Tunnel geological exploration hole while drilling parameter space-time characteristic modeling method and system

Abstract

This invention discloses a method and system for modeling the spatiotemporal characteristics of drilling parameters in tunnel geological exploration, belonging to the field of tunnel exploration technology. Specifically, it includes: continuously acquiring instantaneous torsional angles to generate torsional fluctuation curves; extracting fluctuation curve segments at fixed drilling depths, identifying peaks, and calculating the time intervals between adjacent peaks to form a time interval set; extracting candidate periods and calculating the axial position of reflection points based on the torsional wave propagation velocity; performing exponential decay fitting on the peak amplitude sequence to obtain the decay ratio, and calculating the torque consumption equivalent based on the average torque; distributing the torque consumption equivalent of local peak points to the borehole cross-sectional orientation according to the rotation phase, and superimposing them along the depth to generate a three-dimensional spatiotemporal model; identifying potential borehole collapse sections based on the time difference between the cross-sectional mean and the drill bit depth, and comparing the torque rise depth with the potential borehole collapse section to distinguish between abnormal drill pipe wall friction and abnormal changes in surrounding rock strength, providing borehole wall contact state information for advanced tunnel geological prediction and eliminating false hard rock misjudgments.

Description

Technical Field

[0001] This invention relates to the field of tunnel exploration technology, specifically to a method and system for modeling the spatiotemporal characteristics of drilling parameters in tunnel geological exploration boreholes. Background Technology

[0002] In tunnel engineering exploration, advanced geological borehole parameter monitoring is a commonly used technique. This technique involves installing sensors on the drilling rig to record parameters such as rotational pressure, propulsion pressure, impact pressure, drill bit speed, and drilling velocity in real time during the drilling process. Based on the empirical relationship between these parameters and rock mechanical properties, geological indicators such as the strength, integrity, and abrasion resistance of the rock mass along the borehole are inverted. However, existing monitoring systems typically treat the drilling parameters as the direct mechanical response of the drill bit breaking the rock. Under the assumption of no or negligible energy loss between the drill pipe and the borehole wall, the measured torque and propulsion force are entirely attributed to the interaction between the drill bit and the rock mass, thus establishing a correspondence model between parameters and rock mass quality. Based on this, continuous curves of each parameter are plotted along the borehole depth direction, and the changes in curve shape are used to determine lithological boundaries, fracture zone locations, and surrounding rock grades.

[0003] However, during actual drilling, the drill pipe is not perfectly centered in the borehole. Factors such as localized protrusions on the borehole wall, collapsed debris, rock powder accumulation, and the natural curvature of the borehole trajectory all lead to contact and friction between the outer wall of the drill pipe and the borehole wall. This contact friction consumes some rotational and propulsive energy, resulting in the inclusion of a drill pipe-borehole friction component in the total torque and total propulsion force recorded by the sensor. In actual data processing, either the two are not distinguished, and all torque values ​​are directly used for rock mass inversion, or low-pass filtering or moving average methods are used to smooth the parameter curves in an attempt to eliminate so-called "high-frequency noise."

[0004] Therefore, in practice, it is difficult to identify the specific depth range where drill pipe-hole wall friction occurs, and the magnitude of frictional energy cannot be quantified. Due to the presence of frictional interference, the torque curve, which originally reflects changes in rock hardness, will be distorted, manifesting as abnormal uplift in local contact sections of the hole wall. This uplift may be incorrectly attributed to increased rock strength, resulting in a misjudgment of "false hard rock." Furthermore, since the contact points between the drill pipe and the hole wall are not uniformly distributed throughout the entire drilling depth but concentrated in specific axial intervals, traditional one-dimensional parametric curves present frictional information from these intervals mixed with drill bit information, making it impossible to distinguish the contributions of each. Consequently, it is difficult to accurately determine the true mechanical state of the surrounding rock. Furthermore, as the drill rod rotates at a certain speed inside the borehole, when there are differences in the borehole wall in different circumferential directions (such as the fracture surface appearing only on one side), the friction will change periodically with the rotation phase angle. However, the existing monitoring system only records the time average or peak value of the torque, losing the characteristic information related to the rotation direction. Therefore, the established parameter model can only describe the characteristics of the surrounding rock along the axial direction and cannot distinguish the differences in the rock mass in different directions around the borehole. As a result, even if there are geological anomalies such as local bias pressure, asymmetric fracture zone or unilateral weak interlayer in front of the borehole that seriously threaten the safety of tunnel construction, the system cannot identify their specific location. Due to the lack of directional information, the construction party can only take uniform support measures across the entire cross section, resulting in waste of support materials or insufficient reinforcement in key directions. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for modeling the spatiotemporal characteristics of drilling parameters in tunnel geological exploration boreholes, and to solve the following technical problems: Existing monitoring-while-drilling technology is limited to the overall amplitude analysis of the torque signal at the top of the drill pipe, attributing the abnormal rise in torque to changes in the strength of the surrounding rock. It fails to separate the torque response generated by the friction between the drill pipe and the borehole wall from the torque response generated by the drill bit breaking the rock. This results in the identification results lacking information on borehole stability, making it difficult to meet the information requirements of tunnel advance prediction for accurately distinguishing between borehole collapse risk and surrounding rock strength.

[0006] The objective of this invention can be achieved through the following technical solutions: A method for modeling the spatiotemporal characteristics of drilling parameters in tunnel geological exploration boreholes includes the following steps: S1, a torsional wave sensor is pre-installed on the top of the drill pipe to continuously collect the instantaneous torsional angle of the drill pipe surface and record it in chronological order to establish a torsional wave curve; S2, take each fixed depth drilled by the drill bit as a drilling segment, extract the wave curve segment corresponding to the drilling segment from the torsional wave curve, identify the wave peaks and calculate the time interval between adjacent wave peaks to form a time interval set. S3, extract the interval value with the largest energy accumulation from the time interval set as the characteristic period, determine the torsional wave reflection position that generates the characteristic period as the reflection point, calculate the distance from the reflection point to the torsional wave sensor based on the characteristic period, and determine the axial position of the reflection point as the distance; S4. Based on the amplitude of each wave peak corresponding to the characteristic period, perform exponential decay fitting to obtain the decay ratio. At the same time, obtain the average drill rod torque corresponding to the drilling section. Based on the average torque and the decay ratio, calculate the torque consumption equivalent at the reflection point. S5. Establish a friction feature depth curve with the axial position of the reflection point as the abscissa and the torque consumption equivalent as the ordinate. Determine the local peak points on the curve and obtain the drill pipe rotation phase corresponding to each local peak point. Distribute the torque consumption equivalent of each local peak point to the borehole section orientation according to the rotation phase. S6 superimposes the orientation of the torque consumption equivalent at different depths and rotation phases on the borehole cross section to generate a three-dimensional spatiotemporal feature model of drilling parameters with the borehole depth as the vertical axis and the circumferential direction of the borehole cross section as the horizontal axis.

