Drilling-based methods and systems for identifying adverse geological conditions

Abstract

This invention relates to the field of engineering geological exploration technology, and particularly to a method and system for identifying adverse geological features based on drilling exploration. The method for identifying adverse geological features based on drilling exploration provided by this invention includes the following steps: acquiring first drilling exploration data generated from historical drilling operations; based on the first drilling exploration data, acquiring the correlation between response change data during drilling and the drilling spatial location, and constructing an identification model based on the correlation, the identification model being used to identify the continuity of underground spatial structures; acquiring second drilling exploration data generated from current drilling operations, and extracting a response change dataset during the current drilling process based on the second drilling exploration data; based on the identification model, predicting the continuity of the current underground spatial structure according to the response change dataset, and locating a target spatial region based on the continuity prediction result, the target spatial region being used to indicate the distribution location of adverse geological bodies.

Description

Technical Field

[0001] This invention relates to the field of engineering geological exploration technology, and in particular to a method and system for identifying adverse geological conditions based on drilling exploration. Background Technology

[0002] As urban rail transit, railway, and water conservancy tunnels develop towards deeper burial, larger cross-sections, and more complex geological conditions, the number of construction projects traversing fault fracture zones, karst development areas, and water-rich strata is increasing. Adverse geological bodies are typically characterized by concealed spatial distribution, complex morphology, large scale differences, and high suddenness, making them a major cause of engineering disasters such as sudden water inrush, collapses, and surrounding rock instability, posing a serious threat to tunnel construction safety and project progress. In current engineering practice, the early identification of adverse geological conditions is usually achieved through methods such as core drilling, electrical resistivity tomography, seismic CT, and tunnel seismic prediction (TSP), and relies primarily on manual analysis and judgment of core samples and geophysical exploration results. Summary of the Invention

[0003] In one embodiment of the present invention, a method for identifying adverse geological conditions based on drilling exploration is provided, comprising the following steps: Obtain the first drilling exploration data formed from historical drilling operations; Based on the first drilling exploration data, the correlation between response change data and drilling spatial location during the drilling process is obtained, and based on the correlation, an identification model is constructed. The identification model is used to identify the continuity of underground spatial structures. Acquire the second drilling exploration data generated by the current drilling operation, and extract the response change dataset during the current drilling process based on the second drilling exploration data; Based on the identification model, the continuity of the current underground spatial structure is predicted according to the response change dataset, and the target spatial region is located based on the continuity prediction result. The target spatial region is used to indicate the distribution location of adverse geological bodies.

[0004] In some embodiments, constructing the identification model based on the association relationship includes the following steps: Based on the response change data and corresponding spatial location in the aforementioned relationship, training sample pairs are constructed, wherein each training sample includes a response change data and its corresponding structural continuity label; Multiple training samples are input into a machine learning model for training to obtain a recognition model. The recognition model is configured to receive response change data on the current drilling path and output the prediction result of the underground structure continuity corresponding to the spatial location.

[0005] In some embodiments, locating the target spatial region based on the continuous prediction results includes the following steps: Based on the spatial distribution characteristics of the continuous prediction results, continuous prediction segments that satisfy the preset change pattern are identified, and the spatial location corresponding to the continuous prediction segments is determined as the target spatial region.

[0006] In some embodiments, the drilling exploration-based adverse geological condition identification exploration method further includes the following steps: Based on the target spatial region, perform operation control operations corresponding to the target spatial region. The operation control operations include one or more of the following: adjusting the drilling path, adjusting control parameters during the drilling process, and triggering an early warning signal.

[0007] In some embodiments, the machine learning model is a graph convolutional neural network, and the process of constructing the recognition model based on the graph convolutional neural network includes the following steps: Using each spatial location on the drilling path as a graph node and the spatial adjacency relationship between adjacent locations as graph edges, the response change data corresponding to each graph node is used as the node attribute input to the model. During training, the model weight parameters are optimized by minimizing the loss function between predicted continuous labels.

