A method for constructing a coal reservoir fracture prediction model based on multi-source data body fusion

Abstract

The present application relates to a kind of coal reservoir fracture prediction model construction methods of multi-source data body fusion, it includes: obtaining the coal sample of coal reservoir of certain well in certain study area, the non-overlapping part of fracture intensity obtained from artificial description core fracture intensity and imaging logging data is using the higher fracture intensity value in two, generate coal sample fracture intensity curve;Fractal dimension curve is established;The coal sample is taken and is carried out full diameter CT scanning, and single layer fracture rate quantitative calculation is carried out, and single layer fracture rate curve is obtained;The three kinds of curves obtained are fused, and comprehensive indication curve is established;Fault distance attribute body is constructed;Seismic data body is extracted, the correlation of simulated seismic data body and comprehensive indication curve is selected, and sensitive seismic body is selected;Fracture comprehensive discriminant function body is constructed;Coal reservoir fracture prediction model is constructed, and coal reservoir fracture prediction is carried out.The present application comprehensively takes core data, conventional logging, imaging logging, CT scanning and seismic fracture prediction data, improves the prediction precision of coal reservoir fracture.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of coalbed methane development, in particular to a coal reservoir fracture prediction model construction method based on multi-source data body fusion. BACKGROUND

[0002] Fractures are the main seepage channels of coalbed methane reservoirs, and accurate identification and evaluation of fractures are one of the important contents in the exploration and development process. Fractured reservoirs have the characteristics of heterogeneity and anisotropy, and the resolution of conventional logging in the vertical direction is limited, which cannot provide high-resolution formation information. At the same time, due to the high cost of core and imaging logging, the data is limited, and it is difficult to objectively and truly characterize the fracture distribution in a region only relying on core and imaging logging data. At present, more seismic data are used to evaluate the spatial distribution of fractures, but the results are multiple and uncertain, which may lead to misjudgment. Therefore, it is crucial to find a suitable method to accurately predict the spatial distribution of fractures in coal reservoirs for the development of coalbed methane reservoirs.

[0003] Patent document CN113534247A discloses a fracture quantitative prediction method and device based on post-stack seismic data, which only predicts fractures from the perspective of seismic data. The prediction accuracy for small-scale fractures is poor, and the prediction model itself has uncertainty. SUMMARY

[0004] The present application provides a coal reservoir fracture model construction method based on multi-source data body fusion, which solves the problem of poor prediction accuracy in predicting coal reservoir fractures in the prior art.

[0005] The technical solution adopted by the present application to solve the technical problem is that the coal reservoir fracture prediction model construction method based on multi-source data body fusion comprises the following steps:

[0006] Step one: Obtain a coal sample of a coal reservoir in a well in a certain research area, record the number of end face fractures, fracture dip angle, fracture length, fracture filling degree and fracture properties of the coal sample, calculate the artificial description of core fracture intensity, and use the imaging logging data of the well section to generate the fracture surface intensity. The non-overlapping part of the artificial description of core fracture intensity and the fracture surface intensity obtained from the imaging logging data adopts the higher fracture intensity value in the two, and generates a coal sample fracture intensity curve;

[0007] Step two: considering the sensitivity of conventional logging to fractures in the formation, adopt AC, DEN, GR, RD, RS and SP curves, introduce Hurst index by using R / S analysis method, and establish a fractal dimension curve;

[0008] Step three: full-diameter CT scanning is performed on the coal sample to obtain core fracture image slices, fracture identification and segmentation are performed, and single-layer fracture rate quantitative calculation is performed on the segmented fracture model to obtain a single-layer fracture rate curve;

[0009] Step four: the coal sample fracture strength curve, the fractal dimension curve and the single-layer fracture rate curve are divided at equal intervals, the curve values are normalized, the three types of curves are fused by using the entropy method, and a comprehensive indication curve is established;

[0010] Step five: the medium and large scale fractures are directly identified by using the mutation of the reflection phase axis, the wave group break, the wave system break and the reflection layer occurrence on the profile, the medium and large scale fractures are determined by using Petrel, the Geopmetrical modeling module is adopted to construct a fault distance attribute body;

[0011] Step six: the seismic body and the interpretation data of the research area are obtained, the attribute interpretation of the seismic body is performed, the related algorithm is used to extract the seismic data body, the seismic data body includes chaotic body, variance body, coherence body, structural curvature body, instantaneous amplitude, inclination deviation and ant body attribute, the correlation between the above seismic data body and the comprehensive indication curve is simulated by using the Co-Krigng method, and the sensitive seismic body is selected;

[0012] Step seven: the selected sensitive seismic body is combined with the fault distance attribute body to construct a fracture comprehensive discriminant function body by using an artificial neural network;

[0013] Step eight: based on the constructed fracture comprehensive discriminant function body, a coal reservoir fracture prediction model is constructed by using the Fracturenetwork module in Petrel, and the coal reservoir fracture prediction is performed.

[0014] The step one of the above scheme is: according to the actual situation of the core sample, the sampling is sequentially sampled at intervals of about 0.2m, on the coal sample section, first, the core is homed, the mud on the surface of part of the sample is brushed, the fresh surface is exposed, the number of cracks on the end surface of the coal sample, the crack inclination, the crack length, the crack filling degree and the crack property are recorded, the artificial description of the core fracture strength is obtained, the artificial description of the core fracture strength curve is generated, in addition, the imaging logging data of the well section is used, the CIFlog software is used to generate the fracture surface strength, and the fracture surface strength curve is generated, the two curves are placed together, the non-overlapping part of the two curves adopts the higher fracture strength value of the two, and the coal sample fracture strength curve is generated by connecting.

[0015] In the above scheme, the artificial description of the core fracture strength and the fracture surface strength generated by CIFlog have the same formula f=N / (π*r 2 )(1)

[0016] In the formula, f is the artificial fracture core strength, strip / m2 N is the number of fractures in the core cross-sectional range, strips; r is the core radius, m.

[0017] The method for establishing the fractal dimension curve by introducing the Hurst index in step two of the above scheme is as follows:

[0018] There is a unique conventional logging value at different logging depths, Z = Z(1), Z(2), … Z(n), the logging sequence is denoted as Z(i) and Z(j), i, j = 1, 2, … n, and the full-range range R(n) and the standard deviation S(n) of the logging sequence are represented by the following expressions:

[0019]

[0020] When the logging sequence satisfies statistical self-similarity, there is

[0021]

[0022] lg[R(n) / S(n)] = lgC + H·lgn (5)

[0023] D = 2-H (6)

[0024] In the formula, Z is the logging sequence; R is the full-range range of the logging sequence; S is the full-range standard deviation of the logging sequence; n is the number of logging sequences; u is the number of scales increasing from 0 to n; C is a certain constant; h is the delay distance; H is the Hurst index; and D is the fractal dimension.

[0025] The single-layer fracture quantitative calculation method in step three of the above scheme:

[0026] Since the X-ray CT scanning imaging analysis of the internal structure of the coal sample is realized by the relative density difference, the micro-fracture not filled in the coal sample belongs to the part with the lowest density, and the micro-fracture appears as a black gray color much deeper than the gray value of vitrinite and bright coal in the CT image; the gray boundary of vitrinite and bright coal and micro-fracture is identified by using the watershed algorithm in Avizo, and the volume of 500000 pixels is used as the threshold value for micro-fracture segmentation, the micro-fracture in the sample after denoising and sharpening is segmented and extracted, after the extraction of the fracture, the volume fraction instruction in Avizo is used to calculate the ratio of the extracted fracture voxel and the set background voxel, and the fracture rate of the sample is obtained;

[0027]

[0028] In the formula, F represents the single-layer fracture rate in the core section, S f represents the area of the single-layer fracture gray image, m 2 ; r is the core radius, m, and the higher the F value represents the higher the fracture ratio.

[0029] The method for establishing a comprehensive indicator curve using the entropy method in step 4 of step 4 of the above scheme is as follows:

[0030] (1) Sample data normalization

[0031] Normalize the data of fracture intensity curve, fractal dimension curve and core single layer fracture rate curve;

[0032] (2) Normalized data volume translation

[0033] In the entropy weight method, when x mn When =0, the entropy value is meaningless, and all normalized data are shifted to the right by 0.0001 units;

[0034] (3) Entropy weight calculation

[0035] After data preprocessing is completed, the entropy value and weight of each curve parameter are calculated, and the obtained curve value is multiplied by its obtained weight, and then normalized to obtain the comprehensive indicator curve;

[0036] The entropy calculation formula is:

[0037]

[0038] Q=α·K f +β·K D +λ·K F (11)

[0039] Where Q b is the information entropy value of each ion, ε b is the weight, m ​​is the number of samples, n is the number of indicators, α is the fracture intensity weight, β is the fractal dimension weight, λ is the core fracture rate weight, K f is the crack strength value, K D is the fractal dimension value, K F is the core fracture rate value.

[0040] Step 5 of the above scheme is as follows: Based on the deterministic modeling of the faults in the study area, seismic data is used to directly identify medium- and large-scale fractures in the profile through the sudden changes in reflection phase axes, the offset of wave groups and wave systems, and the sudden changes in the attitude of reflectors. The size of the fault is analyzed, and Petrel is used to deterministically model this part. The scope of the fault influence is delineated according to experience, the seismic fracture attribute characteristics are cut, and the geometrical modeling module is used to construct the fault distance attribute body.

[0041] The specific method of step 7 in the above scheme is that the artificial neural network adopts supervised machine learning, the training supervision points are pre-set, the training supervision points include crack sample points and non-crack sample points, the learning samples include a sensitive data body, a comprehensive indication curve and a crack training set, the artificial neural network is valued by the training supervision points, is extrapolated to a three-dimensional direction by a crack development probability, and finally a comprehensive discrimination body is constructed.

[0042] Beneficial effects:

[0043] 1. The present application comprehensively takes core data, conventional logging, imaging logging, CT scanning and seismic prediction crack data, adopts mathematical analysis and fusion, so as to solve or partially solve the technical problem that the prediction accuracy cannot be ensured when predicting coal reservoir cracks in the prior art.

[0044] 2. The present application extracts crack strength by artificial coring description and imaging logging, identifies conventional logging fractals by R / S analysis method, extracts single-layer crack rate by CT scanning, fuses comprehensive curves by entropy method, finally combines sensitive seismic body and fault distance attribute body by artificial neural network, and constructs a crack model by Petrel, so that multiple data bodies can be effectively fused, the prediction accuracy of cracks is improved, and a basis is provided for actual production.

[0045] 3. The present application analyzes from the angles of seismic and logging, selects seismic attribute bodies, and screens sensitive attribute bodies, so that the predicted crack region is comprehensively determined by artificial, conventional logging and CT scanning, and has the characteristics of accurate prediction.

[0046] 4. The present application constructs a crack strength curve by using gas well coring artificial crack description and combining array sonic logging; considers the sensitivity of conventional logging to cracks in the formation, introduces Hurst index by R / S analysis method, establishes a fractal dimension curve; effectively identifies single-layer crack gray image area based on full-diameter CT scanning, establishes a single-layer crack rate curve; normalizes the values of each curve, fuses three types of curves by entropy method, and establishes a comprehensive indication curve; for medium and large scale cracks, deterministic modeling is adopted, and a fault distance data body is established; multiple sensitive seismic attribute bodies are extracted by Co-Krigng simulation, a comprehensive discrimination function body is constructed by artificial neural network and the comprehensive indication curve, a coal reservoir crack prediction model is established, and the prediction accuracy of coal reservoir cracks is improved by comprehensively considering multiple data bodies. BRIEF DESCRIPTION OF DRAWINGS

[0047] Figure 1 is a flowchart of the present application;

[0048] Figure 2 is a fault crack data body graph in the present application;

[0049] Figure 3The process chart for calculating the single-layer fissure rate of the full-diameter CT scanning in the present application is shown in the figure;

[0050] Figure 4 The fracture prediction model chart for the coal reservoir in the present application is shown in the figure;

[0051] Figure 5 The method chart for generating the fracture intensity curve in step 1 is shown in the figure. DETAILED DESCRIPTION

[0052] The present application is further described below in combination with the accompanying drawings:

[0053] In combination with Figures 1-5 As shown in the figure, the coal reservoir fracture prediction model construction method of the multi-source data body fusion comprises the following steps:

[0054] Step 1: According to the actual situation of the coring sample, the sampling is sequentially sampled at an interval of about 0.2 m. On the coring coal sample section, first, the core is homed, the mud on the surface of part of the sample is brushed, the fresh surface is exposed, the parameters such as the number of fissures, the fissure dip angle, the fissure length, the fissure filling degree and the fissure property of the coal sample end surface are recorded, the fracture intensity is calculated, the imaging logging data exists in part of the well section, the fracture surface intensity is generated in combination with the CIFlog software, the higher fracture intensity value is adopted for the overlapping part of the two, and the coal sample fracture intensity curve is generated. That is, the fracture intensity values are obtained and the fracture intensity curve is generated through the artificial core description and the imaging logging interpretation respectively, the two curves are placed together, the higher fracture intensity value of the two is adopted for the non-overlapping part of the two curves, and the final fracture intensity curve is generated. As shown in the figure, the final fracture intensity curve is the curve formed by the thick solid line. Figure 5 As shown in the figure, the final fracture intensity curve is the curve formed by the thick solid line.

[0055] The formula of the artificial core fracture intensity and the fracture surface intensity generated by the CIFlog is

[0056] f=N / (π*r 2 )(1)

[0057] In the formula, f is the artificial fracture core intensity, strip / m 2 ; N is the number of fractures in the core section range, strip; r is the core radius, m.

[0058] Step 2: Considering the sensitivity of the conventional logging to the fractures in the formation, the AC, DEN, GR, RD, RS and SP curves are mainly adopted, the Hurst index is introduced by using the R / S analysis method, and the fractal dimension curve is established.

[0059] There is a unique conventional logging value at different logging depths, Z=Z(1), Z(2), … Z(n), the sequence can be recorded as Z(i) and Z(j), i, j=1, 2, … n, and the full-section range R(n) and the standard deviation S(n) of the sequence are expressed by the following formula:

[0060]

[0061] When the sequence satisfies statistical self-similarity, there exists

[0062]

[0063] lg[R(n) / S(n)]=lgC+H·lgn (5)

[0064] D=2-H(6)

[0065] Where Z is the logging sequence; R is the full range of the logging sequence; S is the full standard deviation of the logging sequence; n is the number of logging sequences; u is the number of scales increasing from 0 to n; C is a constant; h is the delay distance; H is the Hurst exponent; and D is the fractal dimension.

[0066] For all the well logging data used, starting from the third sampling point, the R(n) / S(n) of each set of well logging data points is obtained using equations (2) to (5). When the horizontal and vertical coordinates are set to logarithmic scales, the linear fit between R(n) / S(n) and N scatter points is very high, and its slope is the index of the well logging sequence. Its fractal dimension D can be calculated by equation (6).

[0067] Step 3: Perform a full-diameter CT scan on the coal sample core to obtain core fracture image slices. Use image processing methods such as threshold segmentation to identify and segment the fractures, and perform quantitative calculation of the single-layer fracture rate of the segmented fracture model.

[0068] The extraction and calculation process of single-layer crack rate is as follows:

[0069] Since X-ray CT scanning analyzes the internal structure of a sample primarily through relative density differences, unfilled microcracks in coal have the lowest density and therefore appear darker in the CT image than the grayscale values ​​of vitrectin and bright coal. A model was reconstructed from the 1,500 scanned photos in Avizo. The watershed algorithm was used to identify the grayscale boundaries between vitrectin and bright coal and the cracks. Using a volume of 500,000 pixels as the threshold for microcrack segmentation, the microcracks in the denoised and sharpened sample were segmented and extracted. After crack extraction, the volumefraction command in Avizo was used to calculate the ratio of the extracted crack voxels in the 1,500 photos to the set background voxels to obtain the crack rate of the sample.

[0070]

[0071] In the formula, F represents the single-layer fracture rate in the core section, S fRepresents the grayscale image area of ​​a single layer of cracks, m 2 ; r is the core radius, m, and the higher the F value, the higher the proportion of fractures.

[0072] Step 4: Divide each curve into equal intervals, normalize the curve values, and use the entropy method to fuse the three types of curves to establish a comprehensive indicator curve.

[0073] The specific measures for entropy method and comprehensive indicator curve calculation are as follows:

[0074] (1) Sample data normalization

[0075] The fracture intensity curve, fractal dimension curve, and core single-layer fracture rate curve data were normalized so that all indicators in the sample were in the same dimension to eliminate the analysis errors caused by large numerical differences between the indicators.

[0076] (2) Normalized data volume translation

[0077] In the entropy weight method, when x mn When =0, the entropy value is meaningless. Based on this, all normalized data are shifted to the right by 0.0001 units.

[0078] (3) Entropy weight calculation

[0079] After data preprocessing is completed, the entropy value and weight of each curve parameter can be calculated, the obtained curve value is multiplied by its obtained weight, and then normalized to obtain the comprehensive indicator curve.

[0080] The entropy calculation formula is:

[0081]

[0082] Q=α·K f +β·K D +λ·K F (11)

[0083] Where Q b is the information entropy value of each ion, ε b is the weight, m ​​is the number of samples, n is the number of indicators, α, β, λ are the weight of fracture intensity, fractal dimension and core fracture rate, K f , K D , K F are the fracture intensity value, fractal dimension value, and core fracture rate value.

[0084] Step 5: From the perspective of seismic interpretation, seismic data is used to directly identify medium- and large-scale fractures in the profile through sudden changes in reflection events, wave group offsets, wave system offsets, and sudden changes in reflector attitude. These medium- and large-scale fractures are deterministically modeled using Petrel. The Geometrical Modeling module is used to construct a fault distance attribute volume. This data volume is mainly used to ensure the presence of small fractures near medium- and large-scale faults, ensuring that the final constructed model conforms to the actual geological characteristics.

[0085] That is, the fault distance attribute body is constructed based on the deterministic modeling of the faults in the study area, the fault size is analyzed, the range of possible impact of the fault is delineated according to experience, the earthquake crack attribute characteristics are cut, and the fault distance attribute body is constructed.

[0086] Step 6: Interpret the attributes of the seismic volume. Use relevant algorithms to extract the chaotic volume, variance volume, coherence volume, structural curvature volume, instantaneous amplitude, dip deviation, and ant volume attributes. Use the Co-Krigng method to simulate the correlation between the above seismic data volume and the comprehensive indicator curve, and select sensitive seismic volumes to proceed to step 7.

[0087] Step 7: Select sensitive earthquake bodies and use artificial neural networks in combination with fault distance attribute bodies to construct a comprehensive fracture discriminant function body. The artificial neural network specifically uses supervised machine learning, that is, the training supervision points are pre-set. The training supervision points include fracture sample points and non-fracture sample points. The learning samples include the sensitive data body acquired in the early stage, the fracture development body (comprehensive indicator curve construction) and the fracture training set. The artificial neural network assigns values ​​to the supervision points and extrapolates the fracture development probability in the three-dimensional direction to finally construct a comprehensive discriminant body.

[0088] Step 8: Finally, based on the constructed comprehensive fracture discriminant function, the FractureNetwork module in Petrel is used to construct a coal reservoir fracture prediction model (i.e., a coal reservoir fracture discrete network model); and the coal reservoir fracture prediction model is used to predict coal reservoir fractures.

Claims

1. A method for constructing a coal reservoir fracture prediction model based on multi-source data fusion, characterized in that The steps include: Step 1: Obtain a coal sample from a coal reservoir in a well in a certain study area, record the number of fractures on the coal sample end face, fracture inclination, fracture length, fracture filling degree, and fracture properties, calculate the artificial core fracture strength, and also generate the fracture surface strength using imaging logging data of the well section. For the non-overlapping portion of the artificial core fracture strength and the fracture surface strength obtained from imaging logging data, the higher fracture strength value is used to generate the coal sample fracture strength curve; Step 2: Considering the sensitivity of conventional logging to fractures in the formation, AC, DEN, GR, RD, RS, and SP curves are used, and the Hurst index is introduced by the R / S analysis method to establish a fractal dimension curve; Step 3: Perform full-diameter CT scanning on the coal sample to obtain core fracture image slices, perform fracture identification and segmentation, and perform quantitative calculation of the single-layer fracture rate on the segmented fracture model to obtain a single-layer fracture rate curve; Step 4: Divide the coal sample crack intensity curve, fractal dimension curve and single layer crack rate curve into equal intervals, normalize the curve values, and use the entropy method to fuse the three types of curves to establish a comprehensive indicator curve; Step 5: Use seismic data to directly identify medium- and large-scale fractures in the profile through sudden changes in reflection events, wave group offsets, wave system offsets, and sudden changes in reflector attitude. Use Petrel to perform deterministic modeling on the identified medium- and large-scale fractures, and use the Geometrical modeling module to construct a fault distance attribute volume. Step 6: Obtain the seismic volume and its interpretation data in the study area, interpret the attributes of the seismic volume, and use relevant algorithms to extract the seismic data volume. The seismic data volume includes chaos volume, variance volume, coherence volume, structural curvature volume, instantaneous amplitude, dip deviation, and ant volume attributes. Use the Co-Krigng method to simulate the correlation between the above seismic data volume and the comprehensive indicator curve to select sensitive seismic volumes; Step 7: Using the selected sensitive earthquake body and the fault distance attribute body, an artificial neural network is used to construct a comprehensive fracture discrimination function body; Step 8: Based on the constructed comprehensive fracture discriminant function, the FractureNetwork module in Petrel is used to construct a coal reservoir fracture prediction model to predict coal reservoir fractures.

2. The method for constructing a coal reservoir fracture prediction model based on multi-source data fusion according to claim 1, characterized in that: The step one is specifically as follows: according to the actual situation of the coring sample, sampling is carried out in an orderly manner at intervals of 0.2m. At the coal sample cross section, the core is first returned to its original position, the mud on the surface of part of the sample is brushed to expose the fresh surface, and the number of cracks on the end face of the coal sample, the crack inclination, the crack length, the crack filling degree and the crack properties are recorded to obtain an artificial description of the core crack strength and generate an artificial description of the core crack strength curve. In addition, the imaging logging data of the well section is used to use CIFlog software to generate the fracture surface strength and generate a fracture surface strength curve. The two curves are placed together, and the non-overlapping part of the two curves adopts the higher fracture strength value of the two, and they are connected to generate a coal sample fracture strength curve.

3. The method for constructing a coal reservoir fracture prediction model based on multi-source data fusion according to claim 2, characterized in that: The formulas for manually describing the core fracture strength and the fracture surface strength generated by CIFlog are: f=N / (π*r 2 ) (1) Where, f is the core strength of artificial fractures, bars / m 2 ; N is the number of cracks within the core cross section, r is the core radius, m.

4. The method for constructing a coal reservoir fracture prediction model based on multi-source data fusion according to claim 3, characterized in that: In step 2, the Hurst exponent is introduced by the R / S analysis method, and the method for establishing the fractal dimension curve is as follows: There is a unique conventional logging value for different logging depths, Z = Z(1), Z(2), ... Z(n), the logging sequence is recorded as Z(i) and Z(j), i, j = 1, 2, ... n, the full range R(n) and standard deviation S(n) of the logging sequence are expressed by the following expressions: When the logging sequence satisfies statistical self-similarity, there exists: lg[R(n) / S(n)]=lgC+H·lgn (5) D=2-H (6) Where Z is the logging sequence; R is the full range of the logging sequence; S is the full standard deviation of the logging sequence; n is the number of logging sequences; u is the number of scales increasing from 0 to n; C is a constant; H is the Hurst exponent; and D is the fractal dimension.

5. The method for constructing a coal reservoir fracture prediction model based on multi-source data fusion according to claim 4, characterized in that: The quantitative calculation method of the single layer crack rate in step 3 is: Since X-ray CT scanning analyzes the internal structure of a sample through its relative density difference, the unfilled microcracks in the coal sample belong to the lowest density part. Microcracks appear as a dark gray color in the CT image, which is much darker than the grayscale values ​​of vitrine and bright coal. In Avizo, the 1500 scanned photos were reconstructed, and the grayscale boundaries between vitrine, bright coal and microcracks were identified using the watershed algorithm. The microcracks in the denoised and sharpened sample were segmented and extracted using a volume of 500,000 pixels as the threshold for microcrack segmentation. After crack extraction, the volume fraction command in Avizo was used to calculate the ratio of the crack voxels extracted from the 1500 photos to the set background voxels to obtain the crack rate of the sample: Where F represents the single-layer fracture rate in the core section, S f Represents the grayscale image area of ​​a single layer of cracks, m 2 ; r is the core radius, m, and the higher the F value, the higher the proportion of fractures.

6. The method for constructing a coal reservoir fracture prediction model based on multi-source data fusion according to claim 5, characterized in that: Specifically, step 5 comprises the following steps: based on deterministic modeling of faults in the study area, using seismic data to directly identify medium- and large-scale fractures in the profile through sudden changes in reflection events, offsets in wave groups and wave systems, and sudden changes in reflector attitudes, analyzing the size of the faults, deterministically modeling these medium- and large-scale fractures using Petrel, empirically delineating the scope of the fault influence, cutting seismic fracture attribute characteristics, and constructing a fault distance attribute body using the Geopmetrical modeling module.

7. The method for constructing a coal reservoir fracture prediction model based on multi-source data fusion according to claim 6, characterized in that: The specific method of step 7 is as follows: the artificial neural network adopts supervised machine learning, the training supervision points are pre-set, the training supervision points include crack sample points and non-crack sample points, the learning samples include sensitive data bodies, comprehensive indicator curves and crack training sets, the artificial neural network assigns values ​​to the training supervision points, and extrapolates the probability of crack development in the three-dimensional direction, and finally constructs a comprehensive discriminant.

Images