Anomaly detection in electrical power distribution systems based on sensors and smart meters data analytics

Abstract

The electrical power distribution systems have been subjected to a modernization process characterized, among other factors, by a higher level of monitoring and measurement. Besides, a telecommunication infrastructure is integrated into these systems allowing, for example, that the data collected from different points of the network to be reliably transferred to data management centers. This modernization allows accessing a greater amount of data that needs to be converted in information and knowledge, adding values to the emerging technologies. The application of these technologies is associated with the expectation of the improvement in quality and reliability of the electrical energy supply and lower operational cost. In this context, the detection of anomalies, characterized by abnormal or irregular condition of the system, draw attention among the functions that can benefit from the structure created in modern electrical power distribution systems. Evaluating the context of the distribution networks, it is possible to highlight two important anomalous conditions that directly impact the revenue of the utilities, also impacting the electricity bill of the consumers: faults and energy thefts. Therefore, this project aims to develop methodologies based on data science that allow detecting and locating faults and irregular connections associated with energy theft. (AU)