Study On The Application Of Hilbert-Huang Transform In Well-logging Radar

The transient pulse well-logging radar is a kind of geophysical exploration radar working in borehole. It emits transient impulses into the strata and uses transmission characteristics of electromagnetic wave to capture strati-graphic information based on the development of borehole radar technology. Becuase of short radial detection ranges and inefficiency, it’s very difficult for conventional well-logging instruments to identify the reservoirs with low porosity and permeability. However, the transient pulse welllogging radar depends on its long detection range and relatively high resolution, and can image the stratigraphic structures surrounding the well. These advantages can make up for flaws of the conventional well-logging instruments, which makes the transient pulse well-logging radar become the focus of research around the world.The signal of transient pulse well-logging radar has no initial condition, no boundary condition and no steady state, so it can not be analyzed in frequency domain by using Fourier transform; And the signal is complex, with a wide frequency range, which means the accurate frequency of the target signal can not be predicted and it is difficult to extract the wavelet; Transient pulse well-logging radar is mainly based on radar original section to judge medium around the well, but for some relatively weak or containing more clutter radar echo signal, the exception information is hard to intuitive shown in the original section, which leads to an inaccurate explanation. The above three points greatly limit the radar data processing and interpretation.Hilbert Huang Transform(HHT) is suitable for non-stationary signal processing.Empirical Mode Decompostion(EMD) is one important part of HHT, it needn’t to know the signal freqency in advance and to extract the wavelet, which can solve the first and second problems very well. And we can get more geological information by using Hilbert Transform to analyze the signal from tree aspects(instantaneous amplitude, instantaneous phase, instantaneous frequency) after EMD. So it is important to study the application of Hilbert Huang Transform in the signal processing of transient pulse well-logging radar.Firstly, we did some studies on cylindrical coordinates Finite Difference Time Domain(FDTD) method. The central singularity problem and the Perfectly Matched Layer(PML) of the cylindrical coordinate system were deduced in detail. After that, we programed to realize the three-dimensional cylindrical coordinates FDTD simulation platform, which lays the foundation for simulation experiment of the transient pulse welllogging radar.Secondly, this paper introduced the data storage form and the conventional A-scan data processing method of the transient pulse well-logging radar. And then these algorithms are implemented by programming. At the same time, this paper also introduces the synthetic aperture migration algorithm used to process the B-scan, including the Kirchhoff migration and the frequency-wavenumber migration, and the simulation data is used to compare the results of the two migration methods.Thirdly, we introduced the Hilbert Huang Tranform method. The simulated data and the real data were processed and the physical meaning of each intrinsic mode were discussed by this method. Then three instantaneous parameters of each mode were extracted and their images were drawn. The result of experiment shown that the EMD had adaptive filtering characteristics and the reconstruction of the intrinsic modes could make the abnormal formation of interest more obvious. In another word the three instantaneous profile can be analyzed from three different aspects on the formation of the anomaly, and can be more effective in the interpretation of the radar signal.The main innovations of this work were applying Hilbert Huang Transform to the data processing of transient pulse well-logging radar, which meant a large number of processing methods were tried to see which was more effective and building the simulation platform of the transient pulse well-logging radar based on 3D cylindrical coordinate FDTD.