超声波测距信号小波阈值去噪参数的选定方法  

The Parameter Selection Method of Wavelet Threshold Denoising in Ultrasonic Ranging Signal Process

作　　者：陈文会[1] 丁晓鸿[1] 陈江宁 刘小民 

机构地区：[1]西北工业大学电子信息学院,西安710129 [2]陕西脉航交通测控有限公司,西安710061

出　　处：《传感技术学报》2017年第3期407-411,共5页Chinese Journal of Sensors and Actuators

基　　金：2015年江苏省科委基金项目(0800/N2015KE0006)

摘　　要：讨论了超声波测距信号小波阈值去噪参数的选定方法和最优选取原则。在超声波测距回波信号中加入10 d B的噪声信号,在MATLAB仿真环境下,用硬阈值函数、软阈值函数、半软阈值函数,分别对超声波回波信号进行处理,通过实验分析得出,在选取2层分解、消失矩阶数为5时,选择Sym7小波基、半软阈值函数降噪时,信号SNR改善较大,为5.35 d B,RMSE最小,为0.094 d B,说明在有效去除信号噪声的同时,重构后的失真最小。Discuss the parameter selection and optimal by researching the de-noising problem of wavelet transform in ultrasonic echo signal. The detailed research about the parameters optimal including how to select the wavelet base, decomposition layers, threshold value and threshold function were fully discussed. Under the MATLAB environment, the hard threshold function, soft threshold function and semi-soft threshold function are utilized to deal with the ultrasonic echo signal of SNR 10 dB respectively. Experimental result shows that the semi-soft threshold function can reach SNR 5.35 dB and RMSE 0.094d B.

关 键 词：超声回波信号 小波去噪 参数选定 

分 类 号：TV698.1[水利工程—水利水电工程]