Moving Objects Tracking in Distributed Maritime Observation Systems

摘要

This paper considers the processes of target tracking complicated by big data gaps in complex media such as Distributed Maritime Observation System (DMOS). The main purpose of DMOS is to support the favorable navigation conditions, monitoring, life-saving on the sea for different ships in harbors, maritime roads and open sea. DMOS can be considered as a heterogeneous distributed computer system, it includes different layers of services at different levels of abstraction: ship, harbor; regional and global levels. Such a framework is based on several satellite and maritime information systems that nowadays favor the integration of maritime data (e.g., AIS, ECDIS, OPTIMARE, GMDSS). Despite of the input data big volume the situations exist when gaps (e.g., time delays from minutes to hours) between target observations’ points are different. Well known algorithms of target tracking do not work properly in such situations. The proposed approach describes synthesis of analytical and simulation methods at tactical hypothesis development for the cases when suitable direct analytics is not applicable. Also, a joined artificial techniques’ scenario approach to tactical situation hypothesis development is proposed.