Stratum distribution interpretation method based on drilling geophysical prospecting data decision-making layer fusion

Abstract

The invention discloses a stratum distribution interpretation method based on drilling geophysical prospecting data decision-making layer fusion, and relates to the technical field of geological survey, the interpretation method takes a geophysical prospecting data interpretation result as soft data (with low credibility), takes a drilling data interpretation result as hard data (with high credibility), and takes a Bayesian probability updating principle as a basis to complete the interpretation of the stratum distribution of the drilling geophysical prospecting data decision-making layer fusion. A neighborhood system optimization method, a corresponding prior probability distribution calculation method, a likelihood function calculation method and an interval division global updating mechanism considering drilling data influence are provided, and two kinds of exploration data are effectively fused at a decision-making layer. High-reliability and high-accuracy interpretation of data information capable of being used for modeling of the two-dimensional layer stratum distribution condition is realized; the method has the advantages of being high in accuracy, good in universality and high in operability, and is suitable for being applied to two-dimensional modeling of stratum distribution on a large scale.

Description

Technical Field

[0001] The present invention relates to the field of geological survey technology, and in particular to a method for interpreting stratum distribution based on drilling geophysical data decision layer fusion. Background Art

[0002] In the field of geotechnical engineering, it is very important to ascertain the engineering geological conditions within the construction area to ensure construction safety. Among them, the distribution of strata is an important basic information of engineering geological conditions, and it is crucial to accurately and reasonably model it based on geological exploration information.

[0003] At present, exploration methods mainly include geophysical exploration and drilling. Drilling is a way to directly obtain stratigraphic information. By drilling and coring the strata, real geological data can be obtained, and the data has a high degree of credibility. Therefore, drilling results are often used as an important reference. Geophysical exploration technology is mainly based on ground penetrating radar, high-density electrical method, seismic wave and other methods. It has the advantages of large coverage, wide exploration depth and relatively low cost. It can also help to effectively identify the distribution of strata. However, according to engineering practice, the shortcomings of both are also very obvious: drilling costs are high and it belongs to one-dimensional data, which may cause the information obtained to be local and one-sided, and it is difficult to fully reflect the large-scale geological characteristics; geophysical exploration can quickly obtain large-scale two-dimensional profile stratigraphic distribution information, but geophysical exploration signals are often limited by the problem of multiple solutions, so the accuracy and credibility of exploration results are relatively low compared to drilling.

[0004] In theory, by fusing and interpreting drilling and geophysical data, the advantages of both technologies can be fully utilized: on the one hand, drilling data can provide reliable calibration and constraints for geophysical exploration, and improve the credibility of geophysical exploration results; on the other hand, geophysical exploration data can expand the application scope of drilling results, making up for the shortcomings of small number of drillings and limited coverage. However, there are still many challenges in practical applications: First, the data of the two exploration methods have dimensional differences. Secondly, the actual engineering geological conditions are often extremely complex, with multi-level and multi-type rock and soil bodies, which makes it difficult for traditional data fusion methods to accurately reflect the actual geological conditions.

[0005] At present, the fusion of exploration data can be roughly divided into three categories, namely data layer fusion, feature layer fusion and decision layer fusion. Data layer fusion refers to the direct fusion of the original data of the data source. This fusion method retains the most original information and is suitable for applications that require high precision and high real-time performance. Feature layer fusion refers to first extracting features from each data source and then merging these features together to improve the accuracy and robustness of the model. This fusion method can better balance the amount of data and computational complexity. Decision layer fusion refers to processing each data source independently and generating their own decisions, and then fusing these decision results. This fusion method improves the accuracy and reliability of the overall decision by integrating the results of each decision.

[0006] Therefore, how to integrate drilling geophysical data at the decision-making level to form highly reliable and accurate information interpretation results that can be used for two-dimensional geological stratigraphic modeling is a key technical issue that urgently needs to be solved in this field. Summary of the invention

[0007] The purpose of the present invention is to overcome the problem that the prior art cannot fuse the drilling and geophysical data at the decision level and form information interpretation results with high reliability and accuracy, and proposes a stratigraphic distribution interpretation method based on the decision level fusion of drilling and geophysical data.

