Scholarlink

Modular multilevel converter (MMC) is commonly applied in flexible DC transmission and high voltage motor drives. However, there are many power switches in MMC which lead to a high probability of switches failure. Rapid fault diagnosis is the key to the safe and reliable operation of MMC. In order to overcome the defect that traditional MMC fault diagnosis only deals with fault text data, MMC topology is constructed for the first time. A fault diagnosis method based on Graph convolutional network (GCN) for open-circuit fault (OCF) in MMC is proposed. Combined with MMC topology, GCN node classification is used for OCF diagnosis. The simulation results show that the proposed method could effectively diagnose OCF and improve diagnostic accuracy.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Switch Open-Circuit Fault Diagnosis for Modular Multilevel Converter via GCN with Strong Scalability

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Haoran Wang,
  • Tianzhen Wang,
  • Fan Zhang

摘要

Modular multilevel converter (MMC) is commonly applied in flexible DC transmission and high voltage motor drives. However, there are many power switches in MMC which lead to a high probability of switches failure. Rapid fault diagnosis is the key to the safe and reliable operation of MMC. In order to overcome the defect that traditional MMC fault diagnosis only deals with fault text data, MMC topology is constructed for the first time. A fault diagnosis method based on Graph convolutional network (GCN) for open-circuit fault (OCF) in MMC is proposed. Combined with MMC topology, GCN node classification is used for OCF diagnosis. The simulation results show that the proposed method could effectively diagnose OCF and improve diagnostic accuracy.