A Synchronizing Approach for Non-Uniform Time-Varying Communication Delays on the Average Consensus of Multi Agent System

This paper presents the average consensus of a Multi Agent Systems (MAS) with non-uniform, asymmetric and time-varying delays in discrete integral systems. The MAS can be multi-electric vehicle system, multi-drone, multi-robot formations or an energy system. As an essential requirement of consensus control in MAS, there is the simultaneities of state quantities shared among agents. However, while the sharing of these state variables is performed via a communication network, there is the communication delay which is non-uniform and time-varying, so that the simultaneity between agents is lost and it becomes impossible to conserve the state variables of the whole system. In this paper, we firstly build a simulation model of the communication delay of an actual energy monitoring system. Next, we analyze the influence of communication delay on the average consensus by this model. From these results, we propose a restricted transmission and synchronization algorithm that can perform a good average consensus even in the presence of asymmetric, non-uniform, and time-varying delays. The proposed method adopts a communication protocol which can measure communication delay by sharing data sets of state variables and their sampling times among agents.