[0008] In order to achieve the above-mentioned object of the invention, the present invention provides a method for interpreting stratum distribution based on decision-layer fusion of drilling geophysical exploration data, comprising the following steps:

[0009] S1. Obtaining interpretation results of geophysical data and drilling data of stratum distribution respectively;

[0010] S2, gridding and standardizing the geophysical data interpretation results and the drilling data interpretation results respectively, to obtain a number of geophysical data grid units and drilling data grid units with the same interpretation space and label;

[0011] S3. Optimize the neighborhood system of the geophysical data grid unit based on spatial constraints; construct a two-dimensional prior probability distribution function of the stratigraphic distribution using the prior probability calculation formula based on the optimized neighborhood system; perform equidistant segmentation and secondary index calculation on the drilling data grid unit based on the grid length of the geophysical data grid unit, and then construct a likelihood function of the stratigraphic distribution using the confusion matrix;

[0012] S4, based on the two-dimensional prior probability distribution function and likelihood function of the formation distribution, the Bayesian probability update mechanism is used to perform fusion interpretation processing to obtain the fusion interpretation result of the formation distribution at the drilling hole location;

[0013] S5. Based on the two-dimensional prior probability distribution function of the stratigraphic distribution and the fused interpretation results of the stratigraphic distribution at the drilling hole position, a fused interpretation process is performed using a global update mechanism based on interval division to obtain a fused interpretation result of the global stratigraphic distribution.

[0014] The present invention discloses a stratigraphic distribution interpretation method based on the fusion of drilling and geophysical data decision layers. The geophysical data interpretation result is regarded as soft data (with lower credibility), and the drilling data interpretation result is regarded as hard data (with higher credibility). Based on the Bayesian probability updating principle, a method for optimizing the neighborhood system, a corresponding prior probability distribution calculation method, a likelihood function calculation method, and a global updating mechanism for interval division considering the influence of drilling data are proposed. The two types of exploration data are effectively fused at the decision layer, and a highly reliable and accurate interpretation of data information that can be used for two-dimensional stratigraphic distribution modeling is achieved. The method has the characteristics of high accuracy, good universality, and strong operability, and is suitable for large-scale application in two-dimensional modeling of stratigraphic distribution.

[0015] Wherein, in step S1, preferably, the geophysical prospecting includes one of electrical prospecting, ground penetrating radar and seismic prospecting; and the drilling is advance drilling.

[0016] Geophysical electromagnetic waves, elastic waves and other signals are affected by various interference factors during propagation, and usually there are various types of noise. Therefore, preprocessing is required to improve the signal-to-noise ratio to better highlight the characteristics of the disease; preferably, in the process of obtaining the interpretation results of geophysical data of stratum distribution, preprocessing of geophysical data is also included; the preprocessing includes at least one of zero point reset, high-pass and low-pass filtering, and exponential gain.

[0017] Among them, in step S2, preferably, the gridding processing of the geophysical interpretation results includes: selecting an appropriate square grid length (the denser the grid, the higher the accuracy of the interpretation results obtained, but the larger the calculation scale, which is reasonably set according to the overall exploration area scale and exploration needs), gridding the two-dimensional geophysical interpretation results, and obtaining a number of geophysical data grid units; the gridding processing of the drilling data interpretation results includes: selecting the same grid length as that of the geophysical interpretation results gridding processing, gridding the drilling data interpretation results of the borehole position, and obtaining a number of borehole data grid units; through gridding processing, it is ensured that the grid units at the borehole position correspond to the grid unit positions of the geophysical interpretation results at the same position.

[0018] Preferably, the standardization process includes: unifying the geophysical data grid unit and the borehole data grid unit, and using the same interpretation space and label to represent them. For example, the geophysical grid unit and the borehole grid unit can be represented as follows:

[0019] S=(label1,label2...label n );

[0020] In the formula, label means label, S represents the grid unit interpretation space, the interpretation space is composed of interpretation labels, and there are n interpretation labels including label1, label2...labeln; the interpretation labels are generally numbers 0, 1, 2, etc., and each data represents a stratigraphic distribution category. If the grid units have the same interpretation space and labels, it means that the stratigraphic distribution categories contained in the two data should be consistent.

[0021] Among them, in step S3, the optimization method of the neighborhood system includes: (A) calculating the discrete degree of its constraint factor based on the interpretation result of the geophysical data grid unit; (B) according to the calculation result of the discrete degree of the constraint factor, taking the direction with the smallest discrete degree of the grid unit constraint factor as the dominant direction.

