CN111983722B - Reservoir logging characteristic parameter extraction method based on target preference coding - Google Patents
Reservoir logging characteristic parameter extraction method based on target preference coding Download PDFInfo
- Publication number
- CN111983722B CN111983722B CN202010905052.9A CN202010905052A CN111983722B CN 111983722 B CN111983722 B CN 111983722B CN 202010905052 A CN202010905052 A CN 202010905052A CN 111983722 B CN111983722 B CN 111983722B
- Authority
- CN
- China
- Prior art keywords
- reservoir
- logging
- target
- characteristic
- depth
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V11/00—Prospecting or detecting by methods combining techniques covered by two or more of main groups G01V1/00 - G01V9/00
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A10/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE at coastal zones; at river basins
- Y02A10/40—Controlling or monitoring, e.g. of flood or hurricane; Forecasting, e.g. risk assessment or mapping
Landscapes
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Geophysics (AREA)
- Geophysics And Detection Of Objects (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a reservoir logging characteristic parameter extraction method based on target preference coding, and belongs to the technical field of petroleum logging. Mainly comprises the following steps: step 1, screening out logging curves capable of reflecting reservoir characteristics according to target reservoir evaluation characteristics, determining priority levels, and sorting for data coding; step 2, eliminating the top and bottom segment data, and statistically analyzing the distribution characteristics of the residual data; step 3, coding the logging data based on the reservoir evaluation target preference; step 4, merging the feature codes, formulating a feature code analysis rule, and finding out feature value sampling depth points; and 5, resampling the sampling depth points, and determining characteristic values of each logging curve of the target reservoir for reservoir parameter modeling and fluid identification plate modeling. Compared with manual extraction of characteristic values, the method has the advantages of high speed and high accuracy, and greatly improves the efficiency of interpretation and evaluation of reservoir logging data through characteristic depth points and characteristic value extraction based on target preference coding.
Description
Technical Field
The invention belongs to the technical field of petroleum logging, and particularly relates to a reservoir logging characteristic parameter extraction method based on target preference coding.
Background
Reservoir logging evaluation is an important content of reservoir evaluation, and reservoir logging characteristic values are core parameters for reservoir physical property parameter model research and fluid identification plate establishment, so that the reservoir logging characteristic values are particularly important to selection. The traditional extraction method mainly depends on expertise with abundant experience to adopt manual extraction, has higher requirements on regional logging interpretation experience, is easy to generate the phenomenon of thousands of people and thousands of sides, and is highly required to be provided for reducing artificial extraction errors and improving the stability of characteristic value extraction, so that the efficiency and stability of the extraction method of the characteristic value of the reservoir can be greatly improved, and the reservoir can be rapidly evaluated and the fluid property can be rapidly identified.
Disclosure of Invention
In order to solve the defects in the prior art, the invention aims to provide a reservoir logging characteristic parameter extraction method based on target preference coding, which has fast data processing speed and high efficiency and can meet the actual requirements of engineering.
The invention is realized by the following technical scheme:
a reservoir logging characteristic parameter extraction method based on target preference coding comprises the following steps:
step 1: according to the target reservoir evaluation characteristics, screening logging curves capable of reflecting reservoir characteristics, determining priority levels according to importance degrees, and sorting;
step 2: removing the top and bottom segment data of the target reservoir, and statistically analyzing the distribution characteristics of the residual data;
step 3: performing coding processing on the residual logging data obtained in the step 2 based on reservoir evaluation target preference;
step 4: combining the feature codes to find out a feature value sampling depth point;
step 5: and (3) extracting characteristic parameters of each logging curve of the whole section of the target reservoir according to the characteristic value sampling depth points obtained in the step (4).
Preferably, the specific steps of step 1 are:
step 1.1: analyzing the geology and oil reservoir characteristics of the research area;
step 1.2: analyzing the weight of the logging curve in the data interpretation;
step 1.3: and (3) determining the arrangement sequence of the logging curves according to the analysis result of the step 1.2.
