CN112034515B - Volcanic channel identification method based on unsupervised neural network - Google Patents
Volcanic channel identification method based on unsupervised neural network Download PDFInfo
- Publication number
- CN112034515B CN112034515B CN202010816088.XA CN202010816088A CN112034515B CN 112034515 B CN112034515 B CN 112034515B CN 202010816088 A CN202010816088 A CN 202010816088A CN 112034515 B CN112034515 B CN 112034515B
- Authority
- CN
- China
- Prior art keywords
- volcanic
- seismic
- sample
- layer
- neural network
- 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
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 24
- 238000012549 training Methods 0.000 claims abstract description 11
- 210000002569 neuron Anatomy 0.000 claims description 27
- 238000010606 normalization Methods 0.000 claims description 14
- 238000012545 processing Methods 0.000 claims description 13
- 230000000694 effects Effects 0.000 claims description 10
- 238000007405 data analysis Methods 0.000 claims description 3
- 210000004205 output neuron Anatomy 0.000 claims description 3
- 230000002860 competitive effect Effects 0.000 claims 1
- 238000011161 development Methods 0.000 abstract description 7
- 238000005516 engineering process Methods 0.000 abstract description 5
- 230000008859 change Effects 0.000 abstract description 4
- 238000011160 research Methods 0.000 abstract description 4
- 239000003208 petroleum Substances 0.000 abstract description 2
- 239000011435 rock Substances 0.000 abstract description 2
- 230000000739 chaotic effect Effects 0.000 abstract 1
- 238000004364 calculation method Methods 0.000 description 5
- 208000035126 Facies Diseases 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 4
- 238000001514 detection method Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000003672 processing method Methods 0.000 description 3
- 230000007480 spreading Effects 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 238000003708 edge detection Methods 0.000 description 2
- 230000003631 expected effect Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V1/00—Seismology; Seismic or acoustic prospecting or detecting
- G01V1/28—Processing seismic data, e.g. for interpretation or for event detection
- G01V1/30—Analysis
- G01V1/306—Analysis for determining physical properties of the subsurface, e.g. impedance, porosity or attenuation profiles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V2210/00—Details of seismic processing or analysis
- G01V2210/60—Analysis
- G01V2210/62—Physical property of subsurface
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Remote Sensing (AREA)
- Acoustics & Sound (AREA)
- Environmental & Geological Engineering (AREA)
- Geology (AREA)
- General Life Sciences & Earth Sciences (AREA)
- Geophysics (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention discloses a volcanic channel identification method based on an unsupervised neural network, wherein the interior seismic data of the volcanic channel has poor quality and very quick transverse change, and is generally represented as a chaotic reflection characteristic with weak amplitude; therefore, the research on the spatial distribution characteristics of the volcanic channels is always a difficult point of geophysical research. With the continuous promotion of petroleum exploration and development work, particularly in the middle and deep oil fields with igneous rock development, the fine identification of volcanic channels is more important; the depiction of the volcanic channel directly influences the deployment of the subsequent development well position. Therefore, a volcanic channel identification technology based on an unsupervised neural network is developed: the method and the opposite propagation neural network are deeply fused to form an unsupervised intelligent learning model, and the model is used for training to obtain new attributes describing the volcanic channel, so that the full three-dimensional automatic interpretation of the volcanic channel is realized.
Description
Technical Field
The invention belongs to the technical field of processing and explaining of petroleum exploration seismic data, and particularly relates to a volcanic channel identification method based on an unsupervised neural network.
Background
Due to the influence of volcanic structure motion, the quality of seismic data in the volcanic channel is poor, and the transverse change is very quick; clutter, which typically manifests as weak amplitudes; therefore, the research on the spatial distribution characteristics of the volcanic channels is always a difficult point of geophysical research. With the continuous advance of oil exploration and development work, particularly in the middle and deep oil fields with igneous rock development, the fine identification of volcanic channels directly influences the deployment of subsequent development well positions.
