Sučasnì Informacìjnì Tehnologìï u Sferì Bezpeki ta Oboroni (Aug 2018)
METHOD OF CLASTERIZATION OF VESSELS ROUTES IN THE SYSTEM OFCOVERING THE SITUATION AT SEA
Abstract
The article solves the problem of classification of data on the routes of the vessels in the system of covering the situation at sea. For the purpose of this classification, the use of clusterization, the method of conducting of which was developed within the framework of this study, is proposed. The peculiarity of the method is to take into account the specificity of presentation the data on the routes of the vessels in the information and telecommunication system of marine guard "Gart-12". This specificity is taken into account in the metric and features of formation of route clusters and definition of their centers. Such an analysis of the data in the system of illumination of the surface environment will solve the problem of risk assessment by identifying spatial anomalies in the routes of the vessels. Described in the work method of clustering routes specified by discrete reports in the system of illumination of the above-ground situation, received algorithmic and software implementation using a programming system Visual Studio Express Edition and a programming language C#. Using the developed program, a classification of the route data was carried out. As a result of analysis of the available data in the information system of the «Hart-12» using the clustering method described in the work, typical routes of movement in the (center of the corresponding cluster are determined). With the aim of classifying routes, a clustering method has been having determined that allows the identified data belonging to individual routes to the cluster to be determined by their centers as routes the total distance from which to the other routes in the cluster is minimal. With the use of a certain metric for new data that comes to the system of lighting the surface situation it is possible to determine their belonging to one of the standard routes and having conducted the assessment of their similarity to carry of their out a risk assessment. Directions of further research is the formation of a methodology for risk assessment based on the analysis of spatial anomalies of vessel traffic.