Previsão de carga para a programação diária eletroenergética do Sistema Interligado Nacional por Redes Neurais Artificiais

• Main Author: Bonvicine, Angelo Antonio de Carvalho
• Publication Date: 2021
• Format: Bachelor thesis
• Language: por
• Source: Repositório Institucional da UNESP
• Download full: http://hdl.handle.net/11449/213945
• topic: Electrical Load forecast; Planning of electrical energy operation; DESSEM model; Time series; Recurrent Neural Networks; Previsão de Carga; Planejamento da operação eletroenergética; Modelo DESSEM; Séries Temporais; Redes Neurais Recorrentes; LSTM

In January 2021, a new model was included in the chain of electro-energy planning models developed by the Electric Energy Research Center, CEPEL. As well as the consolidated medium and short-term models of this chain, the DESSEM model requires a range of input data for its operation, as for example, the inclusion of a forecasted load curve for the horizon of the planning study to be executed, where the National Electric System Operator, ONS, builds daily using its methodologies. The objective of this work is to contribute using concepts of Recurrent Neural Networks with LSTM structures, for the elaboration of a load forecast model, which can be used alternatively to the ONS model for building the load curve associated with the daily electrical energy schedule of the Interconnected System National, the SIN.