Cloud optical depth retrieval via sky's infrared image for solar radiation prediction

• Main Authors: Yee, Lai Kok, Ken, Tan Lit, Asako, Yutaka, Quen, Lee Kee, Liang, Chuan Zun, Syahidah, Wan Nur, Homma, Koji, Arada, Gerald Pacaba, Siang, Gan Yee, Yen, Tey Wah, Sing, Calvin Kong Leng, Kamadinata, Jane Oktavia, Taguchi, Akira
• Format: text
• Published: Animo Repository 2019
• Subjects: Solar radiation—Forecasting; Infrared imaging; Neural networks (Computer science); Electrical and Computer Engineering; Electrical and Electronics; Systems and Communications
• Online Access: https://animorepository.dlsu.edu.ph/faculty_research/1900
• Institution: De La Salle University

Photovoltaic (PV) system is developed to harness solar energy as an alternative energy to reduce the dependency on fossil fuel energy. However, the output of the PV system is not stable due to the fluctuation of solar radiation. Hence, solar radiation prediction in advanced is needed to make sure the tap changer in PV system has enough time to respond. In this research, the cloud base temperature is identified from the sky's thermal image. From the cloud base temperature, cloud optical depth (COD) is calculated. Artificial neural network (ANN) models are established by using different combinations of current solar radiation and COD to predict the solar radiation several minutes in advanced. R-squared value is used to measure the accuracy of the models. For prediction in advanced for every minute, with COD as input, always show the highest R-squared value. The highest R-squared value is 0.8899 for the prediction for 1 minute in advanced and dropped to 0.5415 as the minute of prediction in advanced increase to 5. This shows that the proposed methodology is suitable for prediction of solar radiation for short term in advanced. © 2019 Penerbit Akademia Baru.