相似差值结合小波分析的行波法类反射波抑制研究

Research on the Inhibition of Interferences and Noises Combined Difference with Wavelet Analysis

在电力生产和运行过程中,电线或电缆的精确测长及参数突变点定位,对于电力施工、电缆故障检测、高阻接地故障诊断以及电缆的生产管理等都有着重要意义.而对于脉冲反射法来说,准确找到反射脉冲起点是需要解决的最关键问题.通过对电缆多次、多种情况测量结果的比对分析发现,电线或电缆在同一测试环境下,相邻两次测量结果中噪声及反射波具有极大的相似性,对此,采用相似差值处理可以有效消除大部分干扰和噪声.在此基础上,采用小波软阈值对小波分解系数进行量化处理.对于其中最重要的部分即阈值的确定,采用基于史坦(Stein)的无偏似然估计原理进行阈值选择,得到最优的阈值.仿真和实测表明,相似差值处理结合小波软阈值去噪的新方法,能够有效突出反射信号、消除噪声,尤其是对于影响最大的类反射波的干扰信号有着极强的抑制作用,进而为提高反射脉冲起始点的查找精确度,以及相关仪器的研制提供了理论依据.

In power production and operation process, the accurate length measurement and the mutation point positioning of paramdters of the wire or cable are all of great significance for the construction of electricity, cable fault detection, high impedance ground fault diagnosis and production management of the cable. For the pulse reflection method, the most critical issues need to be resolved is finding the reflect pulse starting point accurately. Under different circumstances, measuring the cables repeatedly and comparing the measurement results found that noise and the reflected wave of adjacent measurement results had a great similarity in the same test environment. Similar difference was proposed to eliminate most of interferences and noises. Wavelet soft threshold was used to quantified wavelet coefficients and Stein unbiased risk estimate was used to select threshold. Simulation and test prove that the new method which is combined similar difference with wavelet soft threshold denoising method not only eliminates the noises effectively but also inhibits the waveform which is similar to reflected wave largely.