即時負載預測在制定經濟、可靠和安全的電力系統運行策略中起著關鍵作用，而不同的時間解析度適合的情境不同，如極短期5分鐘前負載預測主要應用於即時操作，而短期的1小時前負載預測主要應用於短期電力系統調度與電力交易。本研究提出基於深度學習的即時負載預測方法，旨在準確預測未來一段時間內的電力負載需求。該方法以自相關函數與相互資訊作為特徵產生與選取，藉此考慮負載具時間性的規律與變數間的複雜非線性相關性，並使用結合堆疊雙向-單向長短期記憶神經網路和殘差學習，以捕捉時間序列中複雜的模式和趨勢，用以預測下一時間點之負載。
為評估本研究成果應用於不同時間需求的預測成果，本文分別使用新英格蘭獨立系統營運商公開的每小時歷史電力負載數據與沙崙C區資安暨智慧科技研發大樓每5分鐘的歷史電力負載數據，將其各自分為訓練與測試集以驗證準確性。進一步將本研究所提之方法與堆疊式遞歸神經網路、雙向長短期記憶、堆疊式長短期記憶等時間序列模型進行比較，以上述兩個資料集的各個月分數據進行對比。結果證明本研究所提方法在即時負載預測中表現較佳，能夠準確地捕捉到負載的變化趨勢和季節性變化，並且具有較低的預測誤差及較高的穩定性。

Real-time load forecasting plays a key role in formulating economical, reliable, and safe power system operation strategies, and different time resolutions are suitable for different scenarios. Hourly load forecasting is mainly used in short-term power system dispatching and power trading.
This study proposes a real-time load forecasting method based on deep learning, aiming to accurately predict the power load demand for a certain period in the future. This method uses the autocorrelation function and mutual information as feature generation and selection, so as to consider the time-dependent law of the load and the complex nonlinear association between variables and uses a combination of stacked bidirectional-unidirectional long-term short-term memory neural networks and residual learning to capture complex patterns and trends in time series to predict the load at the next point in time.
In order to evaluate the results of this study applied to the prediction results of different time demands. This study uses the hourly historical power load data released by the independent system operator in New England and the 5-min historical power load of the Information Security and Smart Technology R&D Building in Area C of Shalun. And split them into training and testing sets respectively to verify the accuracy.
Further compare the method proposed in this study with time series models such as stacked RNN, bidirectional LSTM, and double layer stacked LSTM, and compare the monthly data of the above two data sets. The results can prove that the method of this study performs well in real-time load forecasting, can accurately capture the changing trend and seasonal changes of load, and has low forecasting error and high stability.

[1] B. F. Hobbs , S. Jitprapaikulsarn, S. Konda, V. Chankong, K. A. Loparo, and D. J. Maratukulam, “Analysis of the Value for Unit Commitment of Improved Load Forecasts,” IEEE Trans Power Syst, vol. 14, no. 4 , pp. 1342-1348, 1999.
[2] Y. LeCun, B. Boser, J. S. Denker, D. Henderson, R. E. Howard, W. Hubbard, and L. D. Jackel, "Backpropagation Applied to Handwritten Zip Code Recognition", Neural Comput., vol. 1, no. 4, pp. 541-551, Dec. 1989.
[3] D. E. Rumelhart, G. E. Hinton, and R. J. Williams, "Learning Representations by Back-propagating Errors" in Neurocomputing: Foundations of Research, Cambridge, MA, USA:MIT Press, pp. 696-699, 1988.
[4] W. Huang, G. Song, H. Hong, and K. Xie, "Deep Architecture for Traffic Flow Prediction: Deep Belief Networks with Multitask Learning", IEEE Trans. Intell. Transp. Syst., vol. 15, no. 5, pp. 2191-2201, Oct. 2014.
[5] I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, , S. Ozair, A. Courville, and Y. Bengio, "Generative Adversarial Networks", Proc. Adv. Neural Inf. Process. Syst., pp. 2672-2680, 2014.
[6] K. Yan, W. Li, Z. Ji, M. Qi, and Y. Du, “A Hybrid LSTM Neural Network for Energy Consumption Forecasting of Individual Households,” IEEE Access, vol. 7, pp. 157633-157642, 2019.
[7] L. Wang, S. Mao, B. M. Wilamowski, and R. M. Nelms, “Ensemble Learning for Load Forecasting,” IEEE Trans. on Green Communications and Networking, vol. 4, no. 2, pp. 616-628, June 2020.
[8] D. Ageng, C. -Y. Huang, and R. -G. Cheng, “A Short-Term Household Load Forecasting Framework Using LSTM and Data Preparation,” IEEE Access, vol. 9, pp. 167911-167919, 2021.
[9] B. Farsi, M. Amayri, N. Bouguila, and U. Eicker, "On Short-Term Load Forecasting Using Machine Learning Techniques and a Novel Parallel Deep LSTM-CNN Approach", IEEE Access, vol. 9, pp. 31191-31212, 2021.
[10] W. Huang, Q. Song, and Y. Huang, “Two-stage Short-term Power Load Forecasting Based on SSA–VMD and Feature Selection,” Appl. Sci., vol. 13, no. 11, pp. 6845, Jun. 2023.
