上海交通大学学报(自然版) ›› 2016, Vol. 50 ›› Issue (04): 534-539.

• 能源与动力工程 • 上一篇    下一篇

基于Sigma点卡尔曼滤波的燃气轮机气路故障诊断

黄宜坤,陈梅珊,张会生,翁史烈   

  1. (上海交通大学 动力机械与工程教育部重点实验室,上海 200240)
  • 收稿日期:2015-04-04 出版日期:2016-04-28 发布日期:2016-04-28

Gas Path Diagnosis Based on Sigma Point Kalman Filter of Gas Turbine

HUANG Yikun,CHEN Meishan,ZHANG Huisheng,WENG Shilie   

  1. (Key Laboratory for Power Machinery and Engineering of the Ministry of Education, Shanghai Jiaotong University, Shanghai 200240, China)
  • Received:2015-04-04 Online:2016-04-28 Published:2016-04-28

摘要: 摘要: 针对燃气机复杂的机械和电子控制系统易发生故障的问题,利用Sigma点卡尔曼滤波进行了燃气轮机的在线非线性故障诊断.首先改进了燃气轮机滤波用离散非线性模型,提高了模型精度.采用单形采样构建Sigma点卡尔曼滤波器,降低了计算量.在Simulink 平台实现了气路故障诊断系统,并进行了单故障、复合故障、渐变故障、突变故障的测试验证.结果表明:设计的诊断系统具有较高的检测、跟踪精度和故障模式、燃气轮机工况适应性.

关键词: 燃气轮机, 气路故障, 卡尔曼滤波, 在线诊断

Abstract: Abstract: In this paper, the Sigma Point Kalman Filter for gas turbine was proposed to estimate health parameters and diagnose gas turbine fault. First, a discrete nonlinear mechanism model of gas turbine was established. Then, an improved Sigma Point Kalman Filter based on simplex sampling was designed and a gas path fault diagnosis system was developed on Simulink platform. The results of verification, including the single fault, compound faults, gradual faults as well as abrupt fault, show that the diagnosis system has accurate detection, stable tracking performance and good adaptability of fault mode and gas turbine operational conditions. This paper is meaningful for achieving online gas path fault diagnosis and advanced model based control.

Key words: Key words: gas turbine, gas path diagnosis, Kalman Filter, online diagnosis

中图分类号: