Cloud Coverage Prediction to Improve Solar Power Management
Cloud movement prediction is a difficult problem that has been studied for many years. As humans, we can easily identify where the clouds are and maybe predict their movement within a couple of minutes; however, we are unable to predict beyond a short period of time as cloud size, altitude, wind velocity, and other weather conditions contribute to it being a very complicated task. But this research is important because cloud prediction gives advance notice to grid managers of coming solar energy shortage or spikes, allowing the system to prepare and buy energy from other sources as needed. In this paper, we continue the work of many researchers over the last decade, finding new ways to process the images and creating new neural networks to handle predictions.