DEVELOPMENT OF A KNOWLEDGE MODEL FOR AN INFORMATION DECISION SUPPORT SYSTEM BY AN AVIATION OPERATOR IN CASE OF SPECIAL CASES IN FLIGHT

Abstract

The subject matter of the article is methods that allow solving the problem of uncertainty in the process of constructing decision support systems for air traffic controllers for special cases in flight. The goal is the analysis and justification of the choice of mathematical tool for constructing a DSS model of an air traffic controller. The tasks are: analysis of a number of well-known methods of data mining, namely: evolutionary algorithms, neural networks, fuzzy logic and Bayesian networks from the point of view of the appropriateness of their use in constructing mathematical models of decision support systems for air traffic controllers in case of special cases in flight. The methods used are: methods of analysis and synthesis of complex information systems, methods of simulation and statistical modeling. The following results were obtained. It has been established that the most effective way to build a DSS model of an air traffic controller is to use a Bayesian network tool. It is a promising probabilistic tool, that allows you to simulate complex hierarchical static and dynamic systems. This is due to the fact that, in contrast to the currently popular “black box” models, the Bayesian network provides an understandable explanation of the findings, has a logical interpretation, makes it possible to take into account the uncertainties of a parametric, static and structural nature and, which is especially important, is based on fundamental provisions of the theory of probability, which has been developed for more than one century. Conclusions. The direction of further research is the construction of a DSS model of an air traffic controller using Bayesian networks and probabilistic programming technology