小波分析在管道泄漏信号识别中的应用

利用多尺度小波变换 ,把管道泄漏产生的负压波信号作为瞬态信号 ,来识别管道的局部泄漏特征。以光滑函数的一阶导数作为小波母函数 ,研究了管道泄漏特征信号拐点区间的敏感性 ,突出小波变换系数的局部极值性。分析表明 ,检测信号的小波变换系数极值的奇异性准确地反映了管道检测信号的泄漏特征 ,并且从局部描述了管道泄漏信号的瞬态正则性。对各级尺度系数进行了S形曲线拟合 ,此曲线能够完整地描述管道泄漏瞬变特征 ,其拐点区间描述了管道发生泄漏时的瞬变过程。

Application of wavelet transform to signal detection of pipeline leakage

Taking the negative pressure wave produced by pipeline leakage as a transient signal, the multi scale wavelet transform can be utilized to identify the pipeline leakage signal in local. The differential coefficient of smooth function was taken as a mother wavelet for researching the sensitivity segment of inflexion of pipeline leakage signal. The local extremum characters of the coefficient of wavelet transform were analyzed. The analytical results show that the oddity of wavelet transform coefficient extremum can reflect accurately the characters of leakage signal and depict the transient regularity of leakage signal in local. The Sigmoid curve fitting to all levels of wavelet coefficients was made. The matched curve can display the characters of transient process of leakage.