Note: Descriptions are shown in the official language in which they were submitted.
CA 03203426 2023-05-26
Calculation Method and Device for Interval Transit Time, and Storage
Medium
Technical Field
Embodiments of the present application relate to, but are not limited to, the
field of
logging data processing, in particular to a method for calculating an interval
transit time, an
apparatus, and a storage medium.
Background
An interval transit time refers to a time of an acoustic wave signal to
propagate per unit
distance in the formation, and is an important parameter in logging
interpretation, and is of
great importance in use, it may be used for calculating porosity, identifying
formation and rock
layer, calculating rock mechanics characteristic parameter, indicating the
overpressured
formation, estimating formation strength, predicting petroleum and sand
production pressure of
formation, estimating reservoir permeability, evaluating formation anisotropy,
analyzing
.. borehole stability, etc., and it is a basis and a key point of array
acoustic logging data
processing and interpretation.
In a related technology, a Slowness Time Coherence (STC) method proposed by
Kimball
in 1984 is adopted to calculate the interval transit time. In this method, by
correlating
waveforms of different receivers, a time slowness correlation diagram is
obtained, and then s a
maximum value of a correlation coefficient is sought to obtain a transmit time
of a target wave
signal.
In a practical application, this method adopts a flow which includes
artificial qualitative
recognition, artificial stratification, stratification parameter artificial
determination, and STC
method quantitative calculation, which has problems such as heavy workload,
high technical
difficulty, and relatively poor timeliness, etc.
Summary
The following is a summary of the subject matter described in detail herein.
This
1
Date regue/Date received 2023-05-26
CA 03203426 2023-05-26
summary is not intended to limit the protection scope of the claims.
An embodiment of the present application provides a method for calculating an
interval
transit time.
In order to achieve a purpose of the embodiments of the present application,
an
embodiment of the present application provides a method for calculating an
interval transit time,
which includes: determining a time domain boundary of different types of wave
signals in an
original signal of a target depth point acquired in advance; extracting a
target wave signal of the
target depth point from the original signal of the target depth point
according to the time
domain boundary of the different types of wave signals in the original signal
of the target depth
point; calculating frequency domain information and time domain information of
the target
wave signal; and calculating an interval transit time of the target wave
signal at the target depth
point by using the frequency domain information and the time domain
information.
A storage medium is provided, having a computer program stored therein,
wherein the
computer program is configured to perform the method described above when
running.
An electronic apparatus is provided, including a memory and a processor,
wherein the
memory has a computer program stored therein, and the processor is configured
to run the
computer program to perform the method described above.
Other aspects will become apparent after reading and understanding the
drawings and
detailed description.
Brief Description of Drawings
Accompanying drawings are used to provide further understanding of technical
solutions
of embodiments of the present application, and constitute a part of the
specification. They are
used together with embodiments of the embodiments of the present application
to explain the
technical solutions of the embodiments of the present application, but do not
constitute a
restriction on the technical solutions of the embodiments of the present
application.
FIG. 1 is a flowchart of a method for calculating an interval transit time
provided
according to an embodiment of the present application.
2
Date regue/Date received 2023-05-26
CA 03203426 2023-05-26
FIG. 2 is a schematic diagram of a signal image of a single receiver of an
array acoustic
logging tool provided according to an embodiment of the present application.
FIG. 3 is a schematic diagram of a result of acoustic wave segmentation on the
signal
image shown in FIG. 2.
FIG. 4 is a structure diagram of a full convolution neural network provided
according to
an embodiment of the present application.
FIG. 5 is a schematic diagram of segmenting different wave signals of a single
depth point
according to a time domain boundary according to an embodiment of the present
application.
FIG. 6 is a schematic diagram of a time-frequency spectrum obtained after
performing a
spectrum calculation on a P wave in FIG. 5.
FIG. 7 is a schematic diagram of performing an extraction operation on a time-
spectrum
diagram in FIG. 6.
Detailed Description
In order to make purposes, technical solutions, and advantages of embodiments
of the
present application clearer, embodiments of the present application will be
described in detail
below in conjunction with accompanying drawings. The embodiments of the
present
application and features in the embodiments may be combined with each other
arbitrarily if
there is no conflict.
An embodiment of the present application provides a method for calculating an
interval
transit time, wherein array acoustic logging data is use to accurately
calculate transmit times of
various wave signals.
