K-Infy Adaptive Super-Twisting Cascade Surface States Observe Scheme for PMSM Sensorless Control

To heighten the observation accuracy and dynamic performance of the permanent magnet synchronous motor (PMSM) sensorless drive system without increasing high frequency noise sensitivity, this paper presents a sensorless control scheme. The scheme involves K-infy adaptive super-twisting sliding mode observer (ASTSMO) connected with cascade surface extended state observer (CSESO) in series. By incorporating the K-infy function into the full order STSMO, K-infy ASTSMO suppresses the chattering and overestimation common in traditional SMOs, allowing the observer to maintain reliable performance in a wide speed range. Subsequently, CSESO is connected in series after K-infy ASTSMO to serve as mechanical states observer. By designing the cascaded error surface, CSESO simplifies the tuning process and shows the ability to suppress sensitivity to high-frequency noise while enhancing the dynamic performance. Stability analysis of the proposed algorithm is provided along with parameter tuning guidelines. Through a series of experiments, the proposed scheme is proved to be capable of delivering improved observation accuracy and dynamic performance over a wide operating range.