Distributed Optimization For Multi-Agent Systems Via Event-Triggered Control

Distributed optimization has been a hot topic in information field due to its wide applications in smart grids,the UAV formation,wireless sensor network and so on.This paper is focused on the distributed optimization problem of continuous-time linear multi-agent systems.Existing results often assume the agents can update the controllers and communicate with other in a continuous manner,which might cost too much computation and communication resources.Thus,we aim at evet-triggered distributed algorithms that can ensure the distributed optimization goal.The contributions of this paper can be summarized as follows:1.We have developed a novel event-triggered rule for each agent to update its own controller.This joint with an established optimal signal generator can solve our distributed optimization problem.It is proven that all the outputs of agents will converge to the expected global optimizer while the Zeno phenomenon does not exist in the closed-loop system.The effectiveness of this event-triggered controller updating mechanism is also verified by the numerical simulation.2.We have developed a novel event-triggered optimal signal generator with discrete communications to solve the formulated distributed optimization problem.Note that only discrete communications are required at the triggering time instants.Both theoretical analysis and numerical simulation show the effectiveness of the eventtriggered communication mechanism.3.We combine the preceding event-triggered controller updating and communication mechanisms and present a novel distributed optimization algorithm.Compared with existing designs,the agents only need to update the tracking controller and communicate with each other at some discrete time instants,which definitely save many computation and communication resources as expected.