[0008] In some embodiments, for a first type of drilling exploration data collected during a single drilling operation, the step of obtaining the correlation between response change data during drilling and the drilling spatial location includes the following steps: This first drilling exploration data is spatially interpolated or resampled according to the drilling path so that each drilling exploration data point corresponds one-to-one with a unique spatial location; Based on the processed first drilling exploration data, at least one variation feature is extracted along the drilling path at different spatial locations; The change features are encoded into response change data feature vectors, and their corresponding spatial locations are used as indices to generate the response change dataset for this drilling operation.

[0009] In some embodiments, response change data at any spatial location includes at least: The first type of change feature is used to characterize the change trend of at least one type of drilling exploration data corresponding to the spatial location on the drilling path; The second type of variation feature is used to characterize the degree of fluctuation of the at least one drilling exploration data within a preset spatial range near this spatial location.

[0010] In some embodiments, response change data at any spatial location includes at least: The third type of variation feature is used to characterize the intensity of variation in the spatial distribution of at least one type of drilling exploration data within this spatial location and a pre-defined spatial range nearby.

[0011] In some embodiments, a third type of variation feature at any spatial location is determined jointly by a first type of variation feature corresponding to a drilling exploration data at that spatial location and a second type of variation feature corresponding to the same drilling exploration data at that spatial location.

[0012] In some embodiments, based on the drilling exploration-based adverse geological identification exploration method proposed in the above embodiments, an drilling exploration-based adverse geological identification exploration system is also provided, which includes a processor and a memory. The memory stores a computer program, and when the computer program is read by the processor, it executes the drilling exploration-based adverse geological identification exploration method proposed in any of the above embodiments.

[0013] The adversity geological identification exploration method and system based on drilling exploration provided by this invention has gains including at least: This invention establishes a correlation between changes in the response of drilling exploration data and spatial location, constructing an identification model for recognizing the continuity of underground structures, thereby achieving structural continuity learning and modeling based on historical data. Compared to traditional methods that rely on geological experience or single-parameter judgments, this invention extracts and learns multiple change patterns from historical drilling exploration data, helping to improve the ability to perceive areas of abnormal underground structures.

[0014] Furthermore, this invention introduces a graph convolutional neural network as the implementation method of the recognition model. It utilizes the graph structure's ability to model the spatial adjacency relationship of the drilling path and uses the response change data as the graph node attribute input to enhance the model's sensitivity to local continuous changes.

[0015] Furthermore, the present invention also provides an exploration system for identifying adverse geological conditions based on drilling exploration. By embedding the above method into a processor in the form of a program, an integrated execution process of exploration data processing, model reasoning and spatial target area positioning is realized. In this way, drilling data can be dynamically loaded and structural continuity prediction results can be output in real time during actual drilling operations, thereby having stronger engineering deployment capabilities and field application value. Attached Figure Description

[0016] From the following description of embodiments in conjunction with the accompanying drawings, aspects, features, and advantages of the present invention will become clearer and more readily understood, in which: Figure 1 A flowchart illustrating the adverse geological condition identification exploration method based on drilling exploration provided by this invention; Figure 2A flowchart illustrating the steps for obtaining the correlation between response change data and drilling spatial position during drilling, as provided in an embodiment of the present invention; Figure 3 A flowchart illustrating the construction of a recognition model provided in one embodiment of the present invention; Figure 4 This is a schematic diagram of the system structure provided by the present invention for performing the drilling-based adverse geological identification exploration method. Detailed Implementation

[0017] To facilitate understanding of the present invention by those skilled in the art, several embodiments are now described in detail with reference to the accompanying drawings. It should be understood that the embodiments are for illustrative purposes only and not for limiting the scope of protection of the present invention; the scope of protection of the present invention is defined by the claims, and includes equivalent schemes and equivalent transformations of the claims.

