Big Data Analytics of Smart Grids using Artificial Intelligence for the Outliers Correction at Demand Measurements

Smart grids are increasingly present to improve the energy efficiency of the electrical network and several equipment are fundamental in this process, among them smart meters. Such devices accumulate an enormous flow of information that can be analyzed to aid in the decision making of the electric system controllers. This makes the concept of Big Data Analytics present, capable of processing various data and correcting atypical values, called outiliers, through algorithms using artificial intelligence, such as Fuzzy logic and Artificial Neural Networks. As a way of improving existing results, the present work suggests the use of a hybrid algorithm, the Adaptive Neuro Fuzzy Inference System (ANFIS). This algorithm showed better performance in the correction of outliers when compared with techniques based on Artificial Neural Networks and Linear Interpolation. Finally, results of the estimations will be presented using real energy demand data for a substation of electricity distribution.