TIME-COMPRESSING MULTI-TRANSMISSION SIGNALS
First Claim
1. The method of time-compressing received multi-transmission signals which are of similar form but differ in peak spectral frequency between n successive transmissions, which method comprises the automatic machine-process steps of:
- for each receiver location, cross-correlating each of said n transmitted signals with each received-signal trace containing it, to produce n correlation functions, adjusting the amplitudes of said correlation functions by applying thereto relative weighting factors that produce, in a summation trace, the simplest form of central wavelets representing corresponding seismic-wave arrivals and minimum amplitudes outside of said central wavelets, summing said weighted correlation functions with their zero-lag times in coincidence to produce said summation trace, and filtering said summation trace by a Wiener-filter operator that reduces said summation-trace amplitudes substantially to zero outside of said central wavelets.
0 Assignments
0 Petitions
Accused Products
Abstract
For each seismic receiver location n multi-transmission signals, which are similar in form but differ in peak frequency within a frequency range of interest, are time-compressed by crosscorrelating the received signals containing it with each transmitted signal alone, weighting the resulting n correlation functions to produce for each event arrival the simplest compressed central wavelet when said n correlation functions are added with zero-lag times in coincidence, summing the weighted correlogram traces, and Wiener filtering the summation trace to reduce its amplitudes substantially to zero outside of said central wavelets.
30 Citations
8 Claims
-
1. The method of time-compressing received multi-transmission signals which are of similar form but differ in peak spectral frequency between n successive transmissions, which method comprises the automatic machine-process steps of:
- for each receiver location, cross-correlating each of said n transmitted signals with each received-signal trace containing it, to produce n correlation functions, adjusting the amplitudes of said correlation functions by applying thereto relative weighting factors that produce, in a summation trace, the simplest form of central wavelets representing corresponding seismic-wave arrivals and minimum amplitudes outside of said central wavelets, summing said weighted correlation functions with their zero-lag times in coincidence to produce said summation trace, and filtering said summation trace by a Wiener-filter operator that reduces said summation-trace amplitudes substantially to zero outside of said central wavelets.
-
2. The method of claim 1 in which said relative weighting factors are determined by the steps which comprise:
- during said n successive transmissions, making a reproducible reference recording of the output of a reference receiver at the seismic-wave transmitting location, said output corresponding to said n multi-transmission signals as they are introduced into the earth at said transmitting location, reproducing said reference recording as corresponding multi-transmission reference signals, cross-correlating each individual one of the n multi-transmission reference signals with the reference-signal trace which contains it, which trace may also contain one or more others of said reference signals transmitted in sequence, to produce n reference correlogram traces, plotting, along a linear frequency scale, at the position of each respective peak spectral frequency, an ordinate proportional to the energy density of each corresponding multi-transmission reference signal, and determining, for each ordinate in succession, a multiplying factor such that the energy increment it then represents causes the total cumulative energy to increase substantially linearly with frequency, said multiplying factors being approximately proportional to said weighting factors.
-
3. The method of claim 2 including the further steps of:
- summing said n reference correlogram traces weighted by said multiplying factors to obtain a first reference summation trace having one central wavelet preceded and followed by waves of substantially reduced amplitudes relative to said wavelet, and calculating a second set of multiplying factors for said reference correlogram traces to produce a second reference summation trace having a least-mean-squares fit to a trace consisting of said central wavelet alone, said second multiplying factors being proportional to optimum values of said weighting factors.
-
4. The method of claim 3 in which said Wiener-filter operator is calculated on the basis of said second reference summation trace, to reduce the wave amplitudes preceding and following said central wavelet substantially to zero.
-
5. The method of claim 1 in which each of said n multi-transmission signals is transmitted by the step of controlling a seismic-wave generating vibrator at a transmitting location by a pilot signal that produces a mono-frequency wave train in the general form of an abruptly terminated constant-frequency sinusoid a plurality of cycles in length.
-
6. The method of claim 5 in which a plurality of said mono-frequency wave trains of different spectral peak frequencies are transmitted in close succession as a group of mono-frequencies.
-
7. The method of claim 6 in which a plurality of said mono-frequency train groups are transmitted and received one group at a time, until said n signals having spectral peak frequencies distributed throughout a desired frequency range have been transmitted and received.
-
8. Apparatus for time-compressing multi-transmission signals received at a receiver location, which signals are of similar form but differ in peak spectral frequency between n successive transmissions, comprising, in combination, means for cross correlating each of said n transmitted signals with each received-signal trace containing it, to produce n correlation functions, means for adjusting the amplitudes of said correlation functions by applying thereto optimum relative weighting factors that produce, in a summation trace, the simplest form of central wavelets representing corresponding seismic-wave arrivals and minimum amplitudes outside of said central wavelets, means for summing said weighted correlation functions with their zero-lag times in coincidence to produce said summation trace, and means for filtering said summation trace by a Wiener-filter operator that reduces said summation-trace amplitudes substantially to zero outside of said central wavelets.
Specification