簡易檢索 / 詳目顯示

研究生: 梁軒賓
LIANG, HSUAN-PIN
論文名稱: 利用衛星數據進行光伏發電預測之深度學習網路
Implementation of deep learning networks for the photovoltaic power generation prediction using satellite data
指導教授: 周碩彥
Shuo-Yan Chou
郭伯勳
Po-Hsun Kuo
口試委員: 周碩彥
Shuo-Yan Chou
郭伯勳
Po-Hsun Kuo
王瑞堂
Jui-Tang Wang
學位類別: 碩士
Master
系所名稱: 管理學院 - 工業管理系
Department of Industrial Management
論文出版年: 2022
畢業學年度: 110
語文別: 英文
論文頁數: 47
中文關鍵詞: 太陽光伏發電預測卷基神經網路機器學習混合網路長短期記憶網路衛星反演數據混和輸入
外文關鍵詞: Solar photovoltaic power prediction, Convolutional neural network, Machine learning, Hybrid networks, Long short-term memory, Satellite inversion data, Mixed inputs
相關次數: 點閱:180下載:1
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報
  • 為了將可再生能源整合到現有發電廠的控制中,準確的太陽能光伏發電預測對於提高系統可靠性至關重要。因此,本研究旨在利用衛星反演數據的隱藏價值來幫助預測10分鐘(提前1個週期)的太陽能光伏發電。為此,研究中利用兩種深度學習網絡對應不同的數據集進行模型訓練,兩者單一模型的太陽能發電預測會做為後續和混和數據模型進行預測性能評估的依據。第一個深度學習網絡是具有衛星反演輻照度輸入的 CNN-LSTM 模型,用於預測未來10分鐘的太陽能發電。第二個是長短期記憶(LSTM)模型,使用歷史天氣數據和案場輻照度數據來預測未來10分鐘的太陽能發電。最後,該研究所提出的混和數據模型會藉由不同輸入數據進行各自的模型訓練,並將訓練產生的特徵結果相連接後再輸入進後續的神經網絡,為太陽能發電提供更準確的預測結果。
    在本研究中所使用的衛星反演數據每10分鐘更新一次,該產品結合地形和氣象因素,同時還考慮雲信息來計算地面輻照度值。使用衛星反演輻照度數據的好處是不會受到太多不確定因素的影響,例如光伏板遮光或灰塵堆積。因此,以不同來源的數據作為輸入,混合數據可以提供更多方面的特徵。而這些大數據可以更準確地發現模型輸入特徵與預測目標之間的關係。
    最後,本研究使用 MAE、MSE、RMSE、MAPE 和 NRMSE 評估指標比較了 LSTM、CNN-LSTM 和提出的混合數據模型之間的預測模型性能。實驗結果表明,對比其他單一模型來說,混和數據模型在絕對平均百分比誤差(MAPE)平均可以降低78%,而正規化均方根誤差(NRMSE)平均可以降低65%。根據預測結果比較,所提出的混合數據模型顯示出更好的預測性能,顯著地降低與實際值的誤差並獲得更集中誤差分布,為太陽能發電提供更穩定的模型預測能力。


    Accurate solar PV power generation forecasts are critical for improving system reliability by integrating renewable energy into the control of existing power plants. Therefore, this study aims to predict one-step solar PV power generation by utilizing the hidden value of satellite inversion data. Moreover, this study proposed two deep learning networks for mixed data inputs to obtain the solar power generation prediction. The first deep learning network is the CNN-LSTM model with satellite inversion irradiance input to predict the one-step-ahead period of solar power generation. The second one is a long short-term memory (LSTM) model to predict the one-step-ahead period of power generation using historical weather data and solar farm observed irradiance data. Then, the outputs from both deep learning networks are combined and produce an output for the final prediction of solar power generation.
    The satellite inversion data is updated every 10 minutes with the terrain and meteorological factors, and cloud information is also considered to calculate the ground irradiance value. The advantage of using satellite inversion is that it will not be affected by too many uncertain factors, such as PV panel shading or dust accumulation. Therefore, by taking mixed data from different sources as inputs, the mixed data can provide more aspects of features. And these big data can discover a more accurate relationship between the model input features and the target.
    Finally, this research compares the prediction model’s performances among the LSTM, the CNN-LSTM, and the proposed mixed data model using evaluation metrics such as MAE, MSE, RMSE, MAPE, and NRMSE. According to the results, the proposed mixed data model shows better prediction performance with better optimization convergence.