The method for describing the volcanic channel distribution characteristics is various, and mainly comprises conventional seismic attribute detection, seismic phase analysis and seismic inversion. The conventional seismic attribute detection mainly uses edge detection technologies such as variance, curvature and the like to depict the distribution range of volcanic channels, the dependence degree of the calculation result of the method on the quality of seismic data is high, and the attribute easily brings many interpretation artifacts due to poor quality of the seismic data around the volcanic channels. The seismic facies analysis technology is developed on the basis of seismic stratigraphy, and mainly comprises two methods such as waveform classification and seismic structure attribute, wherein the change characteristics of sedimentary facies belts are analyzed mainly by distinguishing reflection characteristic differences among seismic channels. Seismic inversion is an important means for predicting longitudinal spread and transverse spread characteristics of a reservoir and an abnormal geologic body, but a conventional inversion method depending on a geologic model has great limitations, and a good inversion result can be obtained only by requiring an accurate geologic model, but the inversion method is difficult to provide in practice, so that the inversion result cannot achieve an expected effect.
Disclosure of Invention
The invention aims to provide a volcanic channel identification method based on an unsupervised neural network, so as to solve the problems of the background art.
In order to achieve the purpose, the specific technical scheme of the volcanic channel identification method based on the unsupervised neural network is as follows:
a volcanic channel identification method based on an unsupervised neural network comprises the following steps:
seismic attribute processing: extracting seismic attributes sensitive to volcanic channels, and normalizing each seismic attribute;
a step of extracting sample points: preferentially selecting an area with obvious volcanic channel seismic reflection characteristics from an original seismic profile, selecting points of the area as sampling points, and extracting normalized seismic attributes corresponding to the sampling points;
acquiring an optimal seismic attribute: classifying and judging sample data types by adopting an iterative self-organizing clustering method, and optimizing sample training data; secondly, the method is fused with an opposite propagation neural network to form an unsupervised intelligent learning model; and obtaining the optimal new attribute of describing the volcanic passage by using the model.
Volcanic passage identification: and applying the newly generated optimal seismic attributes to the whole area, and further depicting the spatial distribution range of the volcanic channels in the whole area.
Further, the seismic attributes include curvature, variance, ant body, dip, azimuth, texture attributes.
Further, the processing method formula in the seismic attribute processing step is as follows;
wherein x is i For each seismic attribute, x i * Normalized seismic attributes.
Further, the step of extracting sample points is as follows: areas with obvious volcanic channel and non-volcanic channel reflection characteristics are selected from the seismic data; meanwhile, P points in the area are selected as sample points, normalized seismic attributes corresponding to the sample points are extracted, and data analysis is carried out by utilizing an iterative self-organizing clustering method. Through multiple iterations and successive updating, the optimal clustering effect of the sample data is obtained, and unsupervised classification and judgment of the sample data are realized.
Further, the processing method in the step of obtaining the optimal seismic attribute comprises the following steps:
and taking sample data of unsupervised classification discrimination as input data of the counter propagation neural network. At this time, it is assumed that the input layer has N neurons, the competition layer has Q neurons, and the output layer has M neurons; the input mode of the sample point is
The corresponding competition layer output vector is
The output layer outputs a vector of
The target output vector is
(where k = g)1, 2.. P); the connection weight between the input layer and the competition layer is W j ＝(w j1 ,w j2 ,...w jN ) j =1,2,. Q; the connection weight between the competition layer and the output layer is V l ＝(v l1, v l2 ,...v lQ I =1,2,. M; the method comprises the following steps:
1. inputting the kth sample point into the pattern
Providing to a network input layer;
2. weight W of connection j Carrying out normalization treatment:
3. the weighted input sum of each neuron in the competition layer is:
4. calculating a connection weight W j And A k Nearest vector W g ：
At this time, neuron g is the winning neuron, whose output is set to 1, and the remaining competing neurons have outputs of 0:
5. correcting connection weight W j The following:
6. handle W g Carrying out normalization processing again;
7. modifying the connection weight vector V from the competition layer to the output layer l The formula is as follows:
v li (t+1)＝v li (t)+βb j (c l -c l ′) l＝1,2,...,M 0＜β＜1,j＝1,2,...Q (7)
from step 4, equation 7 can be simplified to:
v lg (t+1)＝v lg (t)+βb j (c l -c l ') l =1,2,. Multidot.M 0 < beta < 1 (learning rate) (8)
At this time, the weight V for connecting the competition layer neuron g to the output layer neuron is adjusted g ；
8. The actual output values of the output neurons are:
9. returning to the step 1 until the input mode A of P sample points is input k All input is carried out;
10. t = t +1, and A k The neural network training result is provided for the network learning again until T = T and the neural network training result is converged, and the effect is better when T is larger than 5000 generally;
at this time, the best new seismic attributes capable of reflecting the spatial distribution characteristics of the volcanic channels are obtained.
