A method and system for collaborative interpretation and fusion imaging of multi-source data while drilling

Abstract

This invention relates to the field of geological drilling data processing technology, and discloses a method and system for collaborative interpretation and fusion imaging of multi-source data during drilling. It aims to solve the problems of insufficient multi-source data fusion, low computational efficiency, and poor imaging effects in existing technologies. The solution mainly includes: establishing a three-dimensional collaborative detection method system for drilling; classifying detection parameters by attribute; acquiring raw data of each detection parameter through multiple detection methods; obtaining consistent multi-source inversion data through progressive collaborative inversion and unified correction for reliable raw data; achieving deep integration and geological interpretation of multi-source data based on statistical fusion and spatial extrapolation of azimuth overlap areas, obtaining interpretation results reflecting different geological attributes; conducting risk assessment based on the interpretation results, and fusing multi-attribute and risk information to generate three-dimensional images. This invention significantly improves the accuracy, interpretation efficiency, and visualization level of geological exploration.

Description

Technical Field

[0001] This invention relates to the field of geological drilling data processing technology, specifically to a method and system for collaborative interpretation and fusion imaging of multi-source data during drilling. Background Technology

[0002] With the increasing complexity of deep resource development and underground engineering, projects face increasingly severe challenges from geological hazards such as water inrush, borehole collapse, and rock bursts. Detection while drilling (DWD) technology aims to dynamically identify formation lithology, structure, and potential hazards by acquiring multiple parameter data in real time during the drilling process, providing crucial decision support for safe engineering construction. Traditional DWD technologies typically rely on single sensor data (such as electromagnetic waves or gamma rays), and their imaging models have limited ability to characterize complex geological structures, making it difficult to meet the needs of high-resolution geological modeling.

[0003] To improve detection accuracy, collaborative processing and fusion imaging technology of multi-source data has become an important development direction. However, existing technologies still have significant limitations in practical applications: First, existing fusion methods mainly focus on engineering and environmental parameters such as drilling pressure, torque, and temperature, while underutilizing borehole geophysical data from methods such as polarimetric radar, resistivity, and acoustic measurements, resulting in weak comprehensive interpretation capabilities of the geological conditions surrounding the borehole; second, existing collaborative interpretation methods mostly adopt a semi-independent process of independent inversion of each detection data and joint interpretation of the results, which is cumbersome and computationally inefficient; finally, existing imaging results are mostly two-dimensional parametric curves or three-dimensional displays of a single method, failing to achieve multi-attribute, full-element three-dimensional fusion modeling and visualization of the real geological conditions within the borehole measurement range, resulting in poor intuitiveness and decision support.

[0004] Therefore, how to achieve efficient and accurate collaborative interpretation and deep fusion of multi-source heterogeneous data such as engineering parameters and geophysical parameters in drilling exploration, and generate intuitive and comprehensive three-dimensional geological images, is a key technical problem that urgently needs to be solved in this field. Summary of the Invention

[0005] This invention aims to address the problems of insufficient multi-source data fusion, low computational efficiency, and poor imaging effects in existing technologies, and proposes a method and system for collaborative interpretation and fusion imaging of multi-source data during drilling.

[0006] The technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows:

[0007] In a first aspect, the present invention provides a method for collaborative interpretation and fusion imaging of multi-source data while drilling, the method comprising:

[0008] Step 1: Establish a three-dimensional collaborative detection method system for drilling exploration, and determine various detection methods and their corresponding detection parameters;

[0009] Step 2: Classify the detection parameters corresponding to each detection method into three attributes: sensitivity attribute, location attribute, and type attribute. The sensitivity attribute includes correction parameters, indirect parameters, and direct parameters. The location attribute includes borehole wall detection parameters and borehole perimeter detection parameters.

[0010] Step 3: Obtain the raw data of each detection parameter through the various detection methods, evaluate the reliability of the raw data based on Mahalanobis distance, and retain the reliable raw data;

[0011] Step 4: For the retained original data, perform collaborative inversion and correction based on the aforementioned sensitivity and location attributes, specifically including:

[0012] The corrected parameters are inverted and interpreted to obtain the corrected parameter interpretation data;

[0013] Invert the direct parameters to obtain direct parameter inversion data;

[0014] Using the direct parameter inversion data as a priori conditions, the borehole wall detection parameters in the indirect parameters are jointly inverted to obtain the borehole wall indirect parameter inversion data;

[0015] Using the direct parameter inversion data and the borehole wall indirect parameter inversion data, the borehole perimeter detection parameters in the indirect parameters are jointly inverted to obtain borehole perimeter indirect parameter inversion data containing multiple physical properties formed by multiple detection methods;

[0016] The direct parameter inversion data, borehole wall indirect parameter inversion data, and borehole perimeter indirect parameter inversion data are used to perform orientation correction and environmental factor correction on the data interpreted by the correction parameters.

[0017] Step 5: Based on the corrected inversion data, for inversion data with the same type attributes, perform fusion and interpretation based on their azimuth overlap area; for inversion data with different type attributes, perform interpretation separately; and obtain interpretation results reflecting different geological attributes.

[0018] Step 6: Conduct a risk assessment based on the interpretation results, and perform multi-data-attribute 3D imaging based on the interpretation results and risk assessment results, combined with spatial location.