[0022] Preferably, the constraint factor is characterized by: β = (β1, β2, β3, β4), wherein β1, β2, β3 and β4 represent 0, The spatial constraints corresponding to four independent directions, π is the ratio of a circle to a circle.

[0023] Preferably, the discrete degree of the constraint factor is: β i ={Mean[s(label n )]} βi , where i is 1, 2, 3, and 4, representing the calculation of 0, The degree of discreteness in four independent directions; s(label n ) indicates that β is calculated separately i The variance of the grid cell interpretation label in the direction, Mean means taking the average value of the variance calculation results.

[0024] Preferably, the prior probability calculation formula is:

[0025]

[0026] In the formula, represents the initial probability distribution of the interpretation labels in each geophysical data grid cell; represents the initial probability update value considering spatial correlation, where spatial correlation refers to the interpretation label situation in the optimized neighborhood system of the current geophysical data grid unit. The specific value is equal to the frequency of each interpretation label, and finally the result of the addition of the two is renormalized; f x (X) represents the prior probability distribution function of the stratum distribution information X.

[0027] Preferably, the equidistant segmentation includes: segmenting the drilling data according to the grid length of the geophysical data grid unit; the secondary index calculation includes: calculating the mean and variance of the drilling index parameters within the equidistant segmentation sections.

[0028] Preferably, the method of calculating the likelihood function of the stratum distribution using the confusion matrix comprises the steps of:

[0029] (1) Using the optimized classifier model to train the drilling data grid cell data (drilling location) after isometric segmentation and secondary index calculation;

[0030] (2) Use the trained classifier to predict the drilling data grid cell data (undrilled locations);

[0031] (3) The likelihood function is constructed based on the confusion matrix of the predicted output.

[0032] Preferably, in step (1), the classifier model is XGBoost; the model constructs multiple weak evaluators on the data one by one, and gradually accumulates and summarizes the modeling results of multiple weak evaluators through multiple iterations to obtain better regression or classification performance than a single model. This superposition strategy using a single decision tree as a weak evaluator can be expressed in an additive form, that is, the XGBoost classifier model can be expressed as:

[0033]

[0034] In the formula, represents the prediction result of the entire model on this sample, k represents the total number of weak evaluators (decision trees), and f k represents the kth decision tree, x i Represents the eigenvector corresponding to sample i.

[0035] Among them, the optimized XGBoost introduces model complexity to measure the computational efficiency of the algorithm. Therefore, the optimized XGBoost model consists of two parts: the traditional loss function and the model complexity. The specific formula is:

[0036]

[0037] In the formula, O bj represents the objective function of the model, n represents the total amount of data imported into the kth tree, and y i represents the true label, represents the predicted value, f k represents the kth decision tree, Ω represents the transformation value of the tree model, Represents the traditional loss function, measuring the true label y i and predicted values The difference between Represents the complexity of the model.

[0038] Among them, in step (2), the expression of constructing the likelihood function based on the confusion matrix is:

[0039]

[0040] Where TP and FN represent the number of true positive and false negative samples in the prediction confusion matrix, respectively.

[0041] Among them, in step S4, the Bayesian probability update mechanism includes: based on the two-dimensional prior probability distribution function and the likelihood function, using the Bayesian probability estimation formula to obtain the posterior probability distribution function, and then obtaining the fusion interpretation result of the formation distribution at the drilling hole position through the posterior probability distribution function; the Bayesian probability estimation formula is:

[0042] f x ′(X)=aL(X)f x (X);

[0043] In the formula, f x ′(X) represents the posterior probability distribution function of the formation distribution information X; f x (X) represents the prior probability distribution function of the stratum distribution information X; X is the original spatial random input vector (stratum distribution information), X = (X1, X2, ..., X n ) T , n is the number of random parameters; L(X) represents the likelihood function; a is a constant; T is the matrix transposition operator.

[0044] Among them, in step S5, the global update mechanism based on interval division includes: based on the two-dimensional prior probability distribution function of the formation distribution and the fusion interpretation result of the formation distribution at the drilling hole position, using the global update formula to obtain the fusion probability distribution function, and then obtaining the fusion interpretation result of the global formation distribution through the fusion probability distribution function; the global update formula is:

[0045] f x ″(X)=f x (X)+d x (X);

[0046] In the formula, f x ″(X) represents the posterior probability distribution function of the formation distribution information X that needs to be globally updated in addition to the Bayesian probability update position; d x (X) represents the probability update value calculated based on the posterior update probability of the drilling data grid unit that has the greatest impact on the current grid unit and taking into account the distance between the two; f x (X) represents the prior probability distribution function of the stratum distribution information X.

