4.6 Article

Guidance, Navigation, and Control of AUVs for Permanent Underwater Optical Networks

期刊

IEEE JOURNAL OF OCEANIC ENGINEERING
卷 48, 期 1, 页码 43-58

出版社

IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/JOE.2022.3204328

关键词

Oceans; Optical fiber networks; Navigation; Relays; Kinematics; Ad hoc networks; Trajectory; Decentralized model predictive control (D-MPC); ocean exploration; swarm robotics; underwater communication network

向作者/读者索取更多资源

This article presents guidance, navigation, and control systems for driving autonomous underwater vehicles (AUVs) to establish a permanent optical network between the deep sea and the surface. The challenge lies in guaranteeing a path through the AUV network that can relay the signal for an unlimited time within the finite battery life of the AUVs. By combining decentralized model predictive control (D-MPC) with graph theory, the guidance system can calculate the trajectory of each agent relaying the optical signal and form local leader-follower formations with redundant agents. The D-MPC solves a constrained optimization problem to enable immediate re-routing of the ad hoc network if a leader fails to relay the signal, and the combination with graph theory allows the dynamic topology of the swarm to change and relay the optical signal indefinitely.
Ocean exploration is in its infancy with more than 80% of the oceans still unexplored. Contrary to space exploration, electromagnetic waves have seldom been used underwater because of their heavy absorption by seawater. In this article, we present guidance, navigation, and control systems to steer autonomous underwater vehicles (AUVs) such that a permanent optical network can be set up between the deep sea and the surface. The challenge is to guarantee that there is consistently a path through the AUV network to relay the signal for an unlimited time despite the finite battery's lifetime of the AUVs. To achieve this, the guidance system combines a decentralized model predictive control (D-MPC) with graph theory to compute the trajectory of each agent relaying the optical signal as well as of redundant agents forming local leader-follower formations with each relay acting as leader of its own subfleet. The D-MPC solves a constrained optimization problem where the communication and collision avoidance constraints are formulated to enable two successive leaders and one of their followers to remain within optical range and thus allowing to immediately re-route the ad hoc network if a leader fails to relay the signal. Its combination with graph theory allows to change the dynamic topology of the swarm, with additional agents diving from a charging station, such that the network can relay the optical signal for an unlimited time.

作者

我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。

评论

主要评分

4.6
评分不足

次要评分

新颖性
-
重要性
-
科学严谨性
-
评价这篇论文

推荐

暂无数据
暂无数据