Resilient Continuum Deformation Coordination

TL;DR

This paper introduces a resilient multi-agent coordination framework using continuum mechanics principles, enabling robust operation through hybrid modes that adapt to agent failures by reconfiguring group dynamics in real-time.

Contribution

It presents a novel hybrid coordination approach with two modes, HDM and CEM, incorporating leader selection and fluid flow-inspired exclusion strategies for resilient multi-agent systems.

Findings

Effective detection of agent failures in real-time.

Robust coordination maintained despite agent failures.

Fluid flow analogy improves exclusion and reconfiguration.

Abstract

This paper applies the principles of continuum mechanics to safely and resiliently coordinate a multi-agent team. A hybrid automation with two operation modes, Homogeneous Deformation Mode (HDM) and Containment Exclusion Mode (CEM), are developed to robustly manage group coordination in the presence of unpredicted agent failures. HDM becomes active when all agents are healthy, where the group coordination is defined by homogeneous transformation coordination functions. By classifying agents as leaders and followers, a desired n-D homogeneous transformation is uniquely related to the desired trajectories of n+1 leaders and acquired by the remaining followers in real-time through local communication. The paper offers a novel approach for leader selection as well as naturally establishing and reestablishing inter-agent communication whenever the agent team enters the HDM. CEM is activated when at least one agent fails to admit group coordination. This paper applies unique features of decentralized homogeneous transformation coordination to quickly detect each arising anomalous situation and excludes failed agent(s) from group coordination of healthy agents. In CEM, agent coordination is treated as an ideal fluid flow where the desired agents' paths are defined along stream lines inspired by fluid flow field theory to circumvent exclusion spaces surrounding failed agent(s).