[0019] Furthermore, the detection methods include a variety of methods such as drilling parameter determination, core testing, in-hole television, polarimetric radar, resistivity measurement, acoustic measurement, and gamma ray method.

[0020] Furthermore, the correction parameters include temperature, water pressure, location, and angle; the direct parameters are parameters obtained through core testing and borehole television methods, used to directly distinguish formation lithology and fracture distribution; the indirect parameters include drilling speed, drilling pressure, torque, wave velocity, and resistivity; the inversion data of the borehole perimeter indirect parameters includes various physical properties such as wave velocity, resistivity, and gamma ray intensity.

[0021] Furthermore, the type attributes include lithological detection parameters, integrity detection parameters, and water content detection parameters.

[0022] Furthermore, the orientation correction includes: establishing a unified three-dimensional coordinate system by combining inertial navigation and geomagnetic data, and performing a unified orientation description on various types of inversion data based on the three-dimensional coordinate system;

[0023] The environmental factor correction includes temperature correction and pressure correction. The temperature correction includes temperature correction for resistivity, and the formula for temperature correction of resistivity is as follows:

[0024] Corrected resistivity = nominal resistivity × [1 + temperature coefficient × (actual temperature - reference temperature)].

[0025] Furthermore, for inversion data with the same type attributes, fusion and interpretation are performed based on their overlapping azimuth regions, specifically including:

[0026] The inversion data with the same type attributes corresponding to each detection method are standardized, and their respective statistical characteristic parameters are calculated.

[0027] Data fusion is performed on all inversion data within the overlapping area to obtain fused data for that area, and statistical characteristic parameters of the fused data are calculated.

[0028] Determine the trend of change of statistical characteristic parameters of the fused data relative to the data before fusion;

[0029] Based on the aforementioned trend, the inversion data of each detection method within its own detection azimuth but outside the overlapping area of ​​the azimuth are adjusted and extrapolated to obtain data for the non-overlapping area.

[0030] The fused data of the overlapping azimuth regions is combined with the non-overlapping data of each detection method after adjustment and estimation to form complete fused data, and then re-standardized.

[0031] Based on the re-standardized and fully fused data, and combined with the geological significance corresponding to the type attributes, data interpretation is performed to obtain interpretation results reflecting the corresponding geological attributes.

[0032] Furthermore, data fusion is performed on all inversion data within the azimuth overlap region, specifically including:

[0033] The Isomap++ algorithm is used to project the inversion data of the azimuth overlap region onto a low-dimensional manifold space, and the dimensionality-reduced fused data is obtained by compression through an SAE network.

[0034] Furthermore, the statistical characteristic parameters are the normal distribution mean and standard deviation; the trend of change is the difference between the normal distribution mean before and after fusion, and the ratio of the squares of the standard deviations.

[0035] Furthermore, for any detection method, the statistical characteristic parameters of the corresponding non-overlapping region data satisfy the following relationship:

[0036] The difference between the overall mean of the inversion data corresponding to this detection method and the mean of the data in the non-overlapping area is equal to the difference between the mean of the data in the overlapping area before fusion and the mean of the data in the azimuth overlapping area after fusion.

[0037] The ratio of the square of the overall standard deviation of the inversion data corresponding to this detection method to the square of the standard deviation of its non-overlapping region data is equal to the ratio of the square of the standard deviation of its azimuthally overlapping region data before fusion to the square of the standard deviation of the azimuthally overlapping region data after fusion.

[0038] In a second aspect, the present invention provides a drilling multi-source data collaborative interpretation and fusion imaging system for implementing the drilling multi-source data collaborative interpretation and fusion imaging method as described in the first aspect, the system comprising:

[0039] The detection system configuration module is used to establish a three-dimensional collaborative detection method system for drilling-while-drilling detection, and to determine various detection methods and their corresponding detection parameters;

[0040] The parameter attribute classification module is used to classify the detection parameters corresponding to each detection method into three attributes: sensitivity attribute, location attribute, and type attribute. The sensitivity attribute includes correction parameters, indirect parameters, and direct parameters, and the location attribute includes borehole wall detection parameters and borehole perimeter detection parameters.

[0041] The data preprocessing module is used to obtain the raw data of each detection parameter through the various detection methods, evaluate the reliability of the raw data based on Mahalanobis distance, and retain the reliable raw data.

[0042] The collaborative inversion and correction module is used to perform collaborative inversion and correction on the retained original data based on the sensitivity attribute and location attribute, specifically including:

[0043] The corrected parameters are inverted and interpreted to obtain the corrected parameter interpretation data;

[0044] Invert the direct parameters to obtain direct parameter inversion data;

[0045] Using the direct parameter inversion data as a priori conditions, the borehole wall detection parameters in the indirect parameters are jointly inverted to obtain the borehole wall indirect parameter inversion data;

[0046] Using the direct parameter inversion data and the borehole wall indirect parameter inversion data, the borehole perimeter detection parameters in the indirect parameters are jointly inverted to obtain borehole perimeter indirect parameter inversion data containing multiple physical properties formed by multiple detection methods;

[0047] The direct parameter inversion data, borehole wall indirect parameter inversion data, and borehole perimeter indirect parameter inversion data are used to perform orientation correction and environmental factor correction on the data interpreted by the correction parameters.

