3D cloud detection and tracking for solar forecast using multiple sky imagers

Zhenzhou Peng, Shinjae Yoo, Dantong Yu, Dong Huang, Paul Kalb, John Heiser

Abstract: Cloud detection and tracking (CDT) is the most challenging problem in integrating solar energy into the smart grid. In this paper, we present a novel 3D cloud detection and tracking using images from three TSI (Total Sky Imager), and propose to incorporate history into a multi-layer cloud detection pipeline. Our pilot study shows that the new CDT significantly improves the short-term solar irradiance forecasting and enable regional radiation prediction, which is impossible with a single TSI.