Adaptive event-triggered based fault detection filtering for unmanned surface vehicles under DoS attacks

This paper proposes an adaptive event-triggered fault detection filter to warn of faults happening to unmanned surface vehicles in a network. First, the unmanned surface vehicle is modeled as a Markovian jump system which makes allowances for the influences of waves or other disturbances. Then, a fault detection filter is developed to produce crucial residual signals, despite disturbances and denial-of-service attacks. Inputs of the designed filter originate from a resilient adaptive event-triggered mechanism, of which the threshold can be adjusted for saving valuable network resources and mitigating adverse influences of denial-of-service attacks. By using the residual signals, detection evaluation functions are established and a detection logic algorithm is devised. Under this framework, the co-design of both the fault detection filter and the adaptive event-triggered scheme is given. Finally, the proposed method is proved to be applicable by simulating a real unmanned surface vehicle model.