Fault-tolerant consensus control of multi-agent systems under actuator/sensor faults and channel noises: A distributed anti-attack strategy

Abstract

This study investigates the fault-tolerant consensus control problem of multi-agent systems subject to simultaneous actuator/sensor faults and channel noises in physical hierarchy and hostile connectivity-mixed attacks in cyber hierarchy. Actuator/sensor faults are remodeled into unified abrupt-type and incipient-type characteristics, and connectivity-mixed attacks are established with connectivity-maintained and connectivity-paralyzed topologies by a switching and nonoverlapping version. Normalization and estimation-based observer is devised to recollect unknown state and fault observations, and distributed anti-attack fault-tolerant consensus strategy is also developed to achieve the tolerance to faults, resilience to attacks and robustness to noises, respectively, with the novel incorporated sensor fault and output channel noise estimation as well as neighboring output information. Criteria of executing leader-following consensus of multi-agent systems under cyber-physical threats are derived with attack frequency and activation rate technologies. Effectiveness and improvements of the proposed fault-tolerant consensus algorithm are validated on two case studies: 1) multi-machine power system synchronization and 2) multi-aircraft system coordination.