Identification of Atypical Billings of Consumer Units by Artificial Intelligence Techniques to Reduce Non-Technical Losses in Distribution Networks

The article presents the development of a computational platform to identify atypical billings in low voltage consumer units using the Isolation Forest algorithm, aiming to detect peculiar changes in energy consumption and/or evidence of fraud. The algorithm analyzes a large set of historical data (up to 10 years) and identifies “atypical” monthly billings among consumer units with the same type of load, generating a list of candidate units for subsequent inspection in the field. Preliminary tests carried out at a medium-sized distribution concessionaire showed an assertiveness rate of 91.70%, confirming the effectiveness of the proposed algorithm for the large-scale identification of anomalies in the monthly billing data of consumer units.