[0047] Preferably, d x The calculation formula of (X) is:

[0048] dx(X)=f x ′(X)ω d ;

[0049] In the formula, ω d represents the weight, which is related to the distance from the drilling data grid unit. The closer the position is to the drilling data grid unit, the greater the influence of the drilling data is, that is, the greater the weight is; the farther the position is from the drilling data grid unit, the less the influence of the drilling data is, that is, the smaller the weight is; f x ′(X) represents the posterior probability distribution function of the formation distribution information X.

[0050] Preferably, ω d The calculation formula is as follows: d =e -bd ; Where d is the distance from the drilling grid unit, b is a constant term, and e is a constant.

[0051] Preferably, the method for locating the drilling grid unit that has the greatest impact on the current grid includes:

[0052] a. Interval division: The two-dimensional stratigraphic section is divided into intervals according to the number of boreholes. If the number of boreholes is n, it is divided into n+1 intervals, and the grid cells in each interval are calculated independently in the interval;

[0053] b. Interface curve equation fitting: In each interval, the formation interface is fitted using a polynomial equation to obtain the corresponding quantitative curve equation (g i (X), where i = 1, 2, ..., n,, n + 1) and are mapped to two-dimensional grid cells within their respective intervals;

[0054] c. Calculate the vertical distance: Calculate the current grid cell A(X A ,Y A ) and the vertical distance d between the grid cells corresponding to the quantitative curve equation gi(X) A , that is, d A =g(X A )-Y A ;

[0055] d. Locate the drilling grid unit with the greatest impact: In each interval, set the vertical intersection C(X c ,Y c ), then combined with d A The drilling grid unit with the greatest impact on the current grid is obtained as A'(X A' ,YA' ), where X A' =X C , Y A' =Y C -d A .

[0057] Furthermore, the present invention also proposes an interpretation device, comprising at least one processor, and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above-mentioned stratigraphic distribution interpretation method based on drilling geophysical data decision layer fusion.

[0058] Furthermore, the present invention also proposes a computer-readable medium on which instructions executable by a processor are stored. When the instructions are executed by the processor, the processor executes the above-mentioned method for interpreting stratigraphic distribution based on decision layer fusion of drilling geophysical exploration data.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] 1. The present invention discloses a stratigraphic distribution interpretation method based on the fusion of drilling geophysical data decision layers. For the geophysical data results at the two-dimensional level, a priori probability distribution calculation method based on the optimized neighborhood system is proposed. The label interpretation results can be effectively and reasonably converted into the priori probability distribution required for the Bayesian probability update calculation, so that the data fusion update can be performed through the Bayesian probability update principle, overcoming the shortcomings of traditional Kriging interpolation for data distribution form and the inability to consider the credibility of data from different sources, and significantly improving the accuracy of the interpretation results of stratigraphic distribution geological prediction data.

[0061] 2. The present invention provides a method for interpreting stratigraphic distribution by integrating drilling and geophysical data. For the drilling data results at the one-dimensional level, it proposes to train the model and obtain the confusion matrix for each interpretation label by machine learning, so as to reasonably calculate the likelihood function required for the Bayesian probability update calculation, and proposes a global update mechanism that takes into account the influence of drilling results and interval division. By calculating the distance between the current grid unit in the divided interval and the drilling grid unit that has the greatest influence on the current grid and converting it into a weight, the final fusion of the two data is achieved, which significantly improves the accuracy and credibility of the interpretation results of the global stratigraphic distribution geological forecast data, and is conducive to the construction of a stratigraphic distribution model. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 It is a schematic diagram of the process of Example 1 of the method for interpreting formation distribution of the present invention;

[0063] Figure 2 It is a schematic diagram of the constraining direction of the spatial constraint factor in the stratum distribution interpretation method of the present invention;

[0064] Figure 3 Schematic diagram of four typical stratum distributions under an ideal state of spatial constraint factors in the drilling stratum distribution interpretation method of the present invention;

[0065] Figure 4 It is a schematic diagram of a neighborhood system after the spatial constraint factor is optimized under an ideal state in the stratum distribution interpretation method of the present invention;

[0066] Figure 5 It is a schematic diagram of the global update mechanism of interval division in the method for interpreting stratum distribution of the present invention;

[0067] Figure 6 The grid of the drilling geophysical interpretation result and the standardized result diagram in Example 1 of the stratigraphic distribution interpretation method of the present invention;

[0068] Figure 7 This is a graph showing the calculation results of the prior probability of geophysical stratum interpretation labels based on the optimized neighborhood system in Example 1 of the stratum distribution interpretation method of the present invention;

[0069] Figure 8 This is a graph showing the update result of the posterior probability of the drilling position based on the Bayesian conditional probability in Example 1 of the method for interpreting the distribution of drilling formations of the present invention;

[0070] Fig. 9 It is a comparison diagram of the overall edge detection and curve equation fitting of the stratum interface in Example 1 and Comparative Example 1 of the stratum distribution interpretation method of the present invention;

[0071] Fig.10 It is a diagram of overall edge detection and curve equation fitting of the stratum interface of interval 1 in Example 1 of the stratum distribution interpretation method of the present invention;

[0072] Fig.11 It is a diagram of overall edge detection and curve equation fitting of the stratum interface of interval 2 in Example 1 of the drilling stratum distribution interpretation method of the present invention;

[0073] Fig.12 It is a diagram of overall edge detection and curve equation fitting of the stratum interface of interval 3 in Example 1 of the stratum distribution interpretation method of the present invention;

[0074] Fig.13 It is a stratigraphic distribution model diagram constructed using the stratigraphic distribution interpretation results in Example 1 of the stratigraphic distribution interpretation method of the present invention. DETAILED DESCRIPTION

[0075] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. However, it should not be understood that the scope of the above subject matter of the present invention is limited to the following embodiments, and all technologies realized based on the content of the present invention belong to the scope of the present invention.

[0076] Embodiment 1:

[0077] The interpretation of stratigraphic distribution data is based on the geophysical data obtained from an actual geotechnical project and the drilling data under the corresponding profile. The specific interpretation steps (see the flowchart) Figure 1 )include:

[0078] S1. Obtain the interpretation results of geophysical data and drilling data of stratum distribution respectively (the overall profile size is depth Y=40m, survey line length X=79m; from top to bottom, there are strongly weathered sandstone and moderately weathered sandstone respectively, and they are distinguished by two colors; there are two boreholes, located at X=32m and X=61m respectively);

[0079] S2, selecting a suitable square grid length (setting the grid unit size to 1 m for grid division), gridding the two-dimensional geophysical interpretation results, and obtaining a number of geophysical data grid units;

[0080] The gridding processing of the drilling data interpretation results includes: selecting the same grid length (1m) as that of the geophysical exploration interpretation results, gridding the drilling data interpretation results of the borehole position, and obtaining a number of borehole data grid units;

[0081] The standardization process includes: unifying the geophysical data grid units and the borehole data grid units, and using the same interpretation space and labels for representation (such as Figure 6 As shown, it can be seen that compared with the geophysical interpretation results, the drilling results show that the medium-weathered sandstone layer is closer to the upper layer).

[0082] S3. Optimize the neighborhood system of the geophysical data grid unit based on spatial constraints, specifically: (A) Calculate the discrete degree of its constraint factor based on the interpretation result of the geophysical data grid unit; (B) According to the calculation result of the discrete degree of the constraint factor, take the direction with the smallest discrete degree of the grid unit constraint factor as the dominant direction (the calculation result is β = (0.21, 0.46, 0.57, 0.32), which shows that the spatial constraint is dominated by β1).

[0083] The constraint factor is characterized as: β = (β1, β2, β3, β4) (e.g. Figure 2 , Figure 3 As shown), where β1, β2, β3 and β4 represent the neighborhood system 0, The spatial constraints corresponding to four independent directions, π is the ratio of a circle to a circle.

[0084] The discrete degree of constraint factor is: β i ={Mean[s(label n )]} βi , where i is 1, 2, 3, and 4, representing the calculation of 0, The degree of discreteness in four independent directions; s(label n ) indicates that β is calculated separately i The variance of the grid cell interpretation label in the direction, Mean means taking the average value of the variance calculation results.