    Abstract 2 摘要 3 Acknowledgement 4 Table of Contents 5 List of Figures 7 List of Tables 8 1. Chapter 1 Introduction 9 1.1 Background and Motivation 9 1.2 Research Purpose 10 1.3 Research Limitations 10 1.4 Organization of the Thesis 11 2. Chapter 2 Literature Review 12 2.1 Renewable Issues 12 2.2 Solar Energy Issues 12 2.3 Solar Power Generation Prediction 13 3. Chapter 3 METHODOLOGY 17 3.1 Initial Analysis 17 3.1.1 Augmented Dickey-Fuller test(ADF Test) 18 3.1.2 AutoCorrelation Function(ACF) 19 3.1.3 Scaling of input data 19 3.1.4 Timestamp information 19 3.2 Prediction Method 20 3.2.1 Long-Short Term Memory(LSTM) 20 3.2.3 CNN-LSTM network 23 3.3 Mixed data model 24 3.4 Grid search for hyperparameters selection 25 3.5 Model performance evaluation metrics 26 4. Chapter 4 Result and discussion 29 4.1 Data Description 29 4.1.1 Satellite Inversion Data 29 4.1.1.1 Feature Visualization 30 4.1.2 Solar Power Plant Weather Data 31 4.1.2.1 Feature Correlation 32 4.1.2.2 Augmented Dickey-Fuller Test(ADF test) 35 4.1.2.3 Autocorrelation 36 4.2 Prediction Result 37 4.2.1 Solar Power Generation Prediction 37 4.2.2 Result Comparison 39 Chapter 5 Conclusion and Future Research 43 References 44

