Fault diagnosis observer and fault-tolerant control design for unmanned surface vehicles in network environments

Abstract

This study focuses on the network-based fault estimation and fault-tolerant controller designing for an unmanned surface vehicle submit to actuator faults, along with transmission delays, packet dropouts and packet disordering in the communication network channels between sampler and observer, and between controller and actuator. Compared with manned surface vehicles, unmanned surface vehicles in the network bring certain competitive superiorities as well as challenges. By using an intermediate variable, an observer is devised in network environments to estimate the states and actuator faults of the unmanned surface vehicle simultaneously. A sufficient condition is introduced and proved for the fault observer being uniformly ultimately bounded. Based on the fault observer, a fault-tolerant controller is proposed, which can ensure that the network-based closed-loop control system is uniformly ultimately bounded theoretically. Theoretical analysis and simulation results verify the performance of the fault observer and fault-tolerant control in network environments for an unmanned surface vehicle.