Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

USV-AUV Collaboration Framework for Underwater Tasks under Extreme Sea Conditions

Jingzehua Xu, Guanwen Xie, Xinqi Wang, Yimian Ding, Shuai Zhang

TL;DR

This work tackles the challenge of performing coordinated underwater tasks with high reliability in extreme sea conditions by coupling a USV-side path-planned localization scheme with RL-enabled multi-AUV coordination. It advances accurate AUV positioning through maximizing $\det(\mathbf{J}_m)$ in a USBL-based framework and enriches multi-AUV decision-making with environment-aware RL that accounts for ocean currents and USV proximity via customized rewards. The approach is validated through extensive simulations of a multi-AUV data-collection task under both ideal and extreme sea states, demonstrating robust performance and improved positioning accuracy, plus a practical demonstration of scalable coordination. The authors also provide open-source simulation code to encourage further research in USV-AUV collaboration for underwater exploration.

Abstract

Autonomous underwater vehicles (AUVs) are valuable for ocean exploration due to their flexibility and ability to carry communication and detection units. Nevertheless, AUVs alone often face challenges in harsh and extreme sea conditions. This study introduces a unmanned surface vehicle (USV)-AUV collaboration framework, which includes high-precision multi-AUV positioning using USV path planning via Fisher information matrix optimization and reinforcement learning for multi-AUV cooperative tasks. Applied to a multi-AUV underwater data collection task scenario, extensive simulations validate the framework's feasibility and superior performance, highlighting exceptional coordination and robustness under extreme sea conditions. To accelerate relevant research in this field, we have made the simulation code (demo version) available as open-source.

USV-AUV Collaboration Framework for Underwater Tasks under Extreme Sea Conditions

TL;DR

This work tackles the challenge of performing coordinated underwater tasks with high reliability in extreme sea conditions by coupling a USV-side path-planned localization scheme with RL-enabled multi-AUV coordination. It advances accurate AUV positioning through maximizing in a USBL-based framework and enriches multi-AUV decision-making with environment-aware RL that accounts for ocean currents and USV proximity via customized rewards. The approach is validated through extensive simulations of a multi-AUV data-collection task under both ideal and extreme sea states, demonstrating robust performance and improved positioning accuracy, plus a practical demonstration of scalable coordination. The authors also provide open-source simulation code to encourage further research in USV-AUV collaboration for underwater exploration.

Abstract

Autonomous underwater vehicles (AUVs) are valuable for ocean exploration due to their flexibility and ability to carry communication and detection units. Nevertheless, AUVs alone often face challenges in harsh and extreme sea conditions. This study introduces a unmanned surface vehicle (USV)-AUV collaboration framework, which includes high-precision multi-AUV positioning using USV path planning via Fisher information matrix optimization and reinforcement learning for multi-AUV cooperative tasks. Applied to a multi-AUV underwater data collection task scenario, extensive simulations validate the framework's feasibility and superior performance, highlighting exceptional coordination and robustness under extreme sea conditions. To accelerate relevant research in this field, we have made the simulation code (demo version) available as open-source.
Paper Structure (10 sections, 8 equations, 4 figures, 2 tables, 1 algorithm)

This paper contains 10 sections, 8 equations, 4 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Methodology
  3. USV Path Planning Based on Fisher Information Matrix Optimization
  4. Reinforcement Learning Enabled Multi-AUV Collaboration
  5. Simulation of Extreme Sea Conditions
  6. Experiments
  7. Task Description and Settings
  8. Experiment Results and Analysis
  9. Conclusion
  10. Acknoledgments

Figures (4)

  • Figure 1: Illustration of the AUV positioning via USV. The USV on the sea surface uses USBL to locate the underwater AUV.
  • Figure 2: The diagram of USV path planning. We can realize accurate positioning of AUVs via USV path planning through minimizing the determinant of the system’s FIM.
  • Figure 3: The curves of sum data rate, energy consumption and average reward per timestep, using DDPG and SAC for RL training in ideal and extreme sea conditions, respectively. (a) Sum data rate. (b) Energy consumption. (c) Average reward per timestep.
  • Figure 4: Trajectories of AUVs and USV, and positioning error of the AUV with USV fixed at (0,0) and (100, 100), and path planning using FIM optimization, respectively. (a) Trajectories of AUVs and USV. (b) Positioning error of the AUV.