FIG. 1 is a flowchart of a method for calculating an interval transit time
provided by an
embodiment of the present application. As shown in FIG. 1, the illustrated
method includes the
following acts 101 to 104.
In the act 101, a time domain boundary of different types of wave signals in
an original
signal of a target depth point acquired in advance is determined.
In an exemplary embodiment, the different types of wave signals may be
longitudinal
3
Date regue/Date received 2023-05-26
CA 03203426 2023-05-26
waves, transverse waves, or Stoneley waves under a monopole acoustic source;
or, they may be
leakage longitudinal waves or transverse waves under a dipolar acoustic
source.
In an exemplary embodiment, a time domain boundary is time information,
wherein the
time information is a demarcation point between a stop time of reception of
one type of wave
and a start time of reception of another type of wave, that is, a wave signal
in the original signal
in the time before reaching the time domain boundary is one type of wave
signal, and a wave
signal in the original signal in the time after the time domain boundary is
another type of wave
signal.
In the act 102, a target wave signal of the target depth point is extracted
from the original
signal of the target depth point according to the time domain boundary of the
different types of
wave signals in the original signal of the target depth point.
In the act 103, frequency domain information and time domain information of
the target
wave signal are calculated.
In the act 104, an interval transit time of the target wave signal at the
target depth point is
calculated by using the frequency domain information and the time domain
information.
The method for calculating an interval transit time in an embodiment of the
present
disclosure may be performed by a computer.
In the method provided according to an embodiment of the present application,
by
determining the time domain boundary of the different types of wave signals in
the original
signal of the target depth point acquired in advance, extracting the target
wave signal of the
target depth point from the original signal of the target depth point
according to the time
domain boundary of the different types of wave signals in the original signal
of the target depth
point, calculating the frequency domain information and the time domain
information of the
target wave signal, and then calculating the interval transit time of the
target wave signal at the
.. target depth point by using the frequency domain information and the time
domain information,
there are advantages of small workload and good calculation timeliness
compared with a
calculation method in a related technology.
The method provided according to an embodiment of the present application is
described
below.
4
Date regue/Date received 2023-05-26
CA 03203426 2023-05-26
In an exemplary embodiment, determining the time domain boundary of the
different
types of wave signals in the original signal of the target depth point
acquired in advance
includes: using two-dimensional matrix data received by a single receiver of
an array acoustic
logging tool as sample data, segmenting different types of wave signals in the
sample data, and
determining a segmenting line of the different types of wave signals in the
sample data;
acquiring a time domain boundary of the different types of wave signals at
each depth point by
using a position of the segmenting line in the sample data; and determining
the time domain
boundary of the different types of wave signals in the original signal of the
target depth point
from the time domain boundary of the different types of wave signals at the
each depth point.
FIG. 2 is a schematic diagram of a signal image of an array acoustic signal
provided
according to an embodiment of the present application. As shown in FIG. 2,
taking logging data
of a dipolar acoustic source in a soft formation as an example, different
types of waves in the
figure are visible and obvious leakage longitudinal waves (P) or transverse
waves (S).
In FIG. 2, taking the logging data of the dipolar acoustic source in the soft
formation as an
example, in actual situations, there may be following combinations according
to different
acoustic source types and formation types:
1. dipolar acoustic source + soft formation, with visible and obvious leakage
longitudinal
waves (P) and transverse waves (S), see FIG. 3 for details;
2. dipolar acoustic source + hard formation, with visible and weak or
invisible leakage
longitudinal waves (P) and obvious transverse waves (S);
3. monopole acoustic source + soft formation, with visible and obvious
longitudinal
waves (P), weak or invisible transverse waves (S) and Stoneley waves; and
4. monopole acoustic source + hard formation, with visible and obvious
longitudinal
waves (P), transverse waves (S), and Stoneley waves.
Waveform signals of a single receiver may be resampled and cropped, and
converted into
several 512*256 data samples to build a prediction sample database. The sample
database is
predicted by using a full convolution neural network, and time domain
boundaries of different
wave signals in various samples are obtained. Then the time domain boundaries
of the various
sample are converted into times and spliced to obtain a complete wave signal
boundary curve.
5
Date regue/Date received 2023-05-26
CA 03203426 2023-05-26
In an exemplary embodiment, the segmenting line of the different types of wave
signals in
the sample data is obtained by the following way, including: determining a
probability that each
element in the sample data belongs to a wave signal of a target type;
determining an element of
which a probability is greater than a preset probability threshold value in
the sample data as a
target element according to the determined probability; determining a position
of the wave
signal of the target type in the sample data according to a position of the
target element, and
using a boundary of the wave signal of the target type as a segmenting line of
the wave signal
of the target type in the sample data.