[0018] As mentioned above, existing geological exploration mainly relies on methods such as drilling, electrical resistivity tomography, and seismic CT to collect data, and then relies on manual analysis for judgment. Because the exploration data is spatially discrete and unstructured, it is difficult to form a reliable basis for identifying underground spatial structures, resulting in limited effectiveness in identifying highly concealed and complex-shaped adverse geological bodies. This invention provides a method and system for identifying adverse geological formations based on drilling exploration. By constructing a correlation model between changes in drilling parameter response and the continuity of underground structures, and combining the model's prediction results, the target area of ​​adverse geological bodies is automatically located, achieving the identification and zoning judgment of abnormal underground structures.

[0019] See Figure 1 , Figure 1 This is a flowchart illustrating the drilling-based adverse geological condition identification exploration method provided by the present invention. The drilling-based adverse geological condition identification exploration method provided by the present invention includes at least the following: Figure 1 The steps shown are as follows: S01. Obtain the first drilling exploration data formed from historical drilling operations.

[0020] The first drilling exploration data mentioned in this invention refers to various geological information data collected during previous drilling operations and their accompanying stages by setting up boreholes and supporting measurement and acquisition equipment during the construction of underground engineering projects.

[0021] It should be noted that the first drilling exploration data and the second drilling exploration data are historical drilling operation data and current drilling operation data, respectively. Specifically, the first drilling exploration data is data collected during historical drilling operations, which has known structure labels and is used to train the recognition model; the second drilling exploration data is data collected in real time during the current drilling process, which needs to rely on the recognition model for structure prediction and anomaly identification.

[0022] Furthermore, the drilling exploration data described in this invention includes at least drilling-while-drilling parameter data, borehole spatial location information, and geophysical exploration data.

[0023] The drilling parameter data refers to dynamic parameter data used to characterize the drilling conditions and geological resistance characteristics during the drilling operation, including but not limited to drilling torque, drilling pressure, feed rate, rotational speed, pump pressure, and depth of penetration.

[0024] The borehole spatial location information is used to characterize the three-dimensional position and trajectory of the borehole in the underground space. It is usually obtained through GNSS (Global Navigation Satellite System), total station, inertial navigation and other means, including the borehole starting point coordinates, depth, dip angle and azimuth angle.

[0025] The geophysical exploration data refers to data collected by geophysical equipment that reflects the physical characteristics of the underground medium, including seismic wave CT data (such as P-wave / S-wave velocity, reflection coefficient, and energy anomaly zones), microseismic data (microseismic activity frequency, waveform characteristics, etc.), resistivity profiles, and hydrological parameters (water output, seepage rate, etc.).

[0026] In some embodiments, the drilling exploration data may also include core sample images or borehole profile images to provide a more intuitive reference for lithological structure and geological features.

[0027] Specifically, the core sample images are used to provide information such as rock mass color changes, lithological interfaces, weathering degree and dissolution traces, while the borehole profile images are used to provide information such as borehole wall fracture orientation, density, filling condition and structural features.

[0028] In some embodiments, the drilling exploration data may also include supplementary information such as underground radar scanning data, borehole camera data, and gas detection data, to enhance the identification model's ability to identify complex geological bodies, especially water-rich layers, cavities, and gas-bearing structures.

[0029] It should be noted that the above data types for drilling exploration are only some typical examples. In actual engineering applications, the data types and quantities collected can be flexibly configured according to the geological complexity of the construction area, the exploration accuracy requirements, and the stage of the project (such as preliminary exploration, detailed exploration, and construction exploration).

[0030] S02. Based on the first drilling exploration data, obtain the correlation between the response change data during drilling and the drilling spatial location, and construct an identification model based on the correlation. The identification model is used to identify the continuity of the underground spatial structure.

[0031] The response change data described in this invention refers to the change characteristics of various drilling exploration data at the spatial location corresponding to the drilling path. It is used to reflect the response differences when the drill bit and the underground medium interact at different spatial locations, so as to characterize the continuity and anomalies of the underground structure.

[0032] It should be noted that the response change data can be generated from any drilling exploration parameter with spatial distribution characteristics, including but not limited to drilling parameters (such as torque, drill pressure, and advance rate), geophysical parameters (such as wave velocity, reflection intensity, and microseismic signal), and hydrological parameters (such as seepage rate and water output).

[0033] Furthermore, for any type of first drilling exploration data collected during any drilling operation, the relationship between obtaining the response change data during drilling and the drilling spatial location as described in step S02 includes, for example: Figure 2 The steps shown are as follows: S0211. Spatial interpolation or resampling of this first drilling exploration data according to the drilling path is performed so that each drilling exploration data point corresponds one-to-one with a unique spatial location.

[0034] Specifically, by acquiring borehole trajectory information (such as starting coordinates, borehole depth, dip angle, and azimuth), the original drilling data is resampled or interpolated at preset spatial intervals (e.g., every 0.5 meters or 1 meter). For example, the drilling torque data collected over time is resampled at equal intervals according to the borehole advance, so that one sample corresponds to every 1 meter.

[0035] S0212. Based on the processed first drilling exploration data, extract at least one variation feature at different spatial locations along the drilling path.

[0036] Furthermore, the variation characteristics include, but are not limited to, the first derivative, second derivative, moving standard deviation, variation range, abrupt change location, or variation density of this spatial sequence data.

[0037] The aforementioned characteristic indicators can be obtained through existing mathematical processing methods such as sliding window, difference analysis, and extreme value detection, which will not be detailed here.

[0038] S0213. Encode the change features into response change data feature vectors, and use their corresponding spatial locations as indices to generate the response change dataset for this drilling operation.

[0039] Specifically, each spatial location (such as depth location) or three-dimensional coordinate position The multiple change features corresponding to a change feature dataset are aggregated to form a change feature dataset; then each change feature dataset is encoded into a corresponding response change data feature vector.

[0040] In some embodiments, the change characteristics at any spatial location include at least a first type of response change characteristics and a second type of change characteristics.

[0041] Furthermore, the first type of variation feature is used to characterize the variation trend of at least one drilling exploration data corresponding to the spatial location on the drilling path. In this embodiment, the first type of variation feature can reveal the direction and rate characteristics of geological parameters changing in the direction of the drilling path, providing a significant signal of structural transition or trend reversal for the identification model, thereby enhancing the model's sensitivity and fitting ability in identifying the continuous changes in strata.

[0042] Specifically, for example, regarding drilling torque parameters, its position The first type of change features are ,in, Indicates the location The first derivative of the torque spatial sequence data. Indicates the location The second derivative of the torque spatial sequence data.

[0043] Furthermore, the second type of variation feature is used to characterize the degree of fluctuation of the at least one drilling exploration data within a preset spatial range near this spatial location. In this embodiment, the second type of variation feature can characterize the disturbance amplitude and stability of drilling parameters in a local area, providing micro-disturbance information in the spatial domain for the identification model, which helps to improve the model's ability to distinguish and predict abnormal areas such as discontinuous structures and hidden weak zones.

[0044] Specifically, the drilling torque parameters mentioned above, at position The second type of change features are It indicates the position. The sliding standard deviation of the torque spatial sequence data.

[0045] In some embodiments, the response change dataset at any spatial location also includes a third type of change feature, which is used to characterize the intensity of change in the spatial distribution of at least one drilling exploration data in this spatial location and its surrounding preset spatial range. This can be achieved by jointly characterizing the rate of change, fluctuation amplitude, or degree of abnormal concentration of drilling parameters in the spatial sequence.

[0046] In these embodiments, by comprehensively considering the magnitude and degree of variation of drilling parameters in the spatial dimension, quantitative evidence can be provided for the identification model regarding anomalous features such as geological structural changes and interface changes, thereby enhancing the model's ability to perceive adverse geological conditions with greater precision.

[0047] Furthermore, a third type of variation feature at any spatial location is jointly determined by a first type of variation feature corresponding to a drilling exploration data at that spatial location and a second type of variation feature corresponding to the same drilling exploration data at that spatial location.

[0048] Specifically, regarding the drilling torque parameters mentioned above, at a certain location... The third type of change characteristics ,Depend on ,in, This represents a spatial sequence function of drilling torque along the drilling path. Indicates the location The sliding standard deviation of the torque spatial sequence data. and These are the weighting coefficients.

[0049] Furthermore, the aforementioned weighting coefficients are used to adjust the relative influence of the rate of change and fluctuation characteristics in the determination of structural abrupt changes. Their values ​​can be dynamically adjusted based on historical drilling data, target layer identification priority, or data standard deviation. For example, when more sensitive to fracture zone identification, a higher weighting coefficient can be used. : =1:2.

[0050] Furthermore, based on the drilling torque parameters at position Response change dataset ,in, Indicates the first Each spatial location can be used to generate a corresponding response change feature vector to support the generation of a training set for the subsequent recognition model.

[0051] In this embodiment, a response change dataset is constructed by extracting variation features characterizing the response characteristics of underground structures from drilling exploration parameters with spatial distribution attributes and establishing a one-to-one mapping relationship between them and the corresponding drilling spatial location information.

[0052] It should be noted that the method for extracting response change data can be flexibly configured according to the specific type of drilling exploration parameters, the processing system capabilities, and the sensitivity of the target geological structure. This invention is not limited to the specific data processing path shown in the above embodiments, but rather encompasses data feature extraction methods that can reflect the spatial variation relationship between drilling response parameters and underground structures.

[0053] Furthermore, in the process of constructing the recognition model based on the aforementioned association relationship, this embodiment uses methods such as... Figure 3 The process shown implements the recognition model construction: S0221. Based on the response change data and corresponding spatial location in the association relationship, construct training sample pairs, wherein each training sample includes a response change data and its corresponding structural continuity label.

[0054] The structural continuity label is used to indicate whether the underground structure at each spatial location in the drilling path is continuous; specifically, this structural continuity label can be obtained based on geological profiles, core images, or post-hoc manual annotation results.

[0055] In some embodiments, the structural continuity labels can be specifically categorized into "continuous regions," "discontinuous regions," or "abrupt interface regions," such as continuity labels. , where 0 represents structural continuity, 1 represents local discontinuity, and 2 represents structural abrupt change.

[0056] S0222. Multiple training samples are input into a machine learning model for training to obtain a recognition model. The recognition model is configured to receive response change data on the current drilling path and output the prediction result of the underground structure continuity corresponding to the spatial location.

[0057] The machine learning model described in this invention can be a supervised learning model that supports the task of structural continuity recognition, and its type is not limited to a specific model structure. In different embodiments, it can include traditional shallow classification models such as support vector machines, decision trees, and random forests; it can also be deep learning models such as multilayer perceptrons, convolutional neural networks, and recurrent neural networks; and it can also be graph structure models such as graph neural networks that are optimized for spatial structure modeling.

[0058] In one embodiment, the machine learning model is a graph convolutional neural network. The graph convolutional neural network uses each spatial position on the drilling path as a graph node and the spatial adjacency relationship between adjacent positions as graph edges. The feature vector of the response change data corresponding to each node is used as the node attribute input to the model. During training, the model weight parameters are optimized by minimizing the loss function between the predicted continuous label and the true label. Finally, a recognition model that automatically outputs the continuous prediction result based on the response change data of each position on the drilling path is realized.

[0059] It should be noted that the machine learning model can be built using historical drilling datasets. After the model is trained, it can be deployed to an online identification system to receive response change data obtained during the current drilling operation in real time and output the corresponding underground structure continuity identification results to assist in subsequent adverse geological identification or drilling path adjustment decisions.

[0060] S03. Obtain the second drilling exploration data formed by the current drilling operation, and extract the response change dataset during the current drilling process based on the second drilling exploration data.

[0061] Furthermore, the extraction of the response change dataset of any type of drilling exploration data in the second drilling exploration data generated by the current drilling operation can be referred to steps S0211 to S0213 above, and will not be repeated here.

