The concept of building a multi-agent distributed training and training system

Existing training and simulation systems use mathematical and scenario models of the behavior of simulated objects. At the moment, training facilities are being used, which can be implemented both as computer training complexes and operational simulators, using either real elements or their imitation. The novelty of the proposed concept of building a multi-agent distributed training and training system is the decomposition of the model of the modeling object in order to ensure operational changes in both the composition of elements in the modeling environment and the ways it affects the surrounding objects. The implementation of the proposed concept will solve a number of problems related to the construction and operation of training and training systems. The use of intelligent objects will allow the formation of knowledge bases that provide non-standard behavior of objects characteristic of subjective elements of the system.