CN103850679B - method for reconstructing sound wave time difference curve by using various logging curves - Google Patents
method for reconstructing sound wave time difference curve by using various logging curves Download PDFInfo
- Publication number
- CN103850679B CN103850679B CN201410126116.XA CN201410126116A CN103850679B CN 103850679 B CN103850679 B CN 103850679B CN 201410126116 A CN201410126116 A CN 201410126116A CN 103850679 B CN103850679 B CN 103850679B
- Authority
- CN
- China
- Prior art keywords
- curve
- regression
- acoustic
- layers
- decomposition
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 15
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 28
- 239000013598 vector Substances 0.000 claims abstract description 16
- 239000011159 matrix material Substances 0.000 claims abstract description 13
- 238000000611 regression analysis Methods 0.000 claims abstract description 13
- 238000004458 analytical method Methods 0.000 claims abstract description 11
- 230000035945 sensitivity Effects 0.000 claims abstract description 4
- 238000001134 F-test Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 claims description 3
- 238000010219 correlation analysis Methods 0.000 claims description 2
- 230000000694 effects Effects 0.000 claims description 2
- 230000000875 corresponding effect Effects 0.000 claims 10
- 230000002596 correlated effect Effects 0.000 claims 1
- 230000000717 retained effect Effects 0.000 claims 1
- 238000010206 sensitivity analysis Methods 0.000 claims 1
- 238000012360 testing method Methods 0.000 description 5
- 230000001419 dependent effect Effects 0.000 description 3
- 230000008021 deposition Effects 0.000 description 3
- 238000005553 drilling Methods 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 239000011435 rock Substances 0.000 description 1
- 238000013179 statistical model Methods 0.000 description 1
Landscapes
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention relates to a method for reconstructing an acoustic time difference curve by utilizing various logging curves, which mainly comprises the steps of analyzing reservoir sensitivity and correlation of the logging curves, and selecting a plurality of logging curves which obviously respond to reservoir characteristics; carrying out discrete wavelet decomposition on the logging curves and the acoustic curve, wherein the number of decomposition layers is 8; respectively forming high-frequency decomposition results of all layers of other logging curves except acoustic waves into a matrix, solving eigenvalues and eigenvectors corresponding to the matrix, and taking the eigenvectors corresponding to different eigenvalues as new components of signal reconstruction, wherein the components have orthogonality (irrelevance); performing multiple regression analysis on the high-frequency components of the corresponding layer of the wavelet decomposition of the acoustic curve by using the characteristic vectors of each layer, calculating a weighting coefficient corresponding to each vector, returning a regression significance analysis result, and judging regression quality; judging the number of layers of the sound wave high-frequency components reconstructed by using the characteristic vectors according to the regression significance analysis result, selecting a multiple regression result for the part above the selected number of layers as the high-frequency components, and reserving the high-frequency decomposition result of the sound wave logging curve for the part below the selected number of layers; and performing curve reconstruction by using the low-frequency component of the acoustic logging curve and the high-frequency component obtained by regression to obtain a final acoustic reconstruction curve.
Description
The technical field is as follows:
the invention relates to a method for reconstructing an acoustic time difference curve by utilizing various logging curves.
Background art:
Acoustic curve reconstruction is a common method among wave impedance inversions. When the difference between the speed of the underground reservoir and the speed of the mudstone is small, the contact ratio is high; the result of directly performing the wave impedance inversion cannot accurately reflect the difference between the reservoir and the surrounding rock, so that the wave impedance inversion result is inconsistent with the drilling result. Therefore, the acoustic logging curve needs to be reconstructed, lithology information is added to the acoustic logging curve, and the lithology recognition degree of the acoustic logging curve is improved, so that more accurate reservoir prediction can be performed.
