CN106372402B - The parallel method of fuzzy region convolutional neural networks under a kind of big data environment - Google Patents
The parallel method of fuzzy region convolutional neural networks under a kind of big data environment Download PDFInfo
- Publication number
- CN106372402B CN106372402B CN201610762101.1A CN201610762101A CN106372402B CN 106372402 B CN106372402 B CN 106372402B CN 201610762101 A CN201610762101 A CN 201610762101A CN 106372402 B CN106372402 B CN 106372402B
- Authority
- CN
- China
- Prior art keywords
- fuzzy
- convolutional neural
- neural networks
- data
- fuzzy region
- 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
Classifications
-
- 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
- G06N3/043—Architecture, e.g. interconnection topology based on fuzzy logic, fuzzy membership or fuzzy inference, e.g. adaptive neuro-fuzzy inference systems [ANFIS]
-
- 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
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Z—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS, NOT OTHERWISE PROVIDED FOR
- G16Z99/00—Subject matter not provided for in other main groups of this subclass
-
- 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
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Computing Systems (AREA)
- Life Sciences & Earth Sciences (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)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- Automation & Control Theory (AREA)
- Computational Mathematics (AREA)
- Fuzzy Systems (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Image Analysis (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention proposes a kind of parallel methods of fuzzy region convolutional neural networks under big data environment, firstly, building fuzzy region convolutional neural networks, will provide goal hypothesis region and target identification are put into the same network, shared convolutional calculation, a training process update the weight of whole network；Next, the log data collection of input is divided into several small data sets, multiple workflow parallelizations carry out the operation of convolution sum pondization by fuzzy region convolutional neural networks, and each small data set is trained using only gradient decline.The present invention optimizes network structure and parameter, realizes preferably analysis performance and precision；Moreover, the present invention makes the feature extracted preferably reflect the characteristic of oil and gas reservoir itself, can solve log data fuzzy problem for the number of plies of different log data collection adjustment FR-CNN blurrings；The present invention carries out parallel training and the execution of FR-CNN using more GPU, to improve the efficiency of FR-CNN.
Description
Technical field
The present invention relates to petroleum well logging technology fields, more particularly to big data well logging field.
Background technique
Well logging information and deposition are the reflection and governing factor of formation rock physical property, therefore well-log information is all the time
By the important information source as basis in oil and gas reservoir sedimentology research, well logging phase is then well logging information and Reservoir Sedimentological
Bridge between feature.For most oil/gas well, well-log information is the comprehensive letter on only full well section stratum of covering
Source is ceased, therefore well-log facies recognition analysis method is used as always a most important research in oil exploration and exploitation geological research
Means.
However, well logging information has the characteristics that ambiguity, multi-solution and ambiguity with geological Significance.Therefore, it logs well
The identification and analysis of phase must be set up in a large amount of existing deposition characteristics and log parameter relationship (log response) comprehensive depth point
On analysis basis, while referring also to outcrop, core log and earthquake analysis as a result, choosing is suitble to geology characteristic to build
Mould method is just able to achieve accurately identifying for well logging phase.
Further, since lacking effective well logging phase automatic identifying method and technology, current well-log facies recognition is mainly logical
The manual identified realization of geological work personnel is crossed, and due to the System level gray correlation of personnel's experience difference, subjective differences, log data
The factors such as different, the data volume that geological personnel faces is big, workload is heavy.Moreover, the experience difference of geological personnel, it is subjective because
The factors such as plain, different times difference instrument log data systematical difference, so that traditional big discounting of well-log facies recognition accuracy
Button.
It is that solution current oil industry is big that the advanced technologies such as big data analysis, deep learning, which are applied to oil-gas geology research,
The exploration and trial of data analysis resources idle.In recent years, petroleum industry establishes a large amount of cloud data center, but utilization rate is not
Height, resource is by serious waste.One of major reason is just a lack of big data processing platform and corresponding big data technology
To make full use of these calculating, storage resource.
Establish the urgent need that efficient, accurate well-log facies recognition method is present oil-gas geology research.
Summary of the invention
To solve the deficiencies in the prior art, the invention proposes fuzzy region convolutional neural networks under a kind of big data environment
Parallel method.
