Pukas A. Methods and tools for building the mathematical models of complex objects characteristics in conditions of interval uncertainty

In the thesis, the scientific and applied problem of reduction the computational complexity of processes for construction the mathematical models of complex objects characteristics in the conditions of interval uncertainty with simultaneous maintenance of guaranteed accuracy of these models within the limits necessary for decision making is solved. A method of parametric identification of interval models of characteristics of static and dynamic objects based on the analysis of interval data based on procedures of selforganization and self-adaptation of computational procedures by analogy with behavioral models of bee colony is created. The method of structural identification of interval models of characteristics of static and dynamic objects on the basis of the analysis of interval data with procedures of selforganization and self-adaptation of structures of models is developed. The method of ellipsoid estimation of the set of values of parameters of interval models of characteristics of static objects on the basis of the iterative computational scheme of optimum saturated planning of experiment based on parallelization of computational procedures is improved. A software system for constructing interval models of static and dynamic object characteristics has been created, which combines structural and parametric identification methods based on behavioral models of the bee colony, which provided a holistic approach to building models with guaranteed accuracy in interval uncertainty and greatly simplified the use of modeling tools. New and improved methods and the developed software environment for solving the applied problems of construction the models of characteristics of static and dynamic objects in the interval uncertainty conditions, in particular for monitoring the pollution of the atmosphere by motor transport on an example of Ternopil; identification of the reverse laryngeal nerve during thyroid surgery; modeling and forecasting the capacity of the small hydroelectric power station "Topolki"; modeling the visiting web services of providing administrative services, were approved.