A Decentralized Cluster Formation Containment Framework for Multirobot Systems

Junyan Hu,Parijat Bhowmick,Inmo Jang,Alexander Lanzon,Farshad Arvin

doi:10.1109/tro.2021.3071615

Abstract

Cooperative control of multirobot systems (MRSs) has earned significant research interest over the past two decades due to its potential applications in multidisciplinary engineering problems. In contrast to a single specialized robot, the MRS can be designed to offer flexibility, reconfigurability, robustness to faults, and cost-effectiveness in solving complex and challenging tasks. In this article, we aim to develop a unified cluster formation containment coordination framework for networked robots that can be decomposed into two layers containing the leaders and the followers. According to the proposed methodology, the leader robots are first distributed into a set of distinct and nonoverlapping clusters depending on the positions and priorities of the targets exploiting a game-theoretic rule. Then, they are steered to attain the desired formations around the corresponding targets. Subsequently, the follower robots are made to converge into the convex hull spanned by the leaders of the individual clusters. A prototype search and rescue operation is considered to highlight the usefulness of the proposed coordination framework. Furthermore, real-time hardware experiments were conducted on miniature mobile robots to validate the feasibility of the theoretical results.

Highlights

With the recent advancements of computing, communication, sensing and control techniques, cooperative control of multi-robot systems (MRS) has established its worth in the recent years [1]
(Step 2:) A decentralized and adaptive cluster, formation tracking (CFT) protocol has been proposed for the leader robots in Subsection III-B;
(Step 3:) A decentralized and adaptive containment control law is designed for the follower robots in Subsection III-C. Apart from these clustering, formation tracking and containment methodologies, the proposed framework encompasses a set of obstacle and collision avoidance protocols because during real-time implementation, obstacle and/or collision avoidance may frequently be required

Summary

Introduction

With the recent advancements of computing, communication, sensing and control techniques, cooperative control of multi-robot systems (MRS) has established its worth in the recent years [1]. By using this technique, a large number of autonomous robots connected via a communication network can be coordinated to achieve a common objective [2]. The coordinated movements of the robots under the effect of the information flow can be analyzed via graph-theoretic approaches [1], [2]. MRS can offer flexibility, reconfigurability, robustness to faults and cost-effectiveness over a single sophisticated robot in solving complex and challenging tasks. MRS can be potentially utilized in military applications

Objectives

Results

Conclusion