Vessel traffic simulation plays a very important role in enacting laws and regulations pertaining to maritime traffic regulation, and route planning and design. Vessel traffic flow generator is mainly produces all the ship’s navigation behavior (i.e., vessel traffic) in a certain research area. By producing vessel traffic flow to identify traffic problems existing in the sea area, so as to explore the reasonable solution to solve the problem of maritime traffic.In order to overcome the disadvantages of traditional traffic observation method to get the vessel traffic flow characteristics, this paper puts forward the ship traffic flow generator method, the parameter are extract from real AIS data, thus "reproduction" real traffic more accurate. In this paper, the specific contents and research results summarized as follows:First, the research status of ship traffic flow generator and the application overview of Monte Carlo are specified.Second, ship traffic flow composition, characteristics, the characteristics of traffic flow and Monte Carlo in the application of the vessel traffic flow generated is specified.Third, the core part of the vessel traffic flow generator:build the model of vessel traffic network. Using the AIS data and the method of DBSCAN on vessel traffic network (pattern) for automatic identification; Through application of AIS data and Monte Carlo sampling, obtained accurate ship generation time, the information such as type, size and speed of the ship.Fourth, ship traffic flow produced by traffic generator is compared with the real traffic; verify the validity of the model. The results show that the model of vessel traffic flow production is comfort to the actual traffic. |