Scholarlink

Evaporative cooling technology is an effective method for solving the heat dissipation of megawatt level offshore wind turbine. At present, the fault detection and diagnosis of evaporative cooling system is still carried out in an empirical way, which is far from meeting the requirements of safety and reliability. In this paper, a principal component analysis fault detection method based on PCA is proposed. Eigenvalue decomposition is performed by PCA, score matrix is constructed, and T2 statistics and SPE statistics are established in principal element subspace and residual subspace respectively. The control limit of statistics is determined by setting significance level. When the real-time data exceeds the control limit, the system determines that there is a fault and issues an early warning. The method is applied to the experiment, and the result shows that the method can give the fault diagnosis result accurately, and has certain engineering application value.

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

Fault Detection of Stator Evaporative Cooling System for Large-Scale Offshore Wind Turbines Based on Principal Component Analysis

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Yaxing Kang,
  • Lei Li,
  • Bin Wang,
  • Wenbiao Hu,
  • Pengfei Wang,
  • Haifeng Wang,
  • Shun Ye

摘要

Evaporative cooling technology is an effective method for solving the heat dissipation of megawatt level offshore wind turbine. At present, the fault detection and diagnosis of evaporative cooling system is still carried out in an empirical way, which is far from meeting the requirements of safety and reliability. In this paper, a principal component analysis fault detection method based on PCA is proposed. Eigenvalue decomposition is performed by PCA, score matrix is constructed, and T2 statistics and SPE statistics are established in principal element subspace and residual subspace respectively. The control limit of statistics is determined by setting significance level. When the real-time data exceeds the control limit, the system determines that there is a fault and issues an early warning. The method is applied to the experiment, and the result shows that the method can give the fault diagnosis result accurately, and has certain engineering application value.