[0085] Based on the optimized neighborhood system (such as Figure 4 The two-dimensional prior probability distribution function of the stratum distribution is constructed using the prior probability calculation formula ( Figure 7 ); Based on the grid length of the geophysical data grid unit, the drilling data grid unit is equidistantly divided and the secondary index is calculated and processed, and then the likelihood function of the stratigraphic distribution is constructed using the confusion matrix (strongly weathered sandstone (0.95, 0, 0.05), moderately weathered sandstone (0.08, 0.92));

[0086] The prior probability calculation formula is: In the formula, represents the initial probability distribution of the interpretation labels in each geophysical data grid cell; represents the initial probability update value considering spatial correlation, where spatial correlation refers to the interpretation label situation in the optimized neighborhood system of the current geophysical data grid unit. The specific value is equal to the frequency of each interpretation label, and finally the result of the addition of the two is renormalized; f x (X) represents the prior probability distribution function of the stratum distribution information X;

[0087] The method of constructing the likelihood function of stratum distribution using confusion matrix is ​​as follows: (1) using the optimized classifier model In the formula, O bj represents the objective function of the model, n represents the total amount of data imported into the kth tree, and y i represents the true label, represents the predicted value, f k represents the kth decision tree, Ω represents the transformation value of the tree model, Represents the traditional loss function, measuring the true label y i and predicted values The difference between The complexity of the representative model) is trained on the drilling data grid cell data (drilling location) that has been processed by equidistant segmentation and secondary index calculation;

[0088] (2) Use the trained classifier to predict the drilling data grid cell data (undrilled locations);

[0089] (3) The likelihood function is constructed based on the confusion matrix of the predicted output (the expression is: Where TP and FN represent the number of true positive and false negative samples in the prediction confusion matrix, respectively).

[0090] S4, based on the two-dimensional prior probability distribution function and likelihood function of the stratum distribution, using the Bayesian probability update mechanism (the Bayesian probability estimation formula is: f x ′(X)=aL(X)f x (X); where f x ′(X) represents the posterior probability distribution function of the formation distribution information X; f x (X) represents the prior probability distribution function of the stratum distribution information X; X is the original spatial random input vector (stratum distribution information), X = (X1, X2, ..., X n ) T , n is the number of random parameters; L(X) represents the likelihood function; a is a constant; T is the matrix transposition operator) to perform fusion interpretation processing to obtain the fusion interpretation result of the formation distribution at the drilling hole location ( Figure 8 shown);

[0091] S5. Based on the two-dimensional prior probability distribution function of the formation distribution and the fusion interpretation result of the formation distribution at the drilling hole position, the fusion interpretation processing is performed using the global update mechanism based on interval division to obtain the fusion interpretation result of the global formation distribution ( Fig. 9 ).

[0092] The method for data fusion using a global update mechanism based on interval division includes: based on the two-dimensional prior probability distribution function of the formation distribution and the fusion interpretation result of the formation distribution at the drilling hole position, using the global update formula to obtain the fusion probability distribution function, and then obtaining the fusion interpretation result of the global formation distribution through the fusion probability distribution function.

[0093] Among them, the global update formula is: x ″(X)=f x (X)+d x (X); where f x ″(X) represents the posterior probability distribution function of the formation distribution information X that needs to be globally updated in addition to the Bayesian probability update position; d x (X) represents the probability update value calculated based on the posterior update probability of the drilling data grid unit that has the greatest impact on the current grid unit and taking into account the distance between the two; f x (X) represents the prior probability distribution function of the stratum distribution information X.

[0094] Among them, d xThe calculation formula of (X) is: dx(X) = f x ′(X)ω d ; In the formula, ω d represents the weight, which is related to the distance from the drilling data grid unit. The closer the position is to the drilling data grid unit, the greater the influence of the drilling data is, that is, the greater the weight is; the farther the position is from the drilling data grid unit, the less the influence of the drilling data is, that is, the smaller the weight is; f x ′(X) represents the posterior probability distribution function of the formation distribution information X.

[0095] Among them, ω d The calculation formula is as follows: d =e -bd ; Where d is the distance from the drilling grid unit, b is a constant term, and e is a constant.

[0096] Among them, the positioning method of the drilling grid unit that has the greatest impact on the current grid (see the schematic diagram for Figure 5 )include:

[0097] a. Interval division: The two-dimensional stratigraphic profile is divided into three intervals according to the number of boreholes (2), and the grid cells in each interval are calculated independently within the interval;

[0098] b. Interface curve equation fitting: In each interval, the polynomial equation is used to fit the formation interface ( Fig.10 , Fig.11 , Fig.12 ), and the corresponding quantitative curve equation (g i (X), where i = 1, 2, ..., n,, n + 1) and are mapped to two-dimensional grid cells within their respective intervals;

[0099] c. Calculate the vertical distance: Calculate the current grid cell A(X A ,Y A ) and the vertical distance d between the grid cells corresponding to the quantitative curve equation gi(X) A , that is, d A =g(X A )-Y A ;

[0100] d. Locate the drilling grid unit with the greatest impact: In each interval, set the vertical intersection C(X c ,Y c ), then combined with d A The drilling grid unit with the greatest impact on the current grid is obtained as A'(X A' ,Y A' ), where X A' =XC , Y A' =Y C -d A .

[0102] S6. Based on the fusion interpretation result of the global stratum distribution obtained in step S5, a stratum distribution model is constructed ( Fig.13 ).

[0103] Example 2

[0104] An interpretation device includes at least one processor and a memory connected to the at least one processor in communication; the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute a stratigraphic distribution interpretation method for multi-source geological prediction data fusion in Example 1.

[0105] Example 3

[0106] A computer-readable medium stores instructions executable by a processor, and when the instructions are executed by the processor, the processor executes a stratigraphic distribution interpretation method for multi-source geological prediction data fusion in Example 1.

[0107] Comparative Example 1:

[0108] The same geophysical drilling data as in Example 1 is used to interpret the stratigraphic distribution data. The specific interpretation steps include:

[0109] S1 and S2 are the same as in Example 1;

[0110] S3. Based on the conventional neighborhood system (eight-neighborhood system), a two-dimensional prior probability distribution function of the stratigraphic distribution is constructed using the prior probability calculation formula; based on the grid length of the geophysical data grid unit, the drilling data grid unit is equidistantly divided and the secondary index calculation is processed, and then the confusion matrix is ​​used to construct the likelihood function of the stratigraphic distribution (the specific construction steps and principles are the same as those in Example 1).

[0111] S4, same as in Example 1;

[0112] S5, based on the 2D prior probability distribution function of rock mass quality and the fusion interpretation results of rock mass quality at drilling hole location, using the global update formula (global update formula is: f x ″(X)=f x (X)+d x (X); where f x ″(X) represents the posterior probability distribution function of the rock mass quality information X that needs to be globally updated in addition to the Bayesian probability update position; d x(X) represents the probability update value calculated based on the distance from the drilling location; f x (X) represents the prior probability distribution function of rock mass quality information X; d x The calculation formula of (X) is: dx(X) = f x ′(X)ω d ; In the formula, ω d Represents the weight, which is related to the drilling distance. The closer the distance to the drilling location, the greater the impact of the drilling data, that is, the greater the weight; the farther the distance to the drilling location, the less impact of the drilling data, that is, the smaller the weight; f x ′(X) represents the posterior probability distribution function of rock mass quality information X; ω d The calculation formula is as follows: d =e -bd ; where d is the distance from the borehole, b is a constant term, and e is a constant), the fusion probability distribution function is obtained, and then the fusion interpretation result of the global rock mass quality is obtained by fusion probability distribution function ( Fig. 9 The curve in ).

[0113] Finally, it should be noted that the embodiments described in detail above are only the best practices of the invention and cannot be used to limit the scope of rights of the invention. Equivalent replacement of the technical solutions recorded in the aforementioned embodiments does not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the invention, and they should all be included in the scope of the claims and description of the invention.

Claims

1. A method for interpreting stratum distribution based on decision-level fusion of drilling geophysical data, characterized in that: The following steps are involved: S1. Obtaining interpretation results of geophysical data and drilling data of stratum distribution respectively; S2, gridding and standardizing the geophysical data interpretation results and the drilling data interpretation results respectively, to obtain a number of geophysical data grid units and drilling data grid units with the same interpretation space and label; S3. Optimize the neighborhood system of the geophysical data grid unit based on spatial constraints; construct a two-dimensional prior probability distribution function of the stratigraphic distribution using the prior probability calculation formula based on the optimized neighborhood system; perform equidistant segmentation and secondary index calculation on the drilling data grid unit based on the grid length of the geophysical data grid unit, and then construct a likelihood function of the stratigraphic distribution using the confusion matrix; S4, based on the two-dimensional prior probability distribution function and likelihood function of the formation distribution, the Bayesian probability update mechanism is used to perform fusion interpretation processing to obtain the fusion interpretation result of the formation distribution at the drilling hole location; S5. Based on the two-dimensional prior probability distribution function of the stratigraphic distribution and the fused interpretation results of the stratigraphic distribution at the drilling hole position, a fused interpretation process is performed using a global update mechanism based on interval division to obtain a fused interpretation result of the global stratigraphic distribution.

