Modelo de previsão de carga de curtíssimo prazo aplicado na previsão do intercâmbio entre Itaipu e ANDE

• Autor(a) principal: Pimenta, Rodrigo Gonçalves
• Data de Publicação: 2019
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações do UNIOESTE
• Texto Completo: http://tede.unioeste.br/handle/tede/4635
• topic: Operação de hidrelétrica; Previsão de carga no curtíssimo prazo; Séries temporais e SARIMA; Hydroelectric power plant operation; Very short-term load forecasting; Time series and SARIMA; ENGENHARIAS

The electric load forecast is a problem of relevant complexity that integrates the planning and operation process of an electric power system, being influenced by several factors, such as economy, geography position, day, time, seasons and weather conditions. In order to execute the program of an electric power system with efficiency it is very important to be able to anticipate the demand to be supplied. Considering the operation of a hydroelectric power plant, the main objective of the load forecast is to provide input to develop the generation schedule program with economy and safety, taking into account the current operational restrictions, associated with energy availability, active power reserve and control of the upstream and downstream levels. Thus, significant deviations in the generation program of the Itaipu hydroelectric power dam may cause difficulties for the real-time operation teams, impacting the quality requirements for the Brazilian and Paraguayan Electric Power Systems. This paper presents a very short-term load forecasting (VSTLF) methodology applied to solve the problem when significant deviations are observed in the power exchange program between Itaipu e ANDE. The method is based on the Box & Jenkins time series concept in obtaining SARIMA models, taking in advantage of the seasonal characteristics of the time series formed by the hourly observations of the power exchange between Itaipu e ANDE. The proposed methodology is applied in Itaipu´s daily hydroenergetic schedule, presenting satisfactory results with average errors of around 5%, improving the method currently used.