Directional drilling exploration method

Abstract

This invention belongs to the field of geological exploration technology, specifically disclosing a method for directional drilling exploration, including: constructing a three-dimensional fracture network model of the rock mass to be explored; dynamically optimizing the trajectory of the main borehole axis and multiple spiral branch boreholes based on the three-dimensional fracture network model; obtaining the rock mass integrity index through a drilling acoustic testing system; dynamically adjusting drilling parameters in real time according to the rock mass integrity index and uniaxial compressive strength; implementing drilling according to the dynamically adjusted drilling parameters and the dynamically optimized trajectory of the main borehole axis and multiple spiral branch boreholes, and determining the attitude parameters of structural surfaces within the borehole based on the information obtained during the drilling process. This invention reduces the error in the structural surface exposure angle by dynamically optimizing the trajectory of the main borehole axis and spiral branch boreholes through a three-dimensional fracture network model; and significantly improves the density of three-dimensional geological information acquisition and the number of structural surfaces exposed per unit drilling depth through a spatial detection network combining the main borehole and spiral branch boreholes.

Description

Technical Field

[0001] This invention belongs to the field of geological exploration technology, and specifically discloses a method for directional drilling exploration. Background Technology

[0002] In geological exploration, adit exploration technology is traditionally used for detailed investigation of deep ore bodies or geological structures. Although traditional adit exploration technology can directly reveal the rock mass structure, it has disadvantages such as long construction period, high unit cost, and significant ecological disturbance. In addition, it is limited by the terrain slope (≤30°) and is difficult to implement in steep areas.

[0003] In recent years, horizontal directional drilling technology has gradually replaced traditional adit exploration technology, but the following technical problems still exist: 1. The borehole trajectory in horizontal directional drilling often relies on empirical formulas for determination, resulting in significant errors; 2. The failure to couple the borehole trajectory design with the three-dimensional fracture network model leads to large errors in the structural surface exposure angle; 3. Data acquisition is limited to a single technical means, resulting in insufficient completeness of structural surface identification. Summary of the Invention

[0004] The purpose of this invention is to provide a directional drilling exploration method to solve the technical problems of large errors in the structural surface exposure angle and insufficient structural surface identification integrity in existing horizontal directional drilling technology.

[0005] This invention provides a method for directional borehole exploration, comprising:

[0006] Step 1: Construct a three-dimensional fracture network model of the rock mass to be explored;

[0007] Step 2: Based on the three-dimensional fracture network model, dynamically optimize the trajectory of the main borehole axis and multiple spiral branch boreholes;

[0008] Step 3: Obtain the rock mass integrity index using a drilling acoustic testing system;

[0009] Step 4: Dynamically adjust drilling parameters in real time based on the rock mass integrity index and uniaxial compressive strength;

[0010] Step 5: Drilling is carried out according to the dynamically controlled drilling parameters and the dynamically optimized main borehole axis and the trajectory of multiple spiral branch boreholes, and the structural surface attitude parameters inside the borehole are determined based on the borehole information during the drilling process.

[0011] Preferably, step 1 specifically includes:

[0012] A fracture length-spacing probability model is established based on the structural surface occurrence parameters of the rock mass to be explored.

[0013] A three-dimensional fracture network model is constructed based on the fracture length-spacing probability model.

[0014] Preferably, step 2 specifically includes:

[0015] The three-dimensional fracture network model is input into a multi-objective nonlinear constraint optimization model to dynamically optimize the trajectory of the main borehole axis and multiple spiral branch boreholes;

[0016] The optimization objectives of the multi-objective nonlinear constraint optimization model are to maximize the number of exposed advantageous structural surfaces and minimize the total exposure angle error.

[0017] Preferably, the solution algorithm for the multi-objective nonlinear constraint optimization model is a non-dominated sorting genetic algorithm.

[0018] Preferably, the structural surface attitude parameters inside the borehole are determined based on borehole information during the drilling process, specifically as follows:

[0019] Acquire borehole information during the drilling process, wherein the borehole information includes at least one of borehole wall images, acoustic data, and hydrological data;

[0020] The borehole information is input into a CNN-LSTM hybrid neural network model to obtain the structural surface orientation parameters inside the borehole. Steps 1 to 5 are repeated until the drilling is completed.

