Robust Fault-Tolerant Control of an Electro-Hydraulic Actuator With a Novel Nonlinear Unknown Input Observer

Abstract

In this paper, a novel adaptive fault-tolerant controller is proposed for a typical electrohydraulic rotary actuator in the presence of disturbances, internal leakage fault, and sensor fault simultaneously. To construct the suggested controller, a nonlinear unknown input observer is developed to effectively identify the sensor fault, which is unaffected by not only internal leakage fault but also mismatched disturbances/uncertainties. Furthermore, a radial basis function neural network is designed to compensate for the mismatched disturbances/uncertainties caused by payload variation and unknown friction nonlinearities. Besides, an adaptive law based on the projection mapping function is applied to tackle the effect of the internal leakage fault. The integration of the above-mentioned techniques into the adaptive backstepping terminal sliding mode is investigated to obtain high tracking performance, robustness as well as fast convergence. The stability of the closed-loop system is proven by the Lyapunov theory. Finally, the capability and effectiveness of the proposed approach are validated via simulation results under various faulty scenarios.

A Robust Observer for Sensor Faults Estimation on n-DOF Manipulator in Constrained Framework Environment

“…s J s (31) where   F . The external force disturbance in this case is presented by lumped uncertainty ∆, including noise or other external disturbances and modeling error.…”

“…By using the adaptive law (29), virtual torque command (30), and the control signal (31), the influence of the lumped mismatched uncertainty is canceled out. Then, the derivative of the Lyapunov V becomes: .…”

“…As the highly coupled nonlinear characteristics, the actuator fault causes obvious effects on the system dynamics; therefore, they can be easily detected through monitoring abnormal system performances. Following the literature of actuator fault detection and isolation, and estimation, such fault can be considered as matched or mismatched disturbances and observed by using observers such as extended state observers (ESO) [15], [18], [19], disturbance observers (DO) [12], [20]- [22], uncertainty and disturbance estimator (UDE) [23], [24], time-delayed estimation (TDE) [25]- [27], super-twisting algorithm [28], Kalman estimators [29], unknown input observer [30], [31], etc. Then, Fault-tolerant control (FTC) that combines these observers compensation with advanced control algorithms such as active disturbance rejection control (ADRC) [32], [33], adaptive algorithms [34]- [36], or high robust gains can be employed to address the influence of the actuator fault.…”

Design of Robust Fault-Tolerant Control for Target Tracking System With Input Delay and Sensor Fault

“…Many aspects of the active FTC have been covered in the overview studies [20]- [23]. For designing the FDI, fullstate observers [24], [25], unknown input observers [26], [27], time-delay estimation [28], and numerous approaches have been proposed. The main difficulty of designing the FDI remains at decoupling the fault from other unknown factors, such as disturbances, perturbed parameters, and time delay.…”

“…For the FTC law, the issue is to cooperate the fault information in the control law such that the influences of the faults are compensated while the system stability is still preserved. Adaptation [26], switching [29], reconfiguration [24], and robust control laws [25] are popular techniques that have been implemented.…”