The technical scheme of the present invention is realized as follows:
The parallel method of fuzzy region convolutional neural networks under a kind of big data environment, firstly, building fuzzy region volume
Product neural network, will provide goal hypothesis region and target identification is put into the same network, share convolutional calculation, a training
The weight of process update whole network；
Next, the log data collection of input is divided into several small data sets, multiple workflow parallelizations are by fuzzy
Region convolutional neural networks carry out the operation of convolution sum pondization, and each small data set is trained using only gradient decline；Training
After the completion, result is output to waiting list, after the completion of a wheel training, reads output queue, carry out the synchronization of shared weight
Operation is updated, after the completion of update, carries out next round training；In each round training, for the meter of the small data set of each segmentation
It calculates, is all the asynchronous progress on distributed basis, often calculates gradient value, be just appended in list, when all small
After data set all calculates, then the weight and bias of synchronized update fuzzy region convolutional neural networks carry out next round
Training；In terms of parallelization identification, log data is collected by Spout, it is then that data distribution is parallel into each Bolt node
Well-log facies recognition is carried out, recognition result is input in next Bolt node by each Bolt node, counts object letter therein
Breath；
The step of each small data set carries out the operation of convolution sum pondization by fuzzy region convolutional neural networks, it is specific to wrap
It includes: convolutional layer and the interaction of pond layer, in convolutional layer and pond layer progress fuzzy operation, from the of fuzzy region convolutional neural networks
One layer starts, and gradually increases the number of plies of blurring, for the different data set adjustment blurring numbers of plies, fuzzy region convolutional Neural
The last layer of network obtains feature vector, and this feature vector passes through a sliding window for Feature Mapping to a low-dimensional vector
In, two full articulamentums are then input the feature into, a full articulamentum is used to position, another full articulamentum is used to classify.
Optionally, the convolutional layer formula expression are as follows:
The expression of pond layer formula are as follows:
Wherein, it biasesAnd weightIt is fuzzy number, used here as symmetrical triangular fuzzy numbers,For
The vector of fuzzy number composition, j-th of fuzzy numberMembership function are as follows:
Optionally, in the training process of fuzzy region convolutional neural networks, an associated losses function is defined:
Wherein, piIt is the prediction probability that this sample is form of logs,It is the label of sample, is surveyed if it is corresponding
Well tracing pattern,It is 1, otherwiseIt is 0, NclsIt is the loss of two sorted logics；tiIt is the four parameters composition for predicting object boundary
Vector,For the vector of tab area parameter composition, they are respectively as follows:
tx=(x-xa)/wa th=(y-ya)/ha
tw=log (w/wa) th=log (h/ha)
Wherein, x, y, w and h respectively represent the centre coordinate, width and length of object, x, xa, x*Respectively represent Target area
Domain, anchor region and tab area return lossR is smooth loss function
It indicates only when anchor region is positive sampleWhen, it just calculates and returns loss, otherwiseIt does not calculate,
Normalized parameter NclsAnd NregRespectively represent the quantity of the length and anchor region from the low-dimensional vector of maps feature vectors.
Optionally, the standardization for carrying out log data first, is converted into index without dimension assessment value for initial data,
Each index assessment value is all in the same number of levels, then carries out comprehensive test analysis.
Optionally, the standardization for carrying out log data uses following normalization method:
Sx=(x-M)/S, x ∈ GR, AC, DEN, CNL, SDN ... }
Wherein, x indicates the data of every log, and Sx indicates that the borehole log data after standardization, M are corresponding well logging
The mean value of curve data, S are the standard deviation of every borehole log data.
The beneficial effects of the present invention are:
(1) feature fuzzy according to data in well logging big data, incorporates fuzzy theory, proposes fuzzy region convolutional Neural net
Network FR-CNN, and propose progressive fuzzy method, since the first layer of fuzzy region convolutional neural networks, gradually increase fuzzy
The number of plies of change realizes preferably analysis performance and precision to optimize network structure and parameter；
(2) for the number of plies of different log data collection adjustment FR-CNN blurrings, reflect the feature extracted preferably
The characteristic of oil and gas reservoir itself can solve log data fuzzy problem；
(3) present invention carries out parallel training and the execution of FR-CNN using more GPU, to improve the efficiency of FR-CNN.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 is the structural schematic diagram of fuzzy region convolutional neural networks of the present invention；
Fig. 2 is symmetrical triangular fuzzy numbers coordinate schematic diagram；
Fig. 3 is the schematic diagram that fuzzy region convolutional neural networks parallelization of the present invention handles real time data.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
The reason of log data has the characteristics that ambiguity, causes this ambiguity is various, including noise, different
The data space pollution of log data caused by cause property, imperfection etc. also includes different times, different instrument well logging brings
Systematic data difference, these problems bring log data ambiguity all constrain accurately identifying for well logging phase.
