An Automated System for Taking Decisions to Assess the Actual State of Electrical Equipment1

Resumo

An automated system for taking decisions for a complex estimate of the technical state of the electrical network equipment of electric power plants and substations, in order to increase its operating efficiency and to correct repair cycles, is proposed. The main states of the system, which are based on methods of artificial intelligence—artificial neural networks and fuzzy logic methods, which provide the system with self-teaching and self-adjustment properties, are described. An example of the estimate of the technical state of a 110 kV power transformer based on data of a chromatography analysis of the gases dissolved in the oil and using this system is given.