| Title |
DCGAN based Event Detection Scheme Using D-PMU Data in Distribution Systems |
| Authors |
양준혁(June-Hyuck Yang) ; 김태근(Tae-Geun Kim) ; 윤성국(Sung-Guk Yoon) |
| DOI |
https://doi.org/10.5370/KIEE.2022.71.4.555 |
| Keywords |
distribution systems; distribution-phasor measurement units (D-PMU); event detection; machine learning; deep convolutional generative adversarial networks (DCGAN); power quality |
| Abstract |
Distribution-phasor measurement units (D-PMUs) measuring magnitude and phasor angle with high resolution make detailed observations of the distribution system. In this paper, we propose a deep convolutional generative adversarial networks (DCGAN) base event detection method using D-PMU data. GAN is trained through the adversarial process of two models: generator and discriminator. This process helps the discriminator train well without much training data. Also, DCGAN has convolutional layers for better event recognition. After training the proposed DCGAN model using labeled D-PMU data, we use the discriminator to identify distribution system events. The target events to detect are voltage dip, over-voltage, harmonic, and transient. Through a case study with real data from two D-PMUs installed at Soongsil university, the detection performance of the proposed detection method is verified. It is confirmed that the proposed method shows a good detection performance compared to other schemes. |