The invention proposes a kind of parallel methods of fuzzy region convolutional neural networks under big data environment, to well logging number
According to the data space for constructing various dimensions, fuzzy theory is merged with deep learning network R-CNN, proposes to solve fuzzy data feelings
The recognition methods for phase of logging well under condition incorporates fuzzy theory according to the feature that data in well logging big data are fuzzy, proposes fuzzy region
Convolutional neural networks FR-CNN (Fuzzy R-CNN), it is further proposed that progressive blur method, is opened from convolutional neural networks first layer
Begin, gradually increase the number of plies of blurring, optimize network structure and parameter, finally establish the theory and method of FR-CNN, realizes more
Good analysis performance and precision, meanwhile, the present invention carries out parallel training and the execution of FR-CNN using more GPU, to improve FR-
The efficiency of CNN.
Designing suitable fuzzy region convolutional neural networks is emphasis of the invention, below to fuzzy region convolution of the present invention
The building of neural network is described in detail.
Fuzzy region convolutional neural networks FR-CNN is built on the basis of deep learning network R-CNN, such as Fig. 1 institute
Show, FR-CNN will provide goal hypothesis region and target identification is put into the same network, share convolutional calculation, avoid complexity
Calculate step, it is only necessary to which a training process can update the weight of whole network, while also accelerate detection speed, reach fast
The purpose of speed processing.
In Fig. 1, log data passes through fuzzy region convolutional neural networks, carries out the operation of convolution sum pondization.Fuzzy region volume
The core of product neural metwork training is the interaction of convolutional layer and pond layer, therefore carries out fuzzy behaviour in convolutional layer and pond layer
Make.In order to avoid fuzzy excessively caused information loss is excessive, and in view of fuzzy region convolutional neural networks extract feature
Fining degree successively reduce, change blurring of the traditional fuzzy neural network to each layer here, the present invention proposes progressive
Fuzzy method gradually increases the number of plies of blurring, for difference that is, since the first layer of fuzzy region convolutional neural networks
Data set adjustment blurring the number of plies, make extract feature preferably reflect the characteristic of log, to obtain best identified
As a result, and improving recognition efficiency.
The last layer of fuzzy region convolutional neural networks obtains feature vector, and this feature vector passes through a small sliding
Feature Mapping into a low-dimensional vector, is then input the feature into two full articulamentums by window, and a full articulamentum is used to
Positioning, another full articulamentum are used to classify.Several goal hypothesis regions are provided simultaneously at each sliding window, can be referred to as
For anchor region, this region possesses different transverse and longitudinal ratio and scaling centered on sliding window.
The convolutional layer formula of convolutional neural networks R-CNN can be expressed as:
Wherein,What is indicated is the value at position (x, y) of j-th of feature vector of i-th layer of neuron,Table
ShowIt is connected to weight of the convolution kernel of m-th of feature vector on position (p, q).PiAnd QiRespectively indicate the height of convolution kernel
And width, bijFor bias term, f (x) indicates the activation primitive of neuron.
The expression of the pond R-CNN layer formula are as follows:
xij=f (βijdown(xi-1j)+bij) (2)
Down () indicates a down-sampling function, and typical operation is usually to all of the different n*n blocks of input data
Information is summed, and such output data all reduces n times on two dimensions, and each output map corresponding one belongs to certainly
Oneself multiplying property biases β and additivity and biases b.
