Cloud motion tracking system using low-cost sky imager for PV power ramp-rate control

Abstract

Passing cloud results in rapid changes in solar irradiance. The intermittency of PV power output has drawn serious concern as the PV system installation increases significantly. Consequently, power ramp-rate control (PRRC) is introduced as the regulation to avoid significant power fluctuations. These requirements are driving an increasing demand for short-term PV power forecasting. Sky imager has been used as an effective tool to predict the cloud motion, then to forecast the PV power. However, the high cost of sky imager system and long image processing delay are still hindering its application in PRRC. In this paper, a low-cost cloud motion tracking system has been designed and developed. Ultra short-term cloud motion forecasting has been achieved in sub-minute level which can be used in PRRC application. The proposed method improves the forecasting accuracy by multiple cloud centroids tracking. The effectiveness of the proposed method has been verified by the practical experiment results.