Title:Synchronized Task Decomposition for Cooperative Multi-agent Systems

Abstract: The key challenge in cooperative control for multi-agent systems could be how to design the local interaction rules and coordination principles among agents so as to achieve certain desired global behaviors. In this paper, we try to tackle this challenge from the angle of hierarchical control, and propose a divide-and-conquer approach. The basic idea is to decompose the requested global specification into subtasks for individual agents or small clusters of agents. It should be noted that the decomposition is not arbitrary. The global specification should be decomposed in such a way that the fulfilment of these subtasks by each individual agent will imply the satisfaction of the global specification as a team. Formally, a given global specification can be represented as an automaton A, while a multi-agent system can be captured as a set of parallel distributed systems. The first question needs to be answered is whether it is always possible to decompose a given task automaton A into a finite number of sub-automata Ai, where the parallel composition of these sub-automata Ai is bisimilar to the automaton A. By a counter example, it is shown that it is not always possible to do so. Then, a natural follow-up question is what the necessary and sufficient condition should be for the proposed decomposability of a task automaton. The main part of the paper is to answer this question, and it is found that the automaton A is decomposable if it has some symmetric properties. For two agents, this condition is a necessary and sufficient condition. To illustrate the results, a multi-robot system example is given. This result may pave the way towards a new perspective for decentralized cooperative control of multi-agent systems.