2018 Fiscal Year Final Research Report
Development of Irradiance Forecasting Method based on Ensemble Learning in consideration of Satellite Image Information
| Project/Area Number |
16K06217
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Power engineering/Power conversion/Electric machinery
|
|---|
| Research Institution | Nagoya University |
|---|
Principal Investigator |
Kato Takeyoshi 名古屋大学, 未来材料・システム研究所, 教授 (90283465)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
真鍋 勇介 名古屋大学, 未来材料・システム研究所, 寄附研究部門助教 (30751143)
|
|---|
| Research Collaborator |
ENOMOTO Jyunya
KAI Naoto
|
|---|
| Project Period (FY) |
2016-04-01 – 2019-03-31
|
| Keywords | 太陽光発電 / 出力予測 / 日射予測 / 数値気象モデル / 衛星画像 / 機械学習 |
|---|
| Outline of Final Research Achievements |
The accuracy improvement of a day-ahead forecast of irradiance by the combination of GPV(MSM) based method and WRF based method was investigated. It is found that can be tuned by a certain combination of physics models related to irradiance can be chosen so that the forecast error of WRF based model tends to be positive while the forecast error of GPV(MSM) based model tends to be negative in Summer season. Besides, the post processing based on machine learning should not be applied to WRF for the improvement of forecast accuracy of combination method. On the other hand, the accuracy of each model should be improved and combined each other for the accuracy improvement in other seasons. Although it is revealed that the forecast accuracy of GPV(MSM) based model can be improved by using satellite imagery, the contribution of satellite imagery to the accuracy improvement of combination method was not investigated.
|
| Free Research Field |
電力工学
|
| Academic Significance and Societal Importance of the Research Achievements |
太陽光発電の導入は急速に進んでおり，日々の電力需給運用において，その出力予測誤差の影響が顕在化しつつあり，予測精度の向上は急務である。複数の手法の組み合わせによる予測精度の向上については既に適用されているが，本研究では，個々の手法は必ずしも高くないものの，組み合わせた際の予測精度が高くなるように個々の手法を調整できることを明らかにしており，予測精度の向上のための一つの方向性を示すことができたという点で学術的な意義は高い。今回は十分に検討できなかったが，衛星画像情報を組み合わせて更なる精度向上が期待でき，低炭素社会の構築に向け，太陽光発電の導入拡大に資することが期待される。
|