Preferably, the specific steps of step 2 are:
step 2.1: respectively removing 5-8 data points in the top and bottom 0.6-1 meter reservoirs of the target reservoir, and taking the rest data set as a final data analysis sample;
step 2.2: counting the data of each logging curve in sequence to obtain the maximum value and the minimum value of each curve in the reservoir section;
step 2.3: the final data analysis samples are used to draw a frequency distribution histogram.
It is further preferred that in step 2.3, statistics are performed in 3 segments for reservoir thicknesses less than 2.5 meters and in 5 segments for reservoir thicknesses greater than 2.5 meters.
Preferably, the specific steps of the step 3 are as follows:
step 3.1: according to the statistical result of the step 2, counting the number of depth points in each interval;
step 3.2: each interval is encoded sequentially from more to less according to the number of depth points.
Preferably, the specific steps of step 4 are:
step 4.1: the depth points are synthesized and coded according to the logging curve, and the synthesized code value at the j-th depth is C j ,C j =c 1j c 2j *…*c mj M is the number of log curves;
step 4.2: and taking the point with the minimum code value and closest to the center of the reservoir as the characteristic value sampling depth point of the target reservoir.
Further preferably, step 4.2 is specifically: firstly, selecting a code value corresponding to a middle position of a reservoir as an initial layer characteristic value, and recording the position; then, respectively moving up and down a sampling point, taking out the comparison size of the position code value and the current characteristic value, selecting a smaller code value as a reservoir code value, modifying the position information of the code value, and keeping the current code value unchanged if the code values are the same; and pushing until the whole reservoir section is searched, and obtaining the depth position of the characteristic point of the reservoir.
Further preferably, the depth of the sampling point is 0.125 meters.
Compared with the prior art, the invention has the following beneficial technical effects:
the invention discloses a reservoir logging characteristic parameter extraction method based on target preference coding, which mainly aims to improve reservoir parameter rapid modeling efficiency and rapid fluid identification plate establishment efficiency. The characteristic value of the traditional reservoir logging curve depends on expert experience, and is long in time consumption, low in automation degree and dependent on man-machine interaction, so that the requirements of short, flat and rapid scientific research and attack can not be met. The automatic and rapid extraction method provided by the invention is equivalent to expert experience in precision and accuracy, is based on algorithm programming, simulates the manual extraction process by introducing an automatic characteristic value extraction method simulating the human value idea, achieves the same effect, automatically processes in data processing, has higher speed and higher efficiency, and can meet the actual production requirements.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a graph of reservoir characteristic value extraction effects in an embodiment;
FIG. 3 is a graph of reservoir characteristic value processing results in an embodiment;
FIG. 4 is a graph of contrast effect extracted from the time difference features of the acoustic logging in the embodiment.
Detailed Description
The invention will now be described in further detail with reference to the drawings and to specific examples, which are given by way of illustration and not limitation.
In order to quickly extract logging characteristic parameters of a reservoir in a research area, the reservoir lithology needs to be analyzed first. According to the method, a model for predicting the reservoir lithology by using the reservoir characteristic parameters is constructed by using a machine learning method, so that the reservoir lithology in the development area is predicted. To generate the training samples required for modeling, the reservoir logging characteristic parameters and the coring lithology need to be extracted at the reservoir coring interval to form the training samples. The invention discloses a target preference-based reservoir characteristic parameter extraction method, and the implementation flow of the method is shown in figure 1.
The reservoir logging characteristic parameter extraction method based on target preference coding comprises the following steps:
step 1: the geology of the research area and the characteristics of the oil deposit are analyzed, and the analysis is mainly performed based on the characteristics of the lithology of the oil deposit, the type of the oil deposit and the like; weight analysis of logging curves in data interpretation, for example, in sandstone reservoir evaluation, natural gamma is used as lithology recognition standard, the priority level is highest, the density curve reflecting porosity is inferior, the deep lateral resistivity reflecting oil-gas property is third, the acoustic time difference curve reflecting total porosity is fourth, and the natural potential curve reflecting invasion property is last; determining the order of the curve, and arranging the logging curves L according to the order 1 ,L 2 ,L 3 ,L 4 ,L 5 And a corresponding depth profile D.
