Search documents

Browse topics

Document details

Development of the Regional Photovoltaic Power Forecasting Using Irradiation Data from a Few Meteorological Stations
T. Oozeki, J.G.S. Gari da Silva Fonseca Jr., H. Ohtake, K.-I. Shimose, T. Takashima, K. Ogimoto
Prediction, Numerical Weather Prediction, Power Forecast, Regional Forecasting, Power System Operation
Subtopic: PV System Reliability and Availability
Event: 28th European Photovoltaic Solar Energy Conference and Exhibition
Session: 5BV.4.26
3928 - 3930
ISBN: 3-936338-33-7
Paper DOI: 10.4229/28thEUPVSEC2013-5BV.4.26
0,00 EUR
Document(s): paper


High penetration of variable renewable generation systems such as photovoltaic (PV) systems will cause the issue of the supply-demand imbalance in a whole power system as well as voltage and power flow fluctuation in local distribution systems. In the case, the output forecasting of PV system is one of the most important technologies in order to make balance the supply-demand in the power system. The forecasting on the point is not only important, but also the regional forecasting is much important in the case of disseminated large amount of PV systems in the future. In this paper, we proposed regional forecasting using up-scaling method with a few irradiation data. The monitoring data of PV systems in Japan are not enough to use up-scaling method such as SMA does in Germany. However, some meteorological stations have measured horizontal irradiation data in Japan. Therefore, the up-scaling method using only few irradiation data are required for regional forecasting in Japan. We evaluated the impact on the information of azimuth angle and tilt angle for output forecasting because converting technologies from horizontal to tilt angle irradiation are depend on such PV system specification information. The fundamental forecasting method is based on the support vector machine and Japanese Grit point value from NWP. As a results, forecasting error of proposed method were almost same results of the forecasting error which used direct forecasting method of PV system output power for all of PV systems.