A METHODOLOGY FOR CONSTRUCTING A PROBABILISTIC GRAPHICAL MODEL OF NAVIGATIONAL SAFETY ASSESSMENT

Abstract

The purpose of the article is to develop a methodology for constructing a probabilistic graphical model of navigational safety assessment when navigating ships in limited conditions both in marine areas and inland waterways. This goal is achieved by analysing the features of the mathematical apparatus of the Bayesian network and notation of the functional integrity scheme, determining the main events for building a graph of the Bayesian network for navigational safety assessment and the corresponding mathematical expressions. It is established that, despite significant efforts to improve organisational and technical measures to ensure navigation safety, the level of navigation accidents remains high. There is a contradiction between the growth of requirements for shipping safety and the actual situation. One of the main factors hindering the resolution of issues related to navigational safety is the lack of formal methods for describing the navigation process. This is due to the difficulties in modelling the behaviour of complex organisational and technical shipping systems in the probabilistic state space, as the impact on the shipping process is stochastic, and the limitations of theoretical understanding in this area. As a result, there are difficulties in creating models of navigation risks that can be used for a comprehensive assessment of navigation safety within complex organisational and technical shipping systems. The solution to these problems would allow the use of formal methods for managing navigation risks, assessing the significance of system components and their impact on the overall risk. The most significant result is a methodology for constructing a probabilistic graphical model of navigational safety assessment when navigating ships in restricted conditions both in offshore areas and inland waterways using the mathematical apparatus of Bayesian network theory and functional integrity scheme notation. The research has shown that it is possible to use the main features of Bayesian networks and the logical-probabilistic method for the mathematical description of organisational and technical ship management systems. The space of probabilistic events and Kolmogorov's axiomatics together with the rules of Boolean logic algebra can be used to describe complex organisational and technical systems of ship management. This approach makes it possible to consider the peculiarities and integrity of the «ship-shipman-environment» system when assessing navigation risks.