Further, the step W in step 6 of the processing method g And (4) carrying out normalization processing again, wherein the formula of the normalization method is as follows:
compared with the prior art, the invention has the following beneficial effects:
the method provided by the invention has low dependence on the quality of seismic data, and the interpretation artifact of the volcanic channel is reduced to the maximum extent by the calculation result. The plane distribution range of the volcanic channel can be well identified, and the longitudinal distribution characteristics of the volcanic channel are accurately described.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 (a) is a verification diagram of a volcanic channel forward model added with random noise;
FIG. 2 (b) is the result obtained using this method in the validation of the model;
FIG. 3 (a) is an original seismic section of a certain oil field in Bohai sea in the embodiment;
FIG. 3 (b) is a diagram showing a first effect of the process on the volcanic channel in practical application;
fig. 3 (c) is a second effect diagram of the process on the volcanic channel in practical application.
Detailed Description
For a better understanding of the objects, structure and function of the invention, reference should be made to the drawings, FIGS. 1-3, which illustrate the invention.
At present, the most common method for characterizing the spatial distribution of volcanic tunnels at home and abroad mainly comprises the following steps: conventional seismic attribute detection, seismic phase analysis, and seismic inversion.
The conventional seismic attribute detection mainly uses edge detection technologies such as variance, curvature and the like to depict the distribution range of volcanic channels, the dependence degree of the calculation result of the method on the quality of seismic data is high, and the attribute easily brings many interpretation artifacts due to poor quality of the seismic data around the volcanic channels.
The seismic facies analysis technology is developed on the basis of seismic stratigraphy, and mainly comprises two methods such as waveform classification and seismic structure attribute, wherein the change characteristics of sedimentary facies belts are analyzed mainly by distinguishing reflection characteristic differences among seismic channels.
Seismic inversion is an important means for predicting longitudinal spreading and transverse spreading characteristics of reservoirs and abnormal geologic bodies, but a conventional inversion method depending on a geologic model has great limitation, and a good inversion result can be obtained only by an accurate geologic model, but the inversion result is difficult to provide in practice, so that the inversion result cannot achieve the expected effect.
Conventional neural network algorithms require a large amount of labeled data as training samples and specify the desired output and correct input. The data capacity of the sample is large, the marking of the sample needs to be completed in a manual mode, the workload is huge, and a large amount of time needs to be consumed. Due to objective errors caused by subjective human factors and lack of effective constraints, calculation results are not converged frequently, and a good prediction effect cannot be obtained by utilizing a traditional neural network algorithm.
And the iterative self-organizing clustering method is adopted to classify and judge the sample data, the sample training data is optimized, and the problems that a large number of marked training samples are needed in the conventional algorithm and the artificial deviation is large are solved. The method and the opposite propagation neural network are deeply fused to form an unsupervised intelligent learning model; the stability of the new attribute of the depicted volcanic channel obtained by using the model is high; the problems that the conventional volcanic channel prediction method is high in dependence degree on seismic data quality and difficult in longitudinal spread characteristic prediction are solved, and full three-dimensional automatic interpretation of the volcanic channel is realized.
Referring to fig. 1, the main steps of the method of the present invention include the following:
(1) Seismic attributes (curvature, variance, ant body, inclination angle, azimuth angle, texture and other attributes) which are sensitive to volcanic channels are extracted, and normalization processing is carried out on each seismic attribute.
(2) And preferably selecting an area with obvious volcanic channel seismic reflection characteristics from the original seismic profile, selecting points of the area as sample points, and extracting the normalized seismic attributes corresponding to the sample points.
(3) Classifying and judging sample data types by adopting an iterative self-organizing clustering method, and optimizing sample training data; secondly, fusing the method with an opposite propagation neural network to form an unsupervised intelligent learning model; and obtaining the optimal new attribute for describing the volcanic channel by using the model.
(4) And applying the newly generated optimal seismic attributes to the whole area, and further depicting the spatial distribution range of the volcanic tunnels in the whole area.
The method comprises the following specific steps: firstly, extracting seismic attributes (curvature, variance, ant body, inclination angle, azimuth angle, texture and other seismic attributes) sensitive to volcanic channels;
and carrying out normalization processing on each seismic attribute;
wherein x i For each seismic attribute, x i * Is the normalized seismic attribute.
