Robust integration of fault estimation and sliding mode fault‐tolerant control for interconnected systems against sensor fault

Abstract

AbstractThis paper provided a robust strategy for controlling nonlinear interconnected large‐scale systems subject to a sensor fault. The inherent challenge in controlling such systems is to sustain robust closed‐loop stability and performance in nonlinear interactions, faults, and exogenous inputs. A new closed‐loop structure has been devised to tackle this challenge by integrating the sliding mode controller with the unknown input observer. By exploiting the decoupling capability of the controller and observer, this integration ensures the robustness of the closed‐loop system against the simultaneous effect of interaction, fault, and disturbance. In this context, the unknown input observer has been constructed to supply the controller with an accurate sensor fault assessment despite additional unknown inputs. A novel approach is proposed for designing a decentralized fault‐tolerant control system that utilizes sliding mode control and proportional‐integral‐derivative controller tuning via a bacterial foraging optimization algorithm to compensate for the fault effect, leading to a robust output performance. The observer and controller gains are accomplished by utilizing the H∞ performance and linear matrix inequality formulation. The Lyapunov technique is used to demonstrate stability. A power system model emphasizes the proposed approach's robustness and effectiveness. The system performance with and without the proposed controller was compared, where the simulation results show a fast reaction to offset undesirable impacts, whereas the state estimation error approaches to zero in each subsystem.