The multiple regression analysis is a regression analysis method for studying the relationship between a plurality of variables, and is classified into a regression analysis of one dependent variable for a plurality of independent variables (simply referred to as "one-to-many" regression analysis) and a regression analysis of a plurality of dependent variables for a plurality of independent variables (simply referred to as "many-to-many" regression analysis) according to the number correspondence between the dependent variables and the independent variables, and a linear regression analysis and a nonlinear regression analysis according to the type of a regression model.
The F-test (F-test), the most commonly used alias name, is called joint hypothesis test (English), and is also called variance ratio test, and variance homogeneity test. It is a test in which statistical values are subject to F-distribution under the null hypothesis (H0). It is typically used to analyze statistical models that use more than one parameter to determine whether all or a portion of the parameters in the model are suitable for estimating the mother.
In wavelet transform, for non-stationary signals (logging signals are non-stationary signals), it is necessary that the time-frequency window has adjustable property, i.e. it is required to have better time resolution characteristics in the high frequency part and better frequency resolution characteristics in the low frequency part. The wavelet analysis of the logging curve can identify curve loops of different frequencies in the logging curve, the curve loops of high frequencies correspond to high-frequency, short-term, deposition loops, and the curve loops of low frequencies correspond to low-frequency, long-term, deposition loops, so that the logging curve loops of different frequencies can be used for dividing deposition loops of different periods, and the logging curve loops correspond to sequence stratigraphic units of different levels.
The invention content is as follows:
The invention relates to a method for reconstructing an acoustic time difference curve by utilizing various logging curves, which mainly comprises the steps of analyzing reservoir sensitivity and correlation of the logging curves, and selecting a plurality of logging curves which obviously respond to reservoir characteristics; carrying out discrete wavelet decomposition on the logging curves and the acoustic curve, wherein the number of decomposition layers is 8; respectively forming high-frequency decomposition results of all layers of other logging curves except acoustic waves into a matrix, solving eigenvalues and eigenvectors corresponding to the matrix, and taking the eigenvectors corresponding to different eigenvalues as new components of signal reconstruction, wherein the components have orthogonality (irrelevance); performing multiple regression analysis on the high-frequency components of the corresponding layer of the wavelet decomposition of the acoustic curve by using the characteristic vectors of each layer, calculating a weighting coefficient corresponding to each vector, returning a regression significance analysis result, and judging regression quality; judging the number of layers of the sound wave high-frequency components reconstructed by using the characteristic vectors according to the regression significance analysis result, selecting a multiple regression result for the part above the selected number of layers as the high-frequency components, and reserving the high-frequency decomposition result of the sound wave logging curve for the part below the selected number of layers; and performing curve reconstruction by using the low-frequency component of the acoustic logging curve and the high-frequency component obtained by regression to obtain a final acoustic reconstruction curve.
Drawings
FIG. 1 is a flow chart for reconstructing an acoustic moveout curve using multiple well logs
FIG. 2 is a graph showing the significance analysis of feature vector multiple regression
FIG. 3 shows the result of acoustic logging curve reconstruction
The specific implementation mode is as follows:
As shown in fig. 1, the implementation steps of the method are described in detail as follows:
performing correlation analysis by using reservoir porosity and shale content data and a logging curve, and selecting the logging curve with higher correlation with shale content and porosity;
Performing db wavelet decomposition on the logging curves with high correlation with the shale content and the porosity, wherein the number of decomposition layers is 8;
Step three, respectively forming high-frequency decomposition results of all layers of other logging curves except acoustic waves into a matrix, solving eigenvalues and eigenvectors corresponding to the matrix, and taking the eigenvectors corresponding to different eigenvalues as new components of signal reconstruction, wherein the components have orthogonality (irrelevance);
Step four, using the feature vectors to make sound pairsPerforming multiple regression analysis on the high-frequency component of the wavelet decomposition of the wave curve, wherein the coefficient of each item in a regression equation is the weight corresponding to the feature vector, and F detection is used for judging that the regression quality returns two parameters FH and FV; FH is the judgment of the reliability of the multiple regression result; for a given confidence a, F from the F distribution tableα(m, n-m-1) value is compared with FV when FV > Fα(m, n-m-1) indicates that the multivariate regression result is credible, and FH is equal to 1, otherwise FH is equal to 0; the larger the FV value, the more significant the overall regression effect (fig. 2);
And step five, according to the result returned by the F test, selecting the layer with the smallest FV value under the condition that the regression result of each layer is credible, namely FH is equal to 1, and replacing the high-frequency decomposition result of the acoustic wave curve of the layer and all the layers above the layer with a multiple regression result to be used as the high-frequency component of the acoustic wave curve reconstruction.
