The Application Of Wavelet Transform In The Detection Of The NMRL Echo Signal

The echo signal of Nuclear Magnetic Resonance Logging (NMRL) has the characterof weak and wide bandwidth. Traditional filtering method from the frequency domain hasthe the conflict between de-noising and keeping the integrality of the signal. The wavelettransform provides a new idea to the solution of the problem above-mentioned by the wayof analysising signal from the time domain and frequency domain.The NMRL echo signal as an example, this paper firstly analyzed the principle ofDigital Phase Sensitive Detector (DPSD) and its application on NMRL, pointed out thatboth FIR DPSD and average DPSD have the the conflict between de-noising and keepingthe integrality of the echo signal of NMRL in low porosity stratum. Secendly discoursedon the main character and theory of wavelet transform and wavelet thresholdde-noising, particularly concerned about the efficient time-frequency localization ofwavelet transform and excellent performance of de-noising of wavelet thresholdde-noising. Then combined DPSD with wavelet transform, designed and implemented aDPSD based on wavelet threshold de-noising. Lastly implemented wavelet DPSP onFPGA hardware platform based on the Lifting Wavelet Transform (LWT), which can meetthe needs of real-time in NMRL.As the results of experimental data processing show, the wavelet DPSD can keep theintegrality of the peak of echo signal while de-noising, because different characteristicsbetween signal and noise can be extracted from time domain and frequency domainsimultaneously by means of wavelet transform. To some extent, the wavelet DPSD solvesthe conflict the conflict between de-noising and keeping the integrality of the signal, is apowerful tool for detection of the echo signal.