2. The method for interpreting stratum distribution according to claim 1, characterized in that: In step S3, the optimization method of the neighborhood system includes: (A) calculating the discrete degree of its constraint factor based on the interpretation result of the geophysical data grid unit; (B) according to the calculation result of the discrete degree of the constraint factor, taking the direction with the smallest discrete degree of the grid unit constraint factor as the dominant direction.

3. The method for interpreting stratum distribution according to claim 2, characterized in that: The constraint factor is characterized as follows: β = (β1, β2, β3, β4), where β1, β2, β3 and β4 represent the Spatial constraints corresponding to four independent directions.

4. The method for interpreting stratum distribution according to claim 3, characterized in that: The degree of discreteness of the constraint factor is: β i ={Mean[s(label n )]} βi , where i is 1, 2, 3, and 4, representing the calculation The degree of discreteness in four independent directions; s(label n ) indicates that β is calculated separately i The variance of the grid cell interpretation label in the direction, Mean means taking the average value of the variance calculation results.

5. The method for interpreting stratum distribution according to claim 1, characterized in that: In step S3, the prior probability calculation formula is: In the formula, represents the initial probability distribution of the interpretation labels in each geophysical data grid cell; represents the initial probability update value considering spatial correlation; f x (X) represents the prior probability distribution function of the stratum distribution information X.

6. The method for interpreting stratum distribution according to claim 1, characterized in that: In step S3, the method of calculating the likelihood function of the stratum distribution using the confusion matrix includes the following steps: (1) Using an optimized classifier model to train the drilling data grid unit data after isometric segmentation and secondary index calculation; (2) Using the trained classifier to predict the drilling data grid cell data; (3) The likelihood function is constructed based on the confusion matrix of the predicted output.

7. The method for interpreting stratum distribution according to claim 1, characterized in that: In step S5, the global update mechanism based on interval division includes: based on the two-dimensional prior probability distribution function of the formation distribution and the fusion interpretation result of the formation distribution at the drilling hole position, using the global update formula to obtain the fusion probability distribution function, and then obtaining the fusion interpretation result of the global formation distribution by fusing the probability distribution function; the global update formula is: f x ″(X)=f x (X)+d x (X); In the formula, f x ″(X) represents the posterior probability distribution function of the formation distribution information X that needs to be globally updated in addition to the Bayesian probability update position; d x (X) represents the probability update value calculated based on the posterior update probability of the drilling data grid unit that has the greatest impact on the current grid unit and taking into account the distance between the two; f x (X) represents the prior probability distribution function of the stratum distribution information X.

8. The method for interpreting stratum distribution according to claim 7, characterized in that: The methods for locating the drilling grid cells that have the greatest impact on the current grid include: a. Interval division: The two-dimensional stratigraphic section is divided into intervals according to the number of boreholes. If the number of boreholes is n, it is divided into n+1 intervals, and the grid cells in each interval are calculated independently in the interval; b. Interface curve equation fitting: In each interval, the formation interface is fitted using a polynomial equation to obtain the corresponding quantitative curve equation and map it to the two-dimensional grid cells in each interval; c. Calculate the vertical distance: Calculate the current grid cell A(X A ,Y A ) and the vertical distance d between the grid cells corresponding to the quantitative curve equation gi(X) A , that is, d A =g(X A )-Y A ; gi(X) is the quantitative curve equation of the corresponding interval, X A and Y A are the horizontal and vertical coordinates of the grid unit; d. Locate the drilling grid unit with the greatest impact: In each interval, set the vertical intersection C(X c ,Y c ), then combined with d A The drilling grid unit with the greatest impact on the current grid is obtained as A'(X A' ,Y A' ), where X A' =X C , Y A' =Y C -d A .

9. A stratum distribution interpretation device, characterized in that: It includes at least one processor and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the formation distribution interpretation method described in any one of claims 1-8.

10. A computer-readable medium, characterized in that Instructions executable by a processor are stored thereon, and when the instructions are executed by the processor, the processor executes the method for interpreting formation distribution as described in any one of claims 1-8.

Images