And sixthly, performing curve reconstruction by using the low-frequency component of the acoustic logging curve and the high-frequency component obtained by regression to obtain a final acoustic reconstruction curve (figure 3).
Claims (6)
1. a method of reconstructing an acoustic moveout curve using a plurality of well logs, comprising: performing reservoir sensitivity and correlation analysis on the logging curves, and selecting a plurality of logging curves which obviously respond to reservoir characteristics; carrying out discrete wavelet decomposition on the logging curves and the acoustic curve, wherein the number of decomposition layers is 8; respectively forming high-frequency decomposition results of all layers of other logging curves except acoustic waves into a matrix, solving eigenvalues and eigenvectors corresponding to the matrix, and taking the eigenvectors corresponding to different eigenvalues as new components of signal reconstruction, wherein the components have orthogonality at the moment, namely the components are irrelevant; performing multiple regression analysis on the high-frequency components of the corresponding layer of the wavelet decomposition of the acoustic curve by using the characteristic vectors of each layer, calculating a weighting coefficient corresponding to each vector, returning a regression significance analysis result, and judging regression quality by using F detection; judging the number of layers of the sound wave high-frequency components reconstructed by using the characteristic vectors according to the regression significance analysis result, selecting a multiple regression result for the part above the selected number of layers as the high-frequency components, and reserving the high-frequency decomposition result of the sound wave logging curve for the part below the selected number of layers; and performing curve reconstruction by using the low-frequency component of the acoustic logging curve and the high-frequency component obtained by regression to obtain a final acoustic reconstruction curve.
2. the method of claim 1 wherein the plurality of logs are analyzed for reservoir sensitivity and correlation to select a plurality of logs having significant response to reservoir characteristics, wherein correlation is performed using the reservoir porosity, shale content data and the logs to select a log having a higher shale content and porosity correlation.
3. the method of claim 2, wherein the discrete wavelet decomposition is performed on the well logs and the acoustic curve, and the number of decomposed layers is 8, wherein db wavelet decomposition is performed on the well logs with high correlation to shale content and porosity, and the number of decomposed layers is 8.
4. the method according to claim 3, wherein the high frequency decomposition results of the logging curves except the acoustic wave are respectively formed into a matrix, the eigenvalues and eigenvectors corresponding to the matrix are obtained, and the eigenvectors corresponding to different eigenvalues are used as new components for signal reconstruction, wherein the components have orthogonality, i.e. the components are not correlated, and the method is characterized in that the high frequency components of each layer in the wavelet decomposition results are formed into a matrix, and the matrix is formed into 8 matrices; and (3) solving the eigenvalue and the eigenvector corresponding to each matrix, wherein the solved eigenvectors have orthogonality.
5. The method of claim 4, wherein the multiple well-logging curves are used to reconstruct the acoustic moveout curves, the feature vectors of each layer are used to perform multiple regression analysis on the high frequency components of the corresponding layer of the wavelet decomposition of the acoustic curves, the weighting coefficients corresponding to each vector are calculated, and the weighting coefficients are returned to the regression displayJudging regression quality by using the analysis result of the characteristic, and is characterized in that the characteristic vector is used for carrying out multiple regression analysis on the wavelet decomposition high-frequency component of the acoustic curve, the coefficient of each item in a regression equation is the weight corresponding to the characteristic vector, and F detection is used for judging that the regression quality returns two parameters FH and FV; FH is the judgment of the reliability of the multiple regression result; for a given confidence a, F from the F distribution tableα(m, n-m-1) value is compared with FV when FV > Fα(m, n-m-1) indicates that the multivariate regression result is credible, and FH is equal to 1, otherwise FH is equal to 0; the larger the FV value, the more significant the overall regression effect.