The input of convolutional neural networks and calculating process are all real numbers, and obtained result is all deterministic, and for number
According to the situation that the data such as missing are fuzzy, fuzzy theory, improved formula are introduced in fuzzy region convolutional neural networks of the invention
It is as follows:
The expression of convolutional layer formula are as follows:
The expression of pond layer formula are as follows:
Wherein biasAnd weightIt is fuzzy number, used here as symmetrical triangular fuzzy numbers,For mould
Paste array at vector, j-th of fuzzy numberMembership function be
As shown in Fig. 2, wjIt is the symmetrical centre of fuzzy number,It is half length of fuzzy number,Represent the degree of membership at w.
In the training process of fuzzy region convolutional neural networks, an associated losses function is defined:
Wherein piIt is the prediction probability that this sample is form of logs,It is the label of sample, is surveyed if it is corresponding
Well tracing pattern,It is 1, otherwiseIt is 0, NclsIt is the loss of two classification (0 or 1) logic.
tiIt is the vector for predicting four parameters composition of object boundary,For the vector of tab area parameter composition, they divide
Not are as follows:
tx=(x-xa)/wa th=(y-ya)/ha (7)
tw=log (w/wa)th=log (h/ha)
Wherein x, y, w and h respectively represent the centre coordinate, width and length of object, x, xa, x*Estimation range is respectively represented,
Anchor region and tab area (y, w, h are similarly).Return lossR is smooth loss function
It indicates only when anchor region is positive sampleIt just calculates and returns loss, otherwiseNo
It calculates.Normalized parameter NclsAnd NregRespectively represent the number of the length and anchor region from the low-dimensional vector of maps feature vectors
Amount.
Different data can be generated using different well logging means.Such as using natural gamma (GR), compensation sound wave (AC), benefit
Repaying density (DEN), compensated neutron (CNL) and neutron apparent porosity and density apparent porosity poor (SDN) etc. has different dimensions,
Do not have comparativity between data, therefore, the present invention needs to carry out the standardization of log data first, and initial data is converted
For index without dimension assessment value, i.e., each index assessment value is all in the same number of levels, then carries out comprehensive test analysis.
Using following normalization method:
Sx=(x-M)/S, x ∈ GR, AC, DEN, CNL, SDN ... }
Wherein, x indicates the data of every log, and Sx indicates the borehole log data after standardization；M is corresponding well logging
The mean value of curve data, S are the standard deviation of every borehole log data.
The present invention is demarcated from existing log data, establishes the training dataset of FR-CNN, on this basis, by
It is not quite similar in the different revealed information of logging method, so selecting the combination of different log datas as the defeated of FR-CNN
Enter, so that it is determined that FR-CNN it is optimal log data combination, and optimize FR-CNN carry out well-log facies recognition when network parameter
And structure.
FR-CNN two classes full articulamentums more than traditional convolutional neural networks, the operation such as more calculating of area coordinate,
These operation calculation amounts are all very big.Fuzzy operation is present in each layer of network in fuzzy neural network, that is to say, that network
The deeper increased calculation amount of institute is more, this just makes the network for needing heavy calculating originally seem heavy.The increase of calculation amount is led
It causes the training time of network to increase substantially, the period of network model update is extended, when weakening the flexibility of system, while detecting
Between can also lengthen.
The present invention improves FR-CNN training and operational efficiency by parallelization, and the log data collection of input is divided into first
Several small data sets, multiple workflows are run simultaneously, and each section is trained using only gradient decline.After the completion of training,
Result is output to waiting list, after the completion of a wheel training, reads output queue, carries out the synchronized update behaviour of shared weight
Make.After the completion of update, next round training is carried out.
In each round training, calculating for the small data set of each segmentation, be all on distributed basis it is asynchronous into
Capable, gradient value is often calculated, is just appended in list, after all small data sets all calculate, synchronized update
Then the weight and bias of network carry out next round training.
As shown in figure 3, in terms of parallelization identification, the solution taken are as follows: log data is collected by Spout, then
Data distribution is subjected to well-log facies recognition into each Bolt node parallel, recognition result is input to next by each Bolt node
In a Bolt node, object information therein is counted.
