Sistema experto difuso para el diagnóstico del índice de salud de transformadores eléctricos

In the city of Ambato-Ecuador, the company “Liargy” is dedicated among its services for analyzing transformer dielectric oils to determine the insulating oil’s to physical and chemical electrical properties to anticipate possible failures. Currently, the results of the analysis of dielectric oils are diagnosed with the use of classical logic, a formal system of reasoning and argumentation based on well-defined principles and rules based on the parameters of current regulations, which causes slowness in the diagnostic process. The objective of the research is to develop an experimental fuzzy expert system that, based on the results of the physicochemical, chromatographic, and dielectric analysis and compound contents of insulating mineral oils and the current regulations, allows the diagnosis the condition of the transformer based on the calculation of a health index. The research work uses theoretical and empirical methods, where the research design is non-experimental transversal descriptive with an inductive-deductive method. The development methodology adopted is Scrum combined with fuzzy logic and classical analysis operational techniques to develop a system to automate the diagnosis of the state of electrical transformers based on the calculation of the health index with six variables of the physicochemical properties of the insulating oil, through fuzzy inference analysis and expert criteria rules.