Journal of Navigation and Port Research (한국항해항만학회지)

Korean Institute of Navigation and Port Research (한국항해항만학회)
권호 표지
DOI QR코드

DOI QR Code

Comparison and Analysis of Information Exchange Distributed Algorithm Performance Based on a Circular-Based Ship Collision Avoidance Model

원형 기반 선박 충돌 피항 모델에 기반한 정보 교환 분산알고리즘 성능 비교 분석

  • Donggyun Kim (Division of Navigation Science, Mokpo National Maritime University)
  • 김동균 (목포해양대학교 항해학부)
  • Received : 2023.08.16
  • Accepted : 2023.09.14
  • Published : 2023.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

This study compared and analyzed the performance of a distributed area search algorithm and a distributed probability search algorithm based on information exchange between ships. The distributed algorithm is a method that can search for an optimal avoidance route based on information exchange between ships. In the distributed area search algorithm, only a ship with the maximum cost reduction among neighboring ships has priority, so the next expected location can be changed. The distributed stochastic search algorithm allows a non-optimal value to be searched with a certain probability so that a new value can be searched. A circular-based ship collision avoidance model was used for the ship-to-ship collision avoidance experiment. The experimental method simulated the distributed area search algorithm and the distributed stochastic search algorithm while increasing the number of ships from 2 to 50 that were the same distance from the center of the circle. The calculation time required for each algorithm, sailing distance, and number of message exchanges were compared and analyzed. As a result of the experiment, the DSSA(Distributed Stochastic Search Algorithm) recorded a 25%calculation time, 88% navigation distance, and 84% of number of message exchange rate compared to DLSA.

본 연구에서는 선박 간 정보 교환에 기반한 분산지역탐색 알고리즘과 분산확률탐색 알고리즘의 성능을 비교, 분석하고자 한다. 분산알고리즘은 선박 간 정보 교환을 기반으로 하여 최적의 피항 경로를 탐색할 수 있는 방법이다. 분산지역탐색알고리즘은 이웃 선박 중 비용 감소가 최대가 되는 선박만이 다음 예상 위치를 바꿀 수 있도록 해당 선박이 우선권을 가진다. 분산확률탐색알고리즘은 일정 확률로 최적이 아닌 값을 탐색할 수 있도록 하여 새로운 값을 탐색할 수 있도록 한다. 선박 간 충돌 피항 실험은 원형 기반 선박 충돌 피항 모델을 활용하였다. 실험 방법은 원형에 기반하여 원의 중심에서 같은 거리에 떨어진 선박을 2척부터 50척까지 증가시키면서 분산 지역 탐색알고리즘과 분산확률탐색알고리즘을 시뮬레이션 하였다. 실험 평가 방법은 각 알고리즘의 계산 소요 시간, 항행 거리, 메시지 교환 횟수를 비교 분석하였다. 실험 결과 DSSA는 DLSA에 비해 계산시간은 25%, 항행 거리는 88%, 메시지 교환 횟수는 84%를 기록하였다.

