A Measure of Credibility of Solar Power Prediction

Recently, remarkable developments of new energy technologies have been achieved against various energy problems. Photovoltaic (PV) system, one of such technologies, has an advantage of utilizing infinite and clean energy. On the contrary, it also has a disadvantage of unreliable power supply mainly caused by unstable weather. The fluctuation of the power supply of PV systems are considerably large because of rapid insulation changes and rapid weather changes, and in some cases, it seems impossible to realize high-accuracy prediction even with sophisticated prediction models. In this paper, using recently proposed estimator for the Shannon information content, a method to output a measure of credibility for prediction is proposed. With the proposed method, it is possible to judge whether the energy supply at a certain future time is unpredictably fluctuate compared to the current value or not, and it is possible to take measures against the rapid change of solar energy generation in advance. From an experimental result using solar energy supply data, we see that the proposed measure of credibility reflects the difficulty of predicting solar energy supply.