Uma metodologia para monitoramento das condições operativas de transformadores de potência e análise de tendências baseada em lógica fuzzy

Abstract

This paper presents a computational method to monitor the operating state of steady electric power transformers in a real time and perspectives of operation over time trends and their operation. The methodology consists of two main steps. The first step is building composed of a block of fuzzy inference s block that can monitor real-time values of transformer s electrical parameters of the transformer (current, voltage between phases, power, oil temperature, winding temperature), to analyze and map these inputs into a single output that reflects what has been defined as operational status of the transformer. We defined five regions of transformer s operation by which a processor can work: emergency, urgent, warning, stable and great. The second step analyze the trend of increase / decrease of operating state obtained from the fuzzy block in a given period, as one day, a month, one year and / or several years. This trend can be characterized in various ways, such as increasing, decreasing very constant. Statistical methods are employed for this analysis. The methodology was evaluated on a database of a large company of generation and transmission of electric energy from the Brazilian electricity sector. The results were satisfactory in that the development of operation of such equipment was qualitatively mapped and their tendencies well characterized. Conceptually, the analytical model presented can be extended to multiple processing units, electric and other functions for up to a full network of interest, allowing subsidize the operation of this and can indicate the need for any future system reinforcements.