7　Investigation of a method to prorate average solar radiation values to finer temporal granularity using JIT Modeling

Abstract

The use of solar radiation forecast information is essential for the operation of PV systems. Currently, three-hourly averages of solar radiation predictions are available from the Japan Meteorological Agency's MEPS (Meso-scale Ensemble Prediction System). For more detailed energy management, one-hour values with finer temporal granularity are effective.

In this paper, we propose a new method to estimate one-hour values from three-hour MEPS averages by machine learning based on JIT Modeling. The validity of the proposed method is demonstrated by comparing it with a simplified method that uses the prorated amount of exoatmospheric solar radiation.