Imbalanced sample fault diagnosis method for rotating machinery in nuclear power plants based on deep convolutional conditional generative adversarial network

• DOI: 10.1016/j.net.2023.02.036
• Publication Date: 2023-03-03T07:38:08Z
• Authors: Zhichao Wang; Hong Xia; Jiyu Zhang; Bo Yang; Wenzhe Yin
• Keywords: Sample (material)

Rotating machinery is widely applied in important equipment of nuclear power plants (NPPs), such as pumps and valves. The research on intelligent fault diagnosis rotating crucial to ensure the safe operation related NPPs. However, practical applications, data-driven faces problem small imbalanced samples, resulting low model training efficiency poor generalization performance. Therefore, a deep convolutional conditional generative adversarial network (DCCGAN) constructed mitigate impact samples diagnosis. First, designed based neural networks effectively augment samples. original sample features can be extracted by strategy appropriate number filters. In addition, high-quality generated are ensured through visualization process features. Then, (DCNN) extract mixed implement Finally, multi-fault experimental data motor bearing, performance DCCGAN for augmentation verified. proposed method alleviates shows its application value actual