[0021] Preferably, step 3 specifically includes:

[0022] Longitudinal wave velocities were collected at 0.2m intervals along the borehole axis using a drilling acoustic testing system.

[0023] The rock mass integrity index is determined based on the longitudinal wave velocity.

[0024] Preferably, step 4 specifically includes:

[0025] Drilling speed and pump pressure are dynamically adjusted in real time based on the rock mass integrity index;

[0026] The drilling equipment speed is dynamically adjusted in real time based on the rock mass integrity index and uniaxial compressive strength.

[0027] Preferably, the drilling speed is dynamically adjusted in real time based on the rock mass integrity index, specifically as follows:

[0028] When the rock mass integrity index is less than the first threshold, the drilling speed ;

[0029] When the first threshold is less than or equal to the rock mass integrity index and less than the second threshold, the drilling speed is... ;

[0030] When the rock mass integrity index is greater than or equal to the second threshold, the drilling speed ;

[0031] In the formula, For drilling speed, For rock mass integrity index, For the first threshold, This is the second threshold.

[0032] Preferably, the pump pressure is dynamically adjusted in real time based on the rock mass integrity index, specifically as follows:

[0033] The rock mass fragmentation index is determined based on the rock mass integrity index.

[0034] The pump pressure is dynamically adjusted in real time based on the rock mass fracture index.

[0035] Preferably, the drilling equipment rotation speed is dynamically adjusted in real time based on the rock mass integrity index and uniaxial compressive strength, specifically as follows:

[0036] If the rock mass integrity index is less than 50%, the drilling equipment rotation speed will be dynamically adjusted in real time according to the following formula:

[0037]

[0038] In the formula, For drilling equipment rotation speed, It represents the uniaxial compressive strength of the rock mass to be explored.

[0039] The directional drilling exploration method of the present invention has the following advantages compared with the prior art:

[0040] This invention breaks through the limitations of the original design that separates the geological model from the borehole trajectory. It uses the Monte Carlo algorithm to construct a three-dimensional fracture network model. By dynamically optimizing the angle between the borehole trajectory and the orientation of the dominant structural surface through the three-dimensional fracture network model, the error of the structural surface exposure angle can be reduced, which can significantly improve the accuracy compared with traditional exploration methods.

[0041] This invention differs from the single-hole linear drilling mode by designing a spatial detection network that combines the main borehole and spiral branch boreholes, which significantly improves the density of three-dimensional geological information acquisition and the number of structural surfaces revealed per unit drilling depth.

[0042] This invention establishes a real-time correlation between drilling speed, pump pressure, and rock mass integrity index, when the borehole encounters a fractured zone ( The system automatically triggers speed control, and simultaneously reinforces the borehole wall by injecting nano-silicate slurry and activating the borehole wall roller. Combined with the high-frequency vibration rock breaking mode, it significantly reduces the occurrence rate of borehole wall collapse, improving drilling efficiency while ensuring construction safety.

[0043] This invention combines borehole panoramic imaging, array acoustic wave testing, and hydrological monitoring data, and improves the accuracy and effectiveness of extracting borehole structural surface attitude parameters through a CNN-LSTM hybrid neural network model, reducing structural surface attitude identification errors, while also improving the accuracy of the three-dimensional fracture network model.

[0044] The exploration method provided by this invention significantly improves exploration efficiency and reduces overall costs while also significantly improving ecological benefits. By employing spiral branch borehole reuse technology and a waste soil recycling system, the overall exploration cost is greatly reduced, with the cost of acquiring geological information per unit being only 1 / 5 of that of traditional methods. This reduces disturbance to surrounding soil and vegetation, ensuring the accuracy and completeness of geological information acquisition while also significantly reducing the area of ​​ecological disturbance. Attached Figure Description

[0045] Figure 1 This is a flowchart of the directional drilling exploration method according to an embodiment of the present invention. Detailed Implementation

[0046] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of the invention. However, those skilled in the art will understand that the invention can be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0047] This invention provides a method for directional borehole exploration, such as... Figure 1 As shown, it includes:

[0048] Step 1: Construct a three-dimensional fracture network model of the rock mass to be explored, specifically as follows:

[0049] Step 1.1: Establish a fracture length-spacing probability model based on the structural surface occurrence parameters of the rock mass to be explored.