Then, areas with obvious volcanic channel and non-volcanic channel reflection characteristics are selected from the seismic data; meanwhile, P points in the area are selected as sample points, and normalized seismic attributes corresponding to the sample points are extracted to perform data analysis by using an iterative self-organizing clustering method. Through multiple iterations and successive updating, the optimal clustering effect of the sample data is obtained, and the unsupervised classification and judgment of the sample data are realized, and at the moment, the sample data subjected to unsupervised classification and judgment are used as input data of the opposite propagation neural network; assume that the input layer has N neurons, the competition layer has Q neurons, and the output layer has M neurons. The input mode of the sample point is
The corresponding competition layer output vector is
The output layer outputs a vector of
The target output vector is
(where k =1,2.. P). The connection weight between the input layer and the competition layer is W j ＝(w j1 ,w j2 ,...w jN ) j =1,2.. Q; the connection weight between the competition layer and the output layer is V l ＝(v l1 ,v l2 ,...v lQ ,)l＝1,2.. The method has the following learning rule:
1. inputting the kth sample point
Provided to the network input layer.
2. Weight W of connection j Carrying out normalization treatment:
3. the weighted input sum of each neuron in the competition layer is:
4. calculating connection weight W j And A k Nearest vector W g ：
At this time, neuron g is the winning neuron, the output is set to 1, and the outputs of the remaining competing neurons are 0:
5. correcting connection weight W j The following were used:
6. handle W g And (4) carrying out normalization again, wherein the normalization method is the same as the formula (2).
7. Modifying the connection weight vector V from the competition layer to the output layer l The formula is as follows:
v li (t+1)＝v li (t)+βb j (c l ＝c l ′) l＝1,2,...,M 0＜β＜1,j＝1,2,...Q (7)
from step 4, equation 7 can be simplified as:
v lg (t+1)＝v lg (t)+βb j (c l -c l ') l =1,2,. Multidot.M 0 < β < 1 (learning rate) (8)
At this time, the weight V for connecting the competition layer neuron g to the output layer neuron is adjusted g 。
8. The actual output values of the output neurons are:
9. returning to the step 1 until the P sample points are input into the mode A k All inputs are entered.
10. T = t +1, and A k And (4) providing the neural network for learning again until T = T and the result of the neural network training is converged, wherein the effect of T larger than 5000 is better generally. The optimal new seismic attribute capable of reflecting the spatial distribution characteristics of the volcanic channels is obtained.
And finally, applying the newly generated optimal seismic attributes to the whole area, and further depicting the spatial distribution range of the volcanic channels in the whole area.
Wherein, fig. 2 (a) is a volcanic channel forward modeling model added with random noise, and the detail features of the volcanic channel are annihilated by the noise;
FIG. 2 (b) is a result obtained by using the method, which better depicts the spatial distribution characteristics of volcanic channels, and has better identification effect on both volcanic mouths and volcanic channels; the effectiveness of the technical process is verified from a theoretical model.
Fig. 3 (a) is an original seismic section of a certain oil field in the bohai sea, the quality of seismic data of a target layer section in the area is poor, and the identification of the longitudinal spreading characteristics of a volcanic channel is difficult. The longitudinal distribution range of the volcanic channel is well described by utilizing the technical process, and the edge characteristics of the volcanic channel are well identified, such as fig. 3 (b) and fig. 3 (c).
The method provided by the invention has low dependence on the quality of seismic data, and the interpretation artifact of the volcanic channel is reduced to the maximum extent by the calculation result. The plane distribution range of the volcanic channel can be well identified, and the longitudinal distribution characteristics of the volcanic channel are accurately described.
The result calculated by the method accurately describes the spatial distribution range of the ancient and near volcanic tunnels of the 34-9 oil field in the Bohai, effectively avoids the risk area of the volcanic tunnels, and successfully guides the smooth implementation of well position development.
It is to be understood that the present invention has been described with reference to certain embodiments, and that various changes in the features and embodiments, or equivalent substitutions may be made therein by those skilled in the art without departing from the spirit and scope of the invention. In addition, many modifications may be made to adapt a particular situation or material to the teachings of the invention without departing from the essential scope thereof. Therefore, it is intended that the invention not be limited to the particular embodiment disclosed, but that the invention will include all embodiments falling within the scope of the appended claims.