[0048] The data fusion and interpretation module is used to fuse and interpret inversion data with the same type and attributes based on their azimuth overlap area, and to interpret inversion data with different type and attributes separately, so as to obtain interpretation results reflecting different geological attributes.

[0049] The risk assessment and 3D imaging module is used to perform risk assessment based on the interpretation results, and to perform multi-data-attribute 3D imaging based on the interpretation results and risk assessment results, combined with spatial location.

[0050] The beneficial effects of this invention are as follows: The drilling multi-source data collaborative interpretation and fusion imaging method and system provided by this invention achieves full-space, multi-physical attribute data coverage of the borehole wall and surrounding area by constructing a three-dimensional collaborative detection system containing multiple detection methods; using direct parameter inversion data as prior constraints, the indirect parameters of the borehole wall are jointly inverted, and then the borehole wall inversion results are used to constrain the joint inversion of multi-source geophysical data around the borehole, forming a collaborative inversion mechanism with layered constraints and accuracy transfer; subsequently, correction parameters are used to uniformly correct the orientation and environment of all inversion results to eliminate systematic errors. In the fusion interpretation stage, for similar geological attributes, the statistical characteristics of overlapping azimuth areas are fused and trends are extrapolated to achieve consistent integration of multi-source data in spatial and statistical sense; for different geological attributes, multi-dimensional interpretation is performed. Finally, the interpretation results are coupled with risk levels, and a visualization model integrating multiple geological attributes and risk information is generated through three-dimensional imaging technology. This invention effectively overcomes the shortcomings of existing technologies, such as single data source, low degree of integration, cumbersome processing procedures, and unintuitive imaging. It significantly improves the accuracy, efficiency, and comprehensive visualization capabilities of drilling exploration, providing more reliable and intuitive technical support for engineering safety decision-making. Attached Figure Description

[0051] Figure 1 A flowchart illustrating a method for collaborative interpretation and fusion imaging of multi-source data during drilling, provided as an example;

[0052] Figure 2 A schematic flowchart of another method for collaborative interpretation and fusion imaging of multi-source data while drilling, provided as an example;

[0053] Figure 3 A schematic diagram illustrating the attribute classification of detection parameters provided for an embodiment;

[0054] Figure 4 This is a schematic diagram of the structure of the drilling multi-source data collaborative interpretation and fusion imaging system provided in the embodiment. Detailed Implementation

[0055] To address the shortcomings of existing technologies, such as single data sources, low fusion levels, cumbersome processing procedures, and unintuitive imaging, this invention proposes a technical solution. First, a three-dimensional detection system encompassing both borehole wall and periphery information, taking into account both direct and indirect information, is established. Multi-source detection parameters are then classified into three attribute categories based on sensitivity, spatial location, and geological type, constructing a logical framework for data processing. Next, using direct parameters as prior constraints, a progressive joint inversion is implemented from direct parameters to borehole wall parameters and then to periphery parameters. Unified corrections are made through parameter adjustments to ensure consistency in physical meaning and spatial location of multi-source data. Furthermore, for data with similar attributes, statistical characteristics of overlapping azimuth regions are used for fusion and spatial extrapolation, achieving deep integration and geological interpretation of multi-source information. Finally, the interpretation results are coupled with risk levels to achieve full-space three-dimensional visualization imaging.

[0056] The technical solutions in this embodiment will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0057] Please see Figure 1 and Figure 2 The drilling multi-source data collaborative interpretation and fusion imaging method provided in this embodiment includes the following steps:

[0058] Step 1: Establish a three-dimensional collaborative detection method system for drilling exploration, and determine various detection methods and their corresponding detection parameters.

[0059] In this embodiment, the detection methods include multiple methods such as drilling parameter measurement, core testing, borehole television, polarimetric radar, resistivity measurement, acoustic measurement, and gamma-ray method. Detection parameters refer to the data obtained through each detection method.

[0060] This step systematically integrates various detection methods, including drilling parameter measurement, core testing, borehole television, polarimetric radar, resistivity measurement, acoustic measurement, and gamma-ray diffraction, to construct a three-dimensional collaborative detection system that simultaneously covers the borehole wall and periphery and integrates engineering and geophysical parameters. It also clarifies the specific detection parameters (such as drilling speed, resistivity, and wave velocity) for each method, thus laying the foundation for obtaining full-dimensional raw data characterizing different spatial locations, reflecting different physical properties, and exhibiting varying degrees of sensitivity. This ensures the diversity and complementarity of data sources, providing a necessary and sufficient data foundation for subsequent collaborative inversion and deep fusion of multi-source data.

[0061] Step 2: Classify the detection parameters corresponding to each detection method into three attributes: sensitivity attribute, location attribute, and type attribute. The sensitivity attribute includes correction parameters, indirect parameters, and direct parameters. The location attribute includes borehole wall detection parameters and borehole perimeter detection parameters.

[0062] In this embodiment, the correction parameters include temperature, water pressure, location, and angle; the direct parameters are those obtained through core testing and borehole television methods, used to directly distinguish formation lithology and fracture distribution; the indirect parameters include drilling speed, drilling pressure, torque, wave velocity, and resistivity. The type attributes include lithology detection parameters, integrity detection parameters, water content detection parameters, and detection parameters for other type attributes.

[0063] This step establishes a unified framework for the structured processing of multi-source data by systematically classifying all detection parameters into a triple attribute system of sensitivity, location, and type.

