Sun Closeness/Clearance Estimation for 3 Different Days

With the need for clean and sustainable energy, solar energy use has increased significantly in recent years. However, power fluctuations due to solar energy variability are among the main challenges in using photovoltaic systems. One of the most critical factors that cause these fluctuations is the clouds that block the sun’s radiation. In this study, the motions of clouds are tracked, and the motions they could take in the future are estimated. Thus, in the short term, it has been tried to determine the moments when sudden solar radiation changes may occur by estimating the closeness and clearance of the sun. For this purpose, firstly, sky images are collected at regular intervals with the help of a digital camera placed in the main campus area of Afyon Kocatepe University. The Shi-Tomasi algorithm determines the points to be tracked on the collected sky images. Then, the points determined by using the Lucas-Kanade optical flow algorithm are tracked on sequential images. By analyzing the clouds’ motions, the clouds with the risk of obstructing the sun are detected, and the motion estimations are made using the calculated direction and speed information. Finally, due to the experiments performed on datasets obtained from 3 different days, sun closeness/clearance estimations at the 1-second resolution at 120 seconds time horizon are reached. The obtained results showed that the approach could be used successfully in cloud motion estimation in different cloudy conditions.