Litcius/Paper detail

Decoupled Active Disturbance Rejection Control for PMSM Drives to Retain Deadbeat Properties Using Composite Disturbance Observer

Jiewen Lang, Chengde Tong, Yuhong Zheng, Jingang Bai, Ping Zheng

2024IEEE Transactions on Industrial Electronics19 citationsDOI

Abstract

Despite the merit of fast dynamics, deadbeat predictive current control (DPCC) suffers from the dc and ac disturbances. To address the issue, active disturbance rejection control (ADRC) has been proposed, including the conventional and composite ADRC schemes. However, the conventional ADRC cannot suppress the ac disturbance effectively. While in the composite ADRC, the performances of the current control and disturbance suppression loops are coupled, distorting the deadbeat properties of the DPCC. In addition, both schemes are subject to weak capability in tracking the fast time-varying disturbance. To solve the problems, this article proposes a decoupled ADRC based on the composite disturbance observer, where the second-order disturbance extended state observer (SDESO) and quasi-resonant controller (QRC) are employed to suppress the dc and ac disturbances, respectively. Compared with the composite ADRC, the proposed scheme can fully decouple the performances of the current control and disturbance suppression loops, and thus deadbeat properties can be retained. In addition, by the SDESO, the proposed scheme enables superior dynamic performance in tracking the time-varying disturbance. The system stability and anti-disturbance capability of the proposed scheme are analyzed in detail. Finally, the effectiveness of the proposed scheme is validated on a 0.55kWPMSM platform.

Topics & Concepts

Control theory (sociology)Disturbance (geology)Active disturbance rejection controlComposite numberControl engineeringControl (management)Computer scienceEngineeringState observerPhysicsNonlinear systemArtificial intelligenceAlgorithmQuantum mechanicsBiologyPaleontologySensorless Control of Electric MotorsControl Systems in EngineeringIterative Learning Control Systems