Metodologia de previsão de carga de curto prazo multirregional considerando macrorregiões e ponderação por região meteorológica

• Ano de defesa: 2023
• Autor(a) principal: Figueiró, Iuri Castro
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; Brasil; Engenharia Elétrica; UFSM; Programa de Pós-Graduação em Engenharia Elétrica; Centro de Tecnologia
• Palavras-chave em Português: Previsão de carga de curto prazo; Previsão hierárquica; Multirregião; Macrorregiões; Redes neurais artificiais; Short-term load forecasting; Macro-regions; Hierarchical load forecasting; Artificial neural network; CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA
• Link de acesso: http://repositorio.ufsm.br/handle/1/33951

Activities related to the planning and operation of power systems use as premise the load forecasting, which is responsible to provide a load estimative for a given horizon that assists mainly in the electroenergetic operation of an electrical system. The short-term multiregional load forecast becomes an approach used for this purpose, where the overall forecast is performed through system partition in smaller macro regions, and soon after, is aggregated to compose a global forecast. A macro-region is characterized by having a great meteorological diversity throughout its territorial area, and this diversity is considered in this work through the weighting of Meteorological Stations (EMs) that best represent the aggregate demand Macro-Region (MR). In this context, this paper presents a short-term multiregional forecasting approach for macro-regions, with the main contribution being the proposal of an indicator that represents the Average Consumption per Meteorological Region (CERM), to be used as weighting of each EM as their importance for the total demand of the macro-region. In addition, the Variation of Load and Temperature index (IVCT) is proposed, based on the historical variation of temperature and demand and the Fuzzy indicator, responsible for indicating the Expectation of Load Change (EACF) based on the current weighted temperature and scheduled for the next day. These indicators are incorporated into a model of neural network of the Multi-layer perceptron type (MLP) for the load forecasting on the horizon of 7 days ahead with hourly and daily discretization. This methodology is evaluated for MRs belonging to a Multi-region (MTR) through analysis of sensitivities for different test sets. The results showed higher average performance of the variables IVCT and EACF in relation to the other combinations performed, and the best results were used to compose the prediction of the MTR. Finally, the proposed model presented a superior performance compared to an basis aggregate model for MTR, which shows the efficiency of the proposed methodology.