Research On Data Mining And Visualization Of Ship Trajectory

The South China Sea(SCS)is a junction of marine transportation connecting the Indian Ocean and the Pacific Ocean.The channel of SCS has strategic significance of China’s energy security.Moreover,the ship lines of SCS are the most important channel connecting the economic system of Asian-Pacific region and world.With the rapid development of the technology of satellite positioning and mobile communication,trajectory data is increasing day by day and its analysis is applied in many fields.Volunteer Observe Ships(VOS)data is the maritime navigation trajectory data of volunteer observation ships from 2001 to today,which contains a large amount of information about human maritime traffic activities and others affect information.Exploring the information behind the data based on data mining technology is of great significance to sea-route safety and marine ecological environment protection.Therefore,based on the VOS data,using the idea of spatial data mining,this thesis used DBSCAN method and GIS analysis to mine the information as follows: the distribution of shipping lanes,spatial information of port city and the impacts of ecological environment in SCS.Meanwhile,combining the methods of visualization to visualize the excavated information.The main contents and main conclusions in this study were summarized as follows:(1)Shipping lanes information mining and visualization of ship trajectory data.A progressive clustering method based on Minimum Bounding Rectangle(MBR)and buffer similarity was proposed for clustering and then calculated the centerlines to obtain the shipping lanes in SCS by the method of triangulation network.The results showed that there were 16 shipping lanes extracted in the SCS,with 3 first-class shipping lanes,3 second-class shipping lanes and 10 third-class shipping lanes.The distribution of the shipping lanes in the SCS shows some typical local characteristics.Singapore,Guangzhou/Hong Kong,the western intersection of Spratly Islands,Bashi Straits,and the Mindoro Strait are important intersections of the shipping lanes.(2)Port spatial information mining and visualization of ship trajectory data.Port networks based on traffic flow have the characteristics of “small world” network.Singapore,the Pearl River Delta and Kaohsiung have the most connections with other ports,followed by Da Nang,Bangkok and Ho Chi Minh,and Manila,Zamboanga and Belait have fewer connections with other ports.It showed that there was some imbalance in the development of various port cities in the SCS and huge economic difference,which also reflects the differences of economics among different countries to a certain extent.(3)Estimation and visualization of ship’s impact on ecological impact.Based on the VOS trajectory data,kernel density analysis,neighborhood analysis and Inverse Distance Weight(IDW)were used to estimate the pollution of ship emission,noise and invasion of alien in the SCS.Then the weighted superposition method was used to calculate the ecological impact of the ship on the SCS.It was showed that the Malay Peninsula and the Strait of Malacca are the areas with the most serious ecological and environmental impacts.The ecological environment impacts around the Pearl River Estuary and Taiwan Strait are second.The ecological impacts of other areas are basically distributed along the waterway.In the sea area around the waterway,the ecological environment is affected to some extent.It is recommended that relevant departments should strengthen environmental monitoring in the areas around Malay Peninsula,Malacca Strait,Pearl River Estuary and Taiwan Strait,so that when the pollution were detected it can be effectively treatment.This will optimize the marine environment and promote the sustainable development of human beings.