Step 2: cutting out 5-8 data points of the top and bottom data points of the target layer, namely, about 0.6-1 meter reservoir, and taking the rest data set as a final data analysis sample. Counting the data of each logging curve in sequence to obtain the maximum value and the minimum value of each curve in the reservoir section; first, the maximum M of the curve segment is searched i And a minimum value m i Making a statistical histogram of current log curve data;
a frequency distribution histogram is drawn for the current dataset, typically for reservoir thicknesses less than 2.5 meters, in 3 segments, and for reservoir thicknesses greater than 2.5 meters, in 5 segments. The equally divided interval length is denoted as DeltaL i :
ΔL i =(M i -m i ) /K, k=3, or 5.
In this embodiment, the number of data points of the layer segment is greater than 20, the histogram is divided into 5 segments, and the numerical ranges of the segments are respectively:
(-∞,m i +ΔL i ],(m i +ΔL i ,m i +2ΔL i ],(m i +2ΔL i ,m i +3ΔL i ],
(m i +3ΔL i ,m i +4ΔL i ],((m i +4ΔL i ,+∞)。
step 3: logging data encoding based on the target preferences; counting the number of depth points in different areas; encoding is performed in a number of sizes from 1 to 5. Counting the data points falling into each section, and sequentially marking the data points of each section as: n (N) 1 ,N 2 ,N 3 ,N 4 ,N 5 。
Step 4: synthesizing and coding each depth point according to the logging curve; and sequencing the number of data points from large to small, wherein the interval code with the maximum number of data is 1, the code with the number of data at the 2 nd bit is 2, and the like, so as to finish the codes of all logging curves. Thus, at each depth point, each curve has a code value, denoted c, at that point ij Representing the code value of the ith curve at the jth depth. The depth points are synthesized and coded according to the logging curve, and the synthesized code value at the j-th depth is C j C, i.e j =c 1j c 2j c 3j c 4j c 5j This code value is an integer. As shown in fig. 2, after the reservoir GR log is encoded, the corresponding code is 01122. DEN, RT, AC, SP is encoded sequentially according to this method.
And selecting the characteristic points of the reservoir, wherein the point with the minimum code value and closest to the center of the reservoir is taken as the characteristic point of the target reservoir. And selecting the point with the minimum code value and closest to the center of the reservoir as the characteristic point of the target reservoir. The implementation method comprises the following steps: firstly, selecting a code value corresponding to a middle position of a reservoir as an initial layer characteristic value, and recording the position; and then, respectively moving upwards and downwards one, taking out the comparison size of the position code value and the current characteristic value, selecting a smaller code value as a reservoir code value, modifying the position information of the code value, and keeping the current code value unchanged if the code values are the same. The depth position of the characteristic point of the reservoir can be obtained by pushing until the whole reservoir section is searched
Step 5: and extracting characteristic values of the logging curve of the target reservoir section. And (3) extracting characteristic values of each logging curve of the whole section of the target reservoir according to the depth points obtained in the step (4). The depth point natural gamma, acoustic time difference, density, deep lateral direction and natural potential values are extracted, and the depth point natural gamma, acoustic time difference, density, deep lateral direction and natural potential values are the characteristic values of a target reservoir section and participate in reservoir parameter modeling and fluid identification plate modeling.
FIG. 3 is a comparison of depth locations representing target reservoir interval eigenvalues searched by the algorithm of the present invention and depth locations of empirically extracted eigenvalues by engineers. Therefore, the reservoir logging characteristic parameters rapidly extracted through the target preference codes meet the actual production requirements.
FIG. 4 is an AC curve feature value extracted by the algorithm of the present invention, which has high correlation consistency with the curve feature value manually extracted by an experienced engineer.
It is to be understood that the foregoing description is only a part of the embodiments of the present invention, and that the equivalent changes of the system described according to the present invention are included in the protection scope of the present invention. Those skilled in the art can substitute the described specific examples in a similar way without departing from the structure of the invention or exceeding the scope of the invention as defined by the claims, all falling within the scope of protection of the invention.