[11] M. Hall, "Correlation-based Feature Selection for Discrete and Numeric Class Machine Learning", Proc. 17th Int. Conf. Mach. Learn., pp. 359-366, 2000.
[12] S. Lars and W. Svante, “Analysis of variance (ANOVA),” Chemometrics Intell. Lab. Syst., vol. 6, no. 4, pp. 259-272, Jul. 1989.
[13] C. E. Shannon, "A Mathematical Theory of Communication", Bell Syst. Tech. J., vol. 27, no. 3, pp. 379-423, Jul. 1948.
[14] S. Hochreiter and J. Schmidhuber, "Long Short-term Memory", Neural Comput., vol. 9, no. 8, pp. 1735-1780, 1997.
[15] M. Daneshvar and H. Veisi, "Persian Phoneme Recognition Using Long Short-term Memory Neural Network", Information and Knowledge Technology (IKT) 2016 Eighth International Conference on, pp. 111-115, 2016 Sep 7.
[16] S. Wang and J. Jiang, "Learning Natural Language Inference with LSTM", Proc. HLT-NAACL, pp. 1442-1451, 2016.
[17] A. Moghar and M. Hamiche, "Stock Market Prediction Using LSTM Recurrent Neural Network", Procedia Comput Sci, vol. 170, pp. 1168-1173, Jan. 2020.
[18] R. Fu, Z. Zhang, and L. Li, "Using LSTM and GRU Neural Network Methods for Traffic Flow Prediction", Proc. 31st Youth Academic Annu. Conf. Chin. Assoc. Autom. (YAC), pp. 324-328, Nov. 2016.
[19] M. Schuster and K. K. Paliwal, "Bidirectional Recurrent Neural Networks," IEEE Trans. Signal Process., vol. 45, no. 11, pp. 2673-2681, Nov. 1997.
[20] A. Graves and J. Schmidhuber, "Framewise Phoneme Classification with Bidirectional LSTM Networks", Proc. IEEE Int. Joint Conf. Neural Netw., pp. 2047-2052, Jul. 2005.
[21] S. Siami-Namini, N. Tavakoli, and A. S. Namin, "The Performance of LSTM and BiLSTM in Forecasting Time Series", Proc. IEEE Int. Conf. Big Data (Big Data), pp. 3285-3292, Dec. 2019.
[22] D. Balduzzi, M. Frean, L. Leary, J.P. Lewis, K. W. D. Ma, and B. McWilliams, "The Shattered Gradients Problem: If resnets are the answer then what is the question?", Proc. Int. Conf. Mach. Learn., pp. 342-350, 2017.
[23] N. N. Schraudolph, "Accelerated Gradient Descent by Factor-centering Decomposition", 1998.
[24] K. He, X. Zhang, S. Ren, and J. Sun, "Deep Residual Learning for Image Recognition," Proc. IEEE Conf. Comput. Vision Pattern Recognit., pp. 770-778, 2016.
[25] R. J. Hyndman and A. B. Koehler, "Another Look at Measures of Forecast Accuracy", Int. J. Forecasting, vol. 22, no. 4, pp. 679-688, Oct. 2006.
[26] J. S. Armstrong and F. Collopy, "Error Measures for Generalizing about Forecasting Methods: Empirical Comparisons", Int. J. Forecasting, vol. 8, no. 1, pp. 69-80, Jun. 1992.
[27] C. Wang, X. Li, and H. Li, “Role of Input Features in Developing Data-driven Models for Building Thermal Demand Forecast”, Energy Buildings, vol. 277, 2022.
[28] K. Chen, K. Chen, Q. Wang, Z. He, J. Hu, and J. He, “selected_data_ISONE_sample.csv,” Accessed: June 7. 2022 [Online]. Available: https://github.com/yalickj/load-forecasting-resnet/blob/master/ISO-NE/selected_data_ISONE_sample.csv
[29] M. Xia, H. Shao, X. Ma and C. W. de Silva, "A Stacked GRU-RNN-Based Approach for Predicting Renewable Energy and Electricity Load for Smart Grid Operation," IEEE Trans. Ind. Inform., vol. 17, no. 10, pp. 7050-7059, Oct. 2021.
[30] M. Ma, C. Liu, R. Wei, B. Liang and J. Dai. “Predicting Machine's Performance Record Using the Stacked Long Short-term Memory (LSTM) Neural Networks.” J Appl Clin Med Phys., vol. 23, no. 3, Mar. 2022.
[31] G. Xia, Z. Tian and Z. OuYang, “Research on Forecast of Daily Electricity Consumption of Household Air Conditioning Based on Improved Long-short Memory Network.” J. Phys.: Conf. Ser., vol. 1948, no. 3, 2022.
[32] D. Bahdanau, K. Cho, and Y. Bengio, "Neural Machine Translation by Jointly Learning to Align and Translate", Proc. Int. Conf. Learn. Representations, 2015.
[33] G. Chetty and M. Lipton, "Multimodal Feature Fusion for Video Forgery Detection[C]", IEEE 13th International Conference on Information Fusion, pp. 26-29, 2010.
[34] J. Xiao, H. Ye, X. N He, H. W. Zhang, F. Wu, and T. S. Chua, "Attentional Factorization Machines: Learning the Weight of Feature Interactions via Attention Networks", arXiv preprint arXiv:1708.04617, 2017.