Signal Processing For Borehole Radar Well Logging

As a novel and promising geophysical detection technique, borehole radar systems have been widely used in many areas of human production and life, which has arisen the attention all over the world. Different from conventional ground penetrating radar systems, borehole radar systems have their unique working modes and environments. In the practical borehole surveys, there are problems such as the oversized sample set using impulse signals, the bistatic borehole radar imaging that do not follow the exploding reflector model assumption, the low efficiency of imaging with nonuniform samples and poor precision of target imaging in nonuniform layered media. Therefore, there are some special difficulties in the signal processing of borehole radar, especially for the borehole radar imaging. In this thesis, the borehole radar imaging techniques are studied.The main contributions of this thesis are as follows:1. For the contradiction between the range resolution and sampling rate, and the contradiction between data transmission rate and well logging cable in impulse borehole radar systems, a compressive-sensing-based data acquisition and imaging method that is suitable for sparse target spaces is proposed. The proposed method can reduce the requirements on system sampling rate, and relieve the burden of data transmission in the impulse borehole radar surveys.2. For the problem that the distance between transmitting antenna and receiving antenna cannot be ignored in bistatic borehole radar imaging, a novel efficient imaging method of bistatic borehole radar that is not subject to exploding reflector model is proposed. Based on the relationship between radar echo space and target space, and the relationship between angular frequency and wavenumber in bistatic borehole radar, the proposed method is capable of implementing the high resolution bistatic borehole radar imaging.3. Due to the usage of centralizers, the space sampling along the well direction in borehole radar surveys is commonly nonuniform. To solve the imaging problem with nonuniform samples, a novel borehole radar imaging method based on nonuniform-fast-Fourier-transform is proposed. By employing the effective estimation of nonuniform exponent factors, this method is able to obtain the frequency wavenumber spectrum of nonuniform samples in a fast manner and give a fast target space imaging using a large nonuniform sample set with reduced computational complexity, while the imaging quality of target space is guaranteed.4. For the problem of borehole radar imaging in nonuniform layered media with variable wave velocities, a time-gating-based reverse time migration is proposed. By strengthening the cross-correlation between the incident wavefield and scattering wavefield in the dominate reflection window of the same imaging point, the proposed method can suppress the interference signals that outside the observing windows of target dominating reflections in a more effective way, and enhance the imaging quality of borehole radar systems in the nonuniform detection environments.The performance of imaging methods mentioned above have been validated with borehole radar synthetic data and real field data collected by borehole radar prototype. The results have demonstrated the effectiveness of those proposed methods.