In electric power generation and distribution systems, three structures are capable of impacting energy quality: generation, distribution and loads. Currently, the diversity of the latter ranges from purely resistive loads such as heating systems to extremely nonlinear loads such as switched electronic power supplies. This has dramatically affected the quality of electric energy. The operation of these loads contributes to the reduction of the power factor of the installation, the injection of harmonic currents in the electrical network and the imbalance of currents and voltages between the phases in the case of three-phase systems. The implementation of a method capable of recognizing electrical loads and identifying them through algorithms can be used to manage the electric energy consumption, contributing a lot to the improvement of the quality of electric energy. In order to deal with uncertainty in the identification of these loads, probability distributions will be applied, which are able to handle uncertainty through a probabilistic modeling using Bayesian Networks.