The development of an intelligent system based on fuzzy neural networks for diagnosing thermal power equipment

The intelligent system for diagnosing equipment in the Russian fuel and energy industry, developed in this article, has a number of advantages over other similar systems. The authors developed a mechanism, architecture, and digital tools for creating intelligent diagnostic systems for thermal power equipment with the ability to recognize the current state of an object under conditions of dynamic information supply. We analyzed the system using the Bayesian method to assess the states of the thermal power equipment, confirming the applicability of this approach. The digital tools necessary for the functionality of the system were identified. The architecture of the decision-making system and an algorithm for the logical inference mechanism of a fuzzy neural network was developed. The system makes it possible to formalize expert opinions in quantitative metrics, combine data from measuring devices, conduct a diagnostic assessment of the conditions, and propose solutions. The system has high data reliability and allows information to be saved, creating a structured data array, and engineering ontology. The development and integration of such intelligent systems is aimed at increasing efficiency and minimizing resource costs.