Backstepping sliding mode fault-tolerant control for the wind turbine system with disturbance observer. (October 2022)

• Record Type: Journal Article
• Authors: Ma, Chenyuan; He, Xinlin; Xie, Beibei; Sun, Weipeng; Zhao, Daoli; Liao, Weili
• Is Part Of: Proceedings of the Institution of Mechanical Engineers. Volume 236:Number 9(2022)
• Page Start: 1667
• Page End: 1678
• Publication Date: 2022-10
• Subjects: Backstepping sliding mode fault-tolerant control -- nonlinear disturbance observer -- wind turbine system -- actuator faults; Mechanical engineering -- Periodicals; Automatic control -- Periodicals; Systems engineering -- Periodicals; 621.3
• DOI: 10.1177/09596518221103041
• Languages: English
• ISSNs: 0959-6518

Abstract

To guarantee the safe and stable operation of wind farms, this article establishes a backstepping sliding mode fault-tolerant controller for the wind turbine system to surmount uncertain problems, including actuator gain–bias faults, system modeling errors, and external stochastic disturbances. The nonlinear disturbance observer is employed for the stochastic disturbances, which can online estimate and compensate the external disturbance term. In addition, the backstepping control strategy is introduced to reduce the complexity of fault-tolerant controller design. Subsequently, combining the backstepping control algorithm and nonlinear disturbance observer, a disturbance observer-based backstepping sliding mode fault-tolerant control approach is applied for the wind turbine system. Thereinto, the terminal attractor is employed, which is mainly utilized to improve the convergence rate of the sliding surface and reduce the chattering phenomenon. The stability of the wind power closed-loop control system is rigorously verified via Lyapunov stability theory, which can obtain satisfactory control performance. Finally, numerical simulation results demonstrate that the proposed control approach can guarantee that the system state quickly reaches stability within 6–8 s, and the steady-state adjustment time is greatly reduced to 50%–62% when compared with the proportional–integral–derivative control and sliding mode control.