Speed sensorless control of a BL-IM based on an adaptive nonsingular terminal sliding mode observer

Aiming at the speed detection accuracy of the optical encoder in a bearingless induction motor (BL-IM) affected by external environment and other factors, a speed sensorless control strategy based on an adaptive nonsingular terminal sliding mode observer (ANTSMO) is proposed. Firstly, a second-order nonsingular terminal sliding mode surface is constructed based on the stator current error. According to the designed sliding mode control law, the stator current and its common coupling term with the rotor flux are observed. Secondly, in order to avoid the bias effect caused by the pure integrator, the model reference adaptive system (MRAS) identification structure is used to realise the speed closed-loop observation . The adaptive sliding mode gain is designed by the identified speed. The stability and convergence of the observation system are proved through the Lyapunov theorem and Popov superstability theory. Finally, simulation analysis and experimental verification of the proposed strategy are carried out on MATLAB/Simulink and the test prototype. The results show that the proposed strategy not only can effectively reduce the observation chattering, and the speed can be accurately estimated under no-load, sudden load and variable speed conditions, but also the BL-IM can achieve good suspension characteristic of the motor rotor.