Cloud tracking and forecasting method based on optimization model for PV power forecasting

The tracking and forecasting of cloud is very important for ultra-short term photovoltaic power forecast with sky images. To recognize clouds deformation and tracking clouds motion in sky images, an optimization model is constructed in this paper. In a binary sky image, the shape and position of a cloud can be represented by a set of coordinates of cloud border points, and the deformation and motion process can be characterized by the displacement of these point. The motion matrix corresponding to the points´ displacement is obtained by solving the optimization model. Then the shape and position of cloud in the image in the next moment are forecasted by linear extrapolation. The effectiveness of the proposed method is validated by experiment.