Robustness Improvement of Speed Estimation in Speed-Sensorless Induction Motor Drives

In the traditional adaptive full-order observer (AFO) for speed-sensorless induction motor drives, the estimated rotor flux error is neglected in the speed estimation algorithm. In this paper, the estimated rotor flux error and an error coefficient are adopted in the speed estimation algorithm to improve the robustness of speed estimation. Based on the proposed speed estimation algorithm, the design of AFO is divided into two types: 1) designing the error coefficient with existed feedback gains for the stability of speed estimation and 2) designing feedback gains with constant error coefficient for the stability of speed estimation. With both proposed methods, the estimated rotor speed error caused by motor parameter deviation can be reduced especially at low-speed regenerating range. The feasibility of proposed methods is verified by experiments.