[0050] In this embodiment of the invention, an airborne LiDAR system is used to acquire high-precision topographic data of the rock mass to be explored. In conjunction with a ground total station, ≥200 sets of structural surface attitude parameters (dip, strike, dip angle) are collected (accuracy ±0.01°). Then, a fracture length-spacing probability model is established based on the Weibull distribution.

[0051] Step 1.2: Construct a three-dimensional fracture network model based on the fracture length-spacing probability model.

[0052] In this embodiment of the invention, the Monte Carlo algorithm is used to generate a three-dimensional fracture network model, and the key parameters are set as follows: fracture density 3.2-5.8 fractures / m³, mean trace length 8.6m (standard deviation 2.3m), and attitude dispersion ≤12°.

[0053] Step 2: Based on the three-dimensional fracture network model, dynamically optimize the trajectory of the main borehole axis and multiple spiral branch boreholes. Specifically, input the three-dimensional fracture network model into the multi-objective nonlinear constraint optimization model to dynamically optimize the trajectory of the main borehole axis and multiple spiral branch boreholes. The optimization objective of the multi-objective nonlinear constraint optimization model is to maximize the number of exposed advantageous structural surfaces and minimize the total exposure angle error.

[0054] In this embodiment of the invention, the determination rule for branch nodes is as follows: the starting point of the spiral branch borehole is set at every 1m interval along the main borehole axis (determined according to the borehole diameter and height). The development of fractures within a range of 5-10m ahead is comprehensively detected by methods such as borehole panoramic imaging and array acoustic wave testing. The section with higher fracture density is preferentially selected as the starting point of the spiral branch borehole.

[0055] To accurately achieve the optimization objective of "maximizing the number of exposed advantageous structural surfaces and minimizing the total exposure angle error," this embodiment of the invention constructs a multi-objective nonlinear constraint optimization model. This model abstracts the design of the main borehole axis and the spiral branch borehole trajectory into a path planning problem in three-dimensional space. Its input is the three-dimensional fracture network model simulated using the Monte Carlo algorithm, and its output is the optimal set of borehole trajectory parameters.

[0056] This invention employs the Monte Carlo algorithm to construct a three-dimensional fracture network model and dynamically optimizes the angle between the borehole trajectory and the orientation of dominant structural surfaces. Specifically, a multi-objective nonlinear constraint optimization model is established and solved using the Non-dominated Sorting Genetic Algorithm II (NSGA-II). This model aims to maximize the number of exposed dominant structural surfaces. ) and minimizing the total exposure angle error ( To optimize the objectives, and with strict constraints on key engineering parameters such as the angle between the main borehole axis and the dominant structural plane, the ratio of the pitch of the spiral branch borehole to the average fracture spacing (limited to 0.6-0.8), and the minimum radius of curvature, the specific scheme is as follows:

[0057] 1. Model type: Multi-Objective Nonlinear Constrained Optimization Model.

[0058] 2. Solution Algorithm: The Non-dominated Sorting Genetic Algorithm II (NSGA-II) is employed for solving the problem. NSGA-II is particularly suitable for optimization problems with multiple conflicting objectives (such as maximizing the number of exposed advantageous structural surfaces vs. minimizing the total exposure angle error) and complex constraints (such as geometric and engineering constraints). It can efficiently search the Pareto optimal solution set, providing decision-makers with multiple trade-off options. This model sets the number of iterations to 500 to ensure the convergence and robustness of the solution.

[0059] 3. Decision variables ( )

[0060] Initial azimuth angle of the main borehole (unit: degrees).

[0061] Initial inclination angle of the main borehole (unit: degrees).

[0062] Total design length of the main borehole (unit: meters). (Usually preset based on exploration needs, but can also be used as...) )

[0063] The number of spiral branch holes that branch off from the main borehole.

[0064] : No. The starting position of each spiral branch borehole on the axis of the main borehole (unit: meters, m).

[0065] : No. Radius of curvature of each spiral branch borehole (in meters). (Controls the sharpness / slowness of the turn).

[0066] : No. The helical pitch (in meters) of each helical branch borehole. (This controls the density of the helical threads).

[0067] : No. Initial deflection angle of each helical branch borehole (relative to the axis of the main borehole) (unit: degrees).

[0068] : No. The design length of each spiral branch borehole (unit: meters).

[0069] : The rotation direction of the spiral branch drilling helix (left-handed / right-handed, usually preset to right-handed).

[0070] 4. Objective Function (OF)

[0071] The multi-objective nonlinear constrained optimization model aims to simultaneously maximize two key performance indicators:

[0072] OF1: Maximize the number of advantageous structural surfaces revealed (Intersection Count)

[0073] (1)

[0074] In the formula, This represents the total number of all dominant structural surfaces generated in the Monte Carlo algorithm; For the indicator function, if the first The first advantageous structural surface is revealed by the main borehole or any one of the helical branch boreholes (i.e., the trajectory of the main borehole or the helical branch borehole is consistent with the first...). The distance between the planes of the dominant structural surfaces is less than a preset threshold. ,For example (equal to the borehole radius), then =1; otherwise =0.

[0075] OF2: Minimize the total angular deviation sum.

[0076] (2)

[0077] In the formula, For indicator functions; For the first The angle (in degrees) between the normal direction of the exposed dominant structural surface and the direction of the borehole trajectory axis at the exposure point is the actual exposure angle. The desired exposure angle (in degrees) is a preset key parameter (typically within the range of 60°-90°, with near-vertical exposure yielding the best results). This invention optimizes the trajectory to achieve... as close as possible The above. , It is the unit normal vector of the dominant structural surface. It is the unit direction vector of the borehole axis at the point of exposure.

[0078] Optimization goal: To simultaneously optimize within the NSGA-II framework. and Find a set of decision variables This allows for the exposure of as many structural surfaces as possible. To maximize), and for these revealed structural surfaces, the sum of the deviations between the actual angle and the expected angle should be as small as possible. minimize).

[0079] 5. Constraints (C)

[0080] These conditions constrain the space of feasible solutions, ensuring that the trajectory is geometrically and engineeringally realizable and safe:

[0081] C1: Angle constraint between the main borehole axis and the orientation of the dominant structural plane (critical constraint)

[0082] For specific key advantage structural aspects: (Applicable to key advantageous structural aspects that require focused control).

[0083] For all exposed structural surfaces: (More general form, when the main borehole is exposed).

[0084] in, For the first The orientation angle of each dominant structural surface (unit: degrees); , These are the preset minimum and maximum allowable angles (in degrees). For example, to ensure effective exposure, the following settings can be configured: , This avoids the trajectory being nearly parallel to the structural surface, ensuring efficient drilling. This invention optimizes the azimuth angle of the main borehole. Actively satisfy this constraint.

[0085] C2: Minimum radius of curvature constraint (ensures drillability)

[0086] Drilling holes in each spiral branch It must meet the following requirements: .

[0087] in, The minimum radius of curvature (in meters) achievable by the drilling rig and drill string combination is determined by the capabilities of the engineering equipment. For example, A radius of curvature that is too small can lead to the drill bit being unable to bend or the drill getting stuck.

[0088] C3: Constraint on the ratio of branch hole helical pitch to average crack spacing (core innovation constraint)

[0089] Drill holes for each spiral branch The ratio of pitch to average pitch is defined as... .

[0090] The above formula must satisfy the ratio range constraint: .

[0091] in The average spacing (in meters) of the dominant structural planes in the region or target layer is calculated using the Monte Carlo model. , These are the minimum and maximum values ​​of the preset ratio of pitch to average pitch, respectively. For example, , This achieves the goal of matching the rotation cycle of the spiral branch drilling with the periodic distribution of the fractures, maximizing the number of fracture crossings, and avoiding the effects of excessively large pitch (missing fractures) or excessively small pitch (repeatedly exposing the same fracture / increasing ineffective footage).

[0092] C4: Spiral branch drilling deflection angle range constraint

[0093] For each spiral branch borehole i, the initial deflection angle must satisfy: .

[0094] Limiting the deviation angle of the initial direction of the helical branch borehole relative to the main borehole ensures smooth drilling and trajectory control of the helical branch borehole. For example, , .

