Application of GPR Data Deconvolution for Shallow Void Detection with Minimum-phase method

• Main Authors: Cheng-Jyh Shieh, 謝承志
• Other Authors: 康裕明
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/7r877q

碩士 === 國立臺北科技大學 === 土木工程系土木與防災碩士班 === 105 === Among the items of Non-Destructive Testing (NDT), the technique of ground penetrating radar (GPR) was developed and reliable in the recent years. The GPR was performed well on the detection of erosion under the pavement structure of dyke. But in the process of detection, it was difficult to distinguish the signal in the shallow layer due to the low-frequency antenna and the interruption caused by the direct-wave. This thesis focuses on the signal deconvolution of direct-wave for the shallow layer. In the past years, most of the studies on the deconvolution were via the synthetic waves and the field tests; or via the numerical analyses and verified by the field data. In the studies of this thesis, the minimum-phase deconvolution equations were derived from the composed functions and the deconvolution computation of the simulated signals. The efficiency and the accuracy of the results were verified by the signal simulations of different shallow cavities and the tests of sand-box in laboratory. According to the studies, the simulated deconvolution signals could be decomposed by the minimum-phase deconvolution equation. The key-points of the study were the selection of reference-wave and the adjustment of ‘gain’. With the simulation by the 400 MHz antenna, the maximum 9.4 cm-cavity could be distinguished and the error percentage could reduce to 5% or less. The actual applications of ground penetrating radar in the field were different from the theoretic simulations; in this study by the 1600MHz antenna, the maximum 4.0cm-cavity could be distinguished.