Approximate guaranteed cost fault-tolerant control of unknown nonlinear systems with time-varying actuator faults

Abstract

In this paper, the guaranteed cost faulttolerant control problem for unknown multi-input continuous nonlinear systems with loss of actuator effectiveness faults is investigated using the adaptive dynamic programming algorithm. Initially, by modifying the cost function to account for actuator faults, the problem is transformed into an optimal control problem of the nominal system. Subsequently, by using an existing policy iteration (PI) algorithm to solve the corresponding optimal control problem, a guaranteed cost controller is constructed approximately. Furthermore, a rigorous proof is given to show the convergence of the aforementioned PI algorithm while taking the neural network approximation errors into consideration. Finally, simulation examples are provided to show the effectiveness of the proposed approach.

Data‐based fault‐tolerant control for uncertain linear systems with actuator faults

“…In this paper, we discuss the problem of how to design adaptive fault‐tolerant controllers by using the initial stabilising control gain , such that the states converge asymptotically to zero even in the presence of actuator faults. Moreover, it should be pointed out that the initial admissible control is necessary in most of the existing results concerning policy iteration algorithms [26, 27, 30, 32–38].…”

“…However, due to the difference between discrete‐time and continuous‐time systems, these data‐driven approaches cannot be extended to the case of continuous‐time systems. Most recently, by modifying the cost function to account for actuator faults, a data‐based FTC approach is proposed in [26] for continuous‐time non‐linear systems, which guarantees that the system has a guaranteed cost bound. Nevertheless, the type of faults under consideration is loss of actuator effectiveness.…”

Distributed fault‐tolerant control for over‐actuated multi‐agent systems with uncertain perturbations using control allocation

“…Optimal control based methods have also received a lot of attention in recent years. [13][14][15] The significant feature of this robust control method is that it can incorporate uncertainties into the cost function of the optimal control, which means that the system performance can be guaranteed to a certain extent while ensuring the robustness. However, the optimal control of nonlinear systems is very difficult to solve, and thus the reinforcement learning based adaptive optimal control has been developed for solving this challenge.…”

“…Typical robust control methods include H 2/H $$$ \infty $$$ control, 10 sliding mode control, 11 gain assignment technique, 12 and so forth. Optimal control based methods have also received a lot of attention in recent years 13‐15 . The significant feature of this robust control method is that it can incorporate uncertainties into the cost function of the optimal control, which means that the system performance can be guaranteed to a certain extent while ensuring the robustness.…”

Event‐triggered guaranteed cost fault‐tolerant optimal tracking control for uncertain nonlinear system via adaptive dynamic programming

“…As we all know, there are various faults randomly occurring in real engineering systems, such as sensor faults and actuator faults, which can reduce the reliability and robustness of the controlled system. [34][35][36] Therefore, it is necessary to consider the faults in the controller design. 37 While actuator fault is one of the most common faults, many fault-tolerant optimal control methods based on ADP have been developed.…”