Adaptive fault‐tolerant robust control based on radial basis function neural network for a class of mechanical systems with input constraints

Abstract

In this paper, a robust adaptive fault‐tolerant control (FTC) strategy based on hyperbolic tangent function is proposed for a class of mechanical systems with unknown disturbance and unknown nonlinear friction model. The robust adaptive FTC algorithm proposed in this paper can maintain the system performance even after partial failure occurs in the actuator. Different from the traditional fault‐tolerant control algorithms, the proposed algorithm uses the hyperbolic tangent function and a projection algorithm to limit the control input within a known upper bound. In addition, radial basis function neural network is used to approximate the unknown continuous nonlinear function that contains the friction model in the mechanical system. The control target of the system is realized; and the asymptotic stability of the closed‐loop system is proved via the Lyapunov direct method. Finally, the numerical simulation results also demonstrate the effectiveness of the designed controller.

Decentralized adaptive neural fixed‐time tracking control of constrained interconnected nonlinear systems with partially unmeasurable states

“…where 𝜛 i,j0 , 𝜛 i,j1 , and 𝜂 i,j0 are states of system (1), 𝛾 i,j0 > 0, 𝛾 i,j1 > 0, and Ξ i,j is a function. Then, if the initial deviations 𝜛 i,j0 − Ξ i,j (t) and 𝜛 i,j0 − Ξi,j (t) are bounded, the differential term Ξi,j (t) can be approximated by 𝜂 i,j0 .…”

“…During the few decades, the adaptive backstepping control method has attracted much attention in the control field (see References 1‐12). By incorporating fuzzy logic systems (FLSs) 13‐16 or radial basis function neural networks (RBFNNs) 17‐21 into the backstepping control method, the control issue of nonlinear systems containing uncertain functions and unknown parameters can be effectively solved.…”