Decentralized fault-tolerant control of modular robot manipulators with actuator saturation: neural adaptive integral terminal sliding mode-based control approach
Zengpeng Lu,Yuanchun Li,Xirui Fan,Qingyun Wang,Yan Li
DOI: https://doi.org/10.1007/s40747-023-01142-z
IF: 6.7
2023-07-07
Complex & Intelligent Systems
Abstract:Abstract A novel neural adaptive integral terminal sliding mode control for decentralized fault-tolerant control strategy, including the integral terminal sliding mode surface, the nonlinear disturbance observer, the radial basis neural network and robust controller, is presented in this paper to achieve fault-tolerant control of modular robot manipulators. First, the integral terminal sliding mode is designed for the fault-tolerant controller. Then, to boost the performance of the controlled system, the radial basis neural network and disturbance observer are introduced to approximate the nonlinear terms and disturbances. The reconstructed approximate uncertainty term is applied as compensation. Next, the super-twisting technique is employed to compensate for estimation errors to ensure stability. In addition, for the actuator saturation problem, the radial basis function neural network-based compensation control is proposed. Finally, the stability of the closed-loop robotic system is demonstrated based on Lyapunov theory. Computer simulations verified the efficiency and advantages of the presented approach.
computer science, artificial intelligence