Asset management in electrical power companies: the use of a Paraconsistent Specialist System to increase efficiency and quality through prediction

Managers, operators and maintenance workers face challenges daily to raise and maintain the quality of services provided by power companies. This work shows the development and advantages of a Paraconsistent Specialist System (PSS) dedicated to the management of assets in the electrical area and its particularities through the collection of data available in a supervisory and data acquisition system installed in operation centers (CO) of electrical distribution networks. This PSS was built based on paraconsistent annotated logic (PAL) algorithms, configuring decision trees capable of processing data with machine learning and dealing with contradictory or incomplete information. As it runs within a supervisory system, it eliminates the disadvantage of using multiple systems for each purpose, in addition it creates an effective risk index (ERI) that allows operators to work only with this information, integrating operation and asset management. It also provides specialized indicators based on technical standards, communicates in the most common protocols in the electricity sector, and can be integrated into any level of automation. At the end of this article, we present results of the application of the PSS acting in the management of power transformers and high voltage circuit breakers isolated with SF6 using simulated and real data of common and critical faults. The results achieved with the PSS in its first version indicate that the technique used has great potential to increase its scope in decision making, becoming an excellent computational tool to support the management of assets in the power distribution sector.