FIG. 3 is a schematic diagram of a result of acoustic wave segmentation on the
signal
image shown in FIG. 2. As shown in FIG. 3, firstly, acoustic wave logging data
are resampled
and disassembled into several 256*512 fragments to form a prediction sample
database; then
each sample is inputted into the full convolution neural network for
prediction, an edge of a
wave signal is extracted from an output matrix, and an arrival time of a
target wave signal of
each sample is obtained, and finally, arrival times of target wave signals of
various samples are
spliced together to obtain a segmenting line of P wave and S wave in FIG. 3.
For details, refer
to a gray line in a black box region in FIG. 3 as the segmenting line, that
is, a time domain
boundary. As can be seen from FIG. 3, the P wave is at a left side of the
segmenting line, and
the S wave is at a right side of the segmenting line, and each depth point
corresponds to its own
time domain boundary.
In an exemplary embodiment, the segmenting line of the different types of wave
signals in
the sample data is obtained by the following way, including: cutting the
sample data into at
least two segments; identifying a wave type of two-dimensional matrix data in
each segment,
and determining a boundary of wave signals in each segment; converting
boundaries of wave
signals in different segments into time data respectively; and splicing time
data of each segment
together to obtain the segmenting line of the different types of wave signals
in the sample data.
When the volume of sample data is too large, that is, a judgment condition of
a large
amount of data is met, then the sample data is cut; and when the volume of
sample data is
relatively small, that is, the judgment condition of the large amount of data
is not met, then the
sample data is directly processed as a whole, identifying of the wave type is
performed, and a
6
Date regue/Date received 2023-05-26
CA 03203426 2023-05-26
boundary of the wave signals in the sample data is determined without
operations of switching
into segments and splicing the segments.
FIG. 4 is a structure diagram of a full convolution neural network provided
according to
an embodiment of the present application. As shown in FIG. 4, an input of the
model is a matrix
with a size of 512*256, the matrix containing wave signals of 256 sampling
points, five times
of downsampling (a half branch at a left side) and five times of upsampling (a
half branch at a
right side) are performed in the middle, skip connection is adopted in the
middle to ensure that
spatial information is not lost, and a size of an output matrix is consistent
with that of an input
matrix, which is 512*256, each numerical value in the output matrix indicates
a probability that
each point in the input matrix belongs to the target wave signal, and a
segmented wave signal
may be obtained by binaryzation (a threshold value is 0.5).
A processing process of the above full convolution neural network is as
follows: (1)
acoustic wave logging data are resampled and disassembled into several 256*512
fragments to
form a prediction sample database; (2) each sample is inputted into the full
convolution neural
network for prediction and the output matrix is obtained, wherein a value of 1
in the output
matrix is a position of the target wave signal; (3) an edge of a wave signal
is extracted from the
output matrix, at this time an edge is represented by a column number in the
matrix in which it
is located, the edge of the signal may be converted from the column number to
a time according
to the measured start time, stop time, and the total number of columns, and
then the arrival time
of the target wave signal of each sample is obtained; and (4) finally, the
arrival times of the
target wave signals of the various samples are spliced together to obtain the
segmenting line for
the P wave and the S wave in FIG. 3.
In an exemplary embodiment, extracting a target wave signal of the target
depth point
from the original signal of the target depth point according to the time
domain boundary of the
different types of wave signals in the original signal of the target depth
point includes: when the
target wave signal is a signal received earlier in the original signal,
indicating that the target
wave signal is at a left side of the time domain boundary of the different
types of wave signals
in the original signal of the target depth point, then extracting a signal of
the original signal
from a start time to the time domain boundary as the target wave signal; and
when the target
7
Date regue/Date received 2023-05-26
CA 03203426 2023-05-26
wave signal is a signal received later in the original signal, indicating that
a target wave is at a
right side of the time domain boundary of the different types of wave signals
in the original
signal of the target depth point, then extracting a signal of the original
signal from the time
domain boundary to a stop time as the target wave signal.
Taking FIG. 3 as an example for illustration, it can be seen from FIG. 3 that
an arrival
time of the P wave is at the left side of the segmenting line, an arrival time
of the S wave is at
the right side of the segmenting line, and each depth point corresponds to its
own time domain
boundary. The time domain boundary of the different types of wave signals in
the original
signal of the target depth point may be determined from FIG. 3.
FIG. 5 is a schematic diagram of segmenting different wave signals of a single
depth point
according to a time domain boundary according to an embodiment of the present
application.
As shown in FIG. 5, in a coordinate diagram of the original signal, a
horizontal coordinate is
time, a unit is microseconds, and a longitudinal coordinate represents a
signal strength, which is
an amplitude of the signal. The time domain boundary of the original signal
shown in FIG. 5 is
5000 microseconds, so a result after extracting is to split the original
signal into two parts with
the time domain boundary as a segmenting point.
In FIG. 5, an extraction operation under a single depth point in a case of a
dipolar acoustic
source and a soft formation is illustrated as an example. In FIG. 5, taking a
leakage longitudinal
wave (P wave) in the dipolar acoustic source as an example, a same operation
mode will be
adopted for a transverse wave of the dipolar acoustic source, a longitudinal
wave, a transverse
wave, or a Stoneley wave in a monopole acoustic source.
In an exemplary embodiment, calculating frequency domain information and time
domain
information of the target wave signal includes: determining a time-based
spectrum diagram of
the target wave signal, wherein in a coordinate system in which the spectrum
diagram is located,
a horizontal coordinate is time, and a longitudinal coordinate is frequency;
acquiring horizontal
coordinates of positions of left and right boundaries of the spectrum diagram
and longitudinal
coordinates of positions of upper and lower boundaries of the spectrum
diagram; determining
the horizontal coordinates of the positions of the left and right boundaries
as a start time
time start and a stop time time stop of the target wave signal respectively,
and determining the
8
Date regue/Date received 2023-05-26
CA 03203426 2023-05-26
longitudinal coordinates of the positions of the upper and lower boundaries as
a minimum
frequency freq_min and a maximum frequency freq_max of the target wave signal
respectively.
FIG. 6 is a schematic diagram of a time-frequency spectrum obtained after
performing a
spectrum calculation on a P wave in FIG. 5. As shown in FIG. 6, a time-
frequency analysis is
performed on the target wave signal by using wavelet transform to obtain the
time-frequency
spectrum.
FIG. 7 is a schematic diagram of performing an extraction operation on a time-
spectrum
diagram in FIG. 6. As shown in FIG. 7, a wave signal region is segmented from
the
time-frequency spectrum, seeing a white region shown in FIG. 7. The start time
time start and
the stop time time stop, the minimum frequency freq_min, and the maximum
frequency
freq_max of the target wave signal may be calculated by identifying positions
of upper, lower,
left, and right boundary points of the wave signal region.
Different from a way of extracting frequency domain information in a related
technology,
the method provided according to an embodiment of the present application may
automatically
extract the frequency domain information by means of the time-spectrum
diagram, which
improves extraction efficiency.
Based on the above way, required numerical value information may be obtained
quickly
and accurately.
In an exemplary embodiment, calculating the interval transit time of the
target wave
signal at the target depth point by using the frequency domain information and
the time domain
information includes: determining a middle frequency freq_middle of the target
wave signal;
when the target wave signal is a signal received earlier in the original
signal, indicating that a
target wave is at a left side of the time domain boundary of the different
types of wave signals
in the original signal of the target depth point, then correcting the start
time time start of the
target wave signal by using the middle frequency freq_middle to obtain a
corrected start time
time start, and calculating the interval transit time of the target wave
signal at the target depth
point by using the corrected start time time start and the stop time time
stop, the minimum
frequency freq_min, and the maximum frequency freq_max; and when the target
wave signal is
a signal received later in the original signal, indicating that a target wave
is at a right side of the
9
Date regue/Date received 2023-05-26
CA 03203426 2023-05-26
time domain boundary of the different types of wave signals in the original
signal of the target
depth point, then correcting the stop time time stop of the target wave signal
by using the
middle frequency freq_middle to obtain a corrected start time time start, and
calculating the
interval transit time of the target wave signal at the target depth point by
using the corrected
start time time start and the stop time time stop, the minimum frequency
freq_min, and the
maximum frequency freq_max.
For a wave signal at the left side of the boundary, its stop time is
relatively accurate, and
its start time may be corrected; and for a wave signal at the right side of
the boundary, its start
time is relatively accurate, and its stop time may be corrected.
In an exemplary embodiment, when the target wave is at the left side of the
time domain
boundary of the different types of wave signals in the original signal of the
target depth point,
the corrected start time time start is obtained by the following way,
including: determining
len coef according to the type of the target wave signal, wherein a value
range of len coef is
between 2 and 8 according to the different type of target wave signal; time
start = time stop -
106 / freq_ middle * len coef; when the target wave is at the right side of
the time domain
boundary of the different types of wave signals in the original signal of the
target depth point,
the corrected stop time time stop is obtained by the following way, including:
time stop =
time start + 106 / freq_ middle * len coef .
Window length calculation: window length =106/freq_ middle * a; wherein a
value of a is
between 1 and 2, and a specific numerical value is determined according to
experience and may
be set to 1.5.
The following parameters obtained are used as parameters to be brought into
the STC
method for calculating a transmit time, including: when the target wave signal
is a signal
received earlier in the original signal, indicating that a target wave is at a
left side of the time
domain boundary of the different types of wave signals in the original signal
of the target depth
point, then performing the calculation by using the corrected start time time
start and the
corrected stop time time stop, the minimum frequency freq_min, the maximum
frequency
freq_max, and the window length calculation window length; and when the target
wave signal
is a signal received later in the original signal, indicating that a target
wave is at a right side of
Date regue/Date received 2023-05-26
CA 03203426 2023-05-26
the time domain boundary of the different types of wave signals in the
original signal of the
target depth point, then performing the calculation by using the start time
time start and the
corrected stop time time stop, the minimum frequency freq_min, the maximum
frequency
freq_max, and the window length calculation window length.
The method provided according to an embodiment of the present application, in
combination with a deep learning technology and a signal time-frequency
analysis method,
achieves an intelligent calculation of all interpretation parameters in the
STC method, and
finally achieves the calculation of the transmit time by using the STC method,
which has low
workload of data analysis, good timeliness, and low operation difficulty. The
method provided
according to an embodiment of the present application achieves, based on a
wave signal
segmentation technology, accurate segmentation of wave signals obtained by
different
measurement modes and different formation types, and performs, based on a
result of a wave
signal segmentation operation, operations of target wave signal extraction,
target wave signal
time-frequency analysis, target wave signal time domain and frequency domain
information
extraction to automatically complete analysis parameters of the interval
transit time, and finally
achieves the calculation of the transmit time by using a time-slowness
correlation method.
An embodiment of the present application provides a storage medium, having a
computer
program stored therein, wherein the computer program is configured to perform
the method
described in the above when being run.
An embodiment of the present application provides an electronic apparatus,
including a
memory and a processor, wherein the memory has a computer program stored
therein, and the
processor is configured to run the computer program to perform the method
described in the
above.
Those of ordinary skill in the art can understand that all or some of acts in
methods,
systems, functional modules/units in apparatuses disclosed above may be
implemented as
software, firmware, hardware, and appropriate combinations thereof. In a
hardware
embodiment, a division between functional modules/units mentioned in the above
description
does not necessarily correspond to a division of physical components; for
example, one
physical component may have multiple functions, or one function or act may be
performed
11
Date regue/Date received 2023-05-26
CA 03203426 2023-05-26
cooperatively by several physical components. Some or all of components may be
implemented
as software executed by a processor, such as a digital signal processor or a
microprocessor, or
as hardware, or as an integrated circuit, such as an application specific
integrated circuit. Such
software may be distributed on a computer-readable medium, which may include a
computer
storage medium (or non-transient medium) and a communication medium (or a
transient
medium). As is well known to those of ordinary skill in the art, the term
computer storage
medium includes volatile and non-volatile, removable and non-removable media
implemented
in any method or technique for storing information (such as computer-readable
instructions,
data structures, program modules, or other data). The computer storage medium
includes, but is
not limited to, RAM, ROM, EEPROM, a flash memory or another memory technology,
CD-ROM, a digital versatile disk (DVD) or another optical disk storage, a
magnetic cal tlidge, a
magnetic tape, a magnetic disk storage or another magnetic storage apparatus,
or any other
medium that may be configured to store desired information and may be accessed
by a
computer. In addition, it is well known to those of ordinary skill in the art
that a communication
medium typically contains computer readable instructions, data structures,
program modules, or
other data in modulated data signals such as carrier waves or another
transmission mechanism,
and may include any information delivery medium.
12
Date regue/Date received 2023-05-26
