Identification of accident black spots in inland rivers based on DBSCAN with adaptive parameters determination method

Water transport accidents occur frequently in inland rivers, causing negative impact on the safety of waterway transportation. In this study, the statistics of water transport accidents of bulk carriers in the Yangtze River from 2010 to 2019 were collected and analyzed. Based on that, a novel method is proposed to identify the accident black spots of water transport accidents in inland rivers based on DBSCAN with adaptive parameters determination method. In the proposed method, it is used to extract the black point segments of water transport accidents in the lower reaches of the Yangtze River, which can realize the full automation of the clustering process and support the selection of reasonable Eps and MinPts parameters, and obtain high-accuracy clustering results. Then, combining the results with the actual navigation environment conditions, the accident characteristics and accident causes of each typical accident black spot segment are analyzed, and the analysis results are consistent with the actual situation. The research results indicate the feasibility of the proposed method on analyzing the distribution characteristic of water transport accidents in inland rivers. This research can provide a theoretical basis for developing safety management countermeasures of water transportation, and a reference for decision-making of ship navigation in water areas with high risks.