CN107462927A - Seismic facies Forecasting Methodology and device based on Naive Bayes Classification - Google Patents
Seismic facies Forecasting Methodology and device based on Naive Bayes Classification Download PDFInfo
- Publication number
- CN107462927A CN107462927A CN201610391119.5A CN201610391119A CN107462927A CN 107462927 A CN107462927 A CN 107462927A CN 201610391119 A CN201610391119 A CN 201610391119A CN 107462927 A CN107462927 A CN 107462927A
- Authority
- CN
- China
- Prior art keywords
- mrow
- work area
- msub
- predicted
- petrofacies
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V1/00—Seismology; Seismic or acoustic prospecting or detecting
- G01V1/40—Seismology; Seismic or acoustic prospecting or detecting specially adapted for well-logging
- G01V1/44—Seismology; Seismic or acoustic prospecting or detecting specially adapted for well-logging using generators and receivers in the same well
- G01V1/48—Processing data
- G01V1/50—Analysing data
-
- 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
- G01V2210/61—Analysis by combining or comparing a seismic data set with other data
- G01V2210/616—Data from specific type of measurement
- G01V2210/6169—Data from specific type of measurement using well-logging
Landscapes
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Acoustics & Sound (AREA)
- Environmental & Geological Engineering (AREA)
- Geology (AREA)
- Remote Sensing (AREA)
- General Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Geophysics (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The present invention relates to Seismic Reservoir Prediction field, and in particular to seismic facies Forecasting Methodology and device based on Naive Bayes Classification.Bayes's classification is applied in reservoir prediction research by this method, based on Rock physical analysis and elastic parameter inversion, and lithology, fluid etc. are predicted using Naive Bayes Classification statistics, reduce reservoir prediction multi-solution, can also be to the uncertain accuracy for carrying out quantitative assessment, seismic facies prediction being significantly improved of prediction result.
Description
Technical field
The present invention relates to Seismic Reservoir Prediction field, and naive Bayesian is based on more particularly, to one kind
The seismic facies Forecasting Methodology of classification and a kind of seismic facies prediction dress based on Naive Bayes Classification
Put.
Background technology
Petrofacies prediction is carried out using seismic properties, many scholars propose corresponding research method, bag both at home and abroad
Include the analysis of more hierarchical cluster attributes, multi-parameter crosses and neural network prediction etc., but lack rock physicses theory
Support and unsatisfactory to prediction result analysis of uncertainty, its prediction accuracy.
The content of the invention
The present invention proposes a kind of method that can more accurately carry out seismic facies prediction, and the present invention is also
Propose corresponding device.
According to an aspect of the present invention, a kind of seismic facies based on Naive Bayes Classification are disclosed
Forecasting Methodology, this method include：Based on research area's geologic sedimentation feature and features of logging curve, it is determined that
Petrofacies splitting scheme, and crossed with statistical analysis by well logging and to be selected from a variety of elastic parameters pair
The sensitive one or more elastic parameters of lithofacies characteristics；The lithofacies characteristics based on well logging and described one
Individual or multiple elastic parameters carry out Naive Bayes Classification statistics, to determine under various lithofacies characteristics
The conditional probability distribution of one or more of elastic parameters；Based on pre-stack seismic subangle in work area to be predicted
Degree superposition of data inverting obtains one or more of elastic parameters in work area to be predicted；Based on the bar
Part probability distribution and one or more of elastic parameters in work area to be predicted, obtain work area to be predicted
The spatial distribution of petrofacies.
According to another aspect of the present invention, a kind of earthquake rock based on Naive Bayes Classification is also disclosed
Phase prediction meanss, the device include：Log analysis unit, for based on research area's geologic sedimentation feature
And features of logging curve, determine petrofacies splitting scheme, and cross with statistical analysis from a variety of by well logging
The selection one or more elastic parameters sensitive to lithofacies characteristics in elastic parameter；Probability determining unit,
Naive Bayesian is carried out for the lithofacies characteristics based on well logging and one or more of elastic parameters
Statistic of classification, to determine that the condition of one or more of elastic parameters under various lithofacies characteristics is general
Rate is distributed；Work area analytic unit, for anti-based on pre-stack seismic subangle superposition of data in work area to be predicted
Drill to obtain one or more of elastic parameters in work area to be predicted；Work area petrofacies predicting unit, is used for
One or more of elastic parameters based on the conditional probability distribution and work area to be predicted, are treated
Predict the spatial distribution of the petrofacies in work area.
Various aspects of the invention, based on Rock physical analysis and elastic parameter inversion, using Piao
Plain Bayes's classification statistics is predicted to lithology, fluid etc., reduces reservoir prediction multi-solution, can
Realize accurate seismic facies prediction.The present invention is in terms of petrofacies prediction, at least with advantages below：
(1) Bayes statistical method regards elastic parameter attribute as a stochastic variable, takes into full account it
Random character, this is the fundamental difference with the classical theory of statistics；
(2) using information at a small amount of well point, geological data is regarded to the form of likelihood function as, and then push away
Break and the distribution of petrofacies body, be suitable for Small Sample Size, there is very strong applicability；
(3) statistic of classification can be carried out to more attribute, attribute number is unrestricted, and each attribute is joined
With specific distributed arithmetic, single attribute is overcome to predict multi-solution；
(4) infer Posterior distrbutionp using prior information at well point, problem be stacked for different petrofacies,
Classified more science using probability distribution thought, reduce human error, at the same can apply lithology,
In terms of the prediction such as physical property and gas-bearing property；
(5) it is easy to carry out quantitative assessment to the uncertain of prediction result
Brief description of the drawings
Exemplary embodiment of the invention is described in more detail in conjunction with the accompanying drawings, it is of the invention
Above-mentioned and other purposes, feature and advantage will be apparent, wherein, it is exemplary in the present invention
In embodiment, identical reference number typically represents same parts.
Fig. 1 shows the structure chart of Naive Bayes Classification Model.
Fig. 2 shows the earthquake rock according to an embodiment of the invention based on Naive Bayes Classification
The flow chart of phase Forecasting Methodology.
Fig. 3 shows the log-petrofacies and elastic parameter curve block diagram of certain example；
Fig. 4 shows different petrofacies elastic parameter variation characteristic comparison diagrams in the example；
Fig. 5 shows well logging sampling point petrofacies conditional probability distribution in the example；
Fig. 6 (a)~(c) shows the stacked section of different angle in the example；
Fig. 7 (a)~(b) shows the prestack inversion resilient property section in the example；
Fig. 8 (a)~(b) shows different petrofacies and gas sand probability profile；
Fig. 9 shows petrofacies predictor horizon slice.
Embodiment
The preferred embodiment of the present invention is more fully described below with reference to accompanying drawings.Although show in accompanying drawing
The preferred embodiment of the present invention is shown, however, it is to be appreciated that may be realized in various forms the present invention
Without should be limited by embodiments set forth herein.
Herein, first the general principle of the present invention is simply introduced.
Naive Bayes Classifier is a kind of application based on the Bayesian simple probability independently assumed
Grader, more precisely this potential probabilistic model is independent characteristic model.Bayes's classification
Method is counted respectively based on bayesian theory according to the distribution situation of each classification sample in training set
Calculate it and belong to the posterior probability of each classification, the classification of the sample is then judged as maximum a posteriori probability
Corresponding classification.Bayesian formula can be written as form：
Wherein, P (y | x) is Posterior probability distribution, and P (y) is prior distribution, and P (x) is usually constant.
Bayes classification method principle is succinct, but when attribute number is more, amount of calculation is very huge,
Limitation is suffered from actual applications.For simplified operation, a kind of practical simple shellfish is proposed
This sorting algorithm of leaf, it is assumed that influence of any attribute to classification is unrelated with influence of other attributes to classification,
It is this to assume to be referred to as the independent simple hypothesis of class condition.Fig. 1 shows the knot of Naive Bayes Classification Model
Composition, wherein C represent classification to be divided, A1、A2、...AnSample attribute is represented, arrow represents attribute
Dependence between variable and class variable.From figure 1 it appears that in Naive Bayes Classification mould
In type, sample attribute and AiAjRelation of interdependence is not present between (i is not equal to j), they only with
Node classification C is relevant.
Known sample data x=(x1,...,xn) (sample data x shares n attribute A1、A2、...An, its
Middle xiRepresent attribute AiValue) belong to any class y (y ∈ { c1,...,ck) (k classification altogether, cjRepresent numbering
For j classification) probability.A non-classified data sample X is given, using Naive Bayes Classification
Algorithm, forecast sample data X belong to the class with maximum a posteriori probability, and unknown sample X belongs to classification ci
Condition be that and if only if：
P(ci|X)>P(cj|X),1≤i,j≤k,j≠i。 (2)
Therefore, posterior probability P (c will be maximizedi| X) or its logarithmic form be referred to as maximum a posteriori it is assumed that
It is designated as arg maxy P(y|X)。
According to Bayes' theorem, have：
Value is equal in arbitrarily once classifying, that is to say, that probability caused by data sample X is identical
(P (x) is defined as constant), obtained after simplifying：
P(y|X)∝P(X|y)*P(y)。 (4)
According to the conditional attribute of Naive Bayes Classification Algorithm it is separate it is assumed that having：
According to Naive Bayes Classification principle, maximum a posteriori probability is represented by：
Unknown sample X is classified, for each classification ciCalculating P (x | ci)P(ci)；And if only if
P(X|ci)P(ci)>P(X|cj)P(cj), 1≤j≤m, j ≠ i, (7)
Define sample X and belong to classification ci, i.e. X be assigned to P (X | ci)P(ci) maximum classification ci。
Embodiment 1
Fig. 2 shows the earthquake rock according to an embodiment of the invention based on Naive Bayes Classification
The flow chart of phase Forecasting Methodology.In the present embodiment, this method includes：
Step 101, based on research area's geologic sedimentation feature and features of logging curve, petrofacies division side is determined
Case, and crossed with statistical analysis and select from a variety of elastic parameters to lithofacies characteristics sensitivity by well logging
One or more elastic parameters；
Step 102, the lithofacies characteristics based on well logging and one or more of elastic parameters carry out Piao
Plain Bayes's classification statistics, to determine the conditional probability distribution of the elastic parameter under various lithofacies characteristics；
Step 103, work to be predicted is obtained based on work area pre-stack seismic subangle superposition of data inverting to be predicted
One or more of elastic parameters in area；
Step 104, one or more of elasticity based on the conditional probability distribution and work area to be predicted
Parameter, obtain the spatial distribution of the petrofacies in work area to be predicted.
In the present embodiment, based on Rock physical analysis and elastic parameter inversion, using simple pattra leaves
This statistic of classification is predicted to lithology, fluid etc., is reduced reservoir prediction multi-solution, can be realized standard
True seismic facies prediction.
A variety of elastic parameters described above can include p-wave impedance, S-wave impedance, Poisson's ratio, in length and breadth
It is part or all of in wave velocity ratio (Vp/Vs), Lame constants, modulus of shearing and bulk modulus etc..
In one example, the present invention can also include：It can be obtained based on the conditional probability distribution
Petrofacies probability data body, with the uncertainty of qualitative assessment prediction result.
In one example, above-mentioned steps 103 can include：Multiple different angle superpositions can be inputted
Geological data and corresponding wavelet, can provide p-wave impedance, S-wave impedance and density it is longitudinally varying become
Gesture and lateral confinement scope, the AVA analogy methods based on Aki-Richards formula then can be selected to enter
Row prestack Simultaneous Inversion, obtain p-wave impedance data volume, S-wave impedance data volume and P-S wave velocity ratio
Data volume, density data body, it next can be based on inversion result and calculate one or more of elasticity
Other elastic parameters in parameter.
In one example, kernel function can be used to represent the conditional probability density.Main kernel function
Including triangle kernel function, gaussian kernel function, Epanechnikov kernel functions etc..
Further, inventor has found by repetition test and further investigation, in the present invention it is possible to
The conditional probability density is fitted using gaussian kernel function, is advantageous to improve the degree of accuracy of petrofacies prediction.
Following gaussian kernel function can be used to represent the conditional probability distribution：
Wherein, xkRepresent the random value for the elastic parameter that numbering is k, ciRepresent the petrofacies that numbering is i, uci
And σciRepresent average and variance.Can be by adjusting uciAnd σciTo determine the conditional probability distribution.
Embodiment 2
A kind of earthquake rock based on Naive Bayes Classification is also disclosed according to another embodiment of the present invention
Phase prediction meanss, the device include：
Log analysis unit, for based on research area's geologic sedimentation feature and features of logging curve, it is determined that
Petrofacies splitting scheme, and crossed with statistical analysis by well logging and selected from a variety of elastic parameters to described
The sensitive one or more elastic parameters of lithofacies characteristics；
Probability determining unit, for the lithofacies characteristics based on well logging and one or more of elasticity
Parameter carries out Naive Bayes Classification statistics, one or more under various lithofacies characteristics to determine
The conditional probability distribution of individual elastic parameter；
Work area analytic unit, for being obtained based on work area pre-stack seismic subangle superposition of data inverting to be predicted
To one or more of elastic parameters in work area to be predicted；
Work area petrofacies predicting unit, for based on described in the conditional probability distribution and work area to be predicted
One or more elastic parameters, obtain the spatial distribution of the petrofacies in work area to be predicted.
In one example, the device can also include analysis of uncertainty unit, and it can be used for base
Petrofacies probability data body is obtained in the conditional probability distribution, with the uncertain of qualitative assessment prediction result
Property.
In one example, in the work area analytic unit, based on work area pre-stack seismic to be predicted point
One or more of elastic parameters that angular stack data inversion obtains work area to be predicted can include：
The geological data of multiple different angle superpositions and corresponding wavelet can be inputted, compressional wave resistance can be provided
Longitudinally varying trend and the lateral confinement scope of anti-, S-wave impedance and density, then can select to be based on
The AVA analogy methods of Aki-Richards formula carry out prestack Simultaneous Inversion, can obtain p-wave impedance
Data volume, S-wave impedance data volume and P-S wave velocity ratio data volume, density data body, next may be used
To calculate other elastic parameters in one or more of elastic parameters based on inversion result.
In one example, in the probability determining unit, kernel function can be used to represent the bar
Part probability density.
In one example, the kernel function can be gaussian kernel function：
Wherein, xkRepresent the random value for the elastic parameter that numbering is k, ciRepresent the petrofacies that numbering is i, uci
And σciRepresent average and variance., can be by adjusting u in the probability determining unitciAnd σciCome true
The fixed conditional probability distribution.
Application example
For ease of understanding the scheme of the embodiment of the present invention and its effect, a concrete application given below is shown
Example.It will be understood by those skilled in the art that the example, only for the purposes of understanding the present invention, its is any specific
Details, which is not intended to, to be limit the invention in any way.
According to research area's reservoir characteristic and type, different petrofacies are entered using well-log information and interpretation results
Row definition, is divided into gas sand, wherein three kinds of Lithofacies Types of tight sand and mud stone, gas sand pair
Natural gamma GR≤110API, porosity por≤6% and SW≤65% are answered, tight sand corresponds to nature gal
Horse GR≤110API, porosity por<6%, mud stone is natural gamma GR<110API, it is desirable to which petrofacies are again
Define result and result of log interpretation is basically identical；Calculate the elasticity ginseng of each well respectively using theoretical formula
Number curve, including P-wave And S impedance, Poisson's ratio, Vp/Vs, Lame constants and bulk modulus etc., such as
Shown in 3.
Logging response character contrast and the statistics of various elastic parameter variations by different petrofacies,
Can determine study area's gas sand be mainly shown as middle high p-wave impedance, low Vp/Vs, low Poisson's ratio and
The petrophysics properties such as Lame constants, as shown in Figure 4.
In order to overcome the multi-solution of unitary elasticity attribute forecast, preferable p-wave impedance and Vp/Vs attribute
Combine to establish the probability distribution of different petrofacies, corresponding type function selected according to Probability Characteristics,
This selection gaussian kernel function, average and variance etc. are included by the major parameter of adjustment function so that
The function describes section and well logging scatterplot distribution is basically identical, as shown in figure 5, determining gas sand
P-wave impedance average 11300g/cm3*m/s, standard variance 693, accounts for petrofacies ratio 0.15.
Original earthquake prestack CRP trace gathers are pre-processed, including cut off a long way, prestack denoising, height
The processing such as the dynamic correction of rank and amplitude compensation so that prestack trace gather can meet the requirement of prestack inversion.Utilize
Offset gather is converted to angle domain trace gather by earthquake overlap normal-moveout spectrum data, and angularly distribution is folded
Add to obtain subangle superposition of data, Fig. 6 (a) shows the subangle superposition number of nearly (0-10 °) partially
According to the subangle superposition of data of inclined (11-20 °), Fig. 6 (c) are shown during Fig. 6 (b) is shown
The subangle superposition of data of remote (21-30 °) partially.For different angle trace gather extraction sympathetic earthquakes
Ripple is simultaneously demarcated, and then obtains the attributes such as P-wave And S impedance and Vp/Vs by prestack Simultaneous Inversion method
Body, Fig. 7 (a) show p-wave impedance, and Fig. 7 (b) shows Vp/Vs.
The statistical relationship finally established using Fig. 5, different petrofacies are converted to by the attribute volume of prestack inversion
Body and corresponding petrofacies probability volume, from the point of view of petrofacies prediction result, it is coincide substantially at well point, laterally
Change faithful to original seismic data, reflect anisotropism variation characteristic between well, demonstrate the present invention's
Validity, Fig. 8 (a) show lithofacies distribution, and Fig. 8 (b) shows gas sand probability, and Fig. 9 shows
The section of petrofacies predictor layer is gone out.In addition, petrofacies probability data body can be also generated, with further quantitative
The uncertainty of petrofacies prediction result is evaluated, then local reliability is relatively high greatly for probable value.
The present invention can be system, method and/or computer program product.Computer program product can be with
Including computer-readable recording medium, containing for making processor realize various aspects of the invention
Computer-readable program instructions.
Referring herein to method, apparatus (system) according to embodiments of the present invention and computer program product
Flow chart and/or block diagram describe various aspects of the invention.It should be appreciated that flow chart and/or block diagram
Each square frame and flow chart and/or block diagram in each square frame combination, can be by computer-readable journey
Sequence instruction is realized.
Various embodiments of the present invention are described above, described above is exemplary, and exhaustive
Property, and it is also not necessarily limited to disclosed each embodiment.In the model without departing from illustrated each embodiment
Enclose and spirit in the case of, many modifications and changes for those skilled in the art
It will be apparent from.
Claims (10)
1. a kind of seismic facies Forecasting Methodology based on Naive Bayes Classification, this method include：
Based on research area's geologic sedimentation feature and features of logging curve, petrofacies splitting scheme is determined, and lead to
Cross to log well to cross and select one or more sensitive to lithofacies characteristics from a variety of elastic parameters with statistical analysis
Individual elastic parameter；
The lithofacies characteristics and one or more of elastic parameters based on well logging carry out naive Bayesian
Statistic of classification, to determine that the condition of one or more of elastic parameters under various lithofacies characteristics is general
Rate is distributed；
Obtained based on work area pre-stack seismic subangle superposition of data inverting to be predicted described in work area to be predicted
One or more elastic parameters；
One or more of elastic parameters based on the conditional probability distribution and work area to be predicted, are obtained
To the spatial distribution of the petrofacies in work area to be predicted.
2. seismic facies Forecasting Methodology according to claim 1, this method also include：
Petrofacies probability data body is obtained based on the conditional probability distribution, with qualitative assessment prediction result
It is uncertain.
3. seismic facies Forecasting Methodology according to claim 1, wherein, based on work area to be predicted
Pre-stack seismic subangle superposition of data inverting obtains one or more of elastic parameters in work area to be predicted
Including：
The geological data of multiple different angle superpositions and corresponding wavelet are inputted, provides p-wave impedance, shear wave
Impedance and longitudinally varying trend and the lateral confinement scope of density, then selection are based on Aki-Richards
The AVA analogy methods of formula carry out prestack Simultaneous Inversion, obtain p-wave impedance data volume, S-wave impedance
Data volume and P-S wave velocity ratio data volume, density data body, then calculate described one based on inversion result
Other elastic parameters in individual or multiple elastic parameters.
4. seismic facies Forecasting Methodology according to claim 1, wherein, represented using kernel function
The conditional probability density.
5. seismic facies Forecasting Methodology according to claim 4, wherein, the kernel function is height
This kernel function：
<mrow>
<mi>p</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mi>k</mi>
</msub>
<mo>|</mo>
<msub>
<mi>c</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<msqrt>
<mrow>
<mn>2</mn>
<msubsup>
<mi>&pi;&sigma;</mi>
<mrow>
<mi>c</mi>
<mi>i</mi>
</mrow>
<mn>2</mn>
</msubsup>
</mrow>
</msqrt>
</mfrac>
<mi>exp</mi>
<mrow>
<mo>(</mo>
<mfrac>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mi>k</mi>
</msub>
<mo>-</mo>
<msub>
<mi>u</mi>
<mrow>
<mi>c</mi>
<mi>i</mi>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mrow>
<mn>2</mn>
<msubsup>
<mi>&sigma;</mi>
<mrow>
<mi>c</mi>
<mi>i</mi>
</mrow>
<mn>2</mn>
</msubsup>
</mrow>
</mfrac>
<mo>)</mo>
</mrow>
<mo>,</mo>
</mrow>
Wherein, xkRepresent the random value for the elastic parameter that numbering is k, ciThe petrofacies that numbering is i are represented,
uciAnd σciAverage and variance are represented, by adjusting uciAnd σciTo determine the conditional probability distribution.
6. a kind of seismic facies prediction meanss based on Naive Bayes Classification, the device include：
Log analysis unit, for based on research area's geologic sedimentation feature and features of logging curve, it is determined that
Petrofacies splitting scheme, and crossed with statistical analysis by well logging and selected from a variety of elastic parameters to petrofacies
One or more elastic parameters of feature-sensitive；
Probability determining unit, for the lithofacies characteristics based on well logging and one or more of elasticity
Parameter carries out Naive Bayes Classification statistics, one or more under various lithofacies characteristics to determine
The conditional probability distribution of individual elastic parameter；
Work area analytic unit, for being obtained based on work area pre-stack seismic subangle superposition of data inverting to be predicted
To one or more of elastic parameters in work area to be predicted；
Work area petrofacies predicting unit, for based on described in the conditional probability distribution and work area to be predicted
One or more elastic parameters, obtain the spatial distribution of the petrofacies in work area to be predicted.
7. seismic facies prediction meanss according to claim 6, the device also include：
Analysis of uncertainty unit, for obtaining petrofacies probability data body based on the conditional probability distribution,
With the uncertainty of qualitative assessment prediction result.
8. seismic facies prediction meanss according to claim 6, wherein, analyzed in the work area
In unit, work area to be predicted is obtained based on work area pre-stack seismic subangle superposition of data inverting to be predicted
One or more of elastic parameters include：
The geological data of multiple different angle superpositions and corresponding wavelet are inputted, provides p-wave impedance, shear wave
Impedance and longitudinally varying trend and the lateral confinement scope of density, then selection are based on Aki-Richards
The AVA analogy methods of formula carry out prestack Simultaneous Inversion, obtain p-wave impedance data volume, S-wave impedance
Data volume and P-S wave velocity ratio data volume, density data body, then calculate described one based on inversion result
Other elastic parameters in individual or multiple elastic parameters.
9. seismic facies prediction meanss according to claim 6, wherein, in the determine the probability
In unit, the conditional probability density is represented using kernel function.
10. seismic facies prediction meanss according to claim 9, wherein, the kernel function is
Gaussian kernel function：
<mrow>
<mi>p</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mi>k</mi>
</msub>
<mo>|</mo>
<msub>
<mi>c</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<msqrt>
<mrow>
<mn>2</mn>
<msubsup>
<mi>&pi;&sigma;</mi>
<mrow>
<mi>c</mi>
<mi>i</mi>
</mrow>
<mn>2</mn>
</msubsup>
</mrow>
</msqrt>
</mfrac>
<mi>exp</mi>
<mrow>
<mo>(</mo>
<mfrac>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mi>k</mi>
</msub>
<mo>-</mo>
<msub>
<mi>u</mi>
<mrow>
<mi>c</mi>
<mi>i</mi>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mrow>
<mn>2</mn>
<msubsup>
<mi>&sigma;</mi>
<mrow>
<mi>c</mi>
<mi>i</mi>
</mrow>
<mn>2</mn>
</msubsup>
</mrow>
</mfrac>
<mo>)</mo>
</mrow>
<mo>,</mo>
</mrow>
Wherein, xkRepresent the random value for the elastic parameter that numbering is k, ciThe petrofacies that numbering is i are represented,
uciAnd σciAverage and variance are represented, in the probability determining unit, by adjusting uciAnd σciCome true
The fixed conditional probability distribution.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610391119.5A CN107462927A (en) | 2016-06-03 | 2016-06-03 | Seismic facies Forecasting Methodology and device based on Naive Bayes Classification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610391119.5A CN107462927A (en) | 2016-06-03 | 2016-06-03 | Seismic facies Forecasting Methodology and device based on Naive Bayes Classification |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107462927A true CN107462927A (en) | 2017-12-12 |
