Desarrollo de un sistema de predicción con redes neuronales artificiales para determinar la demanda de energía eléctrica en una central hidroeléctrica

The prediction of events has been since ancient times, a phenomenon capable of generating curiosity in human beings, however, to achieve a projection of a future event, a detailed data analysis is required to predict subsequent events, with this idea, the The objective of the research was to develop a prediction system applying artificial neural networks to determine the forecast demand for electrical energy at the ILLUCHI 2 HYDROELECTRIC POWER PLANT. The data used for this study were collected by ELEPCO S.A operators who are based between 2 years 010 to the year 2 019. The input variables that were date and energy consumed were considered to elaborate different cases with different conditions with the purpose of reaching a successful recurrent neural network model (Long Short Term Memory) possible. Once the model variables were understood, the data was divided into two groups: training 80% and validation 20% respectively. For the relevant training, the RMSprop algorithm and the libraries offered by Python 3.8.6 were used. From the data of the demand for electrical energy from the last period of the month of January of the year 2019 to the following week of the month of February of the same year, which correspond to 8 days, it can be affirmed that the proposed model presents the greater adjustment to the behavior of the data series with a mean absolute error (MAE) of 0.0352 and mean absolute percentage error (MAPE) is 3%, thus validating and using the results to predict the demand for electrical energy for the year 2020 and 2021. For future studies it is recommended to carry out the same study applying other computer programs used for data analysis.

• Autor Principal: Bastidas Cashicana, Wilson Roberto (author)
• Outros autores: Moya Cabezas, Angelo Mauricio (author)
• Formato: bachelorThesis
• Idioma: spa
• Publicado: 2022
• Subjects: DEMANDA DE ENERGÍA ELÉCTRICA; CENTRAL HIDROELÉCTRICA; REDES NEURONALES ARTIFICIALES; ELÉCTRICA; POTENCIA
• Acceso en liña: http://repositorio.utc.edu.ec/handle/27000/9340