Speed Sensorless of a Bearingless Induction Motor Based on Super-Twisting-Model Reference Adaptive System

In view of the magnetic chain recognition accuracy problem of the bearingless induction motor (BIM) under direct torque control (DTC), a speed sensorless based on Super-Twisting-model reference adaptive system (MRAS) is proposed. The speed sensoerless makes rotor flux on the foundation of Super-Twisting second-order sliding mode as a reference model. The speed sensorless is constructed by using MRAS to obtain accurate velocity observations. In addition, through the construction of stator resistance adaptive, the stator resistance can be identified and corrected on-line, so as to solve the problem of the speed identification error when the stator resistance changes at the start. The simulation and experimental results show that the speed sensorless based on Super-Twisting-Model reference adaptive system and stator resistance adaptive is test more accurate, this method can not only have strong robustness to the variations of stator resistance, but also have characteristics of wide speed regulation and fast transient torque response simultaneously.