Family
ID=60545785
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610391119.5A Pending CN107462927A (en) | 2016-06-03 | 2016-06-03 | Seismic facies Forecasting Methodology and device based on Naive Bayes Classification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107462927A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108645994A (en) * | 2018-04-25 | 2018-10-12 | 中国石油大学(北京) | A kind of geology stochastic inversion methods and device based on Multiple-Point Geostatistics |
| CN110175721A (en) * | 2019-05-30 | 2019-08-27 | 长江大学 | A kind of quantitative preferred method of potentiality recess and system based on Bayesian statistics prediction |
| CN110609327A (en) * | 2018-06-14 | 2019-12-24 | 中国石油化工股份有限公司 | Carbonate reservoir facies prediction method and device based on pre-stack seismic attributes |
| CN111832636A (en) * | 2020-06-30 | 2020-10-27 | 中国石油大学(北京) | Naive Bayes lithofacies classification method and device based on feature combination |
| CN112649867A (en) * | 2019-10-12 | 2021-04-13 | 中国石油化工股份有限公司 | Virtual well construction method and system |
| CN114185090A (en) * | 2020-09-15 | 2022-03-15 | 中国石油化工股份有限公司 | Lithofacies and elastic parameter synchronous inversion method and device, electronic equipment and medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100017132A1 (en) * | 2008-07-18 | 2010-01-21 | Bhp Billiton Innovation Pty Ltd | Method for evaluating measured electromagnetic data relating to a subsurface region |
| US20110292764A1 (en) * | 2010-04-13 | 2011-12-01 | Spectraseis Ag | Bayesian DHI for Seismic Data |
| CN103176211A (en) * | 2013-03-07 | 2013-06-26 | 中国石油集团东方地球物理勘探有限责任公司 | Method and device, for predicting gas reservoir, based on multi-sensitive elastic parameters |
| CN104297785A (en) * | 2014-09-29 | 2015-01-21 | 中国石油天然气股份有限公司 | Lithofacies constrained reservoir physical property parameter inversion method and device |
| CN104749624A (en) * | 2015-03-03 | 2015-07-01 | 中国石油大学(北京) | Method for synchronously realizing seismic lithofacies identification and quantitative assessment of uncertainty of seismic lithofacies identification |
| CN105044770A (en) * | 2015-07-06 | 2015-11-11 | 成都理工大学 | Compact glutenite gas reservoir quantificational prediction method |
-
2016
- 2016-06-03 CN CN201610391119.5A patent/CN107462927A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100017132A1 (en) * | 2008-07-18 | 2010-01-21 | Bhp Billiton Innovation Pty Ltd | Method for evaluating measured electromagnetic data relating to a subsurface region |
| US20110292764A1 (en) * | 2010-04-13 | 2011-12-01 | Spectraseis Ag | Bayesian DHI for Seismic Data |
| CN103176211A (en) * | 2013-03-07 | 2013-06-26 | 中国石油集团东方地球物理勘探有限责任公司 | Method and device, for predicting gas reservoir, based on multi-sensitive elastic parameters |
| CN104297785A (en) * | 2014-09-29 | 2015-01-21 | 中国石油天然气股份有限公司 | Lithofacies constrained reservoir physical property parameter inversion method and device |
| CN104749624A (en) * | 2015-03-03 | 2015-07-01 | 中国石油大学(北京) | Method for synchronously realizing seismic lithofacies identification and quantitative assessment of uncertainty of seismic lithofacies identification |
| CN105044770A (en) * | 2015-07-06 | 2015-11-11 | 成都理工大学 | Compact glutenite gas reservoir quantificational prediction method |
Non-Patent Citations (3)
| Title |
|---|
| 胡华锋: "基于叠前道集的储层参数反演方法研究", 《中国优秀硕士学位论文全文数据库 基础科学辑》 * |
| 袁成等: "地震岩相识别概率表征方法", 《地球物理学报》 * |
| 袁成等: "基于概率统计的地震岩相识别不确定性定量评价方法", 《地球物理学报》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108645994A (en) * | 2018-04-25 | 2018-10-12 | 中国石油大学(北京) | A kind of geology stochastic inversion methods and device based on Multiple-Point Geostatistics |
| CN110609327A (en) * | 2018-06-14 | 2019-12-24 | 中国石油化工股份有限公司 | Carbonate reservoir facies prediction method and device based on pre-stack seismic attributes |
| CN110609327B (en) * | 2018-06-14 | 2021-04-27 | 中国石油化工股份有限公司 | Carbonate reservoir facies prediction method and device based on pre-stack seismic attributes |
