Distributed fault-tolerant control of modular and reconfigurable robots with consideration of actuator saturation
- PDF / 1,951,491 Bytes
- 14 Pages / 595.276 x 790.866 pts Page_size
- 52 Downloads / 245 Views
(0123456789().,-volV)(0123456789().,-volV)
ORIGINAL ARTICLE
Distributed fault-tolerant control of modular and reconfigurable robots with consideration of actuator saturation Fan Zhou1,2 • Keping Liu1 • Yuanchun Li1 • Guangjun Liu2 Received: 24 June 2019 / Accepted: 3 February 2020 Ó Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract A novel decomposition-based distributed robust fault-tolerant control method is proposed for modular and reconfigurable robots based on joint torque sensing. The designed robust controller compensates for both model uncertainties and a class of actuator faults. In addition, the proposed scheme does not require a fault detection and diagnosis module, avoiding time delay associated with it. Furthermore, a radial basis function neural network-based compensation scheme is proposed to deal with the actuator saturation problem, which is especially critical when actuator fault has to be tolerated by the control system. Simulation results have shown the effectiveness of the presented method. Keywords Modular and reconfigurable robots Decomposition-based control Neural network Distributed fault-tolerant control Joint torque sensing
1 Introduction As modular and reconfigurable robot (MRR) has the advantages of structural flexibility, low cost, easy maintenance and repair, etc., it may be applied to operate in hazardous environments where humans have no access [1–3]. Therefore, the design of fault-tolerant controller is crucial to ensure the stable operation of the robot in such an environment. In the relevant literature, most of the reported fault tolerant methods are based on infinite control input, regardless of the output capability of the actuator [4–6]. However, a robot fault-tolerant controller may encounter the problem of actuator saturation as the originally planned & Yuanchun Li [email protected] Fan Zhou [email protected] Keping Liu [email protected] Guangjun Liu [email protected] 1
Department of Control Science and Engineering, Changchun University of Technology, Changchun, China
2
Department of Aerospace Engineering, Ryerson University, Toronto, ON M5B 2K3, Canada
operation task might need to be carried out with reduced resources and likely extra control efforts upon a fault to compensate for errors due to the fault. Fault-tolerant control (FTC) aims to maintain acceptable performance upon occurrence of fault. FTC strategies are classified in to passive approaches [7, 8] and active approaches [9–11]. Active FTC is based on fault detection and diagnosis (FDD) and has been applied to control of robot manipulators [12, 13]. In [14], a robust training algorithm is proposed and successfully applied to the robotic fault system, which realizes the fault-tolerant control of the robotic system. A finite time active fault-tolerant control method with combined nonsingular fast terminal sliding mode control and high-order sliding mode is proposed in [15], which has high precision and strong robustness and can converge quickly within a limited time. I
Data Loading...
