The Brazilian Association of Energy Distribution Utilities estimates that non-technical losses represent more than 5.5% of the total energy distributed, most coming from fraud and theft. To try to mitigate those losses, the distribution utilities send field crews for the inspection of possible fraudster clients. However, the procedure is expensive and gives no financial return to the utility if it is not focused on areas with high fraud probability. On those locations, there is a correlation between losses and socioeconomic indices. Thus, this work proposes a model able to select clients with high fraud probability, which should be visited by the field crews. The smart grid structure, energy consumption data, clients' registration data and socio-economic indices from the 2010 Brazilian Census are used by the model.
Ao meu orientador, Dr. Erik Eduardo Rego, pelas importantes contribuições, discussões e atenção em todas as etapas do desenvolvimento deste trabalho. Ao Dr. Diogo Mac Cord de Faria pelas sugestões oferecidas no exame de qualificação e na banca de defesa.Ao Dr. Kleber Hashimoto pela participação na banca de defesa e sugestões.Ao Dr. Francisco Anuatti Neto pelas sugestões oferecidas no exame de qualificação.Aos meus professores de graduação, em especial Dr. Luiz Lebensztajn, Dr. Maurício Barbosa Camargo Salles e Dr. Nelson Kagan, pelas orientações ao longo da minha trajetória acadêmica.À secretária de Pós-Graduação, Lídia Nogueira da Silva, pela atenção dispensada em diversas ocasiões.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.