Spatial autocorrelation analysis method of ship conflict grounded in Moran's I and Getis-Ord Gi* indices

To evaluate the spatial autocorrelation of ship conflicts and identify the clustering patterns of ship conflicts in dense traffic waters, this paper effectively combines the Moran’s I index and the Getis-Ord Gi* index. Firstly, the data mining is based on the Automatic Identification System (AIS), and the collision radius of the ship is set by the ship domain model, to extract and identify the collision data of the ship. Secondly, the spatial autocorrelation indices, Moran’s I and Getis-Ord Gi*, are used to measure the spatial autocorrelation of ship conflicts in the port area in Tianjin. Finally, the conflict clustering pattern is identified in light of the conflict clustering advantages of these two indices. Results show that the fusion of the two indices can effectively improve the accuracy of determining the number, degree and type of ship conflicts in the route. By analyzing the ship conflicts relying on the hybrid index, the correlation between the conflict clusters can be effectively recognized, as well as the hot spots of ship conflicts and the traffic characteristics in the waters. It provides support for improving maritime safety and reducing the occurrence of ship collisions.