WIND SPEED FORECAST WITH MULTI-VARIABLE TIME SERIES USING RECURRENT NEURAL NETWORK
DOI:
https://doi.org/10.59627/cbens.2022.1069Keywords:
MV-LSTM, Forecast, Wind EnergyAbstract
Time series analysis is extremely important for control and management of electrical systems. In wind systems, there is a need to predict the wind speed with greater precision, which allows for an intelligent energy dispatch and managing risks that the high volatility of the wind can cause. This paper implements a multivariate Recurrent Neural Network model of short-term memory type (MV-LSTM) for hourly average, maximum and minimum speed data. A single variable LSTM network is also implemented to compare the results. The data used were collected by a sonic anemometer from January 1st to December 31st, 2015 in the native Brazilian cerrado at Fazenda Água Limpa – FAL located in the Federal District. The results show that maximum and minimum speed vectors improve wind speed prediction and that the MV-LSTM model reduces the forecast delay when there is a sudden change in wind speed, which is a problem cited by Xie et al. (2021) his work. The model can be optimized by adding more layers and combining with other machine learning models.
Downloads
References
Cao, Y., Gui, L., 2018. Multi-Step wind power forecasting model Using LSTM networks, Similar Time Series and LightGBM, 5th International Conference on Systems and Informatics (ICSAI), pp. 192-197.
Dubey, A. K., Kumar, A., García-Díaz, V., Sharma, A. K., Kanhaiya, K., 2021. Study and analysis of SARIMA and LSTM in forecasting time series data, Sustainable Energy Technologies and Assessments, vol. 47, n. 10, pp. 114-128.
Du, S., Li, T., Yan Y., Shi-Jinn, H., 2019. Deep Air Quality Forecasting Using Hybrid Deep Learning Framework, arXiv, vol. 47, pp. 114-126.
Hochreiter, S., Schmidhuber, J., 1997. Long short-term memory, Neural Comput, vol. 9, pp. 1735–1780.
Jin, X., Yu, X., Wang, X., Bai, Y., Su, T., Kong, J., 2020. Prediction for Time Series with CNN and LSTM, Springer Singapore, pp. 631-641.
Liang, S., Nguyen, L., Jin, F., 2018. A Multi-variable Stacked Long-Short Term Memory Network for Wind Speed Forecasting, IEEE Int. Conf. Big Data, pp. 4561–4564.
Rashid, M. H., 2018. AMI Smart Meter Big Data Analytics for Time Series of Electricity Consumption, 17th IEEE International Conference On Trust, Security And Privacy In Computing And Communications, pp. 1771-1776.
Xie, A., Yang, H., Chen, J., Sheng, L., Zhang, Q., 2021. A Short-Term Wind Speed Forecasting Model Based on a Multi-Variable Long Short-Term Memory Network, vol. 12, n. 1, pp. 2073-4433.
Yao, W., Huang, P. Jia, Z., 2018. Multidimensional LSTM Networks to Predict Wind Speed, 37th Chinese Control Conference (CCC), pp. 7493-7497.
Zufferey, T., Ulbig, A., Koch, S., Hug, G., 2017. Forecasting of Smart Meter Time Series Based on Neural Networks, Springer International Publishing, pp. 10-21.
