高噪声背景下故障信号自适应检测

Fault Signal Adaptive Detection with High Noisy Environment

为尽早发现电厂锅炉潜在故障,以研究高噪音背景下故障自适应检测系统为途径,提出了高噪音背景下故障信号检测的算法。传统的倒谱分析是对信号基于FFT(Fast Fourier Transformation)变换,而该算法对其进行改进,基于CZT(Chirp Z-Transform)变换进行求逆,得到的故障特征曲线更稳定、更可靠。通过实验证明,该计算方法快速有效,故障报出的正确率在99%以上。

For early detection of boilers potential failure in power plant the high noise background fault self-adaptive detection system is studied.A higher background noise fault signal detection algorithm is provided.Simulation experiment results,and points out that the efficiency of the method.The traditional cepstrum analysis is applied to signal based on FFT（Fast Fourier Transformation） transform,and the algorithm is based on the improvement CZT（Chirp Z-inverse Transform） transform,get the failure characteristics of curve more stable,and more reliable.The experimental results show that this method is fast and effective,fault quoted accuracy over 99%.