[0064] Please see Figure 3 This embodiment categorizes sensitivity attributes into modified parameters, indirect parameters, and direct parameters based on the directness of the parameters in reflecting geological information; it categorizes location attributes into borehole wall detection parameters and borehole perimeter detection parameters based on the spatial location of the parameters; and it categorizes type attributes into lithology, integrity, and water content based on the geological indicative significance of the parameters. This provides precise operational guidelines for subsequent data selection, process control (such as inversion order and correction basis), and fusion interpretation (such as fusion by type and location).

[0065] In practical applications, drilling parameter measurement methods can obtain parameters such as drilling speed, drilling pressure, torque, temperature, and water pressure during the drilling process. Among these, environmental parameters such as temperature and water pressure, and azimuth parameters such as position and angle, can be used to correct the accuracy of other parameters; these are correction parameters. Engineering parameters such as drilling speed, drilling pressure, and torque can indirectly reflect geological information; these are indirect parameters. Core testing and borehole television methods can directly distinguish geological information such as lithology and fracture distribution; the parameters obtained are direct parameters. Polarimetric radar, resistivity methods, acoustic wave measurement, and gamma-ray methods can obtain physical properties such as wave velocity and resistivity in the area surrounding the borehole, indirectly reflecting geological information; these are also indirect parameters. The detection parameters or direct parameters obtained by drilling parameter measurement methods, core testing methods, and borehole television methods only reflect the geological information of the borehole wall; these are borehole wall detection parameters. Detection parameters obtained by geophysical methods reflect the geological information around the borehole; these are borehole perimeter detection parameters. The classification of detection parameters by type and attribute can refer to regulations, standards, and the experience of professionals.

[0066] Step 3: Obtain the raw data of each detection parameter through the various detection methods, evaluate the reliability of the raw data based on Mahalanobis distance, and retain the reliable raw data;

[0067] This step, after simultaneously acquiring raw data for each detection parameter using multiple detection methods, evaluates the reliability of each data point relative to the overall data distribution based on Mahalanobis distance, a multivariate statistical measure. This identifies and eliminates outliers and unreliable data that may be caused by sensor malfunctions or environmental interference, ultimately retaining only the reliable raw data that passes the verification. This ensures the accuracy and consistency of the input data from the source, effectively preventing error propagation and accumulation, and improving the robustness of the entire technical process and the reliability of the final result.

[0068] Step 4: For the retained original data, perform collaborative inversion and correction based on the aforementioned sensitivity and location attributes, specifically including:

[0069] The corrected parameters are inverted and interpreted to obtain the corrected parameter interpretation data;

[0070] Invert the direct parameters to obtain direct parameter inversion data;

[0071] Using the direct parameter inversion data as a priori conditions, the borehole wall detection parameters in the indirect parameters are jointly inverted to obtain the borehole wall indirect parameter inversion data;

[0072] Using the direct parameter inversion data and the borehole wall indirect parameter inversion data, the borehole perimeter detection parameters in the indirect parameters are jointly inverted to obtain borehole perimeter indirect parameter inversion data containing multiple physical properties formed by multiple detection methods; among which, the multiple physical properties include wave velocity, resistivity and gamma ray intensity;

[0073] The direct parameter inversion data, borehole wall indirect parameter inversion data, and borehole perimeter indirect parameter inversion data are subjected to orientation correction and environmental factor correction using the data interpreted by the correction parameters.

[0074] This step establishes a hierarchical processing chain based on the inherent relationships of data attributes to ensure that multi-source data achieve information complementarity and accuracy improvement during the inversion process, and ultimately completes system correction under a unified physical and spatial benchmark.

[0075] In practical applications, firstly, correction parameters such as temperature and water pressure are inverted and interpreted to obtain interpretation data of correction parameters used to describe environmental conditions and instrument attitude. Secondly, direct parameters such as core samples and borehole television data are inverted to obtain direct parameter inversion data that can directly indicate lithology and fractures. This serves as highly reliable geological prior knowledge. Direct parameter inversion data provides a highly deterministic and high-resolution quantitative or qualitative description of geological properties at specific points on the borehole wall. For example, if a fracture is identified and its width (e.g., 2.1 mm) is measured through borehole television images, the output "fracture aperture value is 2.1 mm" after inversion is the direct parameter inversion data. Similarly, if the uniaxial compressive strength of rock is measured through core experiments (e.g., 85 MPa), the inversion yields "the uniaxial compressive strength of the rock mass is 85 MPa".

[0076] Next, using direct parameter inversion data as prior conditions, indirect parameters reflecting borehole wall conditions, such as drilling rate and torque, are jointly inverted to generate borehole wall indirect parameter inversion data, thus achieving the fusion of engineering parameters and direct geological information. Borehole wall indirect parameter inversion data is data used to describe the geological or mechanical state of the borehole wall after establishing a physical correlation model between engineering responses (such as drilling rate changes) and known geological properties (such as lithological strength). Its accuracy and reliability depend on the prior constraints provided by the direct parameter inversion data. For example, given the prior condition that a certain section of the borehole wall is "hard granite" (from direct parameter inversion data), combined with measured drilling parameters (indirect parameters) such as decreased drilling rate and increased torque in that section, joint inversion using a rock-drill bit interaction model can calculate borehole wall indirect parameter inversion data such as the "rock mass drillability grade" or "equivalent shear strength" of that section of the borehole wall.

[0077] Then, by comprehensively utilizing the dual constraints formed by the aforementioned direct parameter inversion data and borehole wall indirect parameter inversion data, the borehole perimeter detection parameters from various detection methods such as polarimetric radar and resistivity methods are jointly inverted, outputting borehole perimeter indirect parameter inversion data that integrates multiple physical properties such as wave velocity, resistivity, and gamma ray intensity. This borehole perimeter indirect parameter inversion data is a geophysical inversion process that extrapolates highly deterministic geological information at points on the borehole wall to the surrounding geophysical volume, thereby significantly improving the accuracy and geological interpretability of imaging complex geological bodies around the borehole (such as fracture zones and aquifers). The result is a three-dimensional data volume integrating multiple physical properties. For example, under the constraint that a certain depth on the borehole wall is a "fractured aquifer fracture zone" (from both direct parameter inversion data and borehole wall indirect parameter inversion data), a joint inversion is performed by combining the P-wave velocity attenuation data obtained by acoustic measurement at that depth, the low-resistivity anomaly data obtained by resistivity method, and the specific response obtained by gamma ray method. The final generated three-dimensional data volume, which shows the spatially overlapping distribution of the "wave velocity reduction zone", "low resistivity zone" and "gamma ray anomaly zone" within a certain range (e.g., a radius of 5 meters) around the borehole, is the borehole perimeter indirect parameter inversion data.

[0078] Finally, using the corrected parameters obtained in the first step to interpret the data, a unified orientation correction and environmental factor correction are performed on all inversion data. The orientation correction includes: establishing a unified three-dimensional coordinate system by combining inertial navigation and geomagnetic data, and performing a unified orientation description on various types of inversion data according to the three-dimensional coordinate system; the environmental factor correction includes temperature correction and pressure correction, and the temperature correction includes temperature correction of resistivity. The temperature correction formula for resistivity is: corrected resistivity = nominal resistivity × [1 + temperature coefficient × (actual temperature - reference temperature)].

[0079] This embodiment significantly improves the accuracy and geological reliability of the inversion results of multi-source geophysical data around the borehole through a progressive constraint process of environmental interpretation, direct geological inversion, borehole wall joint inversion, and borehole perimeter joint inversion. Finally, the system correction eliminates the inconsistencies of various data in spatial and physical benchmarks, providing an accurate, consistent, and directly comparable data foundation for subsequent fusion and imaging.

[0080] Step 5: Based on the corrected inversion data, for inversion data with the same type attributes, perform fusion and interpretation based on their azimuth overlap area; for inversion data with different type attributes, perform interpretation separately; and obtain interpretation results reflecting different geological attributes.

[0081] In this embodiment, for inversion data with the same type attributes, fusion and interpretation are performed based on their overlapping azimuth regions, specifically including:

[0082] Step 501: Standardize the inversion data with the same type attributes corresponding to each detection method, and calculate their respective statistical characteristic parameters; where the statistical characteristic parameters are the mean and standard deviation of the normal distribution.

[0083] Step 502: Perform data fusion on all inversion data within the azimuth overlap region to obtain fused data for the region, and calculate the statistical characteristic parameters of the fused data; wherein, the Isomap++ algorithm is used to project the inversion data of the azimuth overlap region onto a low-dimensional manifold space, and the dimensionality-reduced fused data is obtained by compression through an SAE network.

[0084] Step 503: Determine the trend of change of the statistical characteristic parameters of the fused data relative to the data before fusion; wherein the trend is the difference between the mean of the normal distribution before and after fusion, and the ratio of the square of the standard deviation.

[0085] Step 504: Based on the changing trend, adjust and extrapolate the inversion data of each detection method within its own detection azimuth but outside the overlapping area of ​​the azimuth, to obtain the non-overlapping area data.

[0086] Step 505: Merge the fused data of the azimuth overlapping area with the non-overlapping area data calculated by each detection method after adjustment to form complete fused data, and then re-standardize it.

[0087] Step 506: Based on the re-standardized complete fused data, and combined with the geological significance corresponding to the type attribute, perform data interpretation to obtain interpretation results reflecting the corresponding geological attributes.

[0088] In this embodiment, for any detection method, the statistical characteristic parameters of the corresponding non-overlapping region data satisfy the following relationship:

[0089] The difference between the overall mean of the inversion data corresponding to this detection method and the mean of the data in the non-overlapping area is equal to the difference between the mean of the data in the overlapping area before fusion and the mean of the data in the azimuth overlapping area after fusion.

[0090] The ratio of the square of the overall standard deviation of the inversion data corresponding to this detection method to the square of the standard deviation of its non-overlapping region data is equal to the ratio of the square of the standard deviation of its azimuthally overlapping region data before fusion to the square of the standard deviation of the azimuthally overlapping region data after fusion.

[0091] This step aims to transform various types of inversion data (including direct parameter inversion data, borehole wall indirect parameter inversion data, and borehole perimeter indirect parameter inversion data) that have a unified benchmark but different sources and spatial coverage into interpretation results with clear geological significance based on their geological type attributes.