[0062] Similarly, the method for extracting response change data here can be flexibly configured according to the specific type of drilling exploration parameters, the processing system capabilities, and the sensitivity of the target geological structure for identification.

[0063] S04. Based on the identification model, predict the continuity of the current underground spatial structure according to the response change dataset, and locate the target spatial region based on the continuity prediction result. The target spatial region is used to indicate the distribution location of adverse geological bodies.

[0064] Further, in step S04, the response change dataset on the current drilling path is input into the trained recognition model to obtain the corresponding continuity prediction result of the underground structure.

[0065] In some embodiments, the recognition model is a graph neural network model, in which the real-time collected response change feature vector is used as the node attribute input, and a corresponding graph structure is constructed based on the current drilling depth to output the structural continuity label corresponding to each spatial position on the current path.

[0066] Furthermore, the step S04 of locating the target spatial region based on the continuous prediction results also includes the following steps: identifying continuous prediction segments that satisfy a preset change pattern according to the spatial distribution characteristics of the continuous prediction results, and determining the spatial location corresponding to the continuous prediction segments as the target spatial region.

[0067] Among them, the structural continuity prediction results are used to indicate the degree of integrity of the underground space structure, and their form can be discrete classification labels or continuous real values.

[0068] In some embodiments, structural continuity labels are represented in a discrete classification form, such as a label set {0,1,2}, where 0 indicates structural continuity, representing that the underground structure is intact and undisturbed at this spatial location; 1 indicates local discontinuity, reflecting the presence of slight changes or cracks in the structure; and 2 indicates abrupt structural changes, characterizing the presence of significant structural interfaces or adverse geological units.

[0069] In other embodiments, structural continuity labels are expressed using continuous real values, with a range of [0,1]. Values ​​closer to 1 indicate a more complete structure, while values ​​closer to 0 indicate weaker structural continuity. For example, a continuity value greater than 0.85 can be considered a continuous structural state, while a value less than 0.5 can be identified as a structural abrupt change or an abnormal region.

[0070] For both discrete and continuous labels mentioned above, when identifying the target spatial region, the spatial distribution characteristics of the label sequence can be used as a basis to extract the continuous prediction segment that meets the specific change pattern, and the spatial location corresponding to the segment can be used as the target spatial region.

[0071] For example, for discrete labels, a sudden change can be defined as a label transitioning from "0" to "2" or remaining between "1" and "2" for an extended period. This allows the identification of label transition segments as target segments with significant continuous changes. Similarly, for continuous labels, analyzing the fluctuations in their first derivative or mean value within a sliding window can identify spatial segments where continuous values ​​show a significant decrease or local minima, thus determining the possibility of structural anomalies or continuous abrupt changes within these segments.

[0072] In some embodiments of the drilling exploration-based adverse geological identification exploration method provided by the present invention, the following steps are also included: based on the target spatial area, performing operation control operations corresponding to the target spatial area.

[0073] Specifically, the operation control measures include one or more of the following: adjusting the drilling path to avoid potentially unfavorable geological bodies, adjusting control parameters during the drilling process to adapt to complex geological conditions, and triggering early warning signals to indicate construction risks.

[0074] In some embodiments, if the identification results indicate that there is a target spatial area with structural abrupt changes or poor continuity ahead, the system can dynamically generate an adjusted drilling path based on the spatial position of the area in the drilling coordinate system, or combine the historical evaluation results of the area in the existing geological database to select strategies such as bypassing, deceleration, or multi-angle penetration.

[0075] In other embodiments, the system can also be linked with the drilling rig control module to adjust control parameters such as advance speed, drilling pressure, rotation speed, and grouting volume in real time to enhance the adaptability and stability of the equipment under discontinuous geological conditions.

[0076] In addition, if the target area matches the preset high-risk geological type (such as fault fracture zone, weak interlayer, water and mud inrush zone, etc.), the system can automatically trigger multi-level early warning signals to notify operators to take further safety measures or suspend drilling operations to ensure the safety of on-site personnel and equipment.

