Optimization Design of PMSM Sensorless Control Using Generalized Integrator

The requirements that a generalized integrator (GI) should meet are deduced by analyzing the extended back electromotive force (EEMF) in this article, and then the unified transfer function is obtained forward instead of traditional block diagram. With theoretical, simulation, and experimental analysis, the relationship between the order and pole distribution of GI and the system realizability, harmonic suppression, estimation error, speed range, loading capacity and stability is acquired, and then the optimal third-order GI (TOGI) and optimal fourth-order GI (FOGI) are obtained. Aiming at the system instability caused by the great error in deceleration, tracking differentiator (TD) is introduced to compensate the position error. The low-frequency noise in the no-load position error is significantly weakened by partially adopting given ramp speed in the feedback speed, which expands the low speed range. The experimental results verify the rationality of the optimization scheme, effectiveness of these two solutions, and also prove that optimized TOGI has the best overall performance.