Research On Pavement Layering And Road Defect Detection Algorithms Based On GPR

The development of urbanization in China has driven the construction of highways and urban roads.However,a series of highway quality problems have emerged one after another,such as the short service life caused by the improper construction in the early stage,the discomfort caused by the surface holes and crushing,and the collapse of the road surface caused by the cavity and void under the road surface in the later.Therefore ensuring the normal operation of the road has become an important task in urban construction.Ground Penetrating Radar(GPR)plays an important role in highway detection.It uses emitted and received electromagnetic waves reflected by road stratification or distresses to judge the surface and internal conditions of the road.This paper aims to study the GPR pavement stratification and road defect detection algorithms.The main research work includes the following three aspects:Firstly,the operation mechanism of ground penetrating radar is studied.Electromagnetic waves propagate in the pavement surface layer and the base layer,and reflection and refraction occur only at the interface,which makes the GPR data sparse.Aiming at the propagation mechanism of electromagnetic waves in a discontinuous medium,the relative dielectric constant of the medium was calculated by the intensity of the reflected echo.The basic flow of GPR detection data processing is identified as data editing,zero time correction,DC,data gain,filtering,deconvolution.Secondarily,the algorithms for determining the layer thickness of road pavements are investigated.Road layer detection based on GPR technology has been relatively mature,but it is easy to cause aliasing of echoes when layering thickness is too small,especially for asphalt pavement layer.Estimation of the thickness of the asphalt thin layer is still a problem due to the limitation of GPR resolution and the small difference in dielectric constant between the two asphalt layers.As the reflection coefficient is sparse,the sparse decomposition is used to treat the reflected wave of the asphalt pavement layer and the layer thickness is recognized.The signal aliasing problem is solved and the delay information is accurately extracted.Finally,detection algorithms of subgrade defect of roads are studied.The complex underground medium structure,external electromagneti c waves and strong clutter on the ground have contaminated the radar echo data,r esulting in complex deconvolution results of roadbed defects.For the detection of subgrade distresses,the SVD(singular value decomposition)is used to perform principal component analysis on the radar data,and then the residual signal is “stretched” to highlight the roadbed distresses.Finally,it is proposed to combine SVD with sparse pulse deconvolution to improve the resolution of deconvolution.