[0077] In these embodiments, by linking the identification results with control commands, the present invention not only has a high-precision underground structure identification capability, but also realizes an intelligent drilling control mechanism that integrates "perception-decision-execution".

[0078] See Figure 4 , Figure 4 This is a schematic diagram of the system structure provided by the present invention for performing the drilling-based adverse geological identification exploration method.

[0079] In some embodiments, such as Figure 4 As shown, the drilling exploration-based adverse geological identification exploration system provided by the present invention includes a processor and a memory. The memory stores a computer program. When the computer program is read by the processor, it executes the drilling exploration-based adverse geological identification exploration method proposed in any of the above embodiments.

[0080] Furthermore, the memory is used to store computer-executable program code, identification model parameters, drilling exploration data, and intermediate processing results. The memory may include read-only memory (ROM), random access memory (RAM), flash memory, or other read-write storage devices.

[0081] Furthermore, the processor is used to execute the computer program to implement the drilling exploration-based adverse geological condition identification exploration method described in any embodiment of this specification. Specifically, the processor is capable of: acquiring first drilling exploration data generated from historical drilling operations; constructing an identification model based on the first drilling exploration data; acquiring second drilling exploration data generated from the current drilling operation and extracting the corresponding response change dataset; inputting the response change dataset into the identification model and outputting a structural continuity prediction result; and locating a target spatial region based on the prediction result to indicate the distribution location of adverse geological bodies.

[0082] Furthermore, the processor can also generate operation control signals corresponding to the target spatial area based on the recognition results, so as to realize operations such as adjusting the drilling path, drilling parameters or triggering early warning signals.

[0083] In some embodiments, the adverse geological condition identification exploration system based on drilling exploration provided by the present invention further includes a communication interface for data interaction between the system and external devices. The communication interface may include, but is not limited to, an Ethernet interface, a serial communication interface (RS485 / RS232), a wireless communication module (such as Wi-Fi, 4G, 5G), or an industrial bus communication interface (such as CAN, Modbus, etc.), for receiving real-time data from drilling equipment or field sensors and sending identification results or control commands to the field control system.

[0084] In some embodiments, the system may further include an input / output interface module for connecting to a field display, console, or remote monitoring terminal, so that operators can view the underground structure identification results, early warning information, and control suggestion parameters in real time.

[0085] In some specific applications, the above system can be deployed on engineering vehicle control terminals, drilling rig control cabinets, or rear data processing centers. It integrates and runs the method described in this invention through an industrial computing platform (such as an industrial computer, embedded edge computing device, or portable data terminal), and has good field adaptability and scalability.

[0086] In the above embodiments, the descriptions of different embodiments have different emphases; technical features not detailed or recorded in a certain embodiment can be understood and implemented by referring to the corresponding records of other embodiments. Unless otherwise expressly stated to the contrary: technical features in each embodiment can be substituted or combined with each other without technical conflict; the order of method steps can be adjusted without affecting the function; the device / module / unit can be implemented by hardware, software or a combination thereof, and can be centralized or distributed; parameters, values ​​or ranges include reasonable errors and equivalent values, and the terms "about", "greater than / less than", "between", and range endpoints are all covered without affecting the technical effect; ordinal numbers such as "first / second" are only used for distinction and do not limit the quantity, priority or structural relationship; the reference numerals and names in the specification and drawings are only illustrative and do not limit the structural form, size ratio or installation position; improvements, substitutions or equivalent solutions that are not explicitly stated but can be obtained by those skilled in the art without creative effort should all be included in the protection scope of this invention.

