基於超像素分割之衛星雲圖進行預測與估計日射量之系統

A system for predicting and estimating solar irradiance based on satellite cloud image with superpixel segmentation

郭家宏(Guo Jia-Hong)

指導教授 ： 呂藝光

國立臺灣師範大學/科技與工程學院/電機工程學系/碩士(2022年)

https://doi.org/10.6345/NTNU202201314

摘要

由於日射量容易受天氣因素影響而容易產生變化，進而造成太陽能發電量不穩定，因此，難以將其整合入區域電網當中。本文建立一個以超像素分割衛星雲圖為基礎之日射量估計與預測系統。分析衛星雲圖並萃取其雲層特徵，採用光流法，分析雲層運動，生成預測的衛星雲圖。再將這些影像特徵與一些天氣預報特徵作為長短期記憶(LSTM)之輸入，進行日射量的估計與預測。本文使用幾個效能指標來評估估計與預測的效果，包括平均絕對誤差(MAE)、均方根誤差(RMSE)以及判定係數(R^2)；並設計數個實驗方法進行比較，實驗結果顯示，本文所提出方法有達到預期的成果。

關鍵字

日射量 ； 衛星雲圖 ； 超像素分割 ； 光流法 ； 長短期記憶

並列摘要

Since solar radiation is susceptible to changes due to weather factors, it is difficult to integrate it into the regional power grid because of the instability of solar power generation. In this study, a solar irradiance estimation system based on satellite cloud image superpixel segmentation was established. The satellite cloud image is analyzed, the cloud features are extracted, the satellite cloud image is used as the input, the cloud layer movement is analyzed by the optical flow method, and the predicted satellite cloud image is generated. These features are used as input to long short-term memory (LSTM) to estimate and predict solar irradiance. Several performance metrics are used to evaluate the estimation and prediction, including MAE, RMSE, and R2. Several methods are compared, and the experimental results show that the proposed method performs better.

並列關鍵字

solar irradiance ； satellite cloud image ； super pixel segmentation ； optical flow ； LSTM

參考文獻

[1] T. Hsiao, K. Huang, and Y. Chang, "A model for forecasting of solar power annual electricity and its strategy application," Journal of Taiwan Energy, vol. 4, no. 4, pp. 401-430, 2017.

[2] M. Q. Raza, M. Nadarajah, and C. Ekanayake, "On recent advances in PV output power forecast," Solar Energy, vol. 136, pp. 125-144, 2016.

[3] J. X. Yeo, Y. H. Lee, and J. T. Ong, "Performance of site diversity investigated through RADAR derived results," IEEE Transactions on Antennas and Propagation, vol. 59, no. 10, pp. 3890-3898, 2011.

[4] M. Z. Hassan, M. E. K. Ali, A. S. Ali, and J. Kumar, "Forecasting day-ahead solar radiation using machine learning approach," in 2017 4th Asia-Pacific World Congress on Computer Science and Engineering (APWC on CSE), 2017: IEEE, pp. 252-258.

[5] X. Shao, S. Lu, and H. F. Hamann, "Solar radiation forecast with machine learning," in 2016 23rd International Workshop on Active-Matrix Flatpanel Displays and Devices (AM-FPD), 2016: IEEE, pp. 19-22.