The present invention feature fuzzy according to data in well logging big data, incorporates fuzzy theory, proposes fuzzy region convolution mind
Through network FR-CNN, and propose that progressive fuzzy method is gradually increased since the first layer of fuzzy region convolutional neural networks
The number of plies of blurring realizes preferably analysis performance and precision to optimize network structure and parameter；Moreover, the present invention is directed to
The number of plies of different log data collection adjustment FR-CNN blurrings, makes the feature extracted preferably reflect the spy of oil and gas reservoir itself
Property, it can solve log data fuzzy problem；For computationally intensive problem is operated, the present invention carries out FR- using more GPU
The parallel training of CNN and execution, to improve the efficiency of FR-CNN.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Within mind and principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.
Claims (4)
1. the parallel method of fuzzy region convolutional neural networks under a kind of big data environment, which is characterized in that
Firstly, building fuzzy region convolutional neural networks, will provide goal hypothesis region and target identification are put into the same network
In, convolutional calculation is shared, a training process updates the weight of whole network；
Next, the log data collection of input is divided into several small data sets, fuzzy region is passed through in multiple workflow parallelizations
Convolutional neural networks carry out the operation of convolution sum pondization, and each small data set is trained using only gradient decline；Training is completed
Afterwards, result is output to waiting list, after the completion of a wheel training, reads output queue, carry out the synchronized update of shared weight
Operation after the completion of update, carries out next round training；In each round training, calculating for the small data set of each segmentation, all
It is the asynchronous progress on distributed basis, often calculates gradient value, is just appended in list, when all small data sets
After all calculating, then the weight and bias of synchronized update fuzzy region convolutional neural networks carry out next round training；?
Parallelization identification aspect, collects log data by Spout, then data distribution is logged well parallel into each Bolt node
It mutually identifies, recognition result is input in next Bolt node by each Bolt node, counts object information therein；
The step of each small data set carries out the operation of convolution sum pondization by fuzzy region convolutional neural networks, specifically includes:
Convolutional layer and the interaction of pond layer carry out fuzzy operation in convolutional layer and pond layer, from the first of fuzzy region convolutional neural networks
Layer starts, and gradually increases the number of plies of blurring, for the different data set adjustment blurring numbers of plies, fuzzy region convolutional Neural net
The last layer of network obtains feature vector, and this feature vector passes through a sliding window for Feature Mapping to a low-dimensional vector
In, two full articulamentums are then input the feature into, a full articulamentum is used to position, another full articulamentum is used to classify.
2. the parallel method of fuzzy region convolutional neural networks, special under a kind of big data environment as described in claim 1
Sign is that the convolutional layer formula is expressed are as follows:
The expression of pond layer formula are as follows:
Wherein, it biasesAnd weightIt is fuzzy number, used here as symmetrical triangular fuzzy numbers,It is fuzzy
Array at vector, j-th of fuzzy numberMembership function are as follows:
3. such as the parallelization of fuzzy region convolutional neural networks under a kind of described in any item big data environment of claim 1 to 2
Method, which is characterized in that initial data is converted into index without dimension assessment by the standardization for carrying out log data first
Value, each index assessment value is all in the same number of levels, then carries out comprehensive test analysis.
4. the parallel method of fuzzy region convolutional neural networks, special under a kind of big data environment as claimed in claim 3
Sign is that the standardization for carrying out log data uses following normalization method:
Sx=(x-M)/S, x ∈ GR, AC, DEN, CNL, SDN ... }
Wherein, x indicates the data of every log, and Sx indicates that the borehole log data after standardization, M are corresponding log
