Método combinado na previsão de séries temporais de consumo de energia elétrica

Maria Clara  Palermo de Souza Carvalho Caria; Anna Regina Corbo Costa

doi:10.5585/2024.25125

Authors

Maria Clara Palermo de Souza Carvalho Caria Federal Center for Technological Education Celso Suckow da Fonseca https://orcid.org/0009-0005-7769-544X
Anna Regina Corbo Costa Federal Center for Technological Education Celso Suckow da Fonseca https://orcid.org/0000-0002-6430-8114

DOI:

https://doi.org/10.5585/2024.25125

Keywords:

time series, electrical energy, forecast, combined model

Abstract

The production of electrical energy must be planned in order to optimize its processes and minimize possible failures, which can be helped by analyzing historical consumption series. This article proposes the construction of a combined time series forecasting model, in order to predict electricity consumption per consumer for all Brazilian states. This consists of the linear combination of the TSLMS, TSLMTS and SNAIVE models using three and five years of history. The weights assigned to each model are functions of the errors calculated by the mean absolute deviation of the individual predictions. The combined model presented a mean squared error of 5.7 kWh per consumer and Theil’s U of 0.76, illustrating a more accurate result for three years of history and a performance equivalent to the other individual models for a longer history. Thus, the present proposal is applicable to series with little historical data available, presenting promising results for a forecast year.

Downloads

Download data is not yet available.

Author Biographies

Maria Clara Palermo de Souza Carvalho Caria , Federal Center for Technological Education Celso Suckow da Fonseca

Bachelor in Production Engineering from the Federal Center for Technological Education Celso Suckow da Fonseca - CEFET/RJ (2022).

Anna Regina Corbo Costa, Federal Center for Technological Education Celso Suckow da Fonseca

PhD in Civil Engineering (UFRJ, 2017). Associate Professor at the Department of Mathematics at CEFET/RJ. Area of expertise: Applied Mathematics and Statistical methods.

References

Agostino, I., Araujo, R. S., de Oliveira Noronha, M., Fonseca, J. I. D. S., & Souza, A. M. (2019). Combinação de previsões aplicada à modelagem de operações: um estudo de caso em um terminal portuário. Exacta, 17(1), 99-110. https://doi.org/10.5585/exactaep.v17n1.8327

Aras, S.; Deveci, İ.& Polat, C. (2017). Comparative study on retail sales forecasting between single and combination methods. Journal of Business Economics and Management, 18(5), 803-832. https://doi.org/10.3846/16111699.2017.1367324

Araújo, E. G., Júnior, S. F. A. X., Barbosa, N. F. M., & Oliveira, T. A. (2023). Modelagem e previsao de modelos de séries temporais do consumo de energia elétrica na Regiao Nordeste do Brasil. Sigmae, 12(1), 10-28. https://publicacoes.unifal-mg.edu.br/revistas/index.php/sigmae/article/view/2008

Bilgili, M., & Pinar, E. (2023). Gross electricity consumption forecasting using LSTM and SARIMA approaches: A case study of Türkiye. Energy, 284, 128575. https://doi.org/10.1016/j.energy.2023.128575

Camelo, H., Lucio, P. S., Junior, J. B. V. L., de Carvalho, P. C. M., & dos Santos, D. V. G. (2018). Innovative hybrid models for forecasting time series applied in wind generation based on the combination of time series models with artificial neural networks. Energy, 151, 347-357. https://doi.org/10.1016/j.energy.2018.03.077

Carneiro, A (2014). Previsão do consumo de energia elétrica a curto prazo, usando combinações de métodos univariados (Dissertação de Mestrado). Universidade Federal de Juiz de Fora, Minas Gerais, Brasil.

Chou, J., Tran, D. (2018). Forecasting energy consumption time series using machine learning techniques based on usage patterns of residential householders. Energy, 165(B), 709-726. https://doi.org/10.1016/j.energy.2018.09.144.

