Event-triggered consensus control of continuous-time stochastic multi-agent systems

This paper addresses the event-triggered consensus of stochastic multi-agent systems (MASs), the main breakthroughs are the avoidance of infinitely fast execution and the stochastic stability analysis. Toward this, first, a novel event-triggered mechanism (ETM) based on relative information is proposed to reduce the control effects, and the enforced fixed positive lower bound of the inter-execution times in the ETM can effectively exclude the infinitely fast execution behavior. Moreover, the input-to-state stability (ISS) is analyzed and the execution error is delicately estimated so that the constant control on each inter-execution time possesses enough feedback capability to dominate the effect of execution error. Based on this, the desired consensus is achieved in the almost sure sense. Particularly, the involved stochastic convergence strategy, without using the well-known Lyapunov theorem, is hopeful to provide a potential pattern to achieve an event-triggered consensus for more general stochastic linear/nonlinear MASs. Finally, a simulation example is presented to illustrate the effectiveness of the proposed results.