The mean value of data, S are the standard deviation of every borehole log data.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610762101.1A CN106372402B (en) | 2016-08-30 | 2016-08-30 | The parallel method of fuzzy region convolutional neural networks under a kind of big data environment |
| CN201910042398.8A CN109800863B (en) | 2016-08-30 | 2016-08-30 | Logging phase identification method based on fuzzy theory and neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610762101.1A CN106372402B (en) | 2016-08-30 | 2016-08-30 | The parallel method of fuzzy region convolutional neural networks under a kind of big data environment |
Related Child Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910042398.8A Division CN109800863B (en) | 2016-08-30 | 2016-08-30 | Logging phase identification method based on fuzzy theory and neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106372402A CN106372402A (en) | 2017-02-01 |
| CN106372402B true CN106372402B (en) | 2019-04-30 |
Family
ID=57902416
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610762101.1A Active CN106372402B (en) | 2016-08-30 | 2016-08-30 | The parallel method of fuzzy region convolutional neural networks under a kind of big data environment |
| CN201910042398.8A Active CN109800863B (en) | 2016-08-30 | 2016-08-30 | Logging phase identification method based on fuzzy theory and neural network |
Family Applications After (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910042398.8A Active CN109800863B (en) | 2016-08-30 | 2016-08-30 | Logging phase identification method based on fuzzy theory and neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN106372402B (en) |
Families Citing this family (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106446514A (en) * | 2016-08-31 | 2017-02-22 | 中国石油大学(华东) | Fuzzy theory and neural network-based well-log facies recognition method |
| CN106980895A (en) * | 2017-02-22 | 2017-07-25 | 中国科学院自动化研究所 | Convolutional neural networks Forecasting Methodology based on rotary area |
| CN106980874B (en) * | 2017-03-13 | 2019-09-10 | 南京邮电大学 | A kind of multi-time Scales dimension data fusion method towards distribution big data |
| CN107168527B (en) * | 2017-04-25 | 2019-10-18 | 华南理工大学 | The first visual angle gesture identification and exchange method based on region convolutional neural networks |
| CN109034385A (en) * | 2017-06-12 | 2018-12-18 | 辉达公司 | With the system and method for sparse data training neural network |
| US11244226B2 (en) | 2017-06-12 | 2022-02-08 | Nvidia Corporation | Systems and methods for training neural networks with sparse data |
| CN107678059B (en) * | 2017-09-05 | 2019-06-28 | 中国石油大学(北京) | A kind of method, apparatus and system of reservoir gas-bearing identification |
| US10892953B2 (en) * | 2018-02-23 | 2021-01-12 | Nec Corporation | Network-based application acceleration |
| CN108629072B (en) * | 2018-03-12 | 2019-07-09 | 山东科技大学 | Convolutional neural networks study and prediction technique towards the distribution of earthquake oil and gas reservoir |
| CN108648495A (en) * | 2018-06-08 | 2018-10-12 | 华南理工大学 | A kind of method and system of the intelligence real-time display bus degree of crowding |
| CN109171774B (en) * | 2018-09-30 | 2021-07-06 | 合肥工业大学 | Personality analysis method and system based on multi-channel data |
| CN109920013B (en) * | 2019-01-30 | 2020-11-10 | 北京交通大学 | Image reconstruction method and device based on progressive convolution measurement network |
| CN109919313B (en) * | 2019-01-31 | 2021-06-08 | 华为技术有限公司 | Gradient transmission method and distributed training system |
| CN109816042B (en) * | 2019-02-01 | 2020-11-24 | 北京达佳互联信息技术有限公司 | Data classification model training method and device, electronic equipment and storage medium |
| CN111236932B (en) * | 2020-01-13 | 2022-12-13 | 中国石油天然气股份有限公司大港油田分公司 | Shale oil lithology evaluation method based on indication curve reconstruction |
| CN111462060A (en) * | 2020-03-24 | 2020-07-28 | 湖南大学 | Method and device for detecting standard section image in fetal ultrasonic image |
| CN112784980B (en) * | 2021-01-05 | 2024-05-28 | 中国石油天然气集团有限公司 | Intelligent logging horizon dividing method |
| CN113537458B (en) * | 2021-06-21 | 2022-12-30 | 中国科学技术大学 | Rational function neural network construction method, system and readable storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104657748A (en) * | 2015-02-06 | 2015-05-27 | 中国石油大学(华东) | Vehicle type recognition method based on convolutional neural network |
| CN105354568A (en) * | 2015-08-24 | 2016-02-24 | 西安电子科技大学 | Convolutional neural network based vehicle logo identification method |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1758398A1 (en) * | 2005-08-23 | 2007-02-28 | Syneola SA | Multilevel semiotic and fuzzy logic user and metadata interface means for interactive multimedia system having cognitive adaptive capability |