[0095] C5: Spacing constraint at the starting position of the spiral branch drilling

[0096] Drilling holes for any two different spiral branches The initial position spacing satisfies: This ensures sufficient spacing between adjacent spiral branch boreholes, preventing excessive concentration of branch points that could weaken the main borehole or interfere with construction.

[0097] C6: Total drilling length constraint

[0098] The total drilling length of the entire exploration plan must meet the following requirements: This limits the total drilling length to control costs, among which This is the preset maximum allowable total advance. Total design length of the main borehole (unit: meters). and The scope can be pre-defined according to exploration needs, or it can be used as... ; This refers to the number of spiral branch holes that branch off from the main borehole. For the first The design length of each spiral branch borehole (unit: meters).

[0099] C7: Branch hole length constraint

[0100] For spiral branch drilling i, the length of a single hole must meet the following requirements: This limits the length range of a single spiral branch borehole, ensuring construction feasibility and control precision.

[0101] C8: Target constraint for angle error control

[0102] For all revealed advantageous structural surfaces The angle error must meet the following requirements: .

[0103] in This is the maximum permissible angular error, for example... =5°. Through optimization, this invention ultimately achieves a 95% success rate in revealing the optimal structural plane at the required angle, meaning the error is less than 5°. .

[0104] In this embodiment of the invention, the NSGA-II algorithm, under the premise of satisfying all constraints C1-C8, determines the decision variables... The search is performed within the space, ultimately outputting a Pareto optimal solution set (Pareto Front). Each solution represents a set of feasible borehole trajectory parameters (…). ), and in the objective function and There are different trade-offs involved. Engineers can select the most suitable implementation scheme from this set of optimal solutions based on actual exploration needs (such as prioritizing quantity or accuracy). This invention, through its model and algorithm, significantly reduces the angular error in structural surface exposure, greatly improves the scientific rigor and accuracy of trajectory design, and overcomes the drawback of relying on empirical formulas leading to large errors.

[0105] This invention aims to maximize the number of dominant structural planes revealed. ) and minimizing the total exposure angle error ( To address the dual objectives, a multi-objective nonlinear constrained optimization model was established. The decision variables of this model include: the initial azimuth angle of the main borehole. ,inclination Total design length Number of spiral branch boreholes Starting position of each spiral branch borehole radius of curvature Helix pitch Initial deflection angle Design length The NSGA-II algorithm was used to solve the above model through 500 iterations. The algorithm outputs a Pareto optimal solution set, where each solution represents a set of feasible trajectory parameters that satisfy all constraints, and addresses both objectives (…). and Different trade-offs were achieved between the options. The final optimal solution achieved the main borehole azimuth angle. The design of the right-hand spiral branch drilling ensures that 95% of the advantageous structural surfaces are exposed at the following angles: ).

[0106] Compared with traditional empirical formula prediction methods, this model-driven optimization method significantly improves the accuracy of trajectory design and the effectiveness of geological information acquisition.

[0107] Step 3: Obtain the rock mass integrity index using a drilling acoustic testing system, specifically:

[0108] The longitudinal wave velocity was collected at 0.2m intervals along the borehole axis using a drilling acoustic testing system; then the rock mass integrity index was determined based on the longitudinal wave velocity.

[0109] To better optimize the drilling trajectory, this embodiment of the invention is equipped with a drilling acoustic testing system (MWD) consisting of a triaxial accelerometer (resolution 0.001g), a fluxgate sensor (azimuth accuracy ±0.08°), and a near-bit gamma detector (sampling rate 10Hz) on the drilling equipment.

[0110] This embodiment of the invention employs the aforementioned drilling acoustic testing system, configured with a wideband transmitting probe and channel receiving array, to acquire longitudinal wave velocity (Vp) data at 0.2m intervals along the main borehole axis. The data is obtained through the formula... The rock mass integrity index was calculated, and a Vp-RQD relationship model was established (fit R² ≥ 0.92). A resistivity imager (electrode spacing 0.1 m) was simultaneously installed to acquire formation resistivity. (Ω·m), when When the value is less than 80Ω·m, the auxiliary judgment mechanism for the fracture zone is triggered.

[0111] Step 4: Based on the rock mass integrity index and uniaxial compressive strength, dynamically adjust the drilling parameters in real time. Specifically, dynamically adjust the drilling speed and pump pressure in real time based on the rock mass integrity index; dynamically adjust the rotation speed of the drilling equipment in real time based on the rock mass integrity index and uniaxial compressive strength.

