Ship Automatic Identification System (AIS) data contains rich marine traffic characteristic information. In order to mine the effective and potential information about ship motion law in AIS data and output classical trajectories, an improved spectral clustering algorithm for adaptive mining of ship typical trajectories is proposed. The algorithm takes the trajectory points as the clustering object, and uses Douglas-Peucker (DP) algorithm to compress the ship AIS data on the premise of retaining the characteristic trajectory points; the Calinski-Harabasz index is used to evaluate the spectral clustering algorithm to realize the adaptive selection of clustering parameters; the clustering center point is extracted as the feature point of the trajectory, and the classic trajectory of the ship is output. The actual AIS data of Weihai Port in Shandong Province, China was used as experimental samples to verify the algorithm. The results show that using the algorithm obtains typical trajectory of the ship is consistent with the facts, and the best parameters can be adaptively and accurately selected through the Calinski-Harabasz index. The research results are of great significance for assisting ship trajectory anomaly detection and mining maritime traffic characteristics.
Adaptive mining algorithm of ship typical trajectory based on improved spectral clustering
Seventh International Conference on Electromechanical Control Technology and Transportation (ICECTT 2022) ; 2022 ; Guangzhou,China
Proc. SPIE ; 12302
2022-11-23
Conference paper
Electronic Resource
English
Adaptive mining algorithm of ship typical trajectory based on improved spectral clustering
British Library Conference Proceedings | 2022
|Ship route prediction method based on ship trajectory clustering
European Patent Office | 2021
|