DYNAMIC EVENT-TRIGGERED LEADER-FOLLOWER CONSENSUS CONTROL FOR MULTIAGENT SYSTEMS

A new decentralized event-triggered strategy is proposed to realize consensus of leader-follower multiagent systems under fixed/switching topologies, in which a clock-like variable is designed for each follower agent to determine the event-triggered time. The basic idea of this novel control strategy is to set a positive initial value for each clock-like variable and design its decline rate so that the variable decreases from a preset maximum value to zero. That is, when the value of the clock-like variable is equal to or less than zero, the follower agent corresponding to the clock-like variable will update its feedback control input. Simultaneously, the value of the clock-like variable will be reset to the maximum. Furthermore, it is guaranteed that there exists a strictly positive minimum of interevent time (MIET) for each follower agent; thus the Zeno phenomenon is ruled out. In particular, the positive MIET can be adjusted according to the intrinsic data sampling frequency of the follower agent. It is worth noting that the event-triggered strategy is robust. In other words, if the state of a follower agent is affected by stochastic disturbance or noise, the control strategy can still guarantee that there exists a strictly positive MIET for the follower agent. Therefore, this control strategy is of engineering applicability. Three numerical simulations are provided to verify the correctness and effectiveness of the strategy.