[0112] This invention embodiment dynamically adjusts the drilling speed in real time based on the rock mass integrity index, specifically as follows:

[0113] When the rock mass integrity index is less than the first threshold, the drilling speed .

[0114] When the first threshold is less than or equal to the rock mass integrity index and less than the second threshold, the drilling speed is... .

[0115] When the rock mass integrity index is greater than or equal to the second threshold, the drilling speed .

[0116] In the formula, For drilling speed, For rock mass integrity index, For the first threshold, The second threshold is defined as follows: the first threshold is 50%–55%; and the second threshold is 80%–85%.

[0117] In this embodiment of the invention, the pump pressure is dynamically adjusted in real time according to the rock mass integrity index, specifically: according to the rock mass integrity index... Determine the rock mass fracture index The control accuracy is ±0.5MPa; then, based on the rock mass fracture index... Determine the pump pressure of the drilling equipment corresponding to the main borehole and the spiral branch borehole. (MPa). For example, .

[0118] In this embodiment of the invention, the drilling equipment rotation speed is dynamically adjusted in real time based on the rock mass integrity index and uniaxial compressive strength, as shown in Table 1.

[0119] Table 1 Rotational Speed and Dynamic adjustment relationship

[0120] Step 5: Drilling is carried out according to the dynamically controlled drilling parameters and the dynamically optimized main borehole axis and the trajectory of multiple spiral branch boreholes, and the structural surface attitude parameters inside the borehole are determined based on the borehole information during the drilling process.

[0121] Unlike the single-hole linear drilling mode, the embodiments of the present invention adopt an exploration method that combines a main borehole and a spiral branch borehole.

[0122] For the main borehole construction: Horizontal directional drilling was carried out using a (Φ89mm) diamond impregnated drill bit. For the main borehole, which is relatively long (≥200m), different rock strata sections in the tunnel were classified and constructed as shown in Table 2.

[0123] Table 2 Classification Construction Parameters

[0124] For every 30m of drilling, a Φ114mm casing is installed for borehole wall support, and the casing gap is filled with expansive cement (expansion rate ≥8%) to ensure that the borehole inclination deviation is ≤0.5‰.

[0125] For the construction of helical branch boreholes: Branch starting points are set at 1m intervals along the main borehole axis (determined by the borehole diameter and height). A comprehensive detection method, including in-hole panoramic imaging and array acoustic wave testing, is used to assess fracture development within a 5-10m range ahead. Sections with higher fracture density are preferentially selected as the starting points for the helical branch boreholes. Simultaneously, to achieve steering control of the drilling equipment corresponding to the helical branch boreholes, this embodiment of the invention employs a steerable drill bit in conjunction with a near-bit guide sub. The tool face angle is adjusted in real-time every 1m of drilling to ensure minimal deviation between the actual trajectory and the designed helical parameters.

[0126] In the drilling process of the main borehole and the spiral branch borehole in this embodiment of the invention, a vortex separator and a mud circulation system are configured to achieve drill cuttings particle size classification, and the amount of waste generated is greatly reduced, by 98% compared with traditional adits.

[0127] The embodiments of the present invention utilize a spatially staggered layout of the main borehole and the spiral branch boreholes to significantly increase the number of structural surfaces exposed per unit depth compared to the single-hole linear pattern.

[0128] In this embodiment of the invention, when drilling the main borehole and the spiral branch borehole, as the drilling speed increases... After dynamic adjustment of the rotation speed, to further ensure construction safety and prevent borehole wall collapse, nano-silicate slurry was injected and a borehole wall roller was used to further reinforce the borehole wall.

[0129] When encountering a fault zone during drilling ( When the rock content is less than 40%, a high-frequency vibrator is used to break the rock, and a double-tube coring device is used to ensure the integrity of the core sample, as shown in Table 1.

[0130] To ensure minimal trajectory deviation between the main borehole and the auger branch borehole, this embodiment of the invention utilizes a drilling while drilling (MWD) system to determine the trajectory deviation vector generated for each drilling advance (e.g., 0.5m). When the trajectory deviation vector is greater than the threshold (e.g., 0.3m), the tool face angle is adjusted by the hydraulic guide mechanism to ensure that the trajectory deviation is small.