Claims (1)
1. A volcanic channel identification method based on an unsupervised neural network is characterized by comprising the following steps:
seismic attribute processing: extracting seismic attributes sensitive to volcanic channels, and normalizing each seismic attribute;
a step of extracting sample points: selecting an area with seismic reflection characteristics of an obvious volcanic channel and a non-volcanic channel from an original seismic profile, selecting points in the area as sample points, and extracting seismic attributes after normalization processing corresponding to the sample points;
acquiring an optimal seismic attribute: classifying and judging sample data types by adopting an iterative self-organizing clustering method, and optimizing sample training data; fusing an iterative self-organizing clustering method and an opposite propagation neural network to form an unsupervised intelligent learning model; obtaining the optimal seismic attribute describing the volcanic channel by using the unsupervised intelligent learning model;
volcanic passage identification: applying the newly generated optimal seismic attributes to the whole area, and further drawing the longitudinal section distribution range of the volcanic channel in the whole area;
the seismic attributes comprise curvature, variance, ant body, inclination angle, azimuth angle and texture attributes;
the normalization processing formula in the seismic attribute processing step is as follows;
wherein x is i* For each seismic attribute, x i* * For the normalized seismic attributes, i represents the seismic attribute sample sequence number, and n represents the total sample number;
further comprising: selecting areas with obvious volcanic channel and non-volcanic channel seismic reflection characteristics from the original seismic section; meanwhile, selecting P points in the area as sample points, extracting seismic attributes after normalization processing corresponding to the sample points, and performing data analysis by using an iterative self-organizing clustering method; through multiple iterations and successive updating, the optimal clustering effect of the sample data is obtained, and unsupervised classification and discrimination of the sample data are realized; further comprising: sample data judged by unsupervised classification is used as input data of the opposite propagation neural network; at this time, it is assumed that the input layer has N neurons, the competition layer has Q neurons, and the output layer has M neurons; the input mode of the sample point is
The corresponding competition layer output vector is
The output layer outputs a vector of
The target output vector is
Wherein k =1,2.. P; the connection weight between the input layer and the competition layer is W j ＝(w j1 ,w j2 ,...w jN ) J =1,2,. Q; the connection weight value between the competition layer and the output layer is V l ＝(v l1 ,v l2 ,...v lQ L =1,2,... M; the method comprises the following steps:
(1) Inputting the k-th sample point into the pattern
Providing the data to a network input layer;
(2) Weight W of connection j Carrying out normalization treatment:
(3) The weighted input sum of each neuron in the competitive layer is:
(4) Calculating the connection weight W j And A k Nearest vector W g ：
At this time, neuron g is a winning neuron whose output is set to 1, and the remaining competing neurons output 0:
(5) Correction vector W g The following:
(6) Handle vector W g Carrying out normalization processing again;
(7) Correcting connection weight V from competition layer to output layer l The formula is as follows:
v li (t+1)＝v li (t)+βb j (c l -c' l )，l＝1,2,...,M，0＜β＜1，j＝1,2,...Q(7)
from step (4), equation (7) can be simplified to:
v lg (t+1)＝v lg (t)+βb j (c l -c' l ) L =1, 2.. Beta.0 < 1. Beta. Is the learning rate (8)
At this time, the connecting weight V of the winning neuron g to the output layer neuron is adjusted g ；
(8) The actual output values of the output neurons are:
(9) Returning to the step (1) until the P sample points are input into the mode A k All inputs are carried out;
(10) T = t +1, and A k The neural network training result is provided for the network learning again until T = T, and the effect is better when T > 5000; at this time, the optimal seismic attributes capable of reflecting the spatial distribution characteristics of the volcanic channels are obtained.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010816088.XA CN112034515B (en) | 2020-08-14 | 2020-08-14 | Volcanic channel identification method based on unsupervised neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010816088.XA CN112034515B (en) | 2020-08-14 | 2020-08-14 | Volcanic channel identification method based on unsupervised neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112034515A CN112034515A (en) | 2020-12-04 |
| CN112034515B true CN112034515B (en) | 2023-02-24 |
Family
ID=73578180
