Research activities per year filter by:
Articles
Publication Date: 2025
International Journal of Robust and Nonlinear Control (10498923)35(18)pp. 8131-8143
This paper focuses on the challenge of integrated active fault-tolerant control (IAFTC) for linear discrete-time systems utilizing compatible linear matrix inequality (LMI) techniques. The presence of input nonlinearity, additive faults, external disturbances, and uncertainty in the system matrix makes this problem more applicable to real-world systems while expanding the complexity and functional range. In such a situation, the designer must take an integrated approach instead of synthesizing fault estimation (FE) and fault-tolerant control (FTC) as separate modules; yet, the LMI-based design conditions can be overly conservative. As a result, the reduction of the bi-directional interaction effect between FE and FTC units does not lead to satisfactory improvements in the overall behavior of the control system. To provide an integrated design approach, we propose a dynamic output feedback controller that plays the role of the FTC unit. This scheme incorporates two effective decision parameters and a descriptor observer (As FE), contributing to the IAFTC block's functionality. The design criteria are formulated using tractable LMI constraints in a convex optimization problem, and the (Formula presented.) -stability criterion is proved for the overall closed-loop system. The superiority and effectiveness of the proposed IAFTC are demonstrated through three comparative simulation examples. © 2025 John Wiley & Sons Ltd.
Publication Date: 2024
JVC/Journal of Vibration and Control (10775463)30(9-10)pp. 2251-2270
This paper addresses the problem of maximum power point tracking of photovoltaic (PV) systems in the presence of model uncertainty as well as varying load and atmospheric conditions using techniques based on a Takagi–Sugeno fuzzy model. The proposed approach relies on the linear matrix inequality tool, Lambert W function, and the Newton–Raphson method. First, adopting a quadratic Lyapunov function, an active observer-based fuzzy non-parallel distributed compensation (non-PDC) controller is designed for asymptotic tracking of the desired reference input. Next, to subdue the impact of uncertainty on the PV system, the closed-loop nominal system is regarded as a reference model, and then the main control law is developed using an online lumped uncertainty estimator and keeping the nominal control law within the staple controller. This control law does not require that the bounds on uncertainties be known. The reference voltage is determined by a novel maximum power point-seeking algorithm that is organized based on the one-diode model of PV panel, Lambert W function, and Newton–Raphson method. Finally, simulations are performed for three scenarios to point out the merits and effectiveness of the proposed methodology in the presence of system uncertainties, environmental changes, and load variations. © The Author(s) 2023.
Publication Date: 2023
ISA Transactions (00190578)138pp. 212-225
This paper proposes an active fault-tolerant control (FTC) approach based on the controller management and virtual actuator idea for linear discrete-time systems subject to unknown L2-bounded disturbances, input constraint, and time-varying additive actuator faults. The closed-loop faulty system, which includes the modified nominal controller, the fault and state estimator, and the virtual actuator, suppresses the effects of disturbances and faults, while ensuring input-constraint satisfaction. The management of the nominal controller is performed through an online optimization method – in the form of a standard quadratic programming problem – by manipulating the reference input and intervening in the nominal controller evolution. The proposed method proves the input-to-state stability (ISS) criterion of the overall closed-loop faulty system. The problem of minimizing the ultimate bound of the ISS criterion is formulated in terms of tractable linear matrix inequality (LMI) conditions that allow the fault and state estimation errors to converge to a small neighborhood of the origin. To illustrate the capabilities and advantages of the proposed control strategy, comparative simulation results are presented for a flexible joint robotic system, tracking control of a DC motor's angular velocity, and the multivariable VTOL aircraft. © 2023
Publication Date: 2023
IEEE Transactions on Fuzzy Systems (1063-6706)31(10)pp. 3543-3554
This article addresses the problem of fault-tolerant output tracking control for a class of Takagi-Sugeno (T-S) fuzzy systems with unmeasurable premise variables subject to additive and multiplicative actuator faults and external disturbances. In nominal conditions, utilizing a quadratic Lyapunov function and nonparallel distributed compensation technique, the suggested strategy delivers linear-matrix-inequality-based constraints. Simultaneously design of the proportional-integral (PI)-like state feedback controller and fuzzy antiwindup compensator is achieved with the aim of output tracking. In the faulty case, by considering the nominal system as a reference model, a direct adaptive projection-based approach is developed using the T-S fuzzy modeling and control techniques to supply the adaptive fault-tolerant controller components. An enhanced PI state/fault observer with unmeasurable premise variables is introduced only to provide the estimation of states to be used in the proposed controller. The overall closed-loop system ensures the uniformly ultimately bounded solutions for error dynamics. Two examples, subsuming an inverted pendulum and a chaotic power system, have been used to present the merits and efficiency of the suggested approach persuasively. © 1993-2012 IEEE.
Publication Date: 2022
IEEE Transactions on Fuzzy Systems (1063-6706)30(6)pp. 1914-1928
Active fault-tolerant control strategy aimed at tracking is studied for a class of discrete-time Takagi-Sugeno (T-S) fuzzy systems in the presence of input constraint and additive fault. In the fault-free case, the state space of the system is partitioned based on the presence or absence of inputs. Using the parallel distributed compensation technique, input-to-state stability is proved for the error dynamics. For input constraint satisfaction, a model predictive control based reference-management unit is proposed in the form of an online optimization module that solves a quadratic programming problem. In the faulty case, the fault-free system is used as a reference model that satisfies the input constraint. Then, the concept of virtual actuator is used to make the behavior of the faulty system similar to the behavior of the fault-free system by controlling the performance degradation. In this article, a full-order observer can be used for fault estimation. In order to keep the control input of the faulty system within the admissible range defined for the fault-free system, a constraint-change scenario is considered. After detecting the fault, the constraint on the input of the fault-free system is changed in a way to keep the control input of the faulty system in the specified range. All design criteria are formulated as a linear matrix inequality problem. In order to evaluate the proposed control approach, simulations are performed on two systems: a 3-DoF helicopter and a 2-D overhead crane. © 1993-2012 IEEE.
Publication Date: 2021
Journal of the Franklin Institute (00160032)358(18)pp. 9510-9541
A new control design approach is proposed for a class of nonlinear systems expressed by Takagi–Sugeno (T-S) fuzzy model, considering several objectives including robustness against input time-varying delay, input constraint satisfaction, and reference tracking. The proposed controller is designed on the basis of an augmented model, Lyapunov–Krasovskii functional, linear matrix inequality (LMI) tools, and parallel distributed compensation (PDC) approach. Proof of the input-to-state stability (ISS) criterion is provided for the error dynamics. Input constraint satisfaction is performed using a reference-management algorithm based on the linearized closed-loop system from the reference input to the constrained variables. In order to illustrate the effectiveness of the proposed control approach, simulations are performed on three practical examples, including a flexible-joint robot and a continuous stirred tank reactor (CSTR). © 2021 The Franklin Institute
Publication Date: 2020
pp. 2027-2032
In this paper, a supervisory active fault tolerant control (SAFTC) scheme is presented for linear systems with constraints. The SAFTC framework is composed of a fault diagnosis module, a reconfigurable controller and a pre-designed command governor. In the presence actuator faults, constraints are guaranteed by governing the setpoint to the nearest admissible value. The command governor is adopted in such a way that no recalculation and reconfiguration is needed even in the presence of actuator fault occurrence and reconfiguration of the main controller. The input redundancy assumption, needed to recover the closed-loop tracking performance of system with faulty constrained actuator, is relaxed. A simulation study on VTOL aircraft is performed to evaluate the effectiveness of the proposed method. © 2020 EUCA.
