Research On Imaging Technology Of 3D Ground Penetrating Radar In Expressway Detection

Ground penetrating radar,as an exploration technique for nondestructive testing,has the advantages of high efficiency and high resolution,and is widely used in road subgrade detection and disease identification.Ground penetrating radar data processing is divided into two parts: preprocessing and imaging.Data preprocessing is mainly to remove direct waves with strong echo signal energy,suppress interference and noise.Ground penetrating radar imaging can more intuitively present the internal structure of underground targets,thereby greatly promoting target detection and recognition in engineering applications.In the background of highway subgrade detection,this paper studies the direct wave removal algorithm for ground penetrating radar data and deep weak echo signal gain adjustment algorithm.Combining simulation data and measured data,the two-dimensional ground penetrating radar imaging and three-dimensional ground penetrating radar imaging algorithms are verified respectively.This paper first gives the theoretical basis and working method of the ground penetrating radar.Combined with the three-dimensional array radar used in this paper,the echo signal data format and signal model are given.Data forward simulations were conducted for the deep and shallow target bodies,single target body and multi-target body models.Aiming at the effect of direct reflection wave and direct coupling wave on the ground in the measured data and forward simulation data,the mean-to-direct wave algorithm is introduced.In response to the problem of weak echoes in deep underground targets,exponential gain adjustment is introduced to enhance the amplitude of reflected waves from deep targets,providing high signal-to-interference-to-noise ratio echo signals for subsequent research of ground penetrating radar imaging algorithms.This paper focuses on the two-dimensional imaging algorithm of ground penetrating radar and three-dimensional imaging algorithm of ground penetrating radar.This paper focuses on three kinds of two-dimensional ground penetrating radar imaging algorithms: frequency domain-wave number domain(F-K)wave equation migration imaging,backward projection(BP)imaging,and finite difference(FD)migration imaging.In view of the problem that the running time of the BP imaging algorithm is too long,this paper proposes a down-sampling BP imaging algorithm,which reduces the imaging time under the condition of ensuring the imaging resolution.Aiming at the problem of excessive sidelobes and clutter interference of the classic BP imaging algorithm,this paper proposes a BP imaging algorithm based on the mean correlation coefficient(referred to as MCBP imaging).The measured data and forward data imaging results show that the MCBP imaging algorithm can basically suppress the side lobe and clutter interference,further verifying that the improved algorithm is versatile.Finally,the three-dimensional frequency domain F-K migration imaging algorithm is studied,and the forward imaging data is used to verify the imaging feasibility,which provides the algorithm basis for the development of the three-dimensional imaging software of the host computer in the highway detection of the three-dimensional array radar.