Concepts de base
The core message of this paper is to efficiently restore a biconnected communication network in a multi-robot system after the failure of one or more robots, by minimizing the maximum movement of the robots.
Résumé
The paper focuses on the problem of Fast Biconnectivity Restoration (FBR) in multi-robot systems, where the goal is to restore a biconnected communication network as quickly as possible after the failure of one or more robots.
The key highlights are:
The authors develop a Quadratically Constrained Program (QCP) formulation to optimally solve the FBR problem, but it can only handle small instances due to high computational overhead.
They propose an approximation algorithm for the FBR problem, which divides it into two sub-problems: Graph Topology Optimization (GTO) and Movement Minimization (MM).
The GTO problem aims to find a set of edges to augment the existing communication graph to make it biconnected, while minimizing the maximum cost of the edges. The authors solve this using the Edge Augmentation (EA) algorithm.
The MM problem aims to move the robots to new positions such that the edges found in the GTO problem are realized, while minimizing the maximum movement of the robots. The authors propose two solutions for this: Sequential Cascaded Relocation (SCR) and a QCP-based formulation.
Extensive experiments show that the proposed EA-SCR algorithm significantly outperforms the existing solutions in terms of optimizing the FBR objective, while having comparable running time.
The authors also demonstrate the applicability of their proposed algorithms in the context of a practical multi-robot optimization problem, the Persistent Monitoring task.
Stats
The paper does not contain any explicit numerical data or metrics to support the key logics. The focus is on algorithmic development and empirical evaluation.
Citations
The paper does not contain any striking quotes supporting the key logics.