Enhancement of Low Speed Performance in Sensorless Model Predictive Control for Induction Motors

For sensorless control systems of induction motors based on model predictive control(MPC), the motor model-based sensorless control methods suffer from low-speed regeneration instability issues, which significantly limit the lowspeed operational performance of sensorless systems. To address this problem, this paper proposes an adaptive lawmodified full-order observer-based model predictive control method for induction motors. First, the root cause of low-speed regeneration instability is investigated through analysis of the transfer function characteristics of the full-order observer. Second, an improved adaptive law is developed by incorporating excitation current error into the conventional speed adaptation mechanism. Finally, design of the adaptive law correction coefficients, enhanced low-speed operational characteristics are achieved. The proposed method's effectiveness in achieving stable speed estimation within the low-speed regeneration region is verified through analysis of the system and simulation.