Signal processing method for improved acoustic formation logging system

• US 5,831,934 A
• Filed: 07/24/1997
• Issued: 11/03/1998
• Est. Priority Date: 09/28/1995
• Status: Expired due to Fees

• Current Assignee: Petrol International Inc.
• Original Assignee: Marek Z. Kozak, Stephen P. Gill, Timothy Prowten
• Inventors: Gill, Stephen P., Kozak, Marek Z., Prowten, Timothy
• Primary Examiner(s): Lobo, Ian J.
• Application Number:
  US08/901,230
• Time in Patent Office: 467 Days
• Field of Search: 367/25, 367/43, 367/48, 364/422, 364/724.01, 364/724.06
• US Class Current: 367/25
• CPC Class Codes: G01V 1/48 Processing data
• CPC Class Codes: G01V 1/52 Structural details
• CPC Class Codes: Y10S 367/911 Particular well-logging apparatus

Abstract

A method for analyzing received wave signal data in an acoustic logging system that includes a logging tool adapted to be moved through a borehole in the earth and comprising a transmitter section and receiver section, wherein the transmitter section produces sonic waves at periodic intervals which are received as wave signal data by the receiver section. The method comprises the steps of processing the received wave signal data using Fourier Transform procedures, filtering the processed data using a Heisenberg filter procedure, using a Hilbert transform to refine the filtered data, and using the refined data to provide a phase velocity analysis.

Claims

1. For use in an acoustic logging system that includes a logging tool adapted to be moved through a borehole in the earth and comprising a transmitter section and a receiver section spaced from said transmitter section, wherein said transmitter section produces sonic waves at periodic intervals which are received as wave signal data by said receiver section, a method for analyzing the received wave signal data which comprises the steps of:

processing the received wave digital signal data using Fourier transform procedures;

filtering the processed data using a Heisenberg filter procedures;

using a Hilbert transform to refine filtered data; and

using the refined data to provide a phase velocity analysis.

2. The method as described in claim 1 wherein said Heisenberg filtering step comprises the steps of:

passing real digital signal data through a digital Fourier transform to yield a series of complex coefficients representing the decomposition of each real digital signal into the sum of positive and negative frequencies;

multiplying each complex Fourier coefficient by a corresponding value of the real Heisenberg filter;

reconstituting the filtered Fourier coefficients into an analytic waveform containing both real and imaginary components using said Hilbert transforms.

3. The method as described in claim 1 wherein said Heisenberg filter provides an improved filter frequency response H for a frequency "f" and a center frequency "f₀ " which is defined as follows:

##EQU4## wherein Δ

f is frequency uncertainty, and Δ

t is time uncertainty.

4. The method as described in claim 1 wherein each said Hilbert transform comprises an inverse Fourier transform in which negative frequency components are nulled.