[0131] This invention determines the structural surface attitude parameters within the borehole based on borehole information obtained during the drilling process. Specifically, the borehole information is obtained during the drilling process, including at least one of borehole wall images, acoustic data, and hydrological data. The borehole information is then input into a CNN-LSTM hybrid neural network model to obtain the structural surface attitude parameters within the borehole. These parameters are then input into the optimization model (three-dimensional fracture network model) in step 1 to further improve the model accuracy. Steps 1 to 5 are repeated until drilling is completed.

[0132] The method for acquiring borehole wall images in this embodiment of the invention is as follows: a 360° optical probe is used to acquire borehole wall images every 10m of drilling, and image stitching is achieved through feature matching algorithm. RGB color values ​​(R / G / B channel accuracy ±2) are recorded simultaneously for lithology identification.

[0133] The method for acquiring acoustic data in this embodiment of the invention is as follows: array acoustic testing is used, that is, a drilling acoustic testing system is used, a wideband transmitting probe and a channel receiving array are configured, and longitudinal wave velocity data is collected at 0.2m intervals along the borehole axis.

[0134] The method for acquiring hydrological data in this embodiment of the invention is as follows: installing a distributed optical fiber sensing system to monitor the seepage pressure gradient in real time. (kPa / m) and flow velocity (m / d).

[0135] The obtained information about the pores is input into the trained deep learning model to obtain the structural surface attitude parameters within the pores. In this embodiment of the invention, the deep learning model is preferably a CNN-LSTM hybrid neural network model.

[0136] Input layer: Normalized imaging data, acoustic spectrum (1024-point FFT), hydrological parameters ( ).

[0137] CNN module: 5-layer convolutional structure, extracting structural surface morphological features (length, roughness) ).

[0138] LSTM module: Bidirectional network (128 hidden nodes) to analyze the temporal variation of sound wave velocity.

[0139] Output layer: fused feature maps generate structural surface attitude parameters (tendency) , towards Error control is .

[0140] The training and validation of the CNN-LSTM hybrid neural network model involves constructing a dataset of samples from dozens of typical rock strata structures, adjusting training parameters such as batch size, initial learning rate, and number of iterations, and finally using the accuracy results of the test set as the validation metric, which significantly reduces the error of traditional statistical methods.

[0141] This invention breaks through the limitations of the original design that separates the geological model from the borehole trajectory. It uses the Monte Carlo algorithm to construct a three-dimensional fracture network model. By dynamically optimizing the angle between the borehole trajectory and the orientation of the dominant structural surface through the three-dimensional fracture network model, the error of the structural surface exposure angle can be reduced, which can significantly improve the accuracy compared with traditional exploration methods.

[0142] This invention differs from the single-hole linear drilling mode by designing a spatial detection network that combines the main borehole and spiral branch boreholes, which significantly improves the density of three-dimensional geological information acquisition and the number of structural surfaces revealed per unit drilling depth.

[0143] This invention establishes a real-time correlation between drilling speed, pump pressure, and rock mass integrity index, when the borehole encounters a fractured zone ( When the borehole wall collapse rate is less than 50%, the speed control is automatically triggered. At the same time, the borehole wall is further reinforced by injecting nano-silicate slurry and activating the borehole wall roller. Combined with the high-frequency vibration rock breaking mode, the borehole wall collapse rate is greatly reduced, which improves drilling efficiency and ensures construction safety.

[0144] This invention combines borehole panoramic imaging, array acoustic wave testing, and hydrological monitoring data, and improves the accuracy and effectiveness of extracting borehole structural surface attitude parameters through a CNN-LSTM hybrid neural network model, reducing structural surface attitude identification errors, while also improving the accuracy of the three-dimensional fracture network model.

[0145] The exploration method provided by this invention significantly improves exploration efficiency and reduces overall costs while also significantly improving ecological benefits. By employing spiral branch borehole reuse technology and a spoil recycling system, the overall exploration cost is significantly reduced compared to adits, with the cost per unit of geological information acquisition being only 1 / 5 of that of traditional methods. This reduces disturbance to surrounding soil and vegetation, ensuring the accuracy and completeness of geological information acquisition while also significantly reducing the area of ​​ecological disturbance.

