Previsão de demanda de curto prazo usando aprendizado de máquina com dados por telemetria: Um estudo de caso no Brasil

• Marcos R. de Souza IFES
• Danilo P. Silva IFES
• Thiago M. Paixão IFES

DOI: https://doi.org/10.5753/semish.2025.8804

Resumo

Este estudo avalia o desempenho de modelos de aprendizado de máquina na previsão de demanda de curto prazo em um consumidor com geração solar distribuída. Utilizando um conjunto de dados com padrões multisazonais e alta variabilidade, comparamos modelos como Ridge, KNN, LGBM, XGBoost, CatBoost, Random Forest e MLP. Foi aplicada otimização de hiperparâmetros com validação cruzada temporal durante a fase de ajuste. Os resultados mostraram que o Random Forest superou os demais, alcançando um MAE de 8,94 kW, superando o modelo de base em mais de 51%. O XGBoost, por sua vez, apresentou desempenho 49% superior ao modelo de base e tempo de treinamento 16 vezes menor que o Random Forest.

Referências

Akiba, T., Sano, S., Yanase, T., Ohta, T., and Koyama, M. (2019). Optuna: A nextgeneration hyperparameter optimization framework. In Proceedings of the 25th ACM SIGKDD international conference on knowledge discovery & data mining, pages 2623–2631.

Anh, N. T. N., Anh, N. N., Thang, T. N., Solanki, V. K., Crespo, R. G., and Dat, N. Q. (2024). Online sarima applied for short-term electricity load forecasting. Applied Intelligence, 54:1003–1019.

Dash, S. K., Roccotelli, M., Khansama, R. R., Fanti, M. P., and Mangini, A. M. (2021). Long term household electricity demand forecasting based on rnn-gbrt model and a novel energy theft detection method. Applied Sciences.

Dong, X., Deng, S., and Wang, D. (2022). A short-term power load forecasting method based on k-means and svm. Ambient Intelligence and Humanized Computing, 13:5253–5267.

e Silva, D. P., Félix Salles, J. L., Fardin, J. F., and Rocha Pereira, M. M. (2020). Management of an island and grid-connected microgrid using hybrid economic model predictive control with weather data. Applied Energy, 278:115581.

Guaitolini, T. M., Nascimento, E. B., Breder, R. P., Gobbi, R. E., e Silva, D. P., Silva, F. B. B., and Camargo, R. S. (2023). Load curve based on modeling a photovoltaic system, load and weather data - a case study in brazil. In 2023 15th IEEE International Conference on Industry Applications (INDUSCON), pages 248–253. IEEE.

Guo, W., Che, L., Shahidehpour, M., and Wan, X. (2021). Machine-learning based methods in short-term load forecasting. The Electricity Journal, 34:106884.

Hafeez, G., Alimgeer, K. S., and Khan, I. (2020). Electric load forecasting based on deep learning and optimized by heuristic algorithm in smart grid. Applied Energy, 269:114915.

Hodge, B.-M., Lew, D., and Milligan, M. (2013). Short-term load forecast error distributions and implications for renewable integration studies. In IEEE Green Technologies Conference.

Hwang, J., Kim, J.-S., and Song, H. (2022). Handling load uncertainty during on-peak time via dual ess and lstm with load data augmentation. Energies, 15(9):3001.

Jin, X.-B., Zheng, W.-Z., Kong, J.-L., Wang, X.-Y., Bai, Y.-T., Su, T.-L., and Lin, S. (2021). Deep-learning forecasting method for electric power load via attention-based encoder-decoder with bayesian optimization. Energies, 14(6):1596.

Jin, Y., Guo, H., Wang, J., and Song, A. (2020). A hybrid system based on lstm for short-term power load forecasting. Energies, 13(23):6241.

Koukaras, P., Mustapha, A., Mystakidis, A., and Tjortjis, C. (2024). Optimizing building short-term load forecasting: A comparative analysis of machine learning models. Energies, 17(6).

Kumari, P. and Toshniwal, D. (2021). Deep learning models for solar irradiance forecasting: A comprehensive review. Journal of Cleaner Production, 318:128566.

Mohammad, F., Lee, K.-B., and Kim, Y.-C. (2018). Short term load forecasting using deep neural networks. Energies.

Mori, H. and Ogasawara, T. (1993). A recurrent neural network for short-term load forecasting. IEEE Transactions on Power Systems.

Saksornchai, T., Lee, W.-J., Methaprayoon, K., Liao, J. R., and Ross, R. J. (2005). Improve the unit commitment scheduling by using the neural-network-based short-term load forecasting. IEEE Transactions on Industry Applications, 41(1):169–179.

Sousa, A. R. S., da Silva, C., da Silva, J. S. F., and Oliveira, R. R. (2021). Análise de séries temporais. SAGAH, Porto Alegre.

Tziolis, G., Koumis, A., Makrides, G., Lopez-Lorente, J., Livera, A., Georghiou, G. E., Baka, M.-I., and Theocharides, S. (2022). Comparative analysis of machine learning models for short-term net load forecasting in renewable integrated microgrids. In Proceedings of the International Conference on Energy Transmission in the Mediterranean Area (SyNERGY MED).