Description: The article analyzes the models and methods of construction of multiagent adaptive training systems for adaptive training systems. Based on the analysis of the literature, it has been found that the construction of adaptive training systems for air traffic controllers is one of the most important areas of the learning process automation. Nowadays new principles of building adaptive training systems are being formed, which take into account the individual approach in training and implement the so-called indi-vidual training strategy. However, actual scientific tasks are to develop an intelligent system to support the learning process (formation and correction of tasks, help libraries, knowledge bases of results, expertise and evaluation of students' actions). There are currently no universal methods of exercise correction and process verification that are implemented in adaptive train-ing systems. There are not also common approaches to the structure of intelligen learning support systems. Possible direction of solving these scientific problems is the application of models and methods of multiagent systems. According to the results of the research, it is determined that the empirical nature of studies in the field of improving the training of air traffic controllers with the use of adaptive training systems (simulators) testifies to the prospect of using models and methods of artificial intelligence. The use of a multiagent approach is necessary to adapt the system to the knowledge and skills of the learner and to formulate an individual learning strategy. The proposed method of logical derive is based on precedents and allows solving the problems of classification and decision-making appropriate to the precedent. The areas of further research are the development of methods for the synthesis of intelligent training systems (simulators) for air traffic controllers with embedded mechanisms for generating training programs based on adequate domain models.
Keywords: adaptive training system, air traffic controler, individual training strategy, multiagent decision support system, ontology, simulator, artificial intelligence.