| Title |
Hybrid AC/DC Microgrid Operation Method based on Artificial Neural Network of Two-Step Structure |
| Authors |
Tae-Gyu Kim ; Hoon Lee ; Chang-Gyun An ; Yoon-Seong Lee ; Kyung-Min Kang ; Junsin Yi ; Chung-Yuen Won |
| DOI |
http://doi.org/10.5207/JIEIE.2021.35.3.047 |
| Keywords |
Artificial Neural Network; Distributed Generation(DG); Hybrid AC/DC Microgrid; Interlinking Converter |
| Abstract |
This paper proposes an artificial neural network(ANN) based energy management system(EMS) for hybrid AC/DC microgrid with renewable energy source and distributed generation(DG). The operation of microgrid is divided into grid-connected mode and stand-alone mode according to whether or not the power converters are connected to the main AC grid. In grid-connected mode, the operation of energy storage system(ESS) is usually to maintain an appropriate value of state of charge(SOC). On the other hand, in stand-alone mode, the regenerative operation to the main AC grid is impossible, so additional control of ESS is required. However, in the case of small-scale distribution networks, data on power generation sources and load demands have non-linear characteristics, which makes it difficult to operate the EMS. Therefore, in this paper, ANN based EMS was applied to appropriately control the surplus and shortage of ESS power. The proposed ANN was constructed with a two-step structure to achieve high accuracy even with less repetitive learning. To implement the proposed method, a laboratory-scale interlinking converter based hybrid microgrid was constructed and its validity was verified through the experimental results. |