[0092] In practical applications, the process begins with spatial fusion of multi-source inversion data with similar geological attributes based on statistical characteristics. Then, data with different attributes are interpreted comprehensively from multiple dimensions. Specifically, for inversion data with the same type of attributes, standardization is first used to eliminate dimensional differences. Next, within shared azimuthal overlap areas, the Isomap++ algorithm and SAE network are used for nonlinear dimensionality reduction fusion to generate fused data for that region and extract its statistical characteristics. Then, by comparing the statistical characteristics (mean and standard deviation) of the data before and after fusion, a quantifiable trend is determined. This trend is then used as a transfer function to perform consistency adjustments and extrapolations on data from each method in their unique non-overlapping regions. This extrapolation process is strictly constrained by mathematical relationships to ensure the overall statistical distribution of the data before and after adjustment is coordinated. Finally, the fused data from overlapping areas and the extrapolated data from non-overlapping areas are merged and re-standardized to form spatially continuous and statistically unified complete fused data, which is then interpreted in conjunction with geological knowledge. For data with different types of attributes, interpretation is carried out in parallel.

[0093] The following example illustrates the specific steps for adjusting the extrapolated data for non-overlapping regions:

[0094] First, standardize the inversion data obtained by method A, denoted as data set X, representing a spatial region M. Standardize the inversion data obtained by method B, denoted as data set Y, representing a spatial region N. Let P = M ∩ N, meaning region P is the azimuthal overlap region of region M and N. Data in the X set located within region P are denoted as Xp, and data in the Y set located within region P are denoted as Yp. Calculate the normal distribution mean of X. and standard deviation The mean of Y is a normal distribution. and standard deviation .

[0095] Then, the Xp and Yp data are projected onto a low-dimensional manifold space using the Isomap++ algorithm, and the dimensionality-reduced and fused data Z is obtained by compression using an SAE network. The normal distribution mean of Xp is calculated. and standard deviation The normal distribution mean of Yp and standard deviation The mean of the normal distribution of Z and standard deviation .make and These are the mean and standard deviation of the normal distribution after dimensionality reduction and fusion of X, respectively. and These represent the mean and standard deviation of the normal distribution after dimensionality reduction and fusion of Y, respectively. The trend is as follows: , , , Please refer to Table 1 for the obtained statistical characteristic parameters.

[0096] Table 1. Statistical characteristic parameters before and after inversion data fusion

[0097]

[0098] Finally, using , Calculate the non-overlapping area data of X, i.e., X-Xp, using... , Calculate the non-overlapping area data of Y, i.e., Y-Yp.

[0099] Through the above steps, this embodiment not only overcomes the contradictions and inconsistencies caused by the simple superposition of multi-source data in traditional methods, but also intelligently extends the limited, highly reliable information of overlapping areas to the entire detection space through the mechanism of deep fusion of overlapping areas and statistical trend extrapolation. This generates highly reliable interpretation results that are spatially continuous and directly correspond to specific geological attributes such as lithology, integrity, and water content, providing accurate semantic input for the final risk assessment and three-dimensional imaging.

[0100] Step 6: Conduct a risk assessment based on the interpretation results, and perform multi-data-attribute 3D imaging based on the interpretation results and risk assessment results, combined with spatial location.

[0101] This step transforms the interpretation results reflecting different geological attributes obtained in the previous steps into an intuitively perceptible engineering risk level, and finally couples all geological attribute information and risk information with their actual spatial location to construct a unified three-dimensional visualization model.

[0102] In practical applications, firstly, based on preset evaluation rules and thresholds, the various interpretation results are comprehensively identified and classified into different risk assessment results such as normal, level three risk, level two risk, and level one risk, thus realizing a quantitative mapping from geological description to risk level. Then, the interpretation results of various geological attributes classified into risk levels are combined with their corresponding three-dimensional spatial coordinates, and surface rendering algorithms such as Marching Cubes are used to generate a three-dimensional model that integrates multiple geological attributes such as lithology, integrity, and water content, as well as information on different risk levels.

[0103] Through the above steps, this embodiment changes the traditional complex mode of comprehensive judgment that relies on separate curves, two-dimensional profiles or single attribute models. By generating a three-dimensional fusion image that integrates geological attributes, risk level and spatial location, it presents massive and abstract multi-source data and interpretation results in an intuitive and comprehensive way that decision-makers can directly understand as a geological structure and risk spatial distribution map, which greatly improves the intuitiveness, comprehensiveness and efficiency and accuracy of disaster early warning and decision support.

[0104] In summary, the multi-source data collaborative interpretation and fusion imaging method provided in this embodiment achieves unified characterization and quality control of multi-source data from drilling and geophysical exploration by constructing a three-dimensional collaborative detection system and a three-attribute classification framework. Furthermore, through progressively constrained collaborative inversion and system correction, it significantly improves the accuracy and consistency of borehole geological inversion. Based on this, a spatial fusion method based on statistical trend extrapolation is used to achieve seamless fusion and geological interpretation of multi-source heterogeneous data in three-dimensional space. Finally, through attribute and risk coupling in three-dimensional modeling, an integrated visualization result integrating multiple geological attributes and risk levels is generated. This fundamentally solves the problems of isolated data, superficial fusion, one-sided interpretation, and abstract imaging in traditional drilling techniques, achieving a leap from raw data to comprehensive decision-making knowledge. While improving detection accuracy and efficiency, it greatly enhances the intuitiveness and reliability of geological disaster early warning, providing efficient, intuitive, and reliable technical support for safe engineering construction under complex geological conditions.

