管道阀门泄漏信号仿真分析

Simulation Analysis of Leak Signal of Pipeline Valve

研究管道阀门泄漏检测的技术,针对复杂背景噪声影响,管道泄漏信号不能准确识别及时频特性精确提取。提出了一种基于小波包和时频能量分布的管道阀门声发射泄漏检测的方法,利用小波包重构还原管道阀门泄漏声发射信号,通过计算Choi-Williams时频分布,提取时频特征,实现了泄漏信号识别和时频特性分析。采用PVC管道搭建了管道阀门泄漏声发射检测的实验模型,泄漏状态通过阀门开启度调节。在严重的背景噪声环境下,实时采集泄漏信号,并进行仿真。结果表明,采用小波包和时频能量分布的分析方法可以准确识别严重背景噪声下的泄漏信号,证明了该方法的可行性。

The technical difficult points of pipeline valve leakage detection are the identification of the leak signals and the extracting of its time-frequency property precisely.A method for acoustic emission signal detection of pipeline valve leakage based on wavelet packet and time-frequency energy distribution is presented.The real leak signals were restored by wavelet packet reconstruction,and then the time-frequency characteristics were extracted by calculating the Choi-Williams time-frequency energy distribution,realizing the leakage signal identification and time-frequency analysis.On the basis of theoretical analysis,a detection model to simulate the acoustic emission pipeline leak was constructed,and the leakage state was adjusted by the opening of valve to collect the leak signal in serious noisy environment.Compared with the conventional analysis of wavelet packet energy distribution ratio and power spectrum analysis method,the method based on wavelet packet and time-frequency energy distribution can identify the leak signal accurately and verify the feasibility of this method.