Research map / Multi-agent systems and games / Human-AI and goal-conditioned
UACOF: A USV-AUV Collaboration Framework for Underwater Tasks Under Extreme Sea Conditions (Student Abstract)
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes UACOF, a USV-AUV collaboration framework that combines FIM-determinant-based USV path planning for accurate AUV localization with RL (DDPG/SAC) training of multiple AUVs under simulated extreme sea conditions.
Problem
Accurate positioning and intelligent collaborative control of multiple AUVs are hard under extreme sea conditions, where waves and turbulence degrade performance.
Approach
Use USV path planning that maximizes the determinant of the Fisher information matrix for USBL-based AUV localization, and add environment awareness (ocean current velocity) to the RL state and USV-AUV distance terms to the reward so multi-AUV policies adapt to extreme seas.
Key results
- Under both ideal and extreme sea conditions, DDPG and SAC training curves converge to expert-level performance with comparable SDR, EC and ARPT.
- FIM-optimized USV path planning yields the lowest AUV positioning error among the compared scenarios (path planning vs. USV fixed at (0,0) or (100,100)), reported only qualitatively via a figure.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use FIM-optimized USV path planning instead of keeping the USV fixed | UACOF USV-AUV localization · Underwater data collection under extreme sea conditions | positioning error | Helped |
| Train the multi-AUV system with RL under extreme sea conditions | UACOF · Underwater data collection | performance metrics | No clear effect |
Limitations
- Evaluation is limited to a single simulated data-collection task with two AUVs.
- Results are presented mainly as training curves and error plots without quantitative tables or statistical tests.
Experimental setup
Datasets: simulated underwater data collection task (from Zhang et al. 2024)
Benchmarks: simulated underwater data collection task
Metrics: sum data rate (SDR) energy consumption (EC) average reward per timestep (ARPT) AUV positioning error
Baselines: USV fixed at (0,0) USV fixed at (100,100) DDPG (ideal sea conditions) SAC (ideal sea conditions)
Builds on: DDPG SAC
Closest papers by meaning
- Reducing AUV Energy Consumption Through Dynamic Sensor Directions Switching via Deep Reinforcement Learning (AAAI 2025 · similarity 0.85)
- AoI-MDP: An AoI Optimized Markov Decision Process Dedicated in the Underwater Task (Student Abstract) (AAAI 2025 · similarity 0.83)
- U2UData+: A Scalable Swarm UAVs Autonomous Flight Dataset for Embodied Long-horizon Tasks (AAAI 2026 · similarity 0.81)
- Power-Aware Inverse-Search Machine Learning for Low Resource Multi-Objective Unmanned Underwater Vehicle Control (Student Abstract) (AAAI 2024 · similarity 0.80)
- WebUOT-1M: Advancing Deep Underwater Object Tracking with A Million-Scale Benchmark (NeurIPS 2024 · similarity 0.80)
- Towards Realistic UAV Vision-Language Navigation: Platform, Benchmark, and Methodology (ICLR 2025 · similarity 0.79)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.