| US20080033991A1 (en) * | 2006-08-03 | 2008-02-07 | Jayanta Basak | Prediction of future performance of a dbms |
| US11256982B2 (en) * | 2014-07-18 | 2022-02-22 | University Of Southern California | Noise-enhanced convolutional neural networks |
| CN104217225B (en) * | 2014-09-02 | 2018-04-24 | 中国科学院自动化研究所 | A kind of sensation target detection and mask method |
| CN104573669B (en) * | 2015-01-27 | 2018-09-04 | 中国科学院自动化研究所 | Image object detection method |
| CN105825511B (en) * | 2016-03-18 | 2018-11-02 | 南京邮电大学 | A kind of picture background clarity detection method based on deep learning |
| CN105654729B (en) * | 2016-03-28 | 2018-01-02 | 南京邮电大学 | A kind of short-term traffic flow forecast method based on convolutional neural networks |
-
2016
- 2016-08-30 CN CN201610762101.1A patent/CN106372402B/en active Active
- 2016-08-30 CN CN201910042398.8A patent/CN109800863B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104657748A (en) * | 2015-02-06 | 2015-05-27 | 中国石油大学(华东) | Vehicle type recognition method based on convolutional neural network |
| CN105354568A (en) * | 2015-08-24 | 2016-02-24 | 西安电子科技大学 | Convolutional neural network based vehicle logo identification method |
Non-Patent Citations (2)
| Title |
|---|
| 《Human Action Recognition based on Motion Capture Information using Fuzzy Convolution Neural Networks》;Earnest Paul Ijjina 等;《2015 Eighth International Conference on Advance in Pattern Recognition》;20150302;全文 |
| 《储层测井特征属性反演方法》;徐旺林 等;《天然气工业》;20050331;全文 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109800863A (en) | 2019-05-24 |
| CN106372402A (en) | 2017-02-01 |
| CN109800863B (en) | 2023-05-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106372402B (en) | The parallel method of fuzzy region convolutional neural networks under a kind of big data environment | |
| CN106529667B (en) | Well-log facies recognition and analysis method based on fuzzy deep learning under a kind of big data environment | |
| CN110674841B (en) | Logging curve identification method based on clustering algorithm | |
| CN104747185B (en) | Heterogeneous reservoir reservoir synthetical assortment evaluation method | |
| CN104239666B (en) | A kind of Comprehensive Evaluation of Coal Bed Gas method based on analytic hierarchy process (AHP) | |
| CN110490100A (en) | Ground automatic identification based on deep learning names method and system | |
| CN106372748A (en) | Hard-rock tunnel boring machine boring efficiency prediction method | |
| CN107704970A (en) | A kind of Demand-side load forecasting method based on Spark | |
| CN105760673A (en) | Fluvial facies reservoir earthquake sensitive parameter template analysis method | |
| WO2015187441A2 (en) | Predicting well markers from artificial neural-network-predicted lithostratigraphic facies | |
| CN106446514A (en) | Fuzzy theory and neural network-based well-log facies recognition method | |
| Zhu et al. | Rapid identification of high-quality marine shale gas reservoirs based on the oversampling method and random forest algorithm | |
| Huang et al. | Research on urban modern architectural art based on artificial intelligence and GIS image recognition system | |
| CN103376468A (en) | Reservoir parameter quantitative characterization method based on neural network function approximation algorithm | |
| CN111914478A (en) | Comprehensive geological drilling well logging lithology identification method | |
| Santiago et al. | A methodology for the characterization of flow conductivity through the identification of communities in samples of fractured rocks | |
| CN112861890A (en) | Reservoir evaluation model construction method and reservoir identification method | |
| Guoqiang et al. | Construction of well logging knowledge graph and intelligent identification method of hydrocarbon-bearing formation | |
| Patel et al. | Smart adaptive mesh refinement with NEMoSys | |
| CN114862609A (en) | Oil reservoir exploitation method for ultra-high water-cut oil field | |
| US20240077631A1 (en) | Seismic imaging free gas structure identification method and system | |
| CN112081582A (en) | Prediction method, system and device for dominant channel in water-drive oil reservoir development | |
| Kakouei et al. | Lithological facies identification in Iranian largest gas field: A comparative study of neural network methods | |
| CN106227959B (en) | Method and device for predicting lithologic reservoir favorable area based on four-graph superposition method | |
| de Oliveira et al. | Ensemble of heterogeneous classifiers applied to lithofacies classification using logs from different wells |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | 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 |