TY - GEN
T1 - An Intermittent Fault Detection Method for Three- phase PFC Converters Using EMD-CNN
AU - Liu, Cuiyu
AU - Yang, Zhiming
AU - Xiang, Gang
AU - Yu, Yang
AU - Tian, Jijia
N1 - Publisher Copyright:
© 2022 IEEE.
PY - 2022
Y1 - 2022
N2 - Three-phase PFC converters are highly preferred due to their high power factor. Due to the importance of the 3-phase converters in industry, the need to insure a continuous and safety operation for these power converters is essential. However, the intermittent fault may occur when the 3-phase PFC converters worked for a long time. In order to effectively recognize the intermittent faults for 3-phase PFC converters, an intermittent fault detection method is proposed. The method is a combination of empirical mode decomposition (EMD) and convolutional neural network (CNN). First, the output voltage was decomposed into IMF components. Then the statistical features are extracted from the IMF which is most correlated to the output voltage. Finally, the CNN is used to detect the fault. The proposed fault detection method extracted the features using EMD, so the intermittent fault which lasts for a short time can be detected. Timely intermittent fault detection can avoid further losses. Simulation experimental results validate the practicability and effectiveness of the proposed method.
AB - Three-phase PFC converters are highly preferred due to their high power factor. Due to the importance of the 3-phase converters in industry, the need to insure a continuous and safety operation for these power converters is essential. However, the intermittent fault may occur when the 3-phase PFC converters worked for a long time. In order to effectively recognize the intermittent faults for 3-phase PFC converters, an intermittent fault detection method is proposed. The method is a combination of empirical mode decomposition (EMD) and convolutional neural network (CNN). First, the output voltage was decomposed into IMF components. Then the statistical features are extracted from the IMF which is most correlated to the output voltage. Finally, the CNN is used to detect the fault. The proposed fault detection method extracted the features using EMD, so the intermittent fault which lasts for a short time can be detected. Timely intermittent fault detection can avoid further losses. Simulation experimental results validate the practicability and effectiveness of the proposed method.
KW - 3-phase power factor correction (PFC) converter
KW - convolutional neural network (CNN)
KW - empirical mode decomposition (EMD)
KW - intermittent fault detection
UR - https://www.scopus.com/pages/publications/85143176321
U2 - 10.1109/PHM-Yantai55411.2022.9941904
DO - 10.1109/PHM-Yantai55411.2022.9941904
M3 - 会议稿件
AN - SCOPUS:85143176321
T3 - 2022 Global Reliability and Prognostics and Health Management Conference, PHM-Yantai 2022
BT - 2022 Global Reliability and Prognostics and Health Management Conference, PHM-Yantai 2022
A2 - Guo, Wei
A2 - Li, Steven
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2022 Global Reliability and Prognostics and Health Management Conference, PHM-Yantai 2022
Y2 - 13 October 2022 through 16 October 2022
ER -