[0105] Based on the above technical solutions, this embodiment also proposes a multi-source data collaborative interpretation and fusion imaging system for drilling, used to implement the multi-source data collaborative interpretation and fusion imaging method for drilling described in the embodiment. Please refer to [link to relevant documentation]. Figure 4 The system includes:

[0106] The detection system configuration module is used to establish a three-dimensional collaborative detection method system for drilling-while-drilling detection, and to determine various detection methods and their corresponding detection parameters;

[0107] The parameter attribute classification module is used to classify the detection parameters corresponding to each detection method into three attributes: sensitivity attribute, location attribute, and type attribute. The sensitivity attribute includes correction parameters, indirect parameters, and direct parameters, and the location attribute includes borehole wall detection parameters and borehole perimeter detection parameters.

[0108] The data preprocessing module is used to obtain the raw data of each detection parameter through the various detection methods, evaluate the reliability of the raw data based on Mahalanobis distance, and retain the reliable raw data.

[0109] The collaborative inversion and correction module is used to perform collaborative inversion and correction on the retained original data based on the sensitivity attribute and location attribute, specifically including:

[0110] The corrected parameters are inverted and interpreted to obtain the corrected parameter interpretation data;

[0111] Invert the direct parameters to obtain direct parameter inversion data;

[0112] Using the direct parameter inversion data as a priori conditions, the borehole wall detection parameters in the indirect parameters are jointly inverted to obtain the borehole wall indirect parameter inversion data;

[0113] Using the direct parameter inversion data and the borehole wall indirect parameter inversion data, the borehole perimeter detection parameters in the indirect parameters are jointly inverted to obtain borehole perimeter indirect parameter inversion data containing multiple physical properties formed by multiple detection methods;

[0114] The direct parameter inversion data, borehole wall indirect parameter inversion data, and borehole perimeter indirect parameter inversion data are used to perform orientation correction and environmental factor correction on the data interpreted by the correction parameters.

[0115] The data fusion and interpretation module is used to fuse and interpret inversion data with the same type and attributes based on their azimuth overlap area, and to interpret inversion data with different type and attributes separately, so as to obtain interpretation results reflecting different geological attributes.

[0116] The risk assessment and 3D imaging module is used to perform risk assessment based on the interpretation results, and to perform multi-data-attribute 3D imaging based on the interpretation results and risk assessment results, combined with spatial location.

[0117] It is understood that since the multi-source data collaborative interpretation and fusion imaging system described in this embodiment is a system for implementing the multi-source data collaborative interpretation and fusion imaging method described in the embodiment, the system disclosed in the embodiment is relatively simple to describe because it corresponds to the method disclosed in the embodiment. For relevant parts, please refer to the description of the method, and it will not be repeated here.

Claims

1. A method for collaborative interpretation and fusion imaging of multi-source data during drilling, characterized in that, The method includes: Step 1: Establish a three-dimensional collaborative detection method system for drilling exploration, and determine various detection methods and their corresponding detection parameters; Step 2: Classify the detection parameters corresponding to each detection method into three attributes: sensitivity attribute, location attribute, and type attribute. The sensitivity attribute includes correction parameters, indirect parameters, and direct parameters. The location attribute includes borehole wall detection parameters and borehole perimeter detection parameters. Step 3: Obtain the raw data of each detection parameter through the various detection methods, evaluate the reliability of the raw data based on Mahalanobis distance, and retain the reliable raw data; Step 4: For the retained original data, perform collaborative inversion and correction based on the aforementioned sensitivity and location attributes, specifically including: The corrected parameters are inverted and interpreted to obtain the corrected parameter interpretation data; Invert the direct parameters to obtain direct parameter inversion data; Using the direct parameter inversion data as a priori conditions, the borehole wall detection parameters in the indirect parameters are jointly inverted to obtain the borehole wall indirect parameter inversion data; Using the direct parameter inversion data and the borehole wall indirect parameter inversion data, the borehole perimeter detection parameters in the indirect parameters are jointly inverted to obtain borehole perimeter indirect parameter inversion data containing multiple physical properties formed by multiple detection methods; The direct parameter inversion data, borehole wall indirect parameter inversion data, and borehole perimeter indirect parameter inversion data are used to perform orientation correction and environmental factor correction on the data interpreted by the correction parameters. Step 5: Based on the corrected inversion data, for inversion data with the same type attributes, perform fusion and interpretation based on their azimuth overlap area; for inversion data with different type attributes, perform interpretation separately; and obtain interpretation results reflecting different geological attributes. For inversion data with the same type attributes, fusion and interpretation are performed based on their overlapping azimuth regions, specifically including: The inversion data with the same type attributes corresponding to each detection method are standardized, and their respective statistical characteristic parameters are calculated. Data fusion is performed on all inversion data within the overlapping area to obtain fused data for that area, and statistical characteristic parameters of the fused data are calculated. Determine the trend of change of statistical characteristic parameters of the fused data relative to the data before fusion; Based on the aforementioned trend, the inversion data of each detection method within its own detection azimuth but outside the overlapping area of ​​the azimuth are adjusted and extrapolated to obtain data for the non-overlapping area. The fused data of the overlapping azimuth regions is combined with the non-overlapping data of each detection method after adjustment and estimation to form complete fused data, and then re-standardized. Based on the re-standardized complete fusion data, and combined with the geological significance corresponding to the type attributes, data interpretation is performed to obtain interpretation results reflecting the corresponding geological attributes; Step 6: Conduct a risk assessment based on the interpretation results, and perform multi-data-attribute 3D imaging based on the interpretation results and risk assessment results, combined with spatial location.

