Fraud in electrical energy consumption represents a critical economic burden for utility companies around the world. Despite systematic efforts to mitigate electricity theft, this practice persists mostly in developing countries where companies rely on traditional detection methods. In Brazil it is estimated that around 7% of the total electrical energy available for consumption in 2016 was lost due to frauds. Here we describe an efficient and scalable system to predict fraudulent behavior and guide in loco inspections. We compared the performances of several machine learning algorithms using consumption and inspection data provided by CPFL Energia. We show that proper feature engineering and boosted classification trees trained with XGBoost are able to extract patterns related to fraud occurrence and to achieve predictive power of practical consequences. Moreover, we demonstrate how Shapley additive explanation (SHAP) values can be employed to build user friendly explanations. Together, the proposed model and its explainers contribute not only to reveal potentially fraudulent behavior but also to understand root causes, what can be used to devise robust mitigation strategies.
Energy fraud is a critical economical burden for electric power orga-nizations in Brazil. In this paper we present the application of novel MachineLearning algorithms to boost efficiency in detection of energy frauds. More-over, we also propose a generalized and unsupervised model for fraud detectionbased on consumption anomalies.
This paper aims to validate the performance capabilities of a Pressure Reducing Turbine (PRT) with respect to initial predictions based on analytic calculations. The designed equipment was installed in a beverage facility, located in Brazil. The validation procedure consists of analyzing the data collected in several periods of PRT’s operation, accessed remotely via an online server. The analysis of empirical data identifies the behavior of two key variables: generated power and effective efficiency. However, the observed boundary conditions differed significantly from expected values, forcing the turbine to operate in off-design conditions. The turbine model was hence refined and used to predict the PRT’s performance in such conditions. Results showed satisfactory accuracy for both power and efficiency predictions.
Nos países em desenvolvimento, o roubo de eletricidade é um tipo comum de perdas não técnicas (NTL, isto é, perdas associadas à eletricidade consumida mas não faturada por algum tipo de anomalia), afetando financeiramente não apenas os operadores do sistema de distribuição (DSO), mas também clientes. Da mesma forma que as fraudes em outros contextos, há evidências de que o roubo de eletricidade é altamente influenciado por interações sociais. Aqui nós propomos um modelo de rede multiplexado e heterogêneo para avaliar como as interações sociais e profissionais influenciam no roubo de eletricidade. Particularmente, empregando uma variação do algoritmo de caminhada aleatória com reinicialização, conseguimos obter uma nova pontuação de exposição para discriminar entre fraudadores e clientes regulares.
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.