节点文献

路用地质雷达回波信号处理方法研究

Signal Processing of GPR Used on Road Only

【作者】 杨艳锋

【导师】 郭元术;

【作者基本信息】 长安大学 , 交通信息工程及控制, 2008, 硕士

【摘要】 路用地质雷达是近些年发展起来的一种无损检测设备,与传统公路检测方法相比,路用地质雷达具有不损坏路面,探测速度快,探测过程连续,分辨率高,操作方便灵活,探测费用低,探测范围广等优点。因此,路用地质雷达在国内外的公路工程中已经得到广泛的应用。然而,与地质雷达设备的较快发展相比,路用地质雷达回波信号处理算法发展缓慢,这严重影响了路用地质雷达的推广应用,所以,对路用地质雷达回波信号处理方法进行研究具有重大的现实意义。本文首先分析了电介质中电磁波的传播规律,讨论了地质雷达的基本原理及其体制,创建了获取路用地质雷达原始数据的方法。其次,提出了适合于路用地质雷达信号处理的降噪方法(小波变换与K-L变换相结合降噪法);研究了雷达回波信号的瞬时频率提取方法;应用时域有限差分算法,正演模拟了理想介质条件下不同形状目标体的回波图像。然后,以基于时频分析的瞬时频率提取方法和正演模拟图像为基础,建立了基于回波图像特征的目标空间位置定位方法,和基于回波信号时频特性的目标分类识别方法。最后,本文利用实测数据进行了实验,实验结果表明:所建立的方法成功实现了异常体的目标分类识别和空间位置定位,具有较大的应用前景。

【Abstract】 Ground Penetrating Radar is a kind of equipment with lossless detecting, which has been used comprehensively for roads detecting. Comparing with traditional detecting methods of roads detecting, it has many advantages, such as no destruction to road surface, rapid detection, continuity, high resolution, flexible operation, low cost and wide detecting range. Attribute to above virtues, the road geological radar is widely used in road projects. However, the related signal processing algorithms for echo waves of road radar develop very slowly, so the research of road radar signal processing algorithms has a great meaning in actual application.This paper firstly analyzes the electromagnetic wave transmitting rules in dielectric, illuminates the basic principles and structure of road radar, and proposes method to obtain the original data of road radar. Secondly, adopts a proper methods for noise decreasing-a combination of wavelet transformation and K-L transformation, researchs the instant frequency abstraction algorithms based on time-frequency transforming and simulates in ideal dielectric the echo wave figure of different objects by limited time difference algorithm. And on the basis of above two algorithms, this paper realizes the localization of special objects based on echo wave figure characteristics and objects classification based on echo wave of time-frequency property. Finally, tests the real data and the results show that the methods proposed by this paper are effective.

  • 【网络出版投稿人】 长安大学
  • 【网络出版年期】2009年 08期
节点文献中: 

本文链接的文献网络图示:

本文的引文网络