Event-Triggered Control For Consensus Of Multi-Agent Systems

Recent years, duo to the great potential in applications of multi-agent systems, more and more researchers have begun their research in this field and have obtained many remarkable results. There are not only theoretical research results, but also application results and experiments which were widely validated. The results cover topics including consensus application, formation control of robots, flocking, animal cluster and many other various aspects. Among these directions, the consensus research of multi-agent systems is one of the most important research branches, because it is the basis for multi-agent collaboration between each other, and it has played a very important role as the theoretical support. The consensus algorithm is a protocol to ensure that the status of individual agents get a common level. When the algorithms comes to practical implementation on digital hardware platforms, a typical coordinated control method is to assume that individual agent performs periodic sampling control. However, this implementation in the application usually suffers from the limited hardware and software resources. For this reason, this paper introduces a kind of non-uniform sampling method named event triggering method to solve the problem of resource constraints and studies the coordinated control method based on event trigger mechanism under the consensus control tasks.This paper puts forward the event-triggered consensus algorithm with the minimum sampling size. Then we developed an improved sampling-based self-triggered consensus for multi-agent systems. The purpose of the research is to reduce the events triggering numbers and decrease the inter-agent communication of the multi-agent system. We combine the combinational measurement method with the sampling based detecting scheme in the consensus protocols. Zeno triggering is avoided under these control protocols. As compared with the existing methods, the proposed method can effectively reduce the number of events, controller updates, and communication transmission.We have also proposed a new centralized dynamic event-triggered mechanism for multi-agent systems using an additional internal dynamic variable. The internal variable is used to trigger events and it is chosen to be non-monotone. The triggering conditions of the agents are not only affected by the current states and the error of the agents but also the previous ones. In addition,the triggering condition is more maneuverable.Finally, we have summarized all the results proposed in this paper, and discussed the research directions and possible future applications of the event-triggered consensus control algorithm for multi-agent systems.