Claims (6)
1. The reservoir logging characteristic parameter extraction method based on target preference coding is characterized by comprising the following steps of:
step 1: according to the target reservoir evaluation characteristics, screening logging curves capable of reflecting reservoir characteristics, determining priority levels according to importance degrees, and sorting;
step 2: removing the top and bottom segment data of the target reservoir, and statistically analyzing the distribution characteristics of the residual data;
step 3: performing coding processing on the residual logging data obtained in the step 2 based on reservoir evaluation target preference;
the specific steps of the step 3 are as follows:
step 3.1: according to the statistical result of the step 2, counting the number of depth points in each interval;
step 3.2: sequentially encoding each interval from more to less according to the number of depth points;
step 4: combining the feature codes to find out a feature value sampling depth point;
the specific steps of the step 4 are as follows:
step 4.1: according toThe logging curve synthetically codes each depth point, and the synthetic code value at the j-th depth is C j ,C j =c 1j c 2j *…*c mj M is the number of log curves, c mj Representing the code value of the mth curve at the jth depth;
step 4.2: taking the point with the minimum code value and closest to the center of the reservoir as the characteristic value sampling depth point of the target reservoir;
step 5: and (3) extracting characteristic parameters of each logging curve of the whole section of the target reservoir according to the characteristic value sampling depth points obtained in the step (4).
2. The method for extracting reservoir logging feature parameters based on target preference encoding as claimed in claim 1, wherein the specific steps of step 1 are as follows:
step 1.1: analyzing the geology and oil reservoir characteristics of the research area;
step 1.2: analyzing the weight of the logging curve in the data interpretation;
step 1.3: and (3) determining the arrangement sequence of the logging curves according to the analysis result of the step 1.2.
3. The method for extracting reservoir logging characteristic parameters based on target preference encoding as claimed in claim 1, wherein the specific steps of the step 2 are as follows:
step 2.1: respectively eliminating 5-8 depth points in the top and bottom 0.6-1 m reservoirs of the target reservoir, and taking the residual data set as a final data analysis sample;
step 2.2: counting the data of each logging curve in sequence to obtain the maximum value and the minimum value of each curve in the reservoir section;
step 2.3: the final data analysis samples are used to draw a frequency distribution histogram.
4. The method for extracting reservoir logging characteristic parameters based on target preference encoding as claimed in claim 3, wherein in step 2.3, statistics are performed in 3 segments for reservoir thicknesses less than 2.5 meters and in 5 segments for reservoir thicknesses greater than 2.5 meters.
5. The method for extracting reservoir logging feature parameters based on target preference encoding as claimed in claim 1, wherein the step 4.2 is specifically: firstly, selecting a code value corresponding to a middle position of a reservoir as an initial layer characteristic value, and recording the position; then, respectively moving a depth point upwards and downwards, taking out the comparison size of the position code value and the current characteristic value, selecting a smaller code value as a reservoir code value, modifying the position information of the code value, and keeping the current code value unchanged if the code values are the same; and pushing until the whole reservoir stratum section is searched, and obtaining the depth position of the characteristic value sampling depth point of the reservoir stratum.