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010816088.XA Active CN112034515B (en) | 2020-08-14 | 2020-08-14 | Volcanic channel identification method based on unsupervised neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112034515B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113281810B (en) * | 2021-05-07 | 2022-11-11 | 中国石油天然气股份有限公司 | Volcanic channel identification and division method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102609764A (en) * | 2012-02-01 | 2012-07-25 | 上海电力学院 | CPN neural network-based fault diagnosis method for stream-turbine generator set |
| CN106707340A (en) * | 2016-12-13 | 2017-05-24 | 中国石油天然气股份有限公司大港油田分公司 | Method for predicting volcanic rock facies |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104007468B (en) * | 2014-05-23 | 2016-06-15 | 中国地质大学（武汉） | The method of volcanics space exhibition cloth is portrayed based on amplitude-differential body seismic properties |
| US10139507B2 (en) * | 2015-04-24 | 2018-11-27 | Exxonmobil Upstream Research Company | Seismic stratigraphic surface classification |
| CN111239824A (en) * | 2020-04-03 | 2020-06-05 | 辽宁工程技术大学 | Volcanic lithofacies prediction method |
-
2020
- 2020-08-14 CN CN202010816088.XA patent/CN112034515B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102609764A (en) * | 2012-02-01 | 2012-07-25 | 上海电力学院 | CPN neural network-based fault diagnosis method for stream-turbine generator set |
| CN106707340A (en) * | 2016-12-13 | 2017-05-24 | 中国石油天然气股份有限公司大港油田分公司 | Method for predicting volcanic rock facies |
Non-Patent Citations (7)
| Title |
|---|
| Study of carboniferous volcanic lithology and lithofacies in Zhongguai uplift of Junggar Basin;Chen Yang et al.;《CPS/SEG Beijing 2014 International Geophysical Conference》;20141231;第823-825页 * |
| 反传网络 ( CPN )的改进及其在储层预测中的应用;何平等;《成都理工学院学报》;19980331;第25卷;第76-78页 * |
| 基于对向传播神经网络的水电机组振动故障诊断研究;彭文季等;《西安理工大学学报》;20061231;第22卷(第4期);第365-368页 * |
| 基于样本自组织聚类的 BP 神经网络预测模型;杜晓亮等;《计算机工程与应用》;20091231;第45卷(第21期);第167-169页 * |
| 汽轮机典型非线性振动故障诊断***的开发研究;王亮;《中国优秀博硕士学位论文全文数据库（硕士） 工程科技Ⅱ辑》;20081115;第28-36页 * |
| 火山岩天然气藏评价技术研究;王玲等;《石油地球物理勘探》;20081031;第43卷(第05期);第540-548页 * |
| 用人工神经网络识别火成岩;张世晖等;《石油地球物理勘探》;20031130;第38卷;第84-87页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112034515A (en) | 2020-12-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111783825B (en) | Logging lithology recognition method based on convolutional neural network learning | |
| CN109709603B (en) | Seismic horizon identification and tracking method and system | |
| US6438493B1 (en) | Method for seismic facies interpretation using textural analysis and neural networks | |
| CN112083498B (en) | Multi-wave earthquake oil and gas reservoir prediction method based on deep neural network | |
| CN109799533A (en) | A kind of method for predicting reservoir based on bidirectional circulating neural network | |
| CN107688201B (en) | RBM-based seismic prestack signal clustering method | |
| CN110609320B (en) | Pre-stack seismic reflection pattern recognition method based on multi-scale feature fusion | |
| US20050043891A1 (en) | System and method for enhanced hydrocarbon recovery | |
| US20010037178A1 (en) | System and method for delineating spatially dependent objects, such as hydrocarbon accumulations from seismic data | |
| US20020183932A1 (en) | Method for mapping seismic attributes using neural networks | |
| CN109001801B (en) | Fault variable-scale identification method based on multiple iteration ant colony algorithm | |
| CN109184677A (en) | Reservoir evaluation methods for heterogeneous alternating layers sand body | |
| CN115308799B (en) | Method and system for identifying free gas structure of seismic imaging | |
| CN106443774A (en) | Method for improving pre-stack depth migration imaging precision of irregular earth surface | |
| CN113486591B (en) | Gravity multi-parameter data density weighted inversion method for convolutional neural network result | |
| CN110609327B (en) | Carbonate reservoir facies prediction method and device based on pre-stack seismic attributes | |
| CN111722283B (en) | Stratum velocity model building method | |
| CN114723095A (en) | Missing well logging curve prediction method and device | |
| CN113917527A (en) | Method for detecting gas content based on multiple quantum neural network | |
| CN112034515B (en) | Volcanic channel identification method based on unsupervised neural network | |
| CN105986819B (en) | The method and apparatus with integrated interpretation are automatically processed for well-log information | |
| CN113033637B (en) | Integrated learning lithofacies prediction method based on rock physical sample enhancement | |
| CN111894563A (en) | Method and system for determining classification of fractured reservoir in buried hill section | |
| CN114488297A (en) | Fault identification method and device | |
| Wang et al. | Auto recognition of carbonate microfacies based on an improved back propagation neural network |
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 |