Spatial Markov Chain simulation model of accident risk for marine traffic

In order to control or reduce the risk of water transportation, it is necessary to study the evolution trend of marine traffic risk. A new risk simulation model based on Markov Chain-Monte Carlo algorithm (MCMC) is proposed for marine traffic risk analysis near land. First, the spatial state transfer model of ship traffic risk in the waters adjacent to land was established. Second, the spatial transfer hypothesis matrix was conducted for ships operating in offshore, coastal and port waters; third, the risk value of traffic waters under stochastic process was obtained by using MCMC method to carry out random sampling. Finally, combined with typical coastal marine traffic data, the risk occurrence law and its changing trend were analyzed in different traffic waters through simulation. Empirical data shows that ship traffic risks are 0.193%, 0.563%, and 0.560% respectively in offshore, coastal and port waters, and the risk is high and unstable in restricted waters. MCMC simulation method can be applied to study the characteristics of traffic risk in waters adjacent to land, which provides a theoretical reference for improving safety and accident prevention in certain marine traffic waters.