Font Size: a A A

Extracting layer information using a new inversion algorithm from ground penetrating radar data

Posted on:2008-01-03Degree:Ph.DType:Dissertation
University:University of HoustonCandidate:Kao, Chien-pingFull Text:PDF
GTID:1448390005968412Subject:Engineering
Abstract/Summary:
Ground Penetrating Radar (GPR) is well known as a nondestructive sensing tool and is used in many applications, such as pavement inspection, pipe locating, and damage detection. The objective of this dissertation focuses on the measurement of pavement thickness. In this dissertation, three main goals have been achieved. First, a new algorithm was developed to extract the stratified thickness and permittivity information from GPR data. In general, the Common Middle Point (CMP) method is widely used for estimating the depth and permittivity of the layered media based on the GPR data, however, with increasing antenna separation, the application of the CMP method usually results in noticeable errors in thickness and permittivity readings. In the new mathematical model, we not only considered the effects from the air-ground interface, but also introduced a ray-path-searching process in the GPR measurement using Fermat's shortest path law. The shortest path is then used to precede the inversion of GPR data in order to calculate the thickness and permittivity of every layer. Finally, I applied the Transmission-Line Matrix (TLM) method to simulate the propagation of the GPR signal in the multi-layered formation. A time sequence image, produced by the Finite Difference Time Domain Method (FDTD), was also introduced to explain this new model. Comparing the numerical simulations with the measured results demonstrated that the new model is much more accurate and suitable for GPR applications than earlier model.; Secondly, in order to meet the requirements of high-resolution GPR surveys, several types of antennas were designed, simulated, manufactured, and evaluated. A transverse electromagnetic wave (TEM) horn antenna is usually applied to the air-launching GPR system. Traditionally, the variation of characteristic impedance of a TEM horn antenna is set to range from 50 O (characteristic impedance of a coaxial cable) to 376.7 O (free space wave impedance). However, a difference regularly exists between transmission-line wave characteristic impedance and free space wave impedance. In this dissertation, the simulated and experimental results both show that the performance of the designed [FM horn antenna matching with impedance of 200 O is better than that with impedance of 376.7 O. For full understanding, a discussion was provided to explain the possible reasons. In the final aspects of this dissertation, the other components, such as transmitter, receiver, and control units, would work with the TEM horn antennas to complete an air-launching GPR system. Moreover, field measurements were conducted, and using the inversion algorithm developed in the first section of this dissertation to extract the layer information, the comparison shows that the ratio of the occurrence of errors between the measured data and real data of highway pavement thickness is less than 5%.; As for the third achievement, a modified ridged horn antenna was developed and evaluated. The new modified ridged horn is filled with dielectric materials with a relative dielectric constant of four. The dimension of this new horn is only 7.5 cm * 6.0 cm * 3.9 cm (I. * W * H) and has a good radiation characteristic within the frequency range of 1.0 GHz to 6.0 GHz. The measured result shows that it can be integrated with a pulse transmitter to complete a borehole radar system and is feasible to be used in a confined environment.
Keywords/Search Tags:GPR, Radar, New, Used, Data, TEM, Information, Algorithm
Related items