[0007] As a further aspect of the present invention: the specific construction process of the torsional fluctuation curve in S1 is as follows: A torsional wave sensor is pre-installed on the top of the drill rod. The torsional wave sensor rotates synchronously with the drill bit and continuously collects the instantaneous torsional angle value of the drill rod surface at a preset sampling frequency during the drilling process. The instantaneous torsional angle value corresponding to each sampling moment is recorded in chronological order to form a torsional wave curve with time as the horizontal axis and instantaneous torsional angle value as the vertical axis.

[0008] As a further aspect of the present invention: in step S3, the specific process for determining the reflection point is as follows: For each time interval in the time interval set, starting from each peak in the wave curve segment corresponding to the drilling section, the peaks with adjacent peak time intervals equal to the time interval are selected along the time axis in the positive direction. The amplitude values ​​of the peaks selected from each starting point are squared and summed, and the maximum value is taken as the energy accumulation value corresponding to the time interval, forming a set of energy accumulation values. The set of energy accumulation values ​​is subjected to binary clustering. The set of energy accumulation values ​​is divided into high-energy class and low-energy class based on the maximum inter-class variance. The minimum energy accumulation value in the high-energy class is taken as the adaptive energy threshold. The time interval in which the energy accumulation value exceeds the adaptive energy threshold is determined as the candidate period. Based on the candidate period and the pre-calibrated propagation speed of the torsional wave in the drill pipe, the distance from the reflection point to the torsional wave sensor corresponding to each candidate period is calculated. Reflection points whose distance is greater than the current drilling depth are removed. The remaining reflection points are determined as reflection points within the drilling section, and the corresponding distance of each remaining reflection point is determined as the axial position of the reflection point.

[0009] As a further aspect of the present invention: the specific calculation process for the attenuation ratio in step S4 is as follows: Extract the peak sequence of the characteristic period from the wave curve segment corresponding to the drilling section, and take the amplitude value of each peak in turn to form an amplitude sequence; establish an exponential decay model for the amplitude sequence, use the least squares method to fit the amplitude sequence, determine the decay constant by finding the minimum value of the sum of squared residuals, and determine the decay ratio by taking the inverse power of the natural constant.

[0010] As a further aspect of the present invention: the specific calculation process for the torque consumption equivalent in step S4 is as follows: Obtain the torque value sequence of the drill rod torque changing over time, synchronously recorded by the drill rig torque monitoring system corresponding to the drilling section. The torque value sequence corresponds one-to-one with the fluctuation curve segment in time. Perform an arithmetic mean on all torque values ​​in the torque value sequence, that is, sum the torque values ​​in the torque value sequence and divide by the total number of torque values ​​in the torque value sequence to obtain the average torque value corresponding to the drilling section. Multiply the attenuation ratio by the average torque value, and determine the resulting product as the torque consumption equivalent of the drill rod by the surrounding rock at the reflection point.

[0011] As a further aspect of the present invention: In step S5, the specific process of distributing the torque consumption equivalent of each local peak point to the borehole cross-section orientation according to the rotation phase is as follows: On the friction feature depth curve, the sampling point where the torque consumption equivalent is greater than the torque consumption equivalent of the two adjacent depth sampling points is determined as a local peak point, and the axial position of the reflection point and the torque consumption equivalent corresponding to each local peak point are recorded. The drilling rig monitoring system records the drill rod rotation phase at the sampling time of each local peak point, and establishes a one-to-one correspondence between each local peak point and the drill rod rotation phase. The drill rod rotation phase recorded before drilling begins is obtained as the initial phase, and the geographical orientation of the initial phase measured before drilling begins on the borehole cross section is obtained as the reference orientation. For each local peak point, the drill rod rotation phase corresponding to the local peak point is subtracted from the initial phase to obtain the phase offset of the local peak point relative to the initial phase. Add the phase offset to the reference azimuth to obtain the geographical azimuth of the local peak point on the borehole cross section; take the center of the borehole cross section as the origin and the magnetic north direction as the zero-degree reference direction, and divide the circumference of the borehole cross section into a preset number of azimuth intervals in a clockwise direction. For each local peak point, the azimuth range that the geographical azimuth points to on the borehole section is determined according to the geographical azimuth of the local peak point. The torque consumption equivalent of the local peak point is allocated to the determined azimuth range, and the axial position of the reflection point of the local peak point is recorded as the depth position corresponding to the torque consumption equivalent within the azimuth range.

[0012] As a further aspect of the present invention: in step S6, the specific generation process of the three-dimensional drilling parameter spatiotemporal characteristic model is as follows: A two-dimensional grid coordinate system is established with the axial position of the reflection point as the vertical coordinate axis and the directional intervals divided by the circumferential direction of the borehole section as the horizontal coordinate axis. The unit length of the vertical coordinate axis corresponds to the axial position of a reflection point, and the unit length of the horizontal coordinate axis corresponds to a directional interval. Fill the torque consumption equivalent of each local peak point, which has been allocated to each azimuth interval, into the cell in the two-dimensional grid coordinate system corresponding to the axial position and azimuth interval of the reflection point; The filled two-dimensional meshes corresponding to the axial positions of different reflection points are stacked sequentially along the vertical coordinate axis to obtain a three-dimensional spatiotemporal characteristic model of drilling parameters.

[0013] As a further aspect of the present invention, it also includes extracting the cross-sectional torque consumption equivalent distribution corresponding to the axial position of each reflection point along the vertical coordinate axis from the three-dimensional drilling parameter spatiotemporal feature model, calculating the cross-sectional mean value of the torque consumption equivalent at the axial position of each reflection point, and forming a sequence of cross-sectional mean values ​​changing with the axial position of the reflection point. The cross-sectional mean value sequence is time-aligned with the drill bit depth change sequence recorded during drilling. For each reflection point axial position, the time difference between the moment when the cross-sectional mean value first appears at that reflection point axial position and the moment when the drill bit first arrives at that reflection point axial position is determined. The depth intervals that are continuously distributed along the vertical coordinate axis among the axial positions of reflection points with a time difference greater than zero are defined as potential hole collapse sections.

[0014] As a further aspect of the present invention: S6 further includes generating a torque versus depth curve from the drill rod torque value sequence synchronously recorded by the drilling rig torque monitoring system; comparing the depth positions where torque increases appear on the torque versus depth curve with potential hole collapse sections, determining the depth positions falling within the potential hole collapse section range as torque anomalies caused by friction between the drill rod and the hole wall, and determining the depth positions not falling within the potential hole collapse section range as torque anomalies including contributions from changes in surrounding rock strength.

[0015] This invention also includes a system for modeling the spatiotemporal characteristics of drilling parameters inside tunnel geological exploration boreholes, used to implement the above-mentioned method for modeling the spatiotemporal characteristics of drilling parameters inside tunnel geological exploration boreholes, comprising: The parameter acquisition module is used to pre-install a torsional wave sensor on the top of the drill pipe, continuously collect the instantaneous torsional angle of the drill pipe surface and record it in time sequence to establish a torsional wave curve; The interval acquisition module is used to take a fixed depth of drilling by the drill bit as a drilling segment, extract the wave curve segment corresponding to the drilling segment from the torsional wave curve, identify the wave peaks and calculate the time interval between adjacent wave peaks to form a time interval set. The reflection recognition module is used to extract the interval value with the maximum energy accumulation from the time interval set as the feature period, determine the torsional wave reflection position that generates the feature period as the reflection point, calculate the distance from the reflection point to the torsional wave sensor based on the feature period, and determine the axial position of the reflection point based on the distance. The torque calculation module is used to perform exponential decay fitting based on the amplitude of each peak corresponding to the characteristic period to obtain the decay ratio, and at the same time obtain the average drill rod torque corresponding to the drilling section. Based on the average torque and the decay ratio, the torque consumption equivalent at the reflection point is calculated. The position allocation module is used to establish a friction feature depth curve with the axial position of the reflection point as the abscissa and the torque consumption equivalent as the ordinate, determine the local peak points on the curve and obtain the drill pipe rotation phase corresponding to each local peak point, and allocate the torque consumption equivalent of each local peak point to the borehole section orientation according to the rotation phase. The results generation module is used to superimpose the orientation of torque consumption equivalents at different depths and rotation phases on the borehole cross section, and generate a three-dimensional spatiotemporal feature model of drilling parameters with the borehole depth as the vertical axis and the circumferential direction of the borehole cross section as the horizontal axis.

[0016] The beneficial effects of this invention are: 1) This invention establishes a torsional waveguide monitoring mechanism at the top of the drill rod, constructing a three-dimensional spatiotemporal characteristic model of friction that integrates depth and orientation, achieving a fundamental leap from a one-dimensional torque curve to a three-dimensional hole wall contact strength distribution map. It can be understood that the drill rod can be considered a one-dimensional elastic waveguide. When the drill bit rotates and drills, the local contact points between the drill rod and the hole wall form discontinuities in wave impedance. The incident torsional wave is reflected at these points and transmitted back to the top sensor. The signal received by the sensor is the result of multiple reflections superimposed. Since any measuring point on the drill rod surface periodically sweeps across various circumferential orientations of the borehole cross-section with rotation, when the hole wall has a protrusion, breakage, or bias at a specific orientation, the measuring point on the drill rod encounters stronger frictional resistance when rotating to that orientation, resulting in an instantaneous step change in the torsional angle, which manifests as a peak locked to that orientation on the torsional wave curve. This invention extracts the drill rod rotation phase corresponding to each local peak point and, through the mapping relationship between the initial phase calibrated before drilling and the magnetic north reference azimuth, converts the relative rotation phase into the absolute geographical azimuth on the borehole cross-section, thereby accurately distributing the torque consumption equivalent to the corresponding azimuth interval. The resulting three-dimensional model, with the borehole depth as the vertical axis and the circumferential direction of the borehole cross-section as the horizontal axis, intuitively presents the combined distribution characteristics of borehole wall friction intensity in depth and direction. It can directly locate the main friction direction within the same depth cross-section and automatically divide the depth boundary between intact and broken surrounding rock along the depth direction, providing a geometric decision-making basis for differentiated support such as asymmetric anchor reinforcement and directional grouting.

[0017] 2) This invention utilizes the causal relationship between the propagation timing of torsional elastic waves in the drill pipe waveguide and the drill bit leading event to achieve adaptive identification of potential borehole collapse sections, shifting the identification basis from empirical thresholds to physical causal logic. It can be understood that the torsional wave propagates in the drill pipe at a calibrated wave speed, and its round-trip time precisely corresponds to the distance between the sensor and the reflection point. Therefore, the moment the reflected signal appears on the time axis directly marks the physical moment of the friction event. During drilling, the drill bit is the only leading event that actively breaks the rock and advances along the depth; the frictional contact between the borehole wall and the drill pipe is the subsequent response. If the borehole wall remains stable as the drill bit passes through, the frictional contact is synchronized with the drilling, and the first occurrence of the cross-sectional mean value is basically consistent with the arrival time of the drill bit. If the borehole wall becomes unstable due to the disturbance of the drill bit, the collapsed fragments need a certain amount of time to accumulate and wrap around the drill pipe, and the first occurrence of the cross-sectional mean value will be significantly later than the arrival time of the drill bit. Therefore, this identification logic is entirely based on the physical timing of waveguide propagation and is not affected by external factors such as lithology, borehole diameter, and mud concentration. It has a natural self-adaptive capability and can automatically calibrate the precise range of borehole wall instability under complex and variable geological conditions.

[0018] 3) This invention fundamentally eliminates the misjudgment of "false hard rock" caused by borehole wall friction by establishing the attribution determination of abnormal torque increases on the dual basis of the physical source of torsional wave reflection and the irreversibility of causal timing. It is understandable that the reflected signal received by the torsional wave sensor originates from the friction or collision of the drill rod with the local contact point within the drilled borehole wall section, which is spatially and physically independent of the drill bit's rock-breaking process. The increase in surrounding rock strength is a mechanical response generated the instant the drill bit contacts the hard rock; this response occurs synchronously with the drill bit's arrival and does not trigger torsional wave reflection in the drilled borehole wall section, nor does it produce a time lag. In the potential collapsed borehole section, as verified by timing, the friction signal has been confirmed to occur after the drill bit has passed; its physical cause can only be drill rod-borehole wall friction caused by the delayed collapse of the borehole wall. This invention compares the depth locations where abnormal rises occur on the torque versus depth curve with potential borehole collapse sections. When the rise falls within the range of a potential collapse section, it provides dual proof—both from the signal source and temporal causality—that the torque anomaly occurred after the drill bit passed through and is unrelated to changes in surrounding rock strength. Therefore, it is determined to be a torque anomaly caused by friction between the drill pipe and the borehole wall. This judgment logic severs the inherent erroneous correspondence between torque increase and hard rock in traditional methods, ensuring that only the torque response truly caused by changes in surrounding rock strength is used for rock mechanics parameter inversion, significantly improving the accuracy of tunnel advanced geological prediction. Attached Figure Description

[0019] The invention will now be further described with reference to the accompanying drawings.

[0020] Figure 1 This is a schematic diagram of the method for modeling the spatiotemporal characteristics of drilling parameters in tunnel geological exploration boreholes according to the present invention; Figure 2 This is a schematic diagram of the spatiotemporal characteristic modeling system for in-hole drilling parameters in tunnel geological exploration according to the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Please see Figure 1 As shown, this invention provides a method for modeling the spatiotemporal characteristics of drilling parameters inside a tunnel geological exploration borehole, comprising the following steps: S1, a torsional wave sensor is pre-installed on the top of the drill pipe to continuously collect the instantaneous torsional angle of the drill pipe surface and record it in chronological order to establish a torsional wave curve; A torsional wave sensor is installed at the top of the drill rod. The sensor rotates synchronously with the drill bit, continuously collecting instantaneous torsional angle values ​​on the drill rod surface at a preset sampling frequency during drilling. The instantaneous torsional angle values ​​at each sampling moment are recorded sequentially over time, forming a torsional wave curve with time on the horizontal axis and the instantaneous torsional angle values ​​on the vertical axis. It can be understood that when the drill rod rotates and drills into the hole, if there are local protrusions on the hole wall, collapses of broken sections, or wall contact caused by borehole deviation, friction, scraping, or stick-slip will occur between the outer wall of the drill rod and the hole wall. These contact events apply transient torsional resistance to the drill rod, exciting torsional elastic waves at the contact points. The drill rod itself can be considered a one-dimensional elastic waveguide, with the torsional wave propagating along both ends of the drill rod, the upward propagating portion being captured by the top torsional wave sensor. Since the contact between the drill rod and the hole wall is not continuous and stable, but occurs in the form of stick-slip oscillations or intermittent collisions, each contact event corresponds to a peak on the torsional wave curve. Therefore, each peak of the torsional ripple curve essentially corresponds to an abnormal contact event between the drill pipe and the borehole wall. The amplitude of the peak reflects the intensity of the contact, and the time position of the peak marks the moment when the contact occurred.

[0023] S2, take each fixed depth drilled by the drill bit as a drilling segment, extract the wave curve segment corresponding to the drilling segment from the torsional wave curve, identify the wave peaks and calculate the time interval between adjacent wave peaks to form a time interval set. After each fixed depth, a segment of the torsional undulation curve corresponding to that drilling process is extracted from the torsional undulation curve. For each extracted segment, the positions of all peaks are identified along the time axis, and the amplitude value corresponding to each peak is recorded. The time interval between two adjacent peaks is calculated sequentially along the time axis, and all calculated time intervals are collected into a time interval set for that segment. It can be understood that within a certain short drilling distance, the contact point between the drill pipe and the borehole wall is relatively fixed. If there is a significant contact point with the borehole wall within this segment, the torsional wave propagates back and forth between this contact point and the top sensor, forming a peak on the torsional undulation curve each time it reflects back to the top. The time interval between two adjacent reflections is equal to the time it takes for the torsional wave to travel back and forth between the sensor and the contact point once. If multiple contact points exist, multiple peak sequences with different time intervals will simultaneously exist in this segment of the curve. The time interval set is a candidate set of all potential reflection cycles within this drilling distance, providing a complete candidate space for the next step of screening for true reflection points.

[0024] S3, extract the interval value with the largest energy accumulation from the time interval set as the characteristic period, determine the torsional wave reflection position that generates the characteristic period as the reflection point, calculate the distance from the reflection point to the torsional wave sensor based on the characteristic period, and determine the axial position of the reflection point as the distance; From the fluctuation curve, all peaks appearing at equal intervals are identified. The amplitude values ​​of the extracted peaks are squared and summed to obtain the cumulative energy value corresponding to that time interval, forming a set of cumulative energy values. It can be understood that if a certain time interval corresponds to a true reflection point, then the peaks appearing at equal intervals belong to the same reflection sequence; although the peak amplitude values ​​decrease, the overall energy is concentrated. If a certain time interval is random noise or a minor reflection, then the number of peaks appearing at equal intervals is small, the amplitude is small, and the cumulative energy value is low. By summing the squared amplitudes, the true reflection period and the noise period are significantly separated in the energy domain. Binary clustering is performed on all cumulative energy values ​​in the set, using the maximum inter-class variance as the dividing criterion to divide the set into a high-energy class and a low-energy class. The minimum cumulative energy value in the high-energy class is taken as the adaptive energy threshold. It can be understood that the cumulative energy value corresponding to the true reflection period is significantly higher than the cumulative energy value of the noise period due to energy concentration, naturally forming two clusters in numerical distribution. The maximum inter-class variance criterion automatically finds the optimal split point between two clusters. This split point is entirely determined by the distribution characteristics of the data itself, without the need to manually set any fixed thresholds related to lithology or pore size, and has natural adaptability to geological conditions.

[0025] Time intervals where the accumulated energy value exceeds the adaptive energy threshold are all identified as candidate periods. For each candidate period, the pre-calibrated propagation speed of the torsional wave in the drill pipe is obtained. This propagation speed is multiplied by the candidate period to obtain the round-trip distance of the torsional wave between the sensor and the hole wall contact point. Half of this propagation distance is taken as the distance from the reflection point to the torsional wave sensor corresponding to that candidate period. Each calculated distance is compared with the current drilling depth. Reflection points with distances greater than the current drilling depth are eliminated, and all remaining reflection points are determined as reflection points within the drilling distance interval. The distance corresponding to each reflection point is determined as the axial position of each reflection point. It is understood that reflection points with distances greater than the current drilling depth are physically impossible because torsional waves cannot be reflected from un-drilled locations. This elimination operation is a constraint of physical causality, further eliminating spurious periods in the calculation.

[0026] S4. Based on the amplitude of each wave peak corresponding to the characteristic period, perform exponential decay fitting to obtain the decay ratio. At the same time, obtain the average drill rod torque corresponding to the drilling section. Based on the average torque and the decay ratio, calculate the torque consumption equivalent at the reflection point. All peaks belonging to a candidate period are extracted from the wave curve and arranged in chronological order to form a peak sequence. The amplitude values ​​of each peak are then taken to form an amplitude sequence. It is understandable that each time a torsional wave passes a reflection point, some energy is consumed due to frictional contact, causing the amplitude of the reflected wave to gradually decrease with the number of reflections. The stronger the friction at the reflection point, the greater the energy loss per reflection, and the faster the amplitude decays. An exponential decay model is established for the amplitude sequence, and the least squares method is used to fit the amplitude sequence. The attenuation constant is determined by finding the minimum value of the sum of squared residuals. It is understandable that in a uniform waveguide, the attenuation of the torsional wave amplitude follows an exponential law, and the attenuation constant directly reflects the intensity of energy dissipation at the reflection point. The exponential decay model accurately describes the physical attenuation law of the peak amplitude during multiple reflections. A larger attenuation constant indicates faster energy attenuation and more intense contact friction at the aperture wall. The natural constant is taken as the inverse power of the attenuation constant to determine the attenuation ratio. A smaller attenuation ratio indicates stronger frictional contact at the reflection point.

[0027] Simultaneously, the torque value sequence of the drill rod torque changing over time, synchronously recorded by the drilling rig torque monitoring system during this drilling process, is obtained. The torque value sequence corresponds one-to-one with the fluctuation curve segment in time. The arithmetic mean of all torque values ​​in the torque value sequence is calculated, i.e., the sum of each torque value is divided by the total number of torque values, to obtain the average torque of this drilling process. The average torque represents the total resistance experienced by the drill rod rotation during this drilling process, including the drill bit rock-breaking torque and the sum of the friction torque between various parts of the drill rod and the borehole wall. The attenuation ratio is multiplied by the average torque, and the resulting product is determined as the torque consumption equivalent of the surrounding rock on the drill rod at the reflection point. It can be understood that the attenuation ratio reflects the relative weight of the torsional wave energy loss at the reflection point, and the average torque reflects the absolute level of the total torque. Their product is physically equivalent to an estimate of the torque consumption component allocated to the reflection point within the total torque. The torque consumption equivalent has the same dimensions as the torque and is positively correlated with the dynamic contact interference intensity of the surrounding rock on the drill rod.

[0028] S5. Establish a friction feature depth curve with the axial position of the reflection point as the abscissa and the torque consumption equivalent as the ordinate. Determine the local peak points on the curve and obtain the drill pipe rotation phase corresponding to each local peak point. Distribute the torque consumption equivalent of each local peak point to the borehole section orientation according to the rotation phase. A friction feature depth curve is plotted with the axial position of each reflection point as the x-axis and the torque consumption equivalent corresponding to each reflection point as the y-axis. On this curve, sampling points where the torque consumption equivalent is greater than the torque consumption equivalent of the two adjacent depth sampling points are identified as local peak points. The axial position of the reflection point and the torque consumption equivalent corresponding to each local peak point are recorded. The drill rod rotation phase recorded by the drilling rig monitoring system at the sampling time of each local peak point is obtained, establishing a one-to-one correspondence between each local peak point and the drill rod rotation phase. The drill rod rotation phase recorded before drilling begins is obtained as the initial phase, and the geographical orientation of this initial phase on the borehole cross-section, measured before drilling begins, is obtained as the reference orientation. For each local peak point, the initial phase is subtracted from the drill rod rotation phase corresponding to the local peak point to obtain the phase offset of the local peak point relative to the initial phase. The phase offset is added to the reference orientation to obtain the geographical orientation of the local peak point on the borehole cross-section.

[0029] Using the center of the borehole cross-section as the origin and magnetic north as the zero-degree reference direction, the circumference of the borehole cross-section is divided into a predetermined number of azimuth intervals in a clockwise direction. For each local peak point, the azimuth interval it points to on the borehole cross-section is determined according to the geographical azimuth corresponding to that local peak point. The torque consumption equivalent of that local peak point is allocated to the determined azimuth interval, and the axial position of the reflection point of that local peak point is recorded as the depth position corresponding to that torque consumption equivalent within that azimuth interval. It can be understood that during the rotary drilling process, any measuring point on the surface of the drill pipe periodically sweeps across the various azimuths of the borehole cross-section with rotation. When there is a protrusion, breakage, or bias pressure on the borehole wall at a certain azimuth, the measuring point of the drill pipe encounters stronger frictional resistance when rotating to that azimuth, and the instantaneous torsion angle undergoes a step change, which is represented as a peak locked with that azimuth on the torsion fluctuation curve. Therefore, the rotation phase of the drill pipe at the moment when this peak is generated directly indicates the azimuth position of the friction contact point on the borehole cross-section. The calibration of the initial phase and reference azimuth solves the problem of converting the relative angle of the drill pipe itself to the absolute geographical direction of the rotation phase, enabling friction events recorded at different depths and times to be expressed in a unified borehole cross-section coordinate system.

[0030] S6 superimposes the orientation of the torque consumption equivalent at different depths and rotation phases on the borehole cross section to generate a three-dimensional spatiotemporal feature model of drilling parameters with the borehole depth as the vertical axis and the circumferential direction of the borehole cross section as the horizontal axis.

[0031] A two-dimensional grid coordinate system is established, with the axial position of the reflection point as the vertical coordinate axis and the azimuth intervals divided along the circumference of the borehole section as the horizontal coordinate axes. Each unit length of the vertical coordinate axis corresponds to one axial position of the reflection point, and each unit length of the horizontal coordinate axis corresponds to one azimuth interval. The torque consumption equivalent of each local peak point, already allocated to each azimuth interval, is filled into the cells of the two-dimensional grid coordinate system corresponding to the axial position and azimuth interval of the reflection point. The filled two-dimensional grids corresponding to different axial positions of the reflection points are then stacked sequentially along the vertical axis to obtain a three-dimensional spatiotemporal characteristic model of drilling parameters.

[0032] It is understandable that the vertical axis of this 3D model represents the drilling depth, the horizontal axis represents the circumferential direction of the drilling cross-section, and the values ​​of each cell represent the torque consumption equivalent, intuitively presenting the combined distribution characteristics of borehole wall friction intensity in depth and direction. Traditional drilling technology can only output a one-dimensional curve of torque changing with depth, unable to distinguish whether the torque increase originates from drill bit rock breaking or drill rod friction with the borehole wall, and even less able to obtain the directional information of friction occurrence. This invention, by independently extracting and assigning the torsional wave signal to the cross-sectional orientation, generates a 3D model that simultaneously carries borehole wall contact state information in both depth and orientation dimensions, laying a spatial positioning foundation for achieving the separation of drill bit rock breaking torque and drill rod friction torque.

[0033] It is worth noting that this invention does not simply plot the torque while drilling parameters as a three-dimensional graph. Instead, it first utilizes the propagation and reflection principle of torsional elastic waves in the drill pipe waveguide to locate each abnormal contact event between the drill pipe and the borehole wall as a reflection point with depth and azimuth coordinates. Then, by combining attenuation fitting with the average torque value, the contact intensity of the reflection point is quantified into a torque consumption equivalent with a clear physical meaning. Finally, through cross-sectional azimuth allocation and depth superposition, the discrete reflection points are expanded into a continuous three-dimensional friction feature distribution covering the entire borehole cross-section. This allows the borehole wall friction information, originally hidden in the one-dimensional torque curve, to be peeled away layer by layer and presented independently, providing a data foundation for identifying potential borehole collapse sections and accurately determining the strength of surrounding rock in advanced tunnel geological prediction.

[0034] In a preferred embodiment of the present invention, the specific construction process of the torsional fluctuation curve in step S1 is as follows: A torsional wave sensor is installed on the outer wall of the cylindrical section near the power head at the top of the drill pipe. The torsional wave sensor uses resistance strain gauges, which are attached to the drill pipe surface at a 45-degree angle to the drill pipe axis. The physical basis for this 45-degree attachment angle is that when the drill pipe is subjected to torsion, the maximum normal stress occurs on the inclined section at a 45-degree angle to the axis. Aligning the strain gauge grid length with the direction of the maximum principal strain allows for the acquisition of the strongest response signal to torsional deformation. At the same time, the projection components of bending and tensile strains in directions parallel or perpendicular to the axis are smaller in the 45-degree direction, thus achieving priority acquisition of torsional signals from a physical installation perspective. The torsional wave sensor rotates synchronously with the drill bit and continuously acquires the instantaneous torsional angle value of the drill pipe surface at a preset sampling frequency during drilling. The sampling frequency must satisfy the Nyquist sampling theorem, meaning the sampling frequency must be at least twice the highest effective frequency component of the torsional wave signal in the drill pipe. This ensures that each peak and trough of the torsional wave is captured by at least two sampling points, and that the rising and falling edges of the waveform are not missed. This allows for peak identification and peak time interval calculation in subsequent steps to be accurate to the sampling period level. The instantaneous torsional angle value corresponding to each sampling moment and the precise clock tag for that sampling moment form a data pair. These pairs are stored sequentially in the buffer of the data recording device, with the sampling moment as the horizontal axis and the instantaneous torsional angle value as the vertical axis. Connecting these points forms a torsional wave curve. Each peak on the torsional wave curve corresponds to an abnormal contact event between the drill pipe and the borehole wall. The appearance time of the peak marks the precise time of the event, the amplitude of the peak reflects the intensity of the torsional wave excited by the contact, and the time interval between adjacent peaks carries the propagation time information of the torsional wave during one round trip between the reflection point and the sensor.

[0035] In another preferred embodiment of the present invention, the specific process for determining the reflection point in step S3 is as follows: Extract all peaks that appear at equal intervals from the fluctuation curve. When a torsional wave propagates back and forth between a contact point on the hole wall and the top sensor, each reflection back to the top forms a peak on the fluctuation curve. The time interval between two adjacent reflections is exactly equal to the time it takes for the wave to travel back and forth between the contact point and the sensor once. Therefore, if a certain time interval corresponds to a true reflection point, the peaks extracted at equal intervals should belong to the same reflection sequence. Although the amplitude of these peaks decreases with each energy decay, their occurrence times exhibit a strict equal-interval pattern. Squaring the amplitude values ​​of all peaks belonging to the same time interval and summing them up yields the cumulative energy value corresponding to that time interval. It can be understood that the instantaneous energy density of an elastic wave is proportional to the square of the amplitude. The summation of squared amplitudes essentially estimates the total signal energy contributed by the reflection sequence within the entire time window. The true reflection sequence has concentrated energy due to the large number and amplitude of peaks, while the noise sequence has dispersed energy due to the small number and weak amplitude of peaks. The squaring operation further widens the difference between the two in the energy domain, making it easier to automatically determine the subsequent classification threshold. After performing the above operation on all time interval values, a set of energy accumulation values ​​is formed that corresponds one-to-one with each time interval value and energy accumulation value.

[0036] It is understandable that when a torsional wave signal contains a sequence generated by multiple reflection points and a pseudo-period generated by random noise, the energy accumulation value corresponding to the true reflection sequence is significantly larger due to energy concentration, while the energy accumulation value of the noise pseudo-period is smaller and concentrated. Numerically, the two naturally form two separable clusters. The maximum inter-class variance criterion automatically finds the optimal boundary between the two clusters without the need to manually set any fixed thresholds related to lithology or borehole diameter. It is entirely determined by the statistical distribution characteristics of the data itself and has adaptability under different drilling conditions.

[0037] Time intervals where the accumulated energy exceeds the adaptive energy threshold are all identified as candidate periods. Multiple candidate periods are retained instead of just the single period corresponding to the maximum accumulated energy because the drill pipe may simultaneously contact the borehole wall at multiple different depths within the same drilling distance. The torsional wave signal is a composite waveform formed by the superposition of reflection sequences generated by multiple reflection points; each true reflection point corresponds to a characteristic period. Relying solely on the period with the highest energy would miss other valid reflection points. For each candidate period, the torsional wave propagation velocity, pre-calibrated on the drill pipe, is obtained. This velocity is calibrated by applying an artificial torsional pulse excitation at a known distance on the drill pipe, recording the time difference between the excitation moment and the moment the top sensor receives the reflected signal, and dividing twice the known distance by the time difference to obtain the propagation velocity. The propagation velocity is multiplied by the candidate period; the product equals the total round-trip path length of the torsional wave from the sensor to the reflection point and back to the sensor. Half of this product is taken as the distance from the reflection point to the torsional wave sensor corresponding to that candidate period. This distance is compared with the current drilling depth, and reflection points corresponding to candidate periods with distances greater than the drilling depth are discarded. The physical logic behind this elimination operation is that in the depth range not yet reached by the drill bit, there is no drill rod, and neither hole wall contact nor torsional wave reflection can occur. Any period calculated with a distance greater than the current drilling depth must come from false interference in the signal or calculation errors. All the remaining reflection points are determined as valid reflection points within the drilling distance range, and the distance corresponding to each reflection point is determined as the axial position of each reflection point.

[0038] In another preferred embodiment of the present invention, the specific calculation process of the attenuation ratio in step S4 is as follows: It is understandable that each time the torsional wave is reflected back to the top sensor from the reflection point, some of the wave's energy is consumed due to frictional contact at the reflection point. This causes the amplitude of the reflected wave to decrease gradually with the number of reflections. The first peak in the amplitude sequence corresponds to the amplitude of the torsional wave after one round trip, the second peak corresponds to the amplitude after two round trips, and so on. The attenuation trend of the amplitude sequence directly encodes the intensity of single energy dissipation of the torsional wave at that reflection point.

[0039] The physical basis for using an exponential decay model instead of a nonlinear decay model is that when a torsional wave propagates in a uniform elastic waveguide, the proportion of energy lost due to frictional contact at each reflection point is approximately constant. Therefore, the peak amplitude decreases exponentially with the number of reflections. A larger attenuation constant indicates a larger proportion of energy loss per reflection and stronger friction at the reflection point. Using the attenuation constant and initial amplitude as undetermined parameters, the peak numbers are substituted into the exponential decay model to obtain the theoretical amplitude values ​​corresponding to each number. The sum of squares of the differences between the theoretical and measured amplitude values ​​is calculated to obtain the residual sum of squares function. The attenuation constant and initial amplitude that minimize the residual sum of squares function are then determined.

[0040] It is understandable that the attenuation constant is estimated using the amplitude information of all peaks in the amplitude sequence, rather than relying solely on the amplitude values ​​of the first and last peaks. When the amplitude of a certain peak is abnormally high or low due to instantaneous noise interference, the influence of this abnormal peak on the fitting result is suppressed by the constraint of minimizing the overall error of all peaks, thus making the estimated attenuation constant more robust to random noise. The attenuation ratio is determined by taking the natural constant as the inverse power of the attenuation constant. Physically, the attenuation ratio represents the proportion of the residual amplitude of the torsional wave after each reflection at the reflection point to the incident amplitude. The smaller the attenuation ratio is than 1, the stronger the single energy dissipation at the reflection point and the more intense the frictional contact between the borehole wall and the drill pipe. This attenuation ratio is multiplied by the average drill pipe torque during the same period to obtain the torque consumption equivalent at the reflection point. The attenuation ratio serves as a weighting factor for the contribution of the reflection point to the total torque. Its physical basis is that the energy loss of the torsional wave at the reflection point and the mechanical energy consumed by the drill pipe to overcome friction at that point during rotation both originate from the contact friction between the drill pipe and the borehole wall, and the two are positively correlated.

[0041] In another preferred embodiment of the present invention, the specific calculation process for the torque consumption equivalent in step S4 is as follows: The torque value sequence of the drill rod torque changing over time, synchronously recorded by the drill rig torque monitoring system during the drilling process, is obtained. This torque value sequence corresponds one-to-one with the fluctuation curve segment in time. This temporal correspondence relies on the pulse sequence number tag attached to each sampled data frame by the data recording device in S1. The torque monitoring system and the torsional wave sensor share the same trigger pulse source, and their sampling times are synchronously identified by the same pulse sequence number. Therefore, the torque value sequence recorded within the same time period can be precisely aligned with the fluctuation curve segment on the time axis based on the pulse sequence number. The arithmetic mean of all torque values ​​in the torque value sequence is calculated by summing the torque values ​​at each sampling time and dividing by the total number of torque values ​​in the sequence. This yields the average torque value for that drilling process. Physically, the average torque value represents the total resistance torque experienced by the drill rod rotation within that drilling distance. It includes the torque consumed by the drill bit breaking rock, the torque consumed by friction between the drill rod and the borehole wall at each contact point, and the viscous resistance torque of the drill rod rotating in the mud.

[0042] The understandable attenuation ratio is derived by fitting the attenuation of the torsional wave's amplitude after multiple reflections at the reflection point. It reflects the proportion of energy lost due to frictional contact each time the torsional wave passes the reflection point, which is uniquely determined by the contact state of the borehole wall at that point. The average torque reflects the total torque required for drill pipe rotation during this drilling process, including the torque contributed by friction at the reflection point. Multiplying the two essentially uses the attenuation ratio as a weighting factor to estimate the torque consumption component allocated to the reflection point from the total torque. A smaller attenuation ratio indicates more severe friction at the reflection point, resulting in a larger share of torque consumption in the total torque. Therefore, directly weighting the average torque using the attenuation ratio yields a relative estimate of the torque consumption component at the reflection point. The torque consumption equivalent has the same physical dimensions as torque. A larger value indicates stronger frictional entanglement between the borehole wall and the drill pipe at the reflection point, and more fractured or unstable surrounding rock. Lateral comparisons along the depth and cross-sectional orientation can be made to identify the relative severity of abnormal borehole wall friction zones.

[0043] In another preferred embodiment of the present invention, the specific process of distributing the torque consumption equivalent of each local peak point to the borehole cross-sectional orientation according to the rotation phase in step S5 is as follows: On the friction characteristic depth curve, the torque consumption equivalent of each depth sampling point is compared with the torque consumption equivalent of its two adjacent depth sampling points. If the torque consumption equivalent of the current depth sampling point is greater than the torque consumption equivalent of both its preceding and following depth sampling points, then the current depth sampling point is determined as a local peak point, and the axial position of the reflection point and the torque consumption equivalent corresponding to the local peak point are recorded. The physical meaning of the local peak point is that the frictional contact intensity between the borehole wall and the drill pipe at this depth position is higher than that of the adjacent depth ranges above and below, indicating that there is a relatively independent strong friction source at this position, rather than random fluctuations in a large-area uniform contact background. During drilling, the contact between the drill rod and the borehole wall is not uniformly distributed along the entire depth of the hole, but is concentrated at specific depths such as borehole wall protrusions, accumulation of broken rock blocks, and sections where the drill rod is attached to the wall due to borehole deviation. These locations are represented as local peaks on the friction characteristic depth curve. The local peak points correspond to the most significant friction events, and the correspondence between the drill rod rotation phase at the time of their occurrence and the borehole wall contact orientation is the clearest, with the highest signal-to-noise ratio. This effectively avoids introducing phase noise from weak friction signals into subsequent three-dimensional models.

[0044] The drilling rig monitoring system records the drill rod rotation phase at the sampling moments when each local peak point is generated. The drill rod rotation phase is measured and recorded in real time by a rotary encoder installed on the drilling rig's power head. The resolution of the rotary encoder determines the angular accuracy of the phase. Since the torsional wave sensor rotates synchronously with the drill rod, there is a fixed spatial relationship between the instantaneous torsional angle signal recorded by the sensor and the drill rod rotation phase: the contact state between a certain point on the drill rod surface and the borehole wall is different when it rotates to different orientations. When the point rotates to the orientation of the borehole wall protrusion or breakage, the contact friction is strongest, and the amplitude of the excited torsional wave is the largest. Therefore, the drill rod rotation phase at the moment of wave peak generation directly indicates the orientation position of the borehole wall friction contact point on the borehole cross section. A one-to-one correspondence is established between each local peak point and the drill rod rotation phase, that is, each local peak point has a unique sampling moment, and each sampling moment has a unique drill rod rotation phase value, thereby linking the friction intensity information with the spatial orientation information.

[0045] The initial phase is obtained by recording the drill pipe rotation phase before drilling begins. Before drilling starts, the drill pipe is not yet subjected to drilling resistance torque and is in a freely suspended state. The recorded rotation phase at this time represents the initial orientation of the strain gauges on the drill pipe on the borehole cross-section. The geographic azimuth of this initial phase on the borehole cross-section, determined by a geological compass or gyro inclinometer before drilling begins, is obtained as the reference azimuth. The reference azimuth is measured clockwise with magnetic north as the zero-degree reference direction. It can be understood that the rotary encoder measures the relative angle of the drill pipe relative to a certain mechanical zero point, while subsequent support orientation requires absolute direction information in geographic space. The calibration between the initial phase and the reference azimuth establishes a mapping relationship between the drill pipe's relative coordinate system and the geographic absolute coordinate system, allowing the rotation phase recorded at different times to be converted into a unified geographic azimuth of the borehole cross-section.

[0046] For each local peak point, the initial phase is subtracted from the drill pipe rotation phase corresponding to that local peak point to obtain the phase offset relative to the initial phase. Subtracting the initial phase aims to eliminate the influence of the initial drill pipe stopping angle, ensuring that the rotation phases recorded for different boreholes and drilling shifts are unified under the same angular reference. The phase offset is added to the reference azimuth to obtain the geographical azimuth of that local peak point on the borehole cross-section. When the sum of the phase offset and the reference azimuth exceeds 360 degrees, 360 degrees is subtracted to ensure that the geographical azimuth always falls within the range of 0 to 360 degrees. Thus, the location of the friction event at each local peak point is converted into an absolute geographical azimuth that can be marked on the borehole cross-section diagram.

[0047] Using the center of the borehole cross-section as the origin and magnetic north as the zero-degree reference direction, the circumference of the borehole cross-section is divided into a predetermined number of azimuth intervals in a clockwise direction. For example, 12 azimuth intervals are divided into intervals of 30 degrees each, or 8 azimuth intervals are divided into intervals of 45 degrees each. The purpose of dividing the azimuth intervals into a predetermined number is to discretize the continuous azimuth angles into a finite number of azimuth categories, which facilitates subsequent statistical analysis of the frequency and intensity distribution characteristics of friction events occurring at different depths within the same azimuth interval. For each local peak point, the azimuth interval to which the geographical azimuth corresponds is determined based on the geographical azimuth of the local peak point. For example, a local peak point with a geographical azimuth of 85 degrees should be classified into the 60-90 degree interval or the 60-120 degree interval, depending on the interval spacing. The torque consumption equivalent of the local peak point is then allocated to the determined azimuth interval, meaning that the azimuth interval receives the numerical contribution of the torque consumption equivalent at the local peak point. When multiple local peak points exist within the same azimuth interval at the same depth section, the arithmetic mean of each torque consumption equivalent is taken as the representative value of the torque consumption equivalent at that depth for that azimuth interval. The axial position of the reflection point of the local peak point is recorded as the depth position corresponding to that torque consumption equivalent within that azimuth interval, thereby establishing a joint distribution record of the torque consumption equivalent in both depth and azimuth dimensions.

[0048] In another preferred embodiment of the present invention, the specific generation process of the three-dimensional drilling parameter spatiotemporal characteristic model in step S6 is as follows: A two-dimensional grid coordinate system is established, using the axial position of the reflection point as the vertical coordinate axis and the azimuth intervals divided along the circumference of the borehole cross-section as the horizontal coordinate axes. Each unit length of the vertical coordinate axis corresponds to the axial position of a reflection point, which is the depth coordinate of the contact point with the borehole wall determined by S3 through the torsional wave round-trip time and after being eliminated by drilling depth constraints. Therefore, each coordinate grid point on the vertical coordinate axis represents a cross-sectional position with a definite depth value within the borehole. Each unit length of the horizontal coordinate axis corresponds to an azimuth interval, which is formed in S5 by dividing the circumference of the borehole cross-section at preset angular intervals along a clockwise direction, with magnetic north as the zero-degree reference. For example, a 360-degree circumference is divided into 12 azimuth intervals every 30 degrees, with each interval covering a central angle range of 30 degrees. Each coordinate grid point on the horizontal coordinate axis represents an angular range with a definite geographical orientation on the borehole cross-section. The two-dimensional grid coordinate system established in this way has each cell indexed by the axial position of a reflection point on the vertical coordinate axis and an azimuth interval on the horizontal coordinate axis, uniquely corresponding to a spatial micro-element in a certain direction interval on a certain depth section inside the borehole, providing a standardized coordinate framework for the spatial positioning of borehole wall friction information.

[0049] The torque consumption equivalent of each local peak point, already assigned to its respective azimuth interval, