Sensorless Control of Induction Motor Based on Super-Twisting Sliding Mode Observer With Speed Convergence Improvement

The super twisting sliding-mode observer (ST-SMO) has been proposed to achieve an effective method for alleviating low-order harmonics of measured quantities, issues related to DC drift, and suppression of chattering due to low-frequency sampling. The conventional ST-SMO, on the other hand, suffers from control delay in the convergence trajectory due to the system disturbance, resulting in decreased anti-disturbance capability and impacting the estimation accuracy and energy consumption. This paper proposed an ST-SMO with convergence improvement to address the issue related to the sliding mode controller along the sliding surface. A nonlinear sliding mode manifold is created to achieve the optimal ST-SMO convergence trajectory along the sliding surface. Then, a disturbance compensation term is added to the control law to eliminate the system control delay. In comparison to the conventional ST-SMO, the investigated method can effectively improve the anti-disturbance capability of the induction motor (IM) Observer, resulting in improved speed estimation (rotor flux control under applied load torque disturbances, speed reversal, and zero speed operation), good performance, and stability. The simulation and experimental studies are carried out for an induction motor with a 5.5kW rating. Both simulation and experimental results prove good robustness against the applied load torque disturbances and convergence of rotor speed to its actual value.