[0146] The above descriptions are merely a few embodiments of the present invention and are not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any modifications or alterations made by those skilled in the art without departing from the scope of the technical solution of the present invention using the disclosed technical content are equivalent to equivalent implementation cases and fall within the scope of the technical solution.

Claims

1. A method for directional borehole exploration, characterized in that, include: Step 1: Construct a three-dimensional fracture network model of the rock mass to be explored; Step 2: Based on the three-dimensional fracture network model, dynamically optimize the trajectory of the main borehole axis and multiple spiral branch boreholes; Step 3: Obtain the rock mass integrity index using a drilling acoustic testing system; Step 4: Dynamically adjust drilling parameters in real time based on the rock mass integrity index and uniaxial compressive strength; Step 5: Drilling is carried out according to the dynamically controlled drilling parameters and the dynamically optimized main borehole axis and the trajectory of multiple spiral branch boreholes, and the structural surface attitude parameters inside the borehole are determined based on the borehole information during the drilling process.

2. The directional drilling exploration method according to claim 1, characterized in that, Step 1 is as follows: A fracture length-spacing probability model is established based on the structural surface occurrence parameters of the rock mass to be explored. A three-dimensional fracture network model is constructed based on the fracture length-spacing probability model.

3. The directional drilling exploration method according to claim 2, characterized in that, Step 2 is as follows: The three-dimensional fracture network model is input into a multi-objective nonlinear constraint optimization model to dynamically optimize the trajectory of the main borehole axis and multiple spiral branch boreholes; The optimization objectives of the multi-objective nonlinear constraint optimization model are to maximize the number of exposed advantageous structural surfaces and minimize the total exposure angle error.

4. The directional drilling exploration method according to claim 3, characterized in that, The solution algorithm for the multi-objective nonlinear constraint optimization model is a non-dominated sorting genetic algorithm.

5. The directional drilling exploration method according to claim 3, characterized in that, The attitude parameters of the structural surfaces inside the borehole are determined based on the information obtained during the drilling process, specifically: Acquire borehole information during the drilling process, wherein the borehole information includes at least one of borehole wall images, acoustic data, and hydrological data; The borehole information is input into a CNN-LSTM hybrid neural network model to obtain the structural surface orientation parameters inside the borehole. Steps 1 to 5 are repeated until the drilling is completed.

6. The directional drilling exploration method according to claim 1, characterized in that, Step 3 specifically involves: Longitudinal wave velocities were collected at 0.2m intervals along the borehole axis using a drilling acoustic testing system. The rock mass integrity index is determined based on the longitudinal wave velocity.

7. The directional drilling exploration method according to claim 1, characterized in that, Step 4 is as follows: Drilling speed and pump pressure are dynamically adjusted in real time based on the rock mass integrity index; The drilling equipment speed is dynamically adjusted in real time based on the rock mass integrity index and uniaxial compressive strength.

8. The directional drilling exploration method according to claim 7, characterized in that, The drilling speed is dynamically adjusted in real time based on the rock mass integrity index, specifically as follows: When the rock mass integrity index is less than the first threshold, the drilling speed ; When the first threshold is less than or equal to the rock mass integrity index and less than the second threshold, the drilling speed is... ; When the rock mass integrity index is greater than or equal to the second threshold, the drilling speed ; In the formula, For drilling speed, For rock mass integrity index, For the first threshold, This is the second threshold.

9. The directional drilling exploration method according to claim 7, characterized in that, The pump pressure is dynamically adjusted in real time based on the rock mass integrity index, specifically as follows: The rock mass fragmentation index is determined based on the rock mass integrity index. The pump pressure is dynamically adjusted in real time based on the rock mass fracture index.

10. The directional drilling exploration method according to claim 7, characterized in that, The drilling equipment rotation speed is dynamically adjusted in real time based on the rock mass integrity index and uniaxial compressive strength, specifically as follows: When the rock mass integrity index is less than 50%, the drilling equipment rotation speed is dynamically adjusted in real time according to the following formula: ； In the formula, For drilling equipment rotation speed, It represents the uniaxial compressive strength of the rock mass to be explored.

Images