Keywords

References

  1. Chun, D., Roh, M., Lee, H. Ha, J. and Yu, D.(2021), "Deep reinforcement learning-based collision avoidance for an autonomous ship", Ocean Engineering, Vol. 234, p. 109216.
  2. COLREG(1972), Convention on the International Regulations for Preventing Collisions at Sea, 1972 (COLREGs).
  3. Ferranti, L., Negenborn, R. R., Keviczky, T. and Alonso-Mora, J.(2018), "Coordination of Multiple Vessels Via Distributed Nonlinear Model Predictive Control", 2018 European Control Conference (ECC), Limassol, Cyprus, pp. 2523-2528.
  4. Hornauer, S. and Hahn, A.(2013), "Towards Marine Collision Avoidance Based on Automatic Route Exchange", IFAC Proceedings Volumes, Vol. 46, No. 33, pp. 103-107. https://doi.org/10.3182/20130918-4-JP-3022.00049
  5. Korea Autonomous Surface Ship Project(2023), "Intelligent Navigation System", https://kassproject.org/en/main.php
  6. Korea Maritime Safety Tribunal(2023), "Marine accidents by accident type", https://www.kmst.go.kr/eng/page.do?menuIdx=227.
  7. Kim, D., Hirayama, K. and Okimoto, T.(2015), "Ship Collision Avoidance by Distributed Tabu Search". TransNav, the International Journal on Marine Navigation and Safety of Sea Transportation, Vol. 9, No. 1, pp. 23-29. https://doi.org/10.12716/1001.09.01.03
  8. Kim, D.(2017), "A Study on Factors that Trigger Human Errors Related to Causes of Ship Collisions", Journal of the Korean Society of Marine Environment and Safety Vol. 23 No. 7 pp. 801-809. https://doi.org/10.7837/kosomes.2017.23.7.801
  9. Kim D., Hirayama, K. and Okimoto, T.(2017), "Distributed Stochastic Search Algorithm for Multi-ship Encounter Situations", The Journal of Navigation, Vol. 70, No. 4, pp. 699-718. https://doi.org/10.1017/S037346331700008X
  10. Kim, D.(2019), "Analysis of a Distributed Stochastic Search Algorithm for Ship Collision Avoidance", Journal of the Korean Society of Marine Environment and Safety Vol. 25, No. 2, pp.169-177. https://doi.org/10.7837/kosomes.2019.25.2.169
  11. Kim, J. and Jang H. Sup.(2019), Autonomous ship technology trends and preparations, BULLETIN OF THE SOCIETY OF NAVAL ARCHITECTS OF KOREA, Vol. 56, No. 4, pp. 4-7.
  12. Lee, J. and Song, C(2017), "A Study on the Degree of Collision Risk through Analysing the Risk Attitude of Vessel Traffic Service Operators", Journal of Navigation and Port Research, Vol. 41, No. 3, pp. 93-102. https://doi.org/10.5394/KINPR.2017.41.3.93
  13. Liu, Y., Yang, C. and Du, X.(2007), "A multiagent-based simulation system for ship collision avoidance", In: International Conference on Intelligent Computing. Springer, pp. 316-326.
  14. Lee, J. and Song, C.(2018a), "The Situation Awareness Analysis of VTSOs in the Close Quarters Situation", Journal of Navigation and Port Research, Vol. 42, No. 1, pp. 25-30. https://doi.org/10.5394/KINPR.2018.42.1.25
  15. Lee, J. and Song, C.(2018b), "The Efficient Anchorage Management of VTS through Analysis of Domain Watch", Journal of Navigation and Port Research, Vol. 42, No. 3, pp. 201-206. https://doi.org/10.5394/KINPR.2018.42.3.201
  16. Oh, J., Kim, H. and Park, S.(2018), "Detection of Ship Movement Anomaly using AIS Data: A Study", Journal of Navigation and Port Research, Vol. 42, No. 4, pp. 277-282. https://doi.org/10.5394/KINPR.2018.42.4.277
  17. Park, Y., Jeong, J. Y. and Kim, J. S.(2010), "A Study on the Minimum Safety Distance between Navigation Vessels based on Vessel Operator's Safety Consciousness", Journal of Korean Society of Marine Environment & Safety, Vol. 16, No. 4, pp.401-406.
  18. Pak, C., Jung, C. and Park, S.(2014), "A Study on the Introduction of Effective Route Exchange System in the VTS Area", Journal of Navigation and Port Research, Vol. 38, No. 3, pp. 217-225.
  19. Sun, Z., Fan, Y. and Wang, G.(2023), "An Intelligent Algorithm for USVs Collision Avoidance Based on Deep Reinforcement Learning Approach with Navigation Characteristics", Journal of Marine Science and Engineering, Vol. 11, No. 4.
  20. Szlapczynski, R. and Szlapczynska, J.(2017), "Review of ship safety domains: Models and applications", Vol. 145, pp. 277-289.
  21. Wang, Z. and Zhang, Y.(2022), "Estimation of ship berthing parameters based on Multi-LiDAR and MMW radar data fusion", Ocean Engineering, Vol. 266, No. 5, p. 113155.
  22. Yang, T., C. Han, M. Qin, and Huang, C.(2020), "Learning-Aided Intelligent Cooperative Collision Avoidance Mechanism in Dynamic Vessel Networks", IEEE Transactions on Cognitive Communications and Networking, Vol. 6, No. 1, pp. 74-82. https://doi.org/10.1109/TCCN.2019.2945790
  23. Zhai, P., Zhang, Y. and Shaobo, W.(2022), "Intelligent Ship Collision Avoidance Algorithm Based on DDQN with Prioritized Experience Replay under COLREGs", Journal of Marine Science and Engineering, Vol. 10, No. 5.