Scholarlink

<p>The susceptibility of vibration signals to combustion chamber faults of marine diesel engine varies across different measurement points, increasing difficulty of fault diagnosis. The different fault experiments in the combustion chamber are conducted in this paper, vibration data from different positions are collected. Correlation analysis method with information entropy is studies, the correlations of vibration signal characteristics between different faults at the same point and the same fault at different points are analyzed, and the selection of key points with high accuracy for fault diagnosis is optimized. The one dimensional convolutional neural network is used for combustion chamber faults diagnosis based on the data from optimized points. The study shows that the method can optimally select measurement points with high fault feature discrimination and sensitivity, the accuracy rates of the early fault classification and health status recognition of the combustion chamber reach at least 96.67 % and 97.35 % respectively.</p>

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

Research on sensitivity of measurement points and diagnostic method of combustion chamber health status for marine diesel engine

  • 期刊论文

  • 发表日期：

    2025年5月6日

查看全文
  • Lei Hu,
  • Wentong Wang,
  • Xuekuan Yu,
  • Yonghua Yu,
  • Jia Hu,
  • Bingjie Ma,
  • Jianguo Yang

摘要

The susceptibility of vibration signals to combustion chamber faults of marine diesel engine varies across different measurement points, increasing difficulty of fault diagnosis. The different fault experiments in the combustion chamber are conducted in this paper, vibration data from different positions are collected. Correlation analysis method with information entropy is studies, the correlations of vibration signal characteristics between different faults at the same point and the same fault at different points are analyzed, and the selection of key points with high accuracy for fault diagnosis is optimized. The one dimensional convolutional neural network is used for combustion chamber faults diagnosis based on the data from optimized points. The study shows that the method can optimally select measurement points with high fault feature discrimination and sensitivity, the accuracy rates of the early fault classification and health status recognition of the combustion chamber reach at least 96.67 % and 97.35 % respectively.