6. The method for extracting reservoir logging feature parameters based on target preference encoding as recited in claim 5, wherein the depth of the depth point is 0.125 meters.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010905052.9A CN111983722B (en) | 2020-09-01 | 2020-09-01 | Reservoir logging characteristic parameter extraction method based on target preference coding |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010905052.9A CN111983722B (en) | 2020-09-01 | 2020-09-01 | Reservoir logging characteristic parameter extraction method based on target preference coding |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111983722A CN111983722A (en) | 2020-11-24 |
CN111983722B true CN111983722B (en) | 2023-06-30 |
Family
ID=73448333
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010905052.9A Active CN111983722B (en) | 2020-09-01 | 2020-09-01 | Reservoir logging characteristic parameter extraction method based on target preference coding |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111983722B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114217136B (en) * | 2022-02-22 | 2022-05-06 | 山东卓朗检测股份有限公司 | Lightning protection grounding resistance detection statistical method based on big data |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111562630A (en) * | 2019-02-13 | 2020-08-21 | 中石化石油工程技术服务有限公司 | Reservoir parameter logging evaluation method based on grid division |
Family Cites Families (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6826483B1 (en) * | 1999-10-13 | 2004-11-30 | The Trustees Of Columbia University In The City Of New York | Petroleum reservoir simulation and characterization system and method |
US7249009B2 (en) * | 2002-03-19 | 2007-07-24 | Baker Geomark Llc | Method and apparatus for simulating PVT parameters |
US20070276604A1 (en) * | 2006-05-25 | 2007-11-29 | Williams Ralph A | Method of locating oil and gas exploration prospects by data visualization and organization |
CN105569647A (en) * | 2014-11-06 | 2016-05-11 | 中国石油化工股份有限公司 | Generation method for reservoir characteristic value |
CN105569648A (en) * | 2014-11-07 | 2016-05-11 | 中国石油化工股份有限公司 | Reservoir automatic identification method based on log data |
US10007015B2 (en) * | 2015-02-23 | 2018-06-26 | Nexen Energy Ulc | Methods, systems and devices for predicting reservoir properties |
RU2656303C1 (en) * | 2017-03-06 | 2018-06-04 | Открытое акционерное общество "Сургутнефтегаз" | Method for construction of geological-hydrodynamic models of non-homogeneous reservoirs with thin interlensing of sand-silt and clay rocks |
CN109343115B (en) * | 2018-11-21 | 2019-12-03 | 成都理工大学 | A kind of gas-bearing reservoir depicting method based on Log-constrained |
CN110852527B (en) * | 2019-11-20 | 2022-05-31 | 成都理工大学 | Reservoir physical property parameter prediction method combining deep learning |
-
2020
- 2020-09-01 CN CN202010905052.9A patent/CN111983722B/en active Active
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111562630A (en) * | 2019-02-13 | 2020-08-21 | 中石化石油工程技术服务有限公司 | Reservoir parameter logging evaluation method based on grid division |
Also Published As
Publication number | Publication date |
---|---|
CN111983722A (en) | 2020-11-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112989708B (en) | Well logging lithology identification method and system based on LSTM neural network | |
CN109736790A (en) | Recognition of Weil Logging Lithology method based on deepness belief network | |
CN113344050B (en) | Lithology intelligent recognition method and system based on deep learning | |
CN111749686A (en) | Drill bit rapid optimization method based on stratum drilling resistance parameters | |
CN109447249B (en) | Antagonistic neural network well logging data reconstruction method based on deep convolution | |
CN112761628B (en) | Shale gas yield determination method and device based on long-term and short-term memory neural network | |
CN111983722B (en) | Reservoir logging characteristic parameter extraction method based on target preference coding | |
CN114492521A (en) | Intelligent lithology while drilling identification method and system based on acoustic vibration signals | |
CN114154484B (en) | Construction professional term library intelligent construction method based on mixed depth semantic mining | |
CN116699096A (en) | Water quality detection method and system based on deep learning | |
CN109633748B (en) | Seismic attribute optimization method based on improved genetic algorithm | |
CN111626377A (en) | Lithofacies identification method, device, equipment and storage medium | |
CN111155980A (en) | Water flow dominant channel identification method and device | |
CN115629188A (en) | Rock core productivity simulation experiment system | |
CN114896468A (en) | File type matching method and intelligent data entry method based on neural network | |
CN114202036A (en) | Static sounding automatic soil layer dividing technology based on pattern recognition technology | |
CN116542345A (en) | Machine learning model training method for logging lithology prediction | |
CN115438090A (en) | Borehole wall instability type analysis and prediction method and system based on lost block image characteristics | |
Feder | Machine-learning approach determines spatial variation in shale decline curves | |
CN117407841B (en) | Shale layer seam prediction method based on optimization integration algorithm | |
CN115983088B (en) | Sea shale in-situ gas quantity prediction model established based on ensemble learning | |
CN114676631B (en) | Shale gas reservoir fracturing horizontal well fracturing transformation parameter recommendation method | |
CN110659685B (en) | Well position optimization method based on statistical error active learning | |
CN117391483A (en) | Shale oil well fracturing effect evaluation method based on cooperative training | |
CN116502113A (en) | Sand shale recognition method based on XGBoost algorithm |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |