A lithology prediction method and system fusing while-drilling seismic and drilling geochemistry
Patent Information
- Application Number
- CN202611485506.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-23
- Publication Date
- 2026-10-30
AI Technical Summary
钻进参数能够反映钻具与围岩之间的相互作用关系，与岩石强度、完整性和可钻性具有一定关联，但其变化同时受施工控制参数、刀具磨损状态、设备工况和操作策略等因素影响，单独依靠钻进参数进行岩性判断时，容易出现误判或识别稳定性不足的问题
本公开的一种融合随钻地震与钻进地化的岩性预测方法，首先采集超前钻探过程中产生的被动源地震记录、钻进参数和岩屑地化参数；对被动源地震记录进行干涉重构和瞬时属性提取，获得与前方介质响应相关的瞬时振幅、瞬时频率等指标；同时提取钻进速度、推进压力、扭矩、转速、钻进比能等钻进特征，并融合岩屑地化组成、蚀变指标和岩性指示参数；最后通过多源特征融合模型建立被动源地震响应、钻进力学响应、地化组成特征与岩性类别之间的非线性关系，实现掌子面前方岩性的智能预测。该方法能够减少对传统取芯和室内测试的依赖，提高岩性识别的连续性、实时性和准确性，具有较好的工程应用价值。
Smart Images
Figure CN123010077A_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of drilling lithology identification technology, specifically to a lithology prediction method and system that integrates drilling seismic and drilling geochemical methods. Background Technology
[0002] The statements in this section are merely background information relating to this disclosure and do not necessarily constitute prior art.
[0003] During tunnel construction, geological factors such as changes in lithology, weak interlayers, fault fracture zones, water-rich zones, and abrupt changes in rock mass integrity ahead of the tunnel face directly affect construction safety, tunneling efficiency, selection of support parameters, and proactive risk management. Therefore, timely and accurate identification of the lithology ahead of the tunnel face and its changing characteristics is a crucial foundation for safe and efficient tunnel construction.
[0004] In recent years, some studies have attempted to use drilling parameters such as drilling speed, feed pressure, torque, rotational speed, and drilling energy for lithology identification. Drilling parameters can reflect the interaction between the drill bit and the surrounding rock and are related to rock strength, integrity, and drillability. However, their variations are also affected by factors such as construction control parameters, tool wear, equipment conditions, and operating strategies. Relying solely on drilling parameters for lithology judgment can easily lead to misjudgments or insufficient identification stability.
[0005] However, the existing lithology identification methods mentioned above still have the following limitations: (i) Although traditional coring and indoor testing methods are reliable, they are costly and time-consuming, making it difficult to meet the needs of real-time advance prediction in tunnel construction. (ii) Lithology identification based solely on drilling parameters is easily affected by non-geological factors such as equipment status, construction control parameters, and tool wear, resulting in limited identification accuracy and stability. Summary of the Invention
[0006] To address the aforementioned issues, this disclosure proposes a lithology prediction method and system that integrates seismic data during drilling and drilling geochemistry. Through passive source interferometry reconstruction, instantaneous index extraction, drilling parameter characterization, geochemical parameter lithology indication, and multi-source attention mechanism, it achieves multi-source fusion of wavefield propagation response, rock machine action feedback, and material composition constraints, enabling intelligent, efficient, and continuous prediction of lithology ahead of the tunnel face.
[0007] According to some embodiments, the present disclosure adopts the following technical solutions: A lithology prediction method integrating seismic data during drilling and drilling geochemistry includes: Obtain passive source seismic records, drilling parameters, and cuttings geochemical parameters during the advanced horizontal drilling process of the tunnel; Interferometric reconstruction and instantaneous attribute extraction were performed on passive source seismic records to obtain instantaneous index features related to the response of the preceding medium; Feature extraction was performed on drilling parameters and cuttings geochemical parameters to obtain drilling response features and geochemical lithology indicator features. Instantaneous index features, drilling response features, and geochemical and lithological indicator features are spatially registered in time-distance-mileage to form multi-source features; Multi-source features are input into the multi-source fusion lithology prediction model, and lithology prediction results are output. Engineering interpretation results are then generated based on the lithology prediction results. The multi-source fusion lithology prediction model includes a passive source instantaneous index encoder, a drilling parameter encoder, and a geochemical parameter encoder, which encode the three feature data sources respectively. Based on the data quality, spatial distance, missing rate, and feature uncertainty of each data source, cross-modal fusion weights are generated to generate multi-source fusion features, and finally output lithology category, lithology probability, and uncertainty.
[0008] According to some embodiments, the present disclosure adopts the following technical solutions: A lithology prediction system integrating seismic data during drilling and drilling geochemistry includes: The data acquisition module is used to acquire passive source seismic records, drilling parameters, and cuttings geochemical parameters during the tunnel advanced horizontal drilling process. The passive source interferometric reconstruction module is used to perform interferometric reconstruction of passive source seismic records; The instantaneous index extraction module is used to extract instantaneous attributes from passive source seismic records to obtain instantaneous index features related to the response of the preceding medium. The drilling parameter processing module is used to extract features from drilling parameters to obtain drilling response features; The geochemical parameter processing module is used to extract features from the geochemical parameters of rock cuttings to obtain geochemical lithology indicator features; The spatial registration module is used to perform time-distance-mileage spatial registration of instantaneous index features, drilling response features, and geochemical lithology indicator features to form multi-source features. The multi-source fusion lithology prediction module is used to input multi-source features into the multi-source fusion lithology prediction model and output lithology prediction results. The results output module is used to generate engineering interpretation results based on lithology prediction results; The multi-source fusion lithology prediction model includes a passive source instantaneous index encoder, a drilling parameter encoder, and a geochemical parameter encoder, which encode the three feature data sources respectively. Based on the data quality, spatial distance, missing rate, and feature uncertainty of each data source, cross-modal fusion weights are generated to generate multi-source fusion features, and finally output lithology category, lithology probability, and uncertainty.
[0009] According to some embodiments, the present disclosure adopts the following technical solutions: A computer program product includes a computer program that, when executed by a processor, implements the lithology prediction method that integrates seismic data during drilling and drilling geochemistry.
[0010] According to some embodiments, the present disclosure adopts the following technical solutions: A non-transitory computer-readable storage medium is provided for storing computer instructions, which, when executed by a processor, implement the aforementioned lithology prediction method integrating seismic data during drilling and drilling geochemistry.
[0011] According to some embodiments, the present disclosure adopts the following technical solutions: An electronic device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to perform a lithological prediction method that integrates seismic data while drilling and drilling geochemistry.
[0012] Compared with the prior art, the beneficial effects of this disclosure are as follows: This disclosure presents a lithology prediction method integrating seismic data while drilling and drilling geochemistry. First, it collects passive source seismic records, drilling parameters, and cuttings geochemical parameters generated during advance drilling. Then, it performs interferometric reconstruction and instantaneous attribute extraction on the passive source seismic records to obtain indicators such as instantaneous amplitude and instantaneous frequency related to the response of the underlying medium. Simultaneously, it extracts drilling characteristics such as drilling speed, feed pressure, torque, rotational speed, and drilling specific energy, and integrates cuttings geochemical composition, alteration indices, and lithology indicator parameters. Finally, it establishes a nonlinear relationship between passive source seismic response, drilling mechanical response, geochemical composition characteristics, and lithology category through a multi-source feature fusion model, achieving intelligent prediction of lithology ahead of the drilling face. This method reduces reliance on traditional coring and laboratory testing, improves the continuity, real-time performance, and accuracy of lithology identification, and has significant engineering application value.
[0013] This disclosure presents a lithology prediction method that integrates seismic data while drilling and geochemical data during drilling. It combines transient indices from passive source seismic data, drilling parameters, and geochemical parameters for lithology prediction, achieving multi-source fusion of wavefield propagation response, rock machine action feedback, and material composition constraints, thus avoiding lithology identification bias caused by a single data source.
[0014] This disclosure presents a lithology prediction method that integrates seismic data during drilling and drilling geochemistry. By using robust mutual interference and quality-weighted superposition, it transforms continuous, non-stationary, and uncertain-source passive source vibration records from construction into interpretable virtual source records, thereby improving the availability of passive source information in lithology prediction.
[0015] This disclosure presents a lithology prediction method that integrates seismic data during drilling and drilling geochemistry. It incorporates rock-mechanical characteristics such as drilling specific energy, rock-breaking efficiency, and torque-thrust ratio, and combines them with geochemical elements, mineral composition, and alteration indices. This allows the lithology prediction results to have both mechanical response and material composition basis, thereby enhancing the interpretability of the model. Attached Figure Description
[0016] The accompanying drawings, which form part of this disclosure, are used to provide a further understanding of this disclosure. The illustrative embodiments of this disclosure and their descriptions are used to explain this disclosure and do not constitute an undue limitation of this disclosure.
[0017] Figure 1 This is a schematic diagram of a lithology prediction method that integrates seismic data during drilling and drilling geochemistry, according to an embodiment of this disclosure. Detailed Implementation
[0018] The present disclosure will be further described below with reference to the accompanying drawings and embodiments.
[0019] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this disclosure. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains.
[0020] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this disclosure. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms “comprising” and / or “including” are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.
[0021] Example 1 One embodiment of this disclosure provides a lithology prediction method that integrates seismic data during drilling and drilling geochemistry, the method comprising the following steps: Step 1: Obtain passive source seismic records, drilling parameters, and cuttings geochemical parameters during the tunnel advance horizontal drilling process; Step 2: Perform interferometric reconstruction and instantaneous attribute extraction on passive source seismic records to obtain instantaneous index features related to the response of the preceding medium; Step 3: Extract features from drilling parameters and cuttings geochemical parameters to obtain drilling response features and geochemical lithology indicator features; Step 4: Perform time-distance-mileage spatial registration on the instantaneous index features, drilling response features, and geochemical lithology indicator features to form multi-source features; Step 5: Input the multi-source features into the multi-source fusion lithology prediction model, output the lithology prediction results, and generate engineering interpretation results based on the lithology prediction results; The multi-source fusion lithology prediction model includes a passive source instantaneous index encoder, a drilling parameter encoder, and a geochemical parameter encoder, which encode the three feature data sources respectively. Based on the data quality, spatial distance, missing rate, and feature uncertainty of each data source, cross-modal fusion weights are generated to generate multi-source fusion features, and finally output lithology category, lithology probability, and uncertainty.
[0022] As an example, during tunnel construction, drilling rigs, TBMs, or other construction equipment continuously generate vibration signals. These vibration signals propagate through the surrounding rock and are received by sensors. The seismic records contain rich information closely related to the propagation velocity of the medium ahead, wave field attenuation, interface reflection, and lithological changes. However, there are few methods for using seismic records to predict lithology. Especially in advanced geological drilling, the drilling operation itself can act as a passive source of vibration, eliminating the need for additional active seismic sources and dedicated construction time. This offers advantages such as synchronization with the construction process, continuous data acquisition, and strong engineering adaptability. Furthermore, the rock cuttings generated during advanced drilling can reflect the material composition of the rock mass exposed by the borehole. Geochemical testing of the rock cuttings can yield geochemical indicators such as elemental content, mineral composition, alteration characteristics, and water content or argillaceization indicators. Compared to relying solely on drilling parameters, geochemical parameters can provide lithological identification from the perspective of rock material composition. However, due to the discrete nature of rock cuttings sampling and the difficulty in individually characterizing lithological changes within a continuous spatial range ahead of the tunnel face, it is still necessary to integrate them with other continuous monitoring information.
[0023] Therefore, this disclosure proposes a lithology prediction method integrating seismic data during drilling and drilling geochemistry, using lithology prediction ahead of the tunnel face in drill-and-blast tunnel construction as an application scenario. This tunnel is excavated using the drill-and-blast method, and there are potential adverse geological risks ahead of the tunnel face, such as lithological changes, weak interlayers, fault fracture zones, water-rich structures, or abrupt changes in rock mass integrity. To obtain lithological information ahead of the tunnel face without additional construction time, the drilling disturbance during advanced geological drilling is used as a passive source, combined with drilling parameters and borehole cuttings geochemistry parameters, to intelligently predict the lithology within a certain range ahead of the tunnel face. The specific implementation process is as follows: Step 1: Obtain passive source seismic records, drilling parameters, and cuttings geochemical parameters during the tunnel advance horizontal drilling process; Several geophones are installed at stable locations behind or on the sidewalls of the drill-and-blast tunnel face. These geophones can be single-component or three-component, with a preferred number of 6 to 12. They are linearly arranged along the tunnel sidewalls, arch waist, arch foot, or near the boreholes, with a spacing of 1 to 5 meters, preferably 2 to 3 meters. The geophones are fixed to the surrounding rock or initial support structure using anchoring, expansion bolts, plaster coupling, magnetic coupling, or borehole embedding methods to improve the stability of vibration signal reception.
[0024] During advanced geological drilling, the drill bit continuously interacts with the rock mass in front of the tunnel face, generating continuous or quasi-continuous vibration signals. These vibration signals propagate through the surrounding rock and are received by various geophones, forming multiple passive source seismic records. After each advanced borehole, drilling section, or excavation distance is completed, continuous vibration data is collected for at least 3 to 20 minutes, preferably at least 5 minutes. The passive source vibration signals include vibrations induced by drill bit impact rock breaking, rotary cutting, bottom rock fracturing, drill rod vibration, mechanical disturbance, and micro-fractures in the surrounding rock.
[0025] During advanced geological drilling, drilling parameters of the drilling rig or borehole equipment are collected simultaneously. These drilling parameters include drilling speed. v d 、 propulsion F Drill pipe torque T Rotation speed n Penetration p One or more of the following: drilling pressure fluctuation, torque fluctuation, energy consumption per unit depth, impact frequency, slag discharge volume, hole depth, and borehole rock powder production.
[0026] During advanced drilling, borehole cuttings, rock powder, or slag samples are collected simultaneously according to borehole depth, drilling time, or footage intervals. XRF, portable elemental analysis, and mineral spectral identification are performed on the collected samples to obtain various rock cuttings geochemical parameters, including major elements, trace elements, oxide content, mineral composition, alteration indices, and argillaceization indices.
[0027] Furthermore, for the first i The detector in the first stage q Raw vibration records collected within a time window First, mean removal, trend removal, amplitude normalization, and bandpass filtering are performed to obtain the preprocessed passive source seismic signal. :
[0028] in, This represents the amplitude normalization or amplitude equalization operator. This indicates a bandpass filter operator that limits the operating frequency band. For drill-and-blast tunnel pre-drilling scenarios, the operating frequency band can be determined according to the drilling rig type, surrounding rock conditions, and detector frequency band, preferably 20 Hz to 800 Hz, and more preferably 50 Hz to 500 Hz.
[0029] Step 2: Perform interferometric reconstruction and instantaneous attribute extraction on passive source seismic records to obtain instantaneous index features related to the response of the preceding medium; Because passive source vibrations generated by advanced drilling have characteristics such as unfixed source location, unclear onset time, and wavelet instability, the original vibration records are difficult to use directly for lithology identification. Therefore, this embodiment adopts a robust mutual interference method to reconstruct continuous passive source vibration records into virtual source records similar to active sources.
[0030] Select the first detector, the detector near the working face, or a detector with a high signal-to-noise ratio and stable coupling as the reference channel, denoted as the [number]th channel. r The Way. The First i Daohedi r The Way in the q Preprocessed signals within a time window Transform them to the frequency domain respectively to obtain The frequency domain coherent response between the two channels is calculated using a phase-preserving coherence operator with regularization:
[0031] in, [] indicates complex conjugation. and This is a regularization factor related to the energy or noise level of that time window.
[0032] To mitigate the effects of anomalous time windows and narrowband mechanical line spectra, a time window quality weight is set. :
[0033] in, The proportion of narrowband line spectrum. This represents the proportion of abnormal amplitudes. For multichannel coherence, and This is the adjustment coefficient.
[0034] right Perform inverse Fourier transform.
[0035] Get the first r The passive source virtual source of the Dao is the virtual source of the virtual earthquake. .in, Q The total number of time windows. This is a weighted stable term. For Perform an inverse Fourier transform to obtain the result with the first... r The i-th channel detector is a virtual source, and the i-th channel detector is a virtual source record of the receiving channel. Repeat the above process for all receiving channels to obtain a set of virtual channels for drilling into the passive source.
[0036] Furthermore, the process of obtaining instantaneous index characteristics related to the response of the preceding medium is as follows: In drill-and-blast tunneling, the distance between geophones and the arrival time of the direct wave can be used to estimate the equivalent propagation velocity in the surrounding rock. This is based on the distance between the reference geophone and the other receiving geophones. d i And when the direct wave arrives t i The equivalent direct wave velocity is obtained by using a robust velocity estimation method:
[0037] in, This is the set of receiver channels that can reliably pick up direct waves. Using median estimation can reduce the impact of individual abnormal channels, poorly coupled channels, or picking errors on the velocity estimation results.
[0038] Two-way time travel within the reflection window τ Establish a forward distance mapping relationship:
[0039] in, This is a distance correction coefficient used to correct errors caused by non-uniform velocity, inclined interfaces, borehole direction deviation, or geophone geometry. In practical applications, a geometric correction model can also be established based on the advance borehole trajectory, the face normal direction, and the tunnel axis direction to convert the time-domain virtual record into an advance distance-domain record within the range of 0 to 50 m or 0 to 80 m.
[0040] Effective signals in the range domain Perform adaptive mode decomposition:
[0041] in, Indicates the first i The first distance domain signal k One modal component, For the residual term, The number of modes is denoted as . The adaptive mode decomposition can employ empirical mode decomposition, ensemble empirical mode decomposition, variational mode decomposition, or complete ensemble empirical mode decomposition.
[0042] To preserve the effective propagation response associated with the preceding lithological interface, fracture zone, or weak interlayer, a joint mode preservation score is constructed.
[0043] Establish a joint retention score for each modality:
[0044] in, The percentage of energy in the operating frequency band. The correlation coefficient between the mode and the original signal. For kurtosis or transient concentration, The proportion of narrowband line spectrum. This represents the modal entropy value. Retain values that satisfy... ≥ The modes constitute the effective range domain signal:
[0045] right Perform the Hilbert analytic transform:
[0046]
[0047]
[0048]
[0049] in, For instantaneous amplitude, For instantaneous phase, These passive source instantaneous indicators, which are spatial instantaneous frequency or instantaneous wavenumber indices, constitute a passive source instantaneous indicator vector. S(l) This vector is used to characterize seismic response features such as lithological changes ahead of the tunnel face, medium interfaces, wave field attenuation, and changes in rock mass integrity.
[0050] Step 3: Extract features from drilling parameters and cuttings geochemical parameters to obtain drilling response features and geochemical lithology indicator features; First, during the advanced geological drilling process, drilling parameters of the drilling rig or borehole equipment were collected simultaneously. These drilling parameters included drilling speed. v d Propulsion F Drill pipe torque T Rotation speed n Penetration p Effective rock-breaking area of drill bit or cutterhead A One or more of the following: drilling pressure fluctuation, torque fluctuation, energy consumption per unit depth, impact frequency, slag discharge volume, hole depth, and borehole rock powder production.
[0051] The rock machine's operational characteristics are constructed based on drilling parameters. These characteristics are derived from the synchronously acquired drilling speed. v d Propulsion F Drill pipe torque T Rotation speed n Penetration p Effective rock-breaking area of drill bit or cutterhead A Parameters such as drilling pressure fluctuation, torque fluctuation, energy consumption per unit depth, impact frequency, and slag discharge volume are obtained after outlier suppression, sliding window statistics, and standardization.
[0052] Among them, drilling specific energy SE ( l This can be represented as:
[0053] in, The effective rock-breaking area of the drill bit or cutterhead. To prevent stable terms with excessively small denominators, the drilling response vector is further constructed:
[0054] in, This refers to the rock-breaking efficiency index. , These represent the local rates of change of thrust and torque, respectively.
[0055] Specifically, rock breaking efficiency index The drilling efficiency, which characterizes the drilling footage efficiency per unit input energy, can be obtained from the ratio of drilling speed to drilling specific energy.
[0056] The lower the value, the lower the rock breaking efficiency, the more hard rock there is, or the greater the construction resistance.
[0057] Thrust variation rate It is obtained from the first-order difference of the propulsion force along the depth or time, the local slope, or the change of the sliding window.
[0058]
[0059] Torque change rate It is obtained from the first-order difference of torque along depth or time, local slope or sliding window variation.
[0060]
[0061] The thrust variation rate and torque variation rate are used to reflect load fluctuations caused by abrupt changes in lithology, interlayers, fracture zones, or uneven rock breaking at the bottom of the borehole.
[0062] The drilling response vector It is used to characterize the intensity of the interaction between the drill bit and the surrounding rock, and indirectly reflects the rock strength, integrity and drillability.
[0063] Secondly, during the advanced drilling process, borehole cuttings, rock powder, or slag samples are collected simultaneously according to the borehole depth, drilling time, or footage interval. XRF, portable elemental analysis, and mineral spectral identification are performed on the collected samples to obtain major elements, trace elements, oxide content, mineral composition, alteration indices, and argillaceization indices.
[0064] Geochemical parameters include the content of elements such as Si, Al, Ca, Mg, Fe, K, Na, Mn, and Ti, or the content of oxides such as SiO2, Al2O3, CaO, MgO, Fe2O3, K2O, and Na2O. For the first... j Localized Indicators E j ( l Standardization process:
[0065] in, Indicates the first j The average value of localization indicators in the sample database Indicates the first j The standard deviation of localization indicators in the sample database It is a very small, stable term, preventing the denominator from being 0.
[0066] Constructing geochemical lithology indicator vectors:
[0067] in, , , , and These are the silica-alumina ratio, calcium-magnesium ratio, iron-magnesium index, alteration index, and argillaceization index, respectively. This geochemical lithology indicator vector is used to constrain lithological prediction results from the perspective of rock material composition.
[0068] Step 4: Perform time-distance-mileage spatial registration on the instantaneous index features, drilling response features, and geochemical lithology indicator features to form multi-source features; Specifically, instantaneous index characteristics, drilling response characteristics, and geochemical lithology indicator characteristics will be uniformly mapped to the same advance borehole depth. l .
[0069] For each target spatial unit, construct multi-source fusion features:
[0070] in, For the instantaneous index vector of the passive source, The drilling response vector, This is a geochemical lithology indicator vector.
[0071] Step 5: Input the multi-source features into the multi-source fusion lithology prediction model, output the lithology prediction results, and generate engineering interpretation results based on the lithology prediction results; Multi-source fusion samples are input into a pre-trained multi-source attention lithology prediction model. This model includes a passive source instantaneous index encoder, a drilling parameter encoder, a geochemical parameter encoder, a cross-modal attention fusion layer, and a lithology classification layer.
[0072] First, the three data sources are encoded as follows:
[0073] in, , and These are the passive source instantaneous index encoder, the drilling parameter encoder, and the geochemical parameter encoder, respectively.
[0074] Cross-modal fusion weights are generated based on the data quality, spatial distance, missing rate, and feature uncertainty of each data source:
[0075] in, This indicates different data sources or different modalities. For passive source instantaneous indicators; These are the drilling parameters; g For geodesic parameters; Indicates the first u The credibility score or fusion score of a data source. It can be determined by factors such as the data quality, spatial distance, missing rate, and feature uncertainty of the data source.
[0076] exp( e u This indicates an exponential transformation of the credibility score. This can amplify the reliability differences between different data sources. This represents the summation of index scores across all data sources, used for normalization. α m Indicates the first m The final fusion weight of the data source class. α s + α d + α g =1 The multi-source fusion feature obtained by fusion is represented as follows:
[0077] The final output is the probability of different lithology categories:
[0078] in, y Indicates the lithology category predicted by the model; c This indicates a specific category, such as granite, sandstone, etc. p ( y = c |l) indicates the position l At this point, the model determines that it belongs to category c The probability of; W o This represents the weight matrix of the output layer, which is responsible for fusing the features. h ( l Mapped to a classification space for different lithological categories; b o Indicates the bias term of the output layer; This indicates that the scores for all categories are normalized, and the i-th score is taken. c The probability value of the class.
[0079] The lithological prediction results are as follows:
[0080] Meanwhile, the uncertainty of lithological prediction is characterized by lithological probability entropy:
[0081] in, C This represents the total number of lithological categories. U ( l The larger the value, the higher the uncertainty in lithology determination.
[0082] Furthermore, based on the lithological prediction results, lithological distribution maps, lithological probability profiles, and uncertainty profiles are generated within a range of 0 to 50 m or 0 to 80 m ahead of the tunnel face. For each advance distance unit, the following results are output: lithological category, lithological overview, location of lithological changes, possible location of weak interlayers or fracture zones, water-rich or alteration risk warnings, prediction uncertainty, and recommended construction measures.
[0083] When the prediction results indicate the presence of weak rock layers, fault fracture zones, water-rich zones, or abrupt changes in rock mass integrity ahead, and the uncertainty is low, the construction unit can be advised to adjust blasting parameters in advance, shorten cycle depth, strengthen advanced support, implement advanced grouting, or add supplementary boreholes. When the lithology probability distribution is discrete and the uncertainty is high, it can be advised to increase advanced drilling sampling, extend passive source acquisition time, or increase geochemical detection density. Therefore, this embodiment can fully utilize drilling passive source vibration records, drilling parameters, and rock powder geochemical parameters during the advanced drilling process of drill-and-blast tunnels to achieve continuous, intelligent, and advanced prediction of lithology ahead of the tunnel face.
[0084] Example 2 One embodiment of this disclosure provides a lithology prediction system that integrates seismic data during drilling and drilling geochemistry, comprising: The data acquisition module is used to acquire passive source seismic records, drilling parameters, and cuttings geochemical parameters during the tunnel advanced horizontal drilling process. The passive source interferometric reconstruction module is used to perform interferometric reconstruction of passive source seismic records; The instantaneous index extraction module is used to extract instantaneous attributes from passive source seismic records to obtain instantaneous index features related to the response of the preceding medium. The drilling parameter processing module is used to extract features from drilling parameters to obtain drilling response features; The geochemical parameter processing module is used to extract features from the geochemical parameters of rock cuttings to obtain geochemical lithology indicator features; The spatial registration module is used to perform time-distance-mileage spatial registration of instantaneous index features, drilling response features, and geochemical lithology indicator features to form multi-source features. The multi-source fusion lithology prediction module is used to input multi-source features into the multi-source fusion lithology prediction model and output lithology prediction results. The results output module is used to generate engineering interpretation results based on lithology prediction results; The multi-source fusion lithology prediction model includes a passive source instantaneous index encoder, a drilling parameter encoder, and a geochemical parameter encoder, which encode the three feature data sources respectively. Based on the data quality, spatial distance, missing rate, and feature uncertainty of each data source, cross-modal fusion weights are generated to generate multi-source fusion features, and finally output lithology category, lithology probability, and uncertainty.
[0085] Example 3 One embodiment of this disclosure provides a computer program product, including a computer program that, when executed by a processor, implements the lithology prediction method integrating seismic data during drilling and drilling geochemistry.
[0086] Example 4 One embodiment of this disclosure provides a non-transitory computer-readable storage medium for storing computer instructions. When these computer instructions are executed by a processor, they implement the aforementioned lithology prediction method that integrates seismic data during drilling and drilling geochemistry.
[0087] Example 5 One embodiment of this disclosure provides an electronic device, including a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to perform a lithological prediction method that integrates seismic data while drilling and drilling geochemistry.
[0088] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0089] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0090] While the specific embodiments of this disclosure have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of this disclosure. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of this disclosure are still within the scope of protection of this disclosure.
Claims
1. A lithology prediction method integrating seismic data during drilling and drilling geochemistry, characterized in that, include: Obtain passive source seismic records, drilling parameters, and cuttings geochemical parameters during the advanced horizontal drilling process of the tunnel; Interferometric reconstruction and instantaneous attribute extraction were performed on passive source seismic records to obtain instantaneous index features related to the response of the preceding medium; Feature extraction was performed on drilling parameters and cuttings geochemical parameters to obtain drilling response features and geochemical lithology indicator features. Instantaneous index features, drilling response features, and geochemical and lithological indicator features are spatially registered in time-distance-mileage to form multi-source features; Multi-source features are input into the multi-source fusion lithology prediction model, and lithology prediction results are output. Engineering interpretation results are then generated based on the lithology prediction results. The multi-source fusion lithology prediction model includes a passive source instantaneous index encoder, a drilling parameter encoder, and a geochemical parameter encoder, which encode the three feature data sources respectively. Based on the data quality, spatial distance, missing rate, and feature uncertainty of each data source, cross-modal fusion weights are generated to generate multi-source fusion features, and finally output lithology category, lithology probability, and uncertainty.
2. The lithology prediction method integrating seismic data during drilling and drilling geochemistry as described in claim 1, characterized in that, The process of interferometric reconstruction and transient attribute extraction of passive source seismic records to obtain transient index features related to the response of the preceding medium includes: Several geophones are deployed along the tunnel sidewall, near the working face, in boreholes, or on construction equipment to obtain multiple continuous records. The multiple continuous records are then preprocessed by removing the mean, removing the trend, normalizing the amplitude, and bandpass filtering. Select a reference trace or virtual source trace, transform the preprocessed passive source seismic record to the frequency domain, and perform an inverse Fourier transform to obtain the virtual source record; By estimating the arrival time of the direct wave or the wave velocity of the known surrounding rock, the virtual source record is mapped from the time domain to the distance domain. Through adaptive mode decomposition, the instantaneous amplitude and instantaneous frequency indices related to the response of the preceding medium are extracted.
3. The lithology prediction method integrating seismic data during drilling and drilling geochemistry as described in claim 1, characterized in that, The process involves extracting features from drilling parameters and cuttings geochemical parameters to obtain rock-machine action characteristics and geochemical lithology indication characteristics, including: The drilling parameters include drilling speed, thrust, torque, rotational speed, penetration depth, energy consumption per unit depth, drilling pressure fluctuation, torque fluctuation, and slag discharge. The drilling parameters are converted into rock machine action characteristics, and the drilling response vector is constructed using the rock machine action characteristics.
4. The lithology prediction method integrating seismic data during drilling and drilling geochemistry as described in claim 1, characterized in that, The rock cuttings geochemical parameters include major elements, trace elements, mineral composition, alteration index, argillaceization index, carbonate indicator index, and water-bearing reaction index in borehole cuttings, face rock powder, slag discharge samples, or borehole samples. Geochemical lithological indicator features are constructed after standardizing the geochemical indicators.
5. The lithology prediction method integrating seismic data during drilling and drilling geochemistry as described in claim 1, characterized in that, The instantaneous index features, drilling response features, and geochemical lithology indicator features are spatially registered in time-distance-mileage to form multi-source features, including: By mapping the instantaneous index features of passive sources, drilling response features, and geochemical lithology indicator features to the same advanced borehole depth, multi-source features are constructed for the spatial unit to be predicted.
6. The lithology prediction method integrating seismic data during drilling and drilling geochemistry as described in claim 1, characterized in that, The process of inputting multi-source features into a multi-source fusion lithology prediction model, outputting lithology prediction results, and generating engineering interpretation results based on the lithology prediction results includes: The multi-source fusion lithology prediction model includes a passive source instantaneous index encoder, a drilling parameter encoder, a geochemical parameter encoder, a spatial location encoder, a cross-modal attention fusion layer, and a lithology classification layer; firstly, the three data sources are encoded as follows: in, , and These are, respectively, a passive source instantaneous index encoder, a drilling parameter encoder, and a geochemical parameter encoder; This refers to the instantaneous index characteristics of passive sources; Drilling response characteristics; These are geochemical and lithological indicators; Cross-modal fusion weights are generated based on the data quality, spatial distance, missing rate, and feature uncertainty of each data source, and multi-source fusion features are obtained. Based on the lithology prediction results, a lithology distribution map, a lithology probability profile, and an uncertainty profile are generated within a set distance range in front of the tunnel face.
7. A lithology prediction system integrating seismic data during drilling and drilling geochemistry, characterized in that, include: The data acquisition module is used to acquire passive source seismic records, drilling parameters, and cuttings geochemical parameters during the tunnel advanced horizontal drilling process. The passive source interferometric reconstruction module is used to perform interferometric reconstruction of passive source seismic records; The instantaneous index extraction module is used to extract instantaneous attributes from passive source seismic records to obtain instantaneous index features related to the response of the preceding medium. The drilling parameter processing module is used to extract features from drilling parameters to obtain drilling response features; The geochemical parameter processing module is used to extract features from the geochemical parameters of rock cuttings to obtain geochemical lithology indicator features; The spatial registration module is used to perform time-distance-mileage spatial registration of instantaneous index features, drilling response features, and geochemical lithology indicator features to form multi-source features. The multi-source fusion lithology prediction module is used to input multi-source features into the multi-source fusion lithology prediction model and output lithology prediction results. The results output module is used to generate engineering interpretation results based on lithology prediction results; The multi-source fusion lithology prediction model includes a passive source instantaneous index encoder, a drilling parameter encoder, and a geochemical parameter encoder, which encode the three feature data sources respectively. Based on the data quality, spatial distance, missing rate, and feature uncertainty of each data source, cross-modal fusion weights are generated to generate multi-source fusion features, and finally output lithology category, lithology probability, and uncertainty.
8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the lithology prediction method that integrates seismic data during drilling and drilling geochemistry as described in any one of claims 1-6.
9. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium is used to store computer instructions, which, when executed by a processor, implement a lithology prediction method that integrates seismic data during drilling and drilling geochemistry as described in any one of claims 1-6.
10. An electronic device, characterized in that, include: The device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to perform a lithological prediction method integrating seismic data while drilling and drilling geochemistry as described in any one of claims 1-6.