Scholarlink

Fault diagnosis is crucial for the health management of aero-engines, striving to enhance flight safety, reliability, and cost-effectiveness. First, given the nonlinear interactions among sensor measurements, this paper introduces the kernel principal component analysis (KPCA) method as a means of effectively calculating main components within high-dimensional feature spaces, utilizing integral operators and nonlinear kernel functions. Then, an improved method to calculating the \({\text{T}}^{2}\) and squared prediction error (SPE) statistics is extended to the KPCA. Finally, A sensor diagnostic strategy predicated on KPCA is formulated for the engine's steady state. Simulation results demonstrate the method's efficacy.

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 Diagnosis of Aero-Engine Sensors Based on Kernel Principal Element Analysis

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Huihui Li,
  • Linfeng Gou,
  • Qiang Shen,
  • Meng Zhang

摘要

Fault diagnosis is crucial for the health management of aero-engines, striving to enhance flight safety, reliability, and cost-effectiveness. First, given the nonlinear interactions among sensor measurements, this paper introduces the kernel principal component analysis (KPCA) method as a means of effectively calculating main components within high-dimensional feature spaces, utilizing integral operators and nonlinear kernel functions. Then, an improved method to calculating the \({\text{T}}^{2}\) and squared prediction error (SPE) statistics is extended to the KPCA. Finally, A sensor diagnostic strategy predicated on KPCA is formulated for the engine's steady state. Simulation results demonstrate the method's efficacy.