基于小波分析的燃机排气温度传感器组故障诊断方法研究

Research of fault diagnosis of sensor group based on wavelet analysis

为提高燃机排气温度传感器输出信号的准确性,确保燃机系统安全稳定运行,本文研究了小波分析在传感器故障诊断中的应用。通过检测小波变换的模极大值是否超限判断传感器是否发生故障;分析故障信号的奇异度,求取信号的李氏指数,对故障进行辩识;在缓变信号的辩识中,又借助相关性分析方法进一步寻找信号缓变的真正原因。对温度传感器的故障半物理模拟实验表明了小波分析诊断传感器故障的有效性。

To improve the precision of signals from temperature sensors, and insure the gas turbine running safely and stably, the application of wavelet analysis in fault diagnosis of sensors is studied in this article. From detecting whether the maximum of wavelet coefficient overrun or not, we can judge whether there is something wrong with the sensor. By cumulating the Lipschitz＇ s exponential of fault signal, the type of the fault can be distinguished. Using the method of correlation analysis , the true fact of incipient faults can be found. Fault simulation studies of temperatuer sensors indicate the effectiveness of the proposed method.