IMPROVEMENT OF CONTROL MODELS OF CLOSED SYSTEMS USING NEURAL NETWORKS

Abstract

It is known that the importance of the modeling method in the research of technological systems is increasing. One of the main reasons for this can be attributed to the possibility of simplifying a complex system, as well as the priority aspects of logical consistency and mathematical laws. At the same time, the development of information technology, computing and artificial intelligence has strengthened the practice of using modeling methods. In other words, the modeling apparatus has a good integration feature with these concepts. In this article, the importance and relevance of the development of optimal control models based on a neural network is presented as the main scientific idea on the example of a closed-loop control system. When the control system is formed as an intelligent system, or when the system description is based on the concepts of an intelligent system, the main parameters of modeling are differentiated. It is theoretically and practically based that the results of modeling gain significance depending on these parameters. The effectiveness of the use of neural network models in the optimal control of the activity of the intelligent system is explained against the background of the possibility of minimizing the control error. It has been scientifically proven that modeling closed systems using neural networks has several advantages. Quantitative parameters that need to be paid attention to in the formation of an intelligent system are researched. Conclusions and proposals are presented on the effectiveness of using neural networks in modeling complex technological systems.