Research On Dangerous Encounter Hotspots Mining Based On AIS Big Data

With the rapid development of Chinese economy and society,ship transportation is playing a more and more important role in the integrated transportation system because of its low cost and large volume.In recent years,the number of domestic ship in ports and channel has continued to increase and the type of ship has showed the trend of large scale,specialization and high speed,but it also has produced a series of problems water such as rising traffic flow density,overall navigation environment deterioration,frequent water traffic accidents and serious water pollution problems.The water traffic safety problem brought great threat to the life and property safety of the sailors,which become the focus of shipping management department and shipping enterprises.Ship navigation risk assessment is an important part of water traffic safety research.It is an important means to reduce the navigation risks and ensure the safe navigation of ships.In the aspect of ship navigation risk,ship encounter is closely related to ship navigation risk,especially collision risk.To study on ship encounter risk,it is best able to grasp the ship encounter relatively dynamic ship,but most ships encounter research is carried out through statistical analysis and observation monitoring and failed to clearly understand the real-time dynamic ship encounter due to the lack of a large number of real-time data available.The application and popularization of Automatic Identification System grasps real-time dynamic information of ship.AIS data of ship navigation trajectory provides the conditions for the further understanding of ship navigation risk.The rapid development of cloud computing provides an important technical basis for further AIS data mining.This paper proposes the concept of ships dangerous encounter,builds dangerous encounter ship domain model,and proposes dangerous ship encounter point recognition algorithm to recognize from dangerous ship encounter points from the massive AIS data provided by TY Company.This paper designs and implements parallel DBSCAN algorithm base on MapReudce,which is used to location clustering of dangerous ship encounter points and find out dangerous ship encounter hotspots.in mining.Finally,setting the Pearl River Estuary as the research object,this paper analyze type of ship,ship size,time distribution and encounter situation of dangerous encounter points in Pearl River Estuary,use the parallel DBSCAN algorithm to mine danger point ship hotspots in the Pearl River Estuary,use thermodynamic diagram to analyze of dangerous water region of different thermal situation and determine the dangerous waters region of the Pearl River Estuary.