.png){kind=icon}
.png){kind=icon}
Return
Data-driven power system security assessment using high content database during the COVID-19 pandemic
.png){kind=icon}
PRE
AI
DOI：10.1016/j.ijepes.2023.109077.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
As the coronavirus disease (COVID-19) broke out in late 2019, the electricity sector was significantly impacted. Hence, the effects of the pandemic and restricting measures in power system operation are investigated during pandemic circumstances. The secure operation of the power system is a fundamental requirement. Appropriate procedures should be taken to mitigate these effects and ensure the power system's security. Accordingly, in this study, the authors determine that the COVID-19 pandemic can change the system's operating conditions in the first stage. Since data-driven security assessment methods require the training database to learn about Security constraints, this paper proposes an efficient database generation strategy respecting the consequences of the COVID-19 outbreak. The proposed strategy provides a training set with high information content compatible with the operating conditions. To this end, the method consists of a characteristics extraction approach and updating scheme. The characteristics should be extracted to represent the operating conditions of the system. Further, the similarity of intervals is compared using characteristics in updating scheme. The copula-based sampling approach is provided to generate the random samples. The proposed strategy generates a database for data-driven methods. Therefore, it can be utilized in various applications of security assessment. Real-world data is mapped to the IEEE 39-bus system to illustrate the framework efficiency. The outcomes indicate that a classification using the proposed strategy outperforms conventional methods in terms of evaluation metrics. (c) 2017 Elsevier Inc. All rights reserved.
Keywords:
Security assessment
Decision tree
COVID-19
Copula functions
Machine learning
Journal
I
IF:
5
Papers:
1.1W
Citations:
3.1W
View more organizations