Analysis of approaches to problem structure identification in hybrid intelligent multiagent systems

The complexity of automatic solution of practical problems is largely due to their weak formalization and sometimes, the absence of a clear statement at the time of their occurrence. As a result, significant time and labor costs are required for formulation, identification of the composition and structure of the problem, selection or development of methods relevant to its parts, as well as their synthesis. The paper considers the concept of the problem and develops its macro- and microlevel models, an analysis of the methods of automatic and automated identification of its structure is performed for their possible use in hybrid and synergetic artificial intelligence systems, in particular, hybrid intelligent multiagent systems for solving weakly formalized problems.