Journal of Navigation and Port Research (한국항해항만학회지)

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

DOI QR Code

Research on the Development of Distance Metrics for the Clustering of Vessel Trajectories in Korean Coastal Waters

국내 연안 해역 선박 항적 군집화를 위한 항적 간 거리 척도 개발 연구

  • Received : 2023.11.30
  • Accepted : 2023.12.18
  • Published : 2023.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

This study developed a new distance metric for vessel trajectories, applicable to marine traffic control services in the Korean coastal waters. The proposed metric is designed through the weighted summation of the traditional Hausdorff distance, which measures the similarity between spatiotemporal data and incorporates the differences in the average Speed Over Ground (SOG) and the variance in Course Over Ground (COG) between two trajectories. To validate the effectiveness of this new metric, a comparative analysis was conducted using the actual Automatic Identification System (AIS) trajectory data, in conjunction with an agglomerative clustering algorithm. Data visualizations were used to confirm that the results of trajectory clustering, with the new metric, reflect geographical distances and the distribution of vessel behavioral characteristics more accurately, than conventional metrics such as the Hausdorff distance and Dynamic Time Warping distance. Quantitatively, based on the Davies-Bouldin index, the clustering results were found to be superior or comparable and demonstrated exceptional efficiency in computational distance calculation.

본 연구에서는 국내 연안 해역 환경에서의 해상교통관제 서비스에 기여할 수 있는 항적 간 거리 척도를 개발하였다. 새로운 항적간 거리 척도는 전통적으로 위치 시계열 간의 유사도를 측정하는 데 활용되는 하우스도르프 거리(hausdorff distance)와 두 항적 간의 대지속력(Speed Over Ground, SOG)의 평균 간의 차이, 그리고 대지침로(Course Over Ground)의 분산 간의 차이를 가중합하여 설계되었다. 새로운 척도의 유효성을 검증하기 위하여 실제 AIS 항적 데이터와 병합 군집화 알고리즘을 활용한 기존 항적 간 거리 척도와의 비교 분석이 수행되었으며, 새로운 거리 척도를 활용한 항적 군집화 결과가 하우스도르프 거리(hausdorff distance), 그리고 다이내믹 타임 워핑 거리(Dynamic Time Warping distance) 등 기존 척도에 비해 항적 간 지리적 거리나 대지속도 및 대지침로 등 선박 거동 특성의 분포를 비슷하거나 그 이상의 수준으로 정교하게 반영하고 있음을 데이터 시각화로써 확인하였다. 정량적으로는 Davies-Bouldin 지표를 기준으로, 군집화 결과가 더욱 우수하거나 약간 낮은 수준을 기록한 한편, 거리 계산 효율성에서는 특히 우수함을 실증하였다.

Keywords

Acknowledgement

본 논문은 해양수산부 재원으로 선박해양플랜트연구소의 주요사업인 "선박해양 디지털 전환 지원을 위한 디지털서비스 플랫폼 개발"에 의해 수행되었습니다(1525014879).

References

  1. Alt, H.(2009), "The computational geometry of comparing shapes", Efficient Algorithms: Essays Dedicated to Kurt Mehlhorn on the Occasion of His 60th Birthday, pp. 235-248.
  2. Buchin, M., Dodge, S. and Speckmann, B.(2012), "Context-aware similarity of trajectories", Proc. of the Seventh International Conference on Geographic Information Science, GIScience, pp. 43-56.
  3. Davies, D. L. and Bouldin, D. W.(1979), "A cluster separation measure", IEEE transactions on pattern analysis and machine intelligence, Vol. PAMI-1, No. 2, pp. 224-227. https://doi.org/10.1109/TPAMI.1979.4766909
  4. Dobrkovic, A., Iacob, M. E. and van Hillegersberg, J.(2018), "Maritime pattern extraction and route reconstruction from incomplete AIS data", International Journal of Data science and Analytics, Vol. 5, pp. 111-136. https://doi.org/10.1007/s41060-017-0092-8
  5. Fisher, D.(1996), "Iterative optimization and simplification of hierarchical clusterings", Journal of artificial intelligence research, Vol. 4, pp. 147-178. https://doi.org/10.1613/jair.276
  6. Keogh, E. and Ratanamahatana, C. A.(2005), "Exact indexing of dynamic time warping", Knowledge and information systems, Vol. 7, pp. 358-386. https://doi.org/10.1007/s10115-004-0154-9
  7. Kim, K. I., Jeong, J. S. and Park, G. K.(2014), "A Study on Near-miss Incidents from Maritime Traffic Flow by Clustering Vessel Positions", Journal of Korean Institute of Intelligent Systems, Vol. 24, No. 6, pp. 603-608. https://doi.org/10.5391/JKIIS.2014.24.6.603
  8. Lee, H. K., Chang, S. R., Jeong, G. N. and Park, Y. S.(2010), "A proposal on the marine traffic supporting system in VTS area", Journal of Navigation and Port Research, Vol. 34, No. 9, pp. 693-698. https://doi.org/10.5394/KINPR.2010.34.9.693
  9. Magdy, N., Sakr, M. A., Mostafa, T. and El-Bahnasy, K.(2015), "Review on trajectory similarity measures", Proc. of the Seventh International Conference on Intelligent Computing and Information Systems, ICICIS, pp. 613-619.
  10. Oh, J. Y., Kim, H. J. and Park, S. K.(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
  11. Pallotta, G., Vespe, M. and Bryan, K.(2013), "Vessel pattern knowledge discovery from AIS data: A framework for anomaly detection and route prediction", Entropy, Vol. 15, No. 6, pp. 2218-2245. https://doi.org/10.3390/e15062218
  12. Park, S. Y. and Park, Y. S.(2022), "A basic study on intersection congestion using simulation: focusing on the crossing vessels in Busan New Port caution area", Journal of Korean Society of Transportation, Vol. 40, No. 3, pp. 319-334. https://doi.org/10.7470/jkst.2022.40.3.319
  13. Rong, H., Teixeira, A. P. and Soares, C. G.(2020), "Data mining approach to shipping route characterization and anomaly detection based on AIS data", Ocean Engineering, Vol. 198, 106936.
  14. Zhao, L. and Shi, G.(2019), "A novel similarity measure for clustering vessel trajectories based on dynamic time warping", The Journal of Navigation, Vol. 72, No. 2, pp. 290-306. https://doi.org/10.1017/S0373463318000723
  15. Zhen, R., Jin, Y., Hu, Q., Shao, Z. and Nikitakos, N.(2017), "Maritime anomaly detection within coastal waters based on vessel trajectory clustering and Naive Bayes Classifier", The Journal of Navigation, Vol. 70, No. 3, pp. 648-670. https://doi.org/10.1017/S0373463316000850