Artificial Neural Network Based Losses Model Due to Energy Exchange

Due to the characteristic of the interconnected Brazilian national system and to the location of Mato Grosso do Sul state, the connection between the south and southern power systems has become a fragile link, making this state susceptible to significant changes, as far as losses and damages are concerned, when submitted to different levels of energy exchanges between those regions. The determination of these losses permits the reevaluation of the costs, which are normally transferred to the consumer. This work presents a model that allows the determination of the so called, technical losses related only to the energy internally used by the Mato Grosso do Sul State, and the one related to the energy exchange between south and southern regions. The work presents the methodology of the elaboration of database and the neural network modeling. The results show that the model does match with the actual physical system.