Risk Assessment of Ship Navigation Collision in Inland Waterway Transportation System Based on Bayesian Method

Abstract

Due to the high risk and uncertainty of ship navigation, the identification, early warning and control of ship navigation risks have always been studied by scholars. In recent years, water traffic accidents have occurred frequently, resulting in an incalculable number of casualties and property losses, which makes ship navigation risk management more and more a research hotspot. Ship navigation risk is a complex system consisting of a number of risk factors. The assessment of it is a systematic project aimed at clarifying the risk factors that need to be controlled. The purpose of this paper is to use Bayesian methods to assess the risk of navigational collisions in inland waterway transport systems. In this paper, the Bayesian network and the inland waterway transportation system ship navigation collision risk simulation are combined, and the concept of the ship navigation collision risk analysis model of the Bayesian network inland waterway transportation system is proposed. This method essentially utilizes three important characteristics of Bayesian network itself—complex relation representation ability, probability uncertainty representation ability and causal reasoning ability. Through the expert experience and ship navigation collision risk simulation of inland waterway transportation system the data is learned, the knowledge contained therein is mined, and the ship navigation collision risk analysis model of the inland waterway transportation system is established, and the risk assessment analysis is carried out based on the model. Through a large number of experiments, the proposed method can better model the navigation state of the ship, quantify the navigation risk of the ship, and identify the navigation anomaly data.