Scholarlink

The vibration signal of the transformer core is closely related to its mechanical state, and exploiting this characteristic enables the detection of loosening points in the transformer core. This paper proposes a method for detecting loosening points in the transformer core based on Markov Transition Field (MTF) and Convolutional Neural Network (CNN). Firstly, the initial vibration signal of the transformer core is decomposed using Ensemble Empirical Mode Decomposition (EEMD), and suitable components are selected for signal reconstruction to eliminate interference from the transformer itself and the external environment. Subsequently, the reconstructed signal's MTF image is obtained through MTF transformation. Finally, the MTF image is utilized as the input for CNN to accomplish the detection of loosening points in the transformer core.

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

Detection Method for Loose Points in Transformer Iron Cores Based on MTF and CNN

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Weijie Xu,
  • De Ding,
  • Changgeng Zhang

摘要

The vibration signal of the transformer core is closely related to its mechanical state, and exploiting this characteristic enables the detection of loosening points in the transformer core. This paper proposes a method for detecting loosening points in the transformer core based on Markov Transition Field (MTF) and Convolutional Neural Network (CNN). Firstly, the initial vibration signal of the transformer core is decomposed using Ensemble Empirical Mode Decomposition (EEMD), and suitable components are selected for signal reconstruction to eliminate interference from the transformer itself and the external environment. Subsequently, the reconstructed signal's MTF image is obtained through MTF transformation. Finally, the MTF image is utilized as the input for CNN to accomplish the detection of loosening points in the transformer core.