Meta Navigation Functions: Adaptive Associations for Coordination of Multi-Agent Systems

TL;DR

This paper introduces Meta Navigation Functions (MNFs), a novel class of potential fields that enable adaptive, cooperative coordination among multi-agent systems, improving speed and social behavior in swarm navigation.

Contribution

The paper proposes MNFs, which allow agents to dynamically associate and coordinate more effectively than traditional decentralized navigation functions, with weaker confinement conditions.

Findings

MNFs enable faster multi-agent coordination.

Agents exhibit social behaviors through associations.

Simulations confirm MNFs' efficiency in complex swarms.

Abstract

In this paper, we introduce a new class of potential fields, i.e., meta navigation functions (MNFs) to coordinate multi-agent systems. Thanks to the MNF formulation, agents can contribute to each other's coordination via partial and/or total associations, contrary to traditional decentralized navigation functions (DNFs). In particular, agents may stimulate each other via their MNFs. Moreover, MNFs need to be confined which is a weaker condition compared to the Morse condition of DNFs. An MNF is composed of a confined function and an attraction kernel. The critical points of the former can be confined in a safe region around a target critical point. The collision-free trajectory of an agent and its associations to its peers are governed by a confined function before reaching its safe region. Then, the attraction kernel drives the agent to its target in the safe region. MNFs provide faster coordination compared to DNFs. We illustrate how MNFs may exhibit some social behaviors in the course of partial and total associations among agents. Our simulations verify the efficiency of MNFs to coordinate complex swarms of agents.