| Autor: |
Linjun Shi, Tao Dai, Wenjie Lao, Feng Wu, Keman Lin, Kwang Y. Lee |
| Jazyk: |
angličtina |
| Rok vydání: |
2023 |
| Předmět: |
|
| Zdroj: |
IEEE Access, Vol 11, Pp 74191-74200 (2023) |
| Druh dokumentu: |
article |
| ISSN: |
2169-3536 |
| DOI: |
10.1109/ACCESS.2023.3296533 |
| Popis: |
Quickly recognizing the real-time operating states will be helpful to identify the instantaneous and permanent power loss of the renewable energy station, so as to realize the continuous operation under the influence of the instantaneous disturbances caused by faults. This paper proposes a state recognition method for renewable energy units based on sparse stacked auto-encoder (SSAE) feature extraction and improved k-nearest neighbor (KNN) algorithm. The characteristics of this method is that the electrical parameters of the unit port are collected directly without relying on the unit’s supervisory control and data acquisition (SCADA) system, whose acquisition speed is too slow to meet the recognition accuracy requirement, and that the unit operation states can be recognized quickly and accurately. Firstly, operation states of renewable energy unit are divided, and the framework for the unit’s state recognition is proposed. Moreover, improved strategies for state recognition of renewable energy unit are proposed. Finally, the power system analysis software package (PSASP) is used to obtain the electrical parameters of renewable energy units and the improved KNN algorithm is used to recognize operation states after extracting features based on SSAE. By comparing the method proposed with the traditional KNN algorithm, the effect of the proposed method for states recognition is shown to be the best, with an accuracy of 98.16% and computing time of 50ms. The results show the validity of the proposed method. |
| Databáze: |
Directory of Open Access Journals |
| Externí odkaz: |
|
