load torque observer with different structure. Through the establishment of simulation model, select a different pole placement, analyze factors which effect of reduced-order load torque observer identification, and select the optimal parameters to achieve the effective observation of the load torque. At the same time, using the identification results as the pre-load torque feed-forward compensation to improve the dynamic response capability of the system.Finally, the paper builds a servo motor inertia experimental test platform based on Schneider servo motor, and uses Lab VIEW to develop a PC software interface. Design experiments which simulates practical working environment of servo motor system inertia and load torque appearing changes to verify the validity of the identification algorithm.