2020 Volume 46 Issue 4 Pages 71-77
For photovoltaic power generation prediction, it is necessary to calculate a cloud shadow vector before the arrival of a cloud shadow. The cloud shadow vector is calculated based on the difference of timing for the signal intensity changes of solar radiation sensors arranged on the ground. In this study, signal intensity changes and cloud shadow shapes are measured using some solar radiation sensors and a camera, and the change patterns of the sensor signals are analyzed. The determination value of the sensor signal for the shadow-in and shadow-out of a cloud is proposed based on the change pattern of the signals. The change patterns of sensor signals were categorized into three types. The calculation accuracy of the cloud shadow vector improved by using the 5-min moving-average value of measured solar radiation intensity as the determination value for the shadow-in and shadow-out of a cloud compared with the theoretical solar radiation intensity.