| CN110175721A (en) * | 2019-05-30 | 2019-08-27 | 长江大学 | A kind of quantitative preferred method of potentiality recess and system based on Bayesian statistics prediction |
| CN112649867A (en) * | 2019-10-12 | 2021-04-13 | 中国石油化工股份有限公司 | Virtual well construction method and system |
| CN112649867B (en) * | 2019-10-12 | 2024-04-09 | 中国石油化工股份有限公司 | Virtual well construction method and system |
| CN111832636A (en) * | 2020-06-30 | 2020-10-27 | 中国石油大学(北京) | Naive Bayes lithofacies classification method and device based on feature combination |
| CN111832636B (en) * | 2020-06-30 | 2024-04-12 | 中国石油大学(北京) | Feature combination-based naive Bayesian lithofacies classification method and device |
| CN114185090A (en) * | 2020-09-15 | 2022-03-15 | 中国石油化工股份有限公司 | Lithofacies and elastic parameter synchronous inversion method and device, electronic equipment and medium |
| CN114185090B (en) * | 2020-09-15 | 2024-04-05 | 中国石油化工股份有限公司 | Lithology and elastic parameter synchronous inversion method and device, electronic equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107462927A (en) | Seismic facies Forecasting Methodology and device based on Naive Bayes Classification | |
| US10983233B2 (en) | Method for dynamic calibration and simultaneous closed-loop inversion of simulation models of fractured reservoirs | |
| Thompson et al. | Impediments to predicting site response: Seismic property estimation and modeling simplifications | |
| Rimstad et al. | Hierarchical Bayesian lithology/fluid prediction: A North Sea case study | |
| Spikes et al. | Probabilistic seismic inversion based on rock-physics models | |
| CN104749624B (en) | A kind of seismic facies identification and its uncertain quantitative assessment synchronization realizing method | |
| CN106556867B (en) | Phase-controlled porosity inversion method based on Bayesian classification | |
| Stuart et al. | A two-stage Markov chain Monte Carlo method for seismic inversion and uncertainty quantification | |
| CN103954999B (en) | A kind of S-Wave Velocity Predicted Method being applicable to low-porosity sand-mudstone formation | |
| Downton et al. | Theory-guided data science-based reservoir prediction of a North Sea oil field | |
| Wang et al. | A hybrid inverse method for hydraulic tomography in fractured and karstic media | |
| Khodaverdian et al. | Seismicity parameters and spatially smoothed seismicity model for Iran | |
| Aleardi et al. | A two-step inversion approach for seismic-reservoir characterization and a comparison with a single-loop Markov-chain Monte Carlo algorithm | |
| US10386512B2 (en) | System and method for petro-elastic modeling | |
| Zhang et al. | Prediction of shear wave velocity based on a statistical rock-physics model and Bayesian theory | |
| CN103760600A (en) | Gas saturation inversion method | |
| Sherman et al. | Parameterizing the spatial Markov model from breakthrough curve data alone | |
| Golsanami et al. | Synthesis of capillary pressure curves from post-stack seismic data with the use of intelligent estimators: a case study from the Iranian part of the South Pars gas field, Persian Gulf Basin | |
| Alameedy et al. | Predicting dynamic shear wave slowness from well logs using machine learning methods in the Mishrif Reservoir, Iraq | |
| Shi et al. | Prediction of shear wave velocity using machine learning technique, multiple regression and well logs | |
| Maffioletti et al. | Changing scale and domain of a petrophysical and elastic properties based log-facies classification | |
| Qian et al. | A rock physics driven Bayesian inversion for TOC in the Fuling shale gas reservoir | |
| Barman et al. | Fractured reservoir characterization using streamline-based inverse modeling and artificial intelligence tools | |
| Aleardi et al. | Reservoir characterization through target-oriented AVA-petrophysical inversions with spatial constraints | |
| Wu et al. | Effective parametrization for reliable reservoir performance predictions |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171212 |
|
| RJ01 | Rejection of invention patent application after publication |