ABSTRACT
With the rapid increase in ship traffic volume, the traffic situation in some busy waters becomes more complex, bringing great challenges to the development of shipping industry. To fully understand the ship traffic situations and provide decision-making basis for maritime safety management, we propose a ship traffic risk complexity model based on complex network theory. First, a real-time ship-ship collision risk model that combines a dynamic ship domain model is developed to describe the between-ship proximity degree spatially and temporally. Then four complex indicators is selected and the Projection Pursuit Evaluation model is adopted to quantitatively evaluate the real-time traffic situations by synthesizing the multi-dimension problems to the projection value with one dimension. A case study is finally conducted using the AIS data in the outside waters of Ningbo-Zhoushan Port. Results show that the proposed model can help to identify the structural characteristics of ship traffic situations as well as detect the evolutionary characteristics of traffic situations.