6. the method according to claim 5, wherein the number of layers for reconstructing the acoustic time difference curve using the plurality of well logs is determined according to the regression significance analysis result, the multiple regression results are selected for the parts above the selected number of layers as the high frequency components, and the high frequency decomposition results of the acoustic well log curve are retained for the parts below the selected number of layers, wherein the layer with the smallest FV value is selected according to the returned results of the F-test, and the acoustic curve high frequency decomposition results for the layer and all layers above the layer are replaced with the regression results as the acoustic curve reconstructed high frequency components, in case that the regression results for each layer are credible, that is, FH ═ 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410126116.XA CN103850679B (en) | 2014-04-01 | 2014-04-01 | method for reconstructing sound wave time difference curve by using various logging curves |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410126116.XA CN103850679B (en) | 2014-04-01 | 2014-04-01 | method for reconstructing sound wave time difference curve by using various logging curves |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103850679A CN103850679A (en) | 2014-06-11 |
| CN103850679B true CN103850679B (en) | 2019-12-17 |
Family
ID=50858772
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410126116.XA Expired - Fee Related CN103850679B (en) | 2014-04-01 | 2014-04-01 | method for reconstructing sound wave time difference curve by using various logging curves |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103850679B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104533400B (en) * | 2014-11-12 | 2017-05-17 | 中海油能源发展股份有限公司 | Method for reconstructing logging curve |
| CN108072903A (en) * | 2016-11-09 | 2018-05-25 | 中国石油化工股份有限公司 | A kind of Well logging curve reconstruction method |
| CN106707344B (en) * | 2016-12-12 | 2019-01-18 | 中国石油天然气股份有限公司 | Stratum sequence dividing method and device |
| CN107339099B (en) * | 2017-07-19 | 2020-06-09 | 中国石油天然气集团公司 | Method and device for determining reservoir lithology |
| CN108756867B (en) * | 2018-05-11 | 2021-11-19 | 中国地质调查局油气资源调查中心 | Method for fracturing and selecting layer based on acoustic logging curve and resistivity logging curve |
| CN109061729B (en) * | 2018-08-22 | 2019-11-12 | 西安石油大学 | A kind of high temperature and pressure gas reservoir gassiness sensitivity curve reconstructing method |
| CN109343120B (en) * | 2018-10-17 | 2019-10-01 | 吉林大学 | Incorporate the sound wave curve reconstructing method of constrained sparse spike inversion inverting low-frequency compensation |
| CN109633553B (en) * | 2019-01-18 | 2020-11-13 | 浙江大学 | Mobile sound source arrival time delay estimation method based on dynamic programming algorithm |
| CN112012726B (en) * | 2019-05-30 | 2023-12-12 | 中石化石油工程技术服务有限公司 | Lithology recognition method |
| CN111827966B (en) * | 2020-03-25 | 2022-04-15 | 大庆油田有限责任公司 | Multi-well acoustic logging curve consistency processing method and device and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0609949A1 (en) * | 1993-02-05 | 1994-08-10 | AGIP S.p.A. | Process and device for detecting seismic signals |
| CN201747364U (en) * | 2010-06-01 | 2011-02-16 | 中国石油天然气集团公司 | Interval transit time curve reconstruction equipment |
| CN102707313A (en) * | 2012-04-19 | 2012-10-03 | 电子科技大学 | Pseudo-sonic curve construction method based on pulse coupling neural network |
| CN103485768A (en) * | 2012-06-13 | 2014-01-01 | 中国石油天然气集团公司 | Method for forming acoustic logging curve |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8834624B2 (en) * | 2011-01-26 | 2014-09-16 | Ripi | Modified cement composition, preparation and application thereof |