Claims

1. A method for identifying adverse geological conditions based on drilling exploration, characterized in that, Includes the following steps: Obtain the first drilling exploration data formed from historical drilling operations; Based on the first drilling exploration data, the correlation between response change data and drilling spatial location during the drilling process is obtained, and based on the correlation, an identification model is constructed. The identification model is used to identify the continuity of underground spatial structures. Acquire the second drilling exploration data generated by the current drilling operation, and extract the response change dataset during the current drilling process based on the second drilling exploration data; Based on the identification model, the continuity of the current underground spatial structure is predicted according to the response change dataset, and the target spatial region is located based on the continuity prediction result. The target spatial region is used to indicate the distribution location of adverse geological bodies.

2. The method for identifying adverse geological conditions based on drilling exploration according to claim 1, characterized in that, The process of constructing the recognition model based on the aforementioned association includes the following steps: Based on the response change data and corresponding spatial location in the aforementioned relationship, training sample pairs are constructed, wherein each training sample includes a response change data and its corresponding structural continuity label; Multiple training samples are input into a machine learning model for training to obtain a recognition model. The recognition model is configured to receive response change data on the current drilling path and output the prediction result of the underground structure continuity corresponding to the spatial location.

3. The method for identifying adverse geological conditions based on drilling exploration according to claim 1, characterized in that, The process of locating the target spatial region based on continuous prediction results includes the following steps: Based on the spatial distribution characteristics of the continuous prediction results, continuous prediction segments that satisfy the preset change pattern are identified, and the spatial location corresponding to the continuous prediction segments is determined as the target spatial region.

4. The method for identifying adverse geological conditions based on drilling exploration according to claim 1, characterized in that, It also includes the following steps: Based on the target spatial region, perform operation control operations corresponding to the target spatial region. The operation control operations include one or more of the following: adjusting the drilling path, adjusting control parameters during the drilling process, and triggering an early warning signal.

5. The exploration method for identifying adverse geological conditions based on drilling exploration according to any one of claims 1-4, characterized in that, The machine learning model is a graph convolutional neural network. Based on the graph convolutional neural network, the process of constructing the recognition model specifically includes the following steps: Using each spatial location on the drilling path as a graph node and the spatial adjacency relationship between adjacent locations as graph edges, the response change data corresponding to each graph node is used as the node attribute input to the model. During training, the model weight parameters are optimized by minimizing the loss function between predicted continuous labels.

6. The exploration method for identifying adverse geological conditions based on drilling exploration according to any one of claims 1-4, characterized in that, For a first type of drilling exploration data collected during a drilling operation, the process of obtaining the correlation between response change data and drilling spatial location includes the following steps: This first drilling exploration data is spatially interpolated or resampled according to the drilling path so that each drilling exploration data point corresponds one-to-one with a unique spatial location; Based on the processed first drilling exploration data, at least one variation feature is extracted along the drilling path at different spatial locations; The change features are encoded into response change data feature vectors, and their corresponding spatial locations are used as indices to generate the response change dataset for this drilling operation.

7. The exploration method for identifying adverse geological conditions based on drilling exploration according to any one of claims 1-4, characterized in that, Response change data at any spatial location should include at least: The first type of change feature is used to characterize the change trend of at least one type of drilling exploration data corresponding to the spatial location on the drilling path; The second type of variation feature is used to characterize the degree of fluctuation of the at least one drilling exploration data within a preset spatial range near this spatial location.

8. The exploration method for identifying adverse geological conditions based on drilling exploration according to any one of claims 7, characterized in that, Response change data at any spatial location should include at least: The third type of variation feature is used to characterize the intensity of variation in the spatial distribution of at least one type of drilling exploration data within this spatial location and a pre-defined spatial range nearby.

9. The method for identifying adverse geological conditions based on drilling exploration according to claim 8, characterized in that, A third-type variation feature at any spatial location is determined by a first-type variation feature corresponding to a drilling exploration data at that spatial location, and a second-type variation feature corresponding to the same drilling exploration data at that spatial location.

10. A geological defect identification exploration system based on drilling exploration, characterized in that, It includes a processor and a memory, wherein the memory stores a computer program, and when the computer program is read by the processor, it executes the adverse geological identification exploration method based on drilling exploration as described in any one of claims 1 to 9.

Images