Empresa de Pesquisa Energética – EPE (2022). Consumo Mensal de Energia Elétrica por Classe (regiões e subsistemas). (n.d.). Retrieved August 29, 2022, from https://www.epe.gov.br/pt/publicacoes-dados-abertos/publicacoes/consumo-de-energia-eletrica

Fan, G. F., Wei, X., Li, Y. T., & Hong, W. C. (2020). Forecasting electricity consumption using a novel hybrid model. Sustainable Cities and Society, 61, 102320. https://doi.org/10.1016/j.scs.2020.102320

Farias, L. M., & Sellitto, M. A. (2013). Uso da energia ao longo da história: evolução e perspectivas futuras. Revista Liberato, 12(17), 07–16. https://revista.liberato.com.br/index.php/revista/article/view/164

Gontijo, T. S., Rodrigues, F. D. M., de Cássio Rodrigues, A., da Silva, S. A., & de Azevedo, A. A. (2017). Consumo industrial de energia elétrica: um estudo comparativo entre métodos preditivos. Brazilian Journal of Production Engineering, 3(3), 31-45. https://periodicos.ufes.br/bjpe/article/view/v3n3_03

Hibon, M., & Evgeniou, T. (2005). To combine or not to combine: selecting among forecasts and their combinations. International Journal of Forecasting, 21(1), 15–24. https://doi.org/10.1016/j.ijforecast.2004.05.002

Hu, H., Zhang, J. & Li, T. (2020). A comparative study of VMD-based hybrid forecasting model for nonstationary daily streamflow time series. Complexity, 1-21. https://doi.org/10.1155/2020/4064851

Liu, B., Tang, X., Cheng, J., & Shi, P. (2020). Traffic flow combination forecasting method based on improved LSTM and ARIMA. International Journal of Embedded Systems, 12(1), 22-30. http://arxiv.org/abs/1906.10407

Lorenzo, H. C. (2001). O setor elétrico brasileiro: passado e futuro. Perspectivas: Revista de Ciências Sociais. São Paulo, 24-25: 147-170. https://periodicos.fclar.unesp.br/perspectivas/article/view/406

Marcos, I. P., & Pontes Júnior, A. P. (2021). Previsão do Consumo de Energia Elétrica na Região Nordeste do Brasil. Revista de Engenharia e Pesquisa Aplicada, 6(3), 21–30. https://doi.org/10.25286/repa.v6i3.1684

Martins, V. L. M., & Werner, L. (2014). Comparação de previsões individuais e suas combinações: um estudo com séries industriais. Production. https://lume.ufrgs.br/handle/10183/115339?locale-attribute=es

Mohammed, H. H., Asem, A. & El-Bakry, H. (2024). Advanced Time Series Forecasting Models for Electricity Demand Prediction: A Comparative Study. Fusion: Practice and Applications, 15(1), 19-9. https://doi.org/10.54216/FPA.150102

Morettin, P. A. & Toloi, C. M. D. C. (2018). Análise de séries temporais. 3a ed. São Paulo, Edgard Blucher. 474p.

Neto, A. B., Corrêa, W. L. R., & Perobelli, F. S. (2016). Consumo de Energia e Crescimento Econômico: uma Análise do Brasil no Período 1970-2009. Análise Econômica, 34(65). https://doi.org/10.22456/2176-5456.44622

Reis, L. B. (2011). Geração de energia elétrica. 2a ed. São Paulo, Manole. 203p.

Serrão, F. C. C. (2003). Modelo de Previsão de Carga de Curto Prazo Utilizando Redes Neurais e Lógica Fuzzy (Doctoral dissertation, PUC-Rio). https://www2.dbd.puc-rio.br/pergamum/tesesabertas/0116365_03_pretexto.pdf

Santos, J.A.A. & Chaucoski, Y. (2020). Previsão do Consumo de Energia Elétrica na Região Sudeste: Um Estudo de Caso Usando SARIMA e LSTM. Revista Cereus, v. 12, n. 4, p. 93-104. http://ojs.unirg.edu.br/index.php/1/article/view/3255

Taylor, J. W., de Menezes, L. M., & McSharry, P. E. (2006). A comparison of univariate methods for forecasting electricity demand up to a day ahead. International Journal of Forecasting, 22(1), 1–16. https://doi.org/10.1016/j.ijforecast.2005.06.006

Ueda, R. M., Agostino, Í., Marasca, L., da Silva, R. B., & Souza, A. M. (2018). Combinação de técnicas de previsão de demanda: aplicação em uma agroindústria de nutrição animal. Exacta, 16(2), 83-93. https://doi.org/10.5585/exactaep.v16n2.7295

Zhu, X., Song, Z., Sen, G., Tian, M., Zheng, Y. & Zhu, B. (2022). Prediction study of electric energy production in important power production base, China. Scientific Reports, 12(1), 21472. https://doi.org/10.1038/s41598-022-25885-w