-
2014
- 2014-04-01 CN CN201410126116.XA patent/CN103850679B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0609949A1 (en) * | 1993-02-05 | 1994-08-10 | AGIP S.p.A. | Process and device for detecting seismic signals |
| CN201747364U (en) * | 2010-06-01 | 2011-02-16 | 中国石油天然气集团公司 | Interval transit time curve reconstruction equipment |
| CN102707313A (en) * | 2012-04-19 | 2012-10-03 | 电子科技大学 | Pseudo-sonic curve construction method based on pulse coupling neural network |
| CN103485768A (en) * | 2012-06-13 | 2014-01-01 | 中国石油天然气集团公司 | Method for forming acoustic logging curve |
Non-Patent Citations (5)
| Title |
|---|
| 储层特征曲线重构技术在储层预测中的应用研究;张静等;《天然气地球科学》;20080630;第19卷(第3期);全文 * |
| 分频重构技术在砂泥岩薄互层储层预测中的应用;熊冉等;《油气地球物理》;20131031;第11卷(第4期);全文 * |
| 小波分解在测井解释中的应用;李兴龙等;《内江科技》;20111231;全文 * |
| 拟声波曲线构建的意义及应用;姜传金 等;《大庆石油地质与开发》;20040229;第23卷(第1期);全文 * |
| 测井声波时差反演重构技术研究及应用;宋维琪等;《地震地质》;20090331;第31卷(第1期);第134页第1.1节、第135页第3节第1段、第137页第5节第3-4段 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103850679A (en) | 2014-06-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103850679B (en) | method for reconstructing sound wave time difference curve by using various logging curves | |
| CN111208561B (en) | Seismic acoustic wave impedance inversion method based on time-varying wavelet and curvelet transformation constraint | |
| CN111551992B (en) | Rock reservoir structure characterization method and device, computer-readable storage medium and electronic equipment | |
| Wang | Multichannel matching pursuit for seismic trace decomposition | |
| CN110361778B (en) | Seismic data reconstruction method based on generation countermeasure network | |
| US5687138A (en) | Methods of analyzing stoneley waveforms and characterizing underground formations | |
| US11353616B2 (en) | Methods and systems for processing slowness values from borehole sonic data | |
| US20220244419A1 (en) | Machine learning enhanced borehole sonic data interpretation | |
| CN113109289A (en) | Selection method of optimal wavelet denoising combination and THz spectrum denoising method | |
| US20110166842A1 (en) | Layer stripping method | |
| Saad et al. | Unsupervised deep learning for single-channel earthquake data denoising and its applications in event detection and fully automatic location | |
| Qian et al. | Unsupervised seismic footprint removal with physical prior augmented deep autoencoder | |
| CN111060961B (en) | Quality factor determination method, device and system based on multi-information constraint inversion | |
| Wu et al. | A high‐resolution nonlinear inversion method of reservoir parameters and its application to oil/gas exploration | |
| Fedi | Global and local multiscale analysis of magnetic susceptibility data | |
| Alvarez et al. | Lithologic characterization of a reservoir using continuous-wavelet transforms | |
| CN113655522A (en) | Frequency domain seismic weak signal enhancement method | |
| US10775525B2 (en) | Identifying and visually presenting formation slowness based on low-frequency dispersion asymptotes | |
| Walker et al. | Wavelength-based axial resolution limitations of flexural wave dispersion sonic logging | |
| CN107678065B (en) | The guarantor for improving seismic resolution constructs well control space the Method of Deconvolution and device | |
| Nose-Filho et al. | Algorithms for sparse multichannel blind deconvolution | |
| He et al. | A reliable online dictionary learning denoising strategy for noisy microseismic data | |
| US20160170066A1 (en) | Probability Distribution Based Logging Tool Data Compression | |
| Zhang et al. | Q estimation via the discriminant method based on error modeling | |
| Zhou et al. | Flattening the seismic data for optimal noise attenuation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20191217 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |