2022 Fiscal Year Final Research Report
Optimal EV charging algorithm with probabilistic EV demand forecasting
Project/Area Number |
21K14150
|
Research Category |
Grant-in-Aid for Early-Career Scientists
|
Allocation Type | Multi-year Fund |
Review Section |
Basic Section 21010:Power engineering-related
|
Research Institution | University of Tsukuba (2022) Tokyo University of Science (2021) |
Principal Investigator |
|
Project Period (FY) |
2021-04-01 – 2023-03-31
|
Keywords | 電力需要予測 / EV充放電最適化 / 電気自動車 / 最適化 / 配電系統 |
Outline of Final Research Achievements |
There are two main results. The first is the construction of a theoretical framework for predicting the charging demand of a group of electric vehicles (EVs). We developed a machine learning method to generate probabilistic prediction intervals and demonstrated its effectiveness. Additionally, we confirmed that the accuracy of the forecast is critical for optimal charging behavior. The second result is the development of a theory to reduce computational time in optimal charge-discharge algorithms for a large number of EVs. We developed a method that treats multiple EVs as a single large-capacity EV in a virtual sense, enabling the calculation of charge-discharge schedules in a practical computational time. Furthermore, we constructed an optimization algorithm for charge-discharge to maximize profits from selling electricity.
|
Free Research Field |
スマートグリッド
|
Academic Significance and Societal Importance of the Research Achievements |
太陽光発電の普及が電力需給バランスの維持を困難にしている。低コストでこの問題を解決する方法として電気自動車(EV)を用いた制御が注目されているが、多数のEVの制御には不確定性の高い充電需要の予測と計算量の増加という課題がある。本研究では、これらの課題に対応するため、EVの充電需要を確率的に予測し、その誤差を考慮する新たな理論を構築した。また、充電制御の計算量を低減する方法を開発した。これらの理論は、電力系統の調整力を確保するためのEV充電制御技術の社会実装に貢献する。
|