基于注意力机制的VMD-BiLSTM风力发电预测
VMD-BiLSTM Wind Power Prediction Based on Attention Mechanism
摘要: 针对风力发电数据噪音多、波动大等特点，本文提出了基于注意机制的VMD-BiLSTM超短期风力发电预测模型。首先，对数据进行变分模态分解，降低序列的非平稳性和强非线性；其次，将各子模态固定在同一时间滑动窗口上，分别使用BiLSTM进行进一步的双向序列特征提取，然后引入注意机制，加强重要特征信息的影响，提高模型的最终预测效果。最后通过NREL网站的美国西南地区的发电数据集来进行验证，最终结果显示，无论是从预测的准确性还是评判指标方面，本文所提出的模型在风力发电预测上都优于其他5类模型。
Abstract: Aiming at the characteristics of large noise and large fluctuations in wind power generation data, an attention mechanism-based VMD-BiLSTM ultra-short-term wind power generation prediction model is proposed. Firstly, the data is decomposed using the variational mode (VMD) to reduce the se-quence non-stationarity and strong nonlinearity. Secondly, each sub-mode is fixed on the sliding window at the same time, and BiLSTM is used to further extract bidirectional sequence features, and then the attention mechanism is introduced to strengthen the influence of important feature information and improve the final performance of the model. Finally, the power generation data set in the Southwest of the United States on the NREL website was verified. The final results show that the model proposed in this paper is superior to the other 5 models in terms of wind power predic-tion accuracy and evaluation indicators.
文章引用：邓文武, 丁咏梅. 基于注意力机制的VMD-BiLSTM风力发电预测[J]. 应用数学进展, 2023, 12(1): 153-165. https://doi.org/10.12677/AAM.2023.121019

参考文献

[1] 张爱枫, 段新宇, 何枭峰. 基于CNN和LightGBM的新型风电功率预测模型[J]. 电测与仪表, 2021, 58(11): 121-127.
[2] Nepal, B., Yamaha, M., Yokoe, A., et al. (2020) Electricity Load Forecasting Using Clustering and ARIMA Model for Energy Management in Buildings. Japan Architectural Review, 3, 62-76. [Google Scholar] [CrossRef] 
[3] 佟佳弘, 武志刚, 谢钰. 缺少气象数据场景下的超短期风电预测[J]. 电力系统及其自动化学报, 2022, 34(3): 142-150.
[4] Li, L.L., Zhao, X., Tseng, M., et al. (2020) Short-Term Wind Power Forecasting Based on Support Vector Machine with Improved Dragonfly Algorithm. Journal of Cleaner Production, 242, Article ID: 118447. [Google Scholar] [CrossRef] 
[5] 阳曾, 丁施尹, 叶萌, 李晶, 薛书倩, 吴昊天. 基于变分模态分解和深度学习的短期电力负荷预测模型[J]. 电测与仪表, 2022, 1-8.
[6] 朱伟, 孙运全, 钱尧, 金浩, 杨海晶. 基于CEEMD-GRU模型的短期电力负荷预测方法[J]. 电测与仪表, 2023, 60(1): 16-22.
[7] 赵兵, 王增平, 纪维佳, 高欣, 李晓兵. 基于注意力机制的CNN-GRU短期电力负荷预测方法[J]. 电网技术, 2019, 43(12): 4370-4376.
[8] Meng, Y., Chang, C., Huo, J., et al. (2022) Research on Ultra-Short-Term Prediction Model of Wind Power Based on Attention Mechanism and CNN-BiGRU Combined. Frontiers in Energy Research, 10, Article ID: 920835. [Google Scholar] [CrossRef] 
[9] 梁智, 孙国强, 李虎成, 卫志农, 臧海祥, 周亦洲, 陈霜. 基于VMD与PSO优化深度信念网络的短期负荷预测[J]. 电网技术, 2018, 42(2): 598-606.
[10] Dragomiretskiy, K. and Zosso, D. (2014) Variational Mode Decomposition. IEEE Transactions on Signal Processing, 62, 531-544. [Google Scholar] [CrossRef] 
[11] Wang, S., Liu, C., Liang, K., et al. (2022) Wind Speed Prediction Model Based on Improved VMD and Sudden Change of Wind Speed. Sustainability, 14, 8705. [Google Scholar] [CrossRef] 
[12] Tan, M., Yuan, S., Li, S., et al. (2022) Ultra-Short-Term Industrial Power Demand Forecasting Using LSTM Based Hybrid Ensemble Learning. IEEE Transactions on Power Systems, 35, 2937-2948. [Google Scholar] [CrossRef] 
[13] 肖浩逸, 何晓霞, 梁佳佳, 李春丽. 一种基于模态分解和机器学习的锂电池寿命预测方法[J/OL]. 储能科学与技术, 2022, 11(12): 3999-4009.
[14] 潘洛华. 基于SSA-CNN-BILSTM的电力短期负荷数据预测方法[D]: [硕士学位论文]. 杭州: 浙江大学, 2022.
[15] Tian, C., Niu, T. and Wei, W. (2022) Developing a Wind Power Forecasting System Based on Deep Learning with Attention Mechanism. Energy, 257, Article ID: 124750. [Google Scholar] [CrossRef] 
[16] 刘国海, 孙文卿, 吴振飞, 等. 基于Attention-GRU的短期光伏发电功率预测[J]. 太阳能学报, 2022, 43(2): 226-232.
[17] Zhou, H., Zhang, Y., Yang, L., et al. (2019) Short-Term Photovoltaic Power Forecasting Based on Long Short Term Memory Neural Network and Attention Mechanism. IEEE Access, 7, 78063-78074. [Google Scholar] [CrossRef] 
[18] 刘杰, 金勇杰, 田明. 基于VMD和TCN的多尺度短期电力负荷预测[J]. 电子科技大学学报, 2022, 51(4): 550-557.
[19] 王金玉, 金宏哲, 王海生, 张忠伟. ISSA优化Attention双向LSTM的短期电力负荷预测[J]. 电力系统及其自动化学报, 2022, 34(5): 111-117.
[20] 姜旭初, 许宇澄, 宋超. 短期风力发电负荷预测的新方法[J]. 北京师范大学学报(自然科学的版), 2022, 58(1): 39-46.

为你推荐