    [1] M. H. R. Nasrin, N. A. Rahim, “Effect of high irradiation on photovoltaic power and energy,” International Journal of Energy Research, 2017.
    [2] S. N. Zhang and S. Y. Tian, “The institution of the hourly solar radiation model,” Acta Energiate Solais Sinica, vol. 18, no. 3, pp. 273–277, Mar. 1997
    [3] David Roberts, “Getting to 100% Renewable Energy Requires Cheap Energy Storage. But How Cheap?” Vox, September 20, 2019, https://www.vox.com/energy-and environment/2019/8/9/20767886/renewable-energy-storage-cost-electricity.
    [4] M. K. Behera, I. Majumder, and N. Nayak, ‘‘Solar photovoltaic power forecasting using optimized modified extreme learning machine technique,’’ Eng. Sci. Technol. Int. J., vol. 21, no. 3, pp. 428–438, 2018.
    [5] C. B. Martinez-Anido, B. Botor, A. R. Florita, C. Draxl, S. Lu, H. F. Hamann, and B.-M. Hodge, ‘‘The value of day-ahead solar power forecasting improvement,’’ Sol. Energy, vol. 129, pp. 192–203, May 2016.
    [6] S.-G. Kim, J.-Y. Jung, and K. M. Sim, ‘‘A two-step approach to solar power generation prediction based on weather data using machine learning,’’ Sustainability, vol. 11, no. 5, p. 1501, 2019.
    [7] S. Alessandrini, L. D. Monache, S. Sperati, and G. Cervone, ‘‘An analog ensemble for short-term probabilistic solar power forecast,’’ Appl. Energy, vol. 157, pp. 95–110, Nov. 2015.
    [8] Das, U. K., Tey, K. S., Seyedmahmoudian, M., Mekhilef, S., Idris, M. Y. I., Van Deventer, W., ... & Stojcevski, A. (2018). Forecasting of photovoltaic power generation and model optimization: A review. Renewable and Sustainable Energy Reviews, 81, 912-928.
    [9] Shi, J., Lee, W. J., Liu, Y., Yang, Y., & Wang, P. (2012). Forecasting power output of photovoltaic systems based on weather classification and support vector machines. IEEE Transactions on Industry Applications, 48(3), 1064-1069.
    [10] Persson, C., Bacher, P., Shiga, T., & Madsen, H. (2017). Multi-site solar power forecasting using gradient boosted regression trees. Solar Energy, 150, 423-436.
    [11] Si, Z., Yang, M., Yu, Y., & Ding, T. (2021). Photovoltaic power forecast based on satellite images considering effects of solar position. Applied Energy, 302, 117514.
    [12] Nguyen, N. Q., Bui, L. D., Van Doan, B., Sanseverino, E. R., Di Cara, D., & Nguyen, Q. D. (2021). A new method for forecasting energy output of a large-scale solar power plant based on long short-term memory networks a case study in Vietnam. Electric Power Systems Research, 199, 107427.
    [13] Wang, K., Qi, X., & Liu, H. (2019). A comparison of day-ahead photovoltaic power forecasting models based on deep learning neural network. Applied Energy, 251, 113315.
    [14] Khan, W., Walker, S., & Zeiler, W. (2022). Improved solar photovoltaic energy generation forecast using deep learning-based ensemble stacking approach. Energy, 240, 122812.
    [15] Qu, J., Qian, Z., & Pei, Y. (2021). Day-ahead hourly photovoltaic power forecasting using attention-based CNN-LSTM neural network embedded with multiple relevant and target variables prediction pattern. Energy, 232, 120996.
    [16] Gensler, A., Henze, J., Sick, B., & Raabe, N. (2016, October). Deep Learning for solar power forecasting—An approach using AutoEncoder and LSTM Neural Networks. In 2016 IEEE international conference on systems, man, and cybernetics (SMC) (pp. 002858-002865). IEEE.
    [17] P. Vasant, I. Zelinka, and G.-W. Weber, Intelligent Computing & Optimization, Proceedings of the 2nd International Conference on Intelligent Computing and Optimization 2019 (ICO 2019). 2019.
    [18] Greff, K.; Srivastava, R.; Koutnik, J.; Steunebrink, B.; Schmidhuber, J. LSTM: A Search Space Odyssey. IEEE Trans. Neural Netw. Learn. Syst. 2017, 28, 2222–2232.
    [19] Galvez, R. L., Bandala, A. A., Dadios, E. P., Vicerra, R. R. P., & Maningo, J. M. Z. (2018, October). Object detection using convolutional neural networks. In TENCON 2018-2018 IEEE Region 10 Conference (pp. 2023-2027). IEEE.
    [20] Tajbakhsh, N., Shin, J. Y., Gurudu, S. R., Hurst, R. T., Kendall, C. B., Gotway, M. B., & Liang, J. (2016). Convolutional neural networks for medical image analysis: Full training or fine tuning?. IEEE transactions on medical imaging, 35(5), 1299-1312. https://doi.org/10.1109/ TMI.2016.2535302.
    [21] Hasan, A. M., Jalab, H. A., Meziane, F., Kahtan, H., & Al-Ahmad, A. S. (2019). Combining deep and handcrafted image features for MRI brain scan classification. IEEE Access, 7, 79959-79967.
    [22] Gu, J., Wang, Z., Kuen, J., Ma, L., Shahroudy, A., Shuai, B., ... & Chen, T. (2018). Recent advances in convolutional neural networks. Pattern recognition, 77, 354-377.
    [23] Kutlu, H., & Avcı, E. (2019). A novel method for classifying liver and brain tumors using convolutional neural networks, discrete wavelet transform and long short-term memory networks. Sensors, 19(9), 1992.
    [24] Singh, D., Kumar, V., & Kaur, M. (2020). Classification of COVID-19 patients from chest CT images using multi-objective differential evolution–based convolutional neural networks. European Journal of Clinical Microbiology & Infectious Diseases, 39(7), 1379-1389.
    [25] Lundervold AS, Lundervold A. An overview of deep learning in medical imaging focusing on MRI. Z Med Phys 2019;29:102–27.
    [26] Chang P, Grinband J, Weinberg BD, Bardis M, Khy M, Cadena G, Su MY, Cha S, Filippi CG, Bota D, Baldi P, Poisson XLM, Jain XR, Chow XD. Deep-learning convolutional neural networks accurately classify genetic mutations in gliomas. Am J Neuroradiol 2018;39:1201–7.
    [27] K. Simonyan, A. Zisserman, Very Deep Convolutional Networks for Large-Scale Image Recognition, arXiv:1409.1556 [cs] (2014). ArXiv:1409.1556.
    [28] Islam, M. Z., Islam, M. M., & Asraf, A. (2020). A combined deep CNN-LSTM network for the detection of novel coronavirus (COVID-19) using X-ray images. Informatics in medicine unlocked, 20, 100412.
    [29] Yuan, Z., Jiang, Y., Li, J., & Huang, H. (2020). Hybrid-DNNs: Hybrid deep neural networks for mixed inputs. arXiv preprint arXiv:2005.08419.
    [30] Hutter, F., Kotthoff, L., & Vanschoren, J. (2019). Automated machine learning: methods, systems, challenges (p. 219). Springer Nature.

    QR CODE