Research On Signal Processing Technology Of Underground Pipe Network Based On Ground Penetrating Radar

Urban underground pipe network is the infrastructure of the city.And also is one of the important basic data of urban planning,construction and management.It is particularly necessary that a set of effective underground pipeline detection method for the urban planning,transformation,construction and development to provide relevant information.Ground penetrating radar technology with nondestructive,rapid and high resolution characteristics is a hot spot in research of underground pipeline detection technology in recent years,has been widely used in underground pipeline detection,GPR technology can quickly detect underground pipeline and identify,and has great significance in practice application.But at present there are still some complicated problems such as echo image interpretation difficult,pipe diameter accuracy improvement,complex piping material identification method and so on.So this topic is based on GPR to study the underground pipe network detection technology and found new harvest.Based on simple introduction of the basic principle of GPR detection underground pipeline and a brief introduction to the three basic forms of data,this paper analyzing the echo imaging characteristics model of the underground pipeline.And the effectiveness of using GPR nondestructive detection technology to obtain underground pipeline characteristic parameters is discussed.This paper also introduces the basic process of GPR data processing.In the ground penetrating radar clutter suppression method,an improved linear prediction method is proposed based on analyzed the common clutter signal composition and characteristics of different clutter signal.This method based on linear variation characteristics between adjacent word clutters,by using the linear prediction method to calculate the clutter,coupled with PCA method,to highlight the main characteristics of the target signal and extract the weak signals in the target signals.Compared with several common clutter suppression methods,it has high signal to noise.At the same time,the improved linear prediction method ensures the integrity of target hyperbolic information on the basis of effectively suppresses clutter.For edge extraction method,the paper puts forward an edge extraction method based on gradient change.This method comprehensive considers the change of hyperbolic gradient of underground pipeline imaging characteristics model,and the actual situation that underground pipeline echo data in proportion to the echo image in the measured data.Compared with common edge extraction method,this method overcomes the shortage of the hyperbolic vertex poor continuity of the previous algorithms,and for more accurate extraction of underground pipeline echo hyperbola.In view of the underground pipeline diameter identifying,the least squares method is proposed by using extract hyperbolic to calculate diameter,and the effectiveness and accuracy of the identification method is verified by simulation data and the measured data.In addition,according to the characteristics of electromagnetic wave,introducing the phase method for the recognition of underground pipeline material.This method can effectively distinguish metal and nonmetal material pipeline from echo images.Finally,this paper developed processing software of underground pipeline echo data based on GPR.It realizes some functions that the common data processing methods of GPR signal and other methods are studied in the paper.