2. The method for collaborative interpretation and fusion imaging of multi-source data during drilling according to claim 1, characterized in that, The detection methods include a variety of methods such as drilling parameter determination, core testing, in-hole television, polarimetric radar, resistivity, acoustic measurement, and gamma ray method.

3. The method for collaborative interpretation and fusion imaging of multi-source data during drilling according to claim 2, characterized in that, The correction parameters include temperature, water pressure, location, and angle; the direct parameters are those obtained through core testing and borehole television methods, used to directly distinguish formation lithology and fracture distribution; the indirect parameters include drilling speed, drilling pressure, torque, wave velocity, and resistivity; the borehole perimeter indirect parameter inversion data contains various physical properties including wave velocity, resistivity, and gamma ray intensity.

4. The method for collaborative interpretation and fusion imaging of multi-source data during drilling according to claim 1, characterized in that, The type attributes include lithological detection parameters, integrity detection parameters, and water content detection parameters.

5. The method for collaborative interpretation and fusion imaging of multi-source data during drilling according to claim 1, characterized in that, The orientation correction includes: establishing a unified three-dimensional coordinate system by combining inertial navigation and geomagnetic data, and performing a unified orientation description on various types of inversion data based on the three-dimensional coordinate system; The environmental factor correction includes temperature correction and pressure correction. The temperature correction includes temperature correction for resistivity, and the formula for temperature correction of resistivity is as follows: Corrected resistivity = nominal resistivity × [1 + temperature coefficient × (actual temperature - reference temperature)].

6. The method for collaborative interpretation and fusion imaging of multi-source data during drilling according to claim 1, characterized in that, Data fusion is performed on all inversion data within the overlapping area, specifically including: The Isomap++ algorithm is used to project the inversion data of the azimuth overlap region onto a low-dimensional manifold space, and the dimensionality-reduced fused data is obtained by compression through an SAE network.

7. The method for collaborative interpretation and fusion imaging of multi-source data during drilling according to claim 1, characterized in that, The statistical characteristic parameters are the mean and standard deviation of the normal distribution; the trend of change is the difference between the mean of the normal distribution before and after fusion, and the ratio of the squares of the standard deviations.

8. The method for collaborative interpretation and fusion imaging of multi-source data during drilling according to claim 7, characterized in that, For any detection method, the statistical characteristic parameters of the corresponding non-overlapping region data satisfy the following relationship: The difference between the overall mean of the inversion data corresponding to this detection method and the mean of the data in the non-overlapping area is equal to the difference between the mean of the data in the overlapping area before fusion and the mean of the data in the azimuth overlapping area after fusion. The ratio of the square of the overall standard deviation of the inversion data corresponding to this detection method to the square of the standard deviation of its non-overlapping region data is equal to the ratio of the square of the standard deviation of its azimuthally overlapping region data before fusion to the square of the standard deviation of the azimuthally overlapping region data after fusion.

9. A collaborative interpretation and fusion imaging system for multi-source data during drilling, characterized in that, The system is used to implement the collaborative interpretation and fusion imaging method for multi-source data during drilling as described in any one of claims 1 to 8, the system comprising: The detection system configuration module is used to establish a three-dimensional collaborative detection method system for drilling-while-drilling detection, and to determine various detection methods and their corresponding detection parameters; The parameter attribute classification module is used to classify the detection parameters corresponding to each detection method into three attributes: sensitivity attribute, location attribute, and type attribute. The sensitivity attribute includes correction parameters, indirect parameters, and direct parameters, and the location attribute includes borehole wall detection parameters and borehole perimeter detection parameters. The data preprocessing module is used to obtain the raw data of each detection parameter through the various detection methods, evaluate the reliability of the raw data based on Mahalanobis distance, and retain the reliable raw data. The collaborative inversion and correction module is used to perform collaborative inversion and correction on the retained original data based on the sensitivity attribute and location attribute, specifically including: The corrected parameters are inverted and interpreted to obtain the corrected parameter interpretation data; Invert the direct parameters to obtain direct parameter inversion data; Using the direct parameter inversion data as a priori conditions, the borehole wall detection parameters in the indirect parameters are jointly inverted to obtain the borehole wall indirect parameter inversion data; Using the direct parameter inversion data and the borehole wall indirect parameter inversion data, the borehole perimeter detection parameters in the indirect parameters are jointly inverted to obtain borehole perimeter indirect parameter inversion data containing multiple physical properties formed by multiple detection methods; The direct parameter inversion data, borehole wall indirect parameter inversion data, and borehole perimeter indirect parameter inversion data are used to perform orientation correction and environmental factor correction on the data interpreted by the correction parameters. The data fusion and interpretation module is used to fuse and interpret inversion data with the same type and attributes based on their azimuth overlap area, and to interpret inversion data with different type and attributes separately, so as to obtain interpretation results reflecting different geological attributes. The risk assessment and 3D imaging module is used to perform risk assessment based on the interpretation results, and to perform multi-data-attribute 3D imaging based on the interpretation results and risk assessment results, combined with spatial location.

Images