경로계획과 PPO 기반 심도제어를 통합한 수중자율운동체의 해저지형 회피 알고리즘 개발

김현승¹^{, *} ; 현철² ; 이성균² ; 고진용² ; 김창환²

1LIG넥스원 해양연구소 선임연구원
2LIG넥스원 해양연구소 수석연구원

Development of Integrated Path Planning and PPO-based Depth Control for Seafloor Avoidance of Underwater Autonomous Vehicle

Hyunseung Kim¹^{, *} ; Chul Hyun² ; Sungkyun Lee² ; Jinyong Go² ; Changhwan Kim²

1Research engineer, Marine R&D Center, LIG Nex1
2Chief research engineer, Marine R&D Center, LIG Nex1

Correspondence to: ^*Hyunseung Kim Marine R&D Center, LIG Nex1 333 Pangyo-ro, Bundang-gu, Seongnam-si, Gyeonggi-do, 13488, Republic of Korea Tel: +82-31-5179-7272 Fax: +82-31-5179-7086 E-mail: hyunseung.kim2@lignex1.com

초록

본 논문에서는 수중 자율 운동체의 안전한 주행을 위해 경로계획과 심도제어를 통합한 해저지형 회피 알고리즘을 개발하였다. 성능 검증을 위해 1차원의 수심 지형 환경에서 기존 A* 알고리즘 또는 PPO 알고리즘 단독 방식과의 비교를 수행하였고, 충돌률과 추종 오차 측면에서의 우수함을 확인하였다. 본 논문은 복잡한 해저 환경에서도 심도 안정성과 에너지 효율을 극대화 하기 위한 경로 수립에 활용할 수 있을 것으로 기대된다.

Abstract

This paper proposed a hybrid algorithm that integrates A* path planning and PPO(proximal Policy optimization)-based depth control for seafloor avoidance of underwater autonomous vehicles. To evaluate a performance in 1D depth scenario, the proposed hybrid algorithm outperformed standalone A* algorithm and PPO approaches in terms of collision rate and tracking accuracy. This study is expected to be used to establish a route for stable and energy-efficient navigation in complex underwater environments.

Keywords:

Underwater Autonomous Vehicle, Path Planning, Depth Control, Proximal Policy Optimization Algorithm

키워드:

수중자율운동체, 경로계획, 심도제어, PPO 알고리즘

References

H. S. Kim, C. Hyun, S. K. Lee, J. Y. Go and C. H. Kim, “Research on Terrain Tracking Algorithm for Seafloor Topography Avoidance of Underwater Autonomous Vehicle,” Journal of the Korea Society for Naval Science and Technology, Vol. 7, No. 3, pp. 304-308, Sep. 2024. [https://doi.org/10.31818/JKNST.2024.9.7.3.304]
J. Du, D. Zhou and W. Wang, “Reference Model-based Deterministic Policy for Pitch and Depth Control of Autonomous Underwater Vehicle,” Journal of Marine Science and Engineering, Vol. 11, No. 3, pp. 1-23, Mar. 2023. [https://doi.org/10.3390/jmse11030588]
A. Zhang, W. Wang, W. Bi and Z. Huang, “A Path Planning Method based on Deep Reinforcement Learning for AUV in Complex Marine Environment,” Ocean Engineering, Vol. 313, No. 1, pp. 1-9, Dec. 2024. [https://doi.org/10.1016/j.oceaneng.2024.119354]
H. Wu, S. Song, K. You and C. Wu, “Depth Control of Model-free AUV via Reinforcement Learning,” IEEE Transactions on Systems, Man, and Cybernetics, Vol. 49, No. 12, pp. 2499-2510, Dec. 2019. [https://doi.org/10.1109/TSMC.2017.2785794]
H. S. Kim, C. Hyun, S. K. Lee and J. Y. Go, “A Study on Underwater Autonomous Vehicle Terrain-following Depth Control Algorithm Using Reinforcement Learning Method,” Journal of the Korean Institute of Defense Technology, Vol. 7, No. 1, pp. 7-12, Jul. 2025. [https://doi.org/10.52682/jkidt.2025.7.1.7]
Li. Y, He. X, Lu. Z, Jing. P and Su. Y, “Comprehensive Ocean Information Enabled AUV Motion Planning based on Reinforcement Learning,” Remote Sensing, Vol. 15, No. 12, pp. 1-22, Dec. 2023. [https://doi.org/10.3390/rs15123077]
B. Hadi, A. Khosravi and P. Sarhadi, “Deep Reinforcement Learning for Adaptive Path Planning and Control of an Autonomous Underwater Vehicle,” Applied Ocean Research, Vol. 129, No. 9, pp. 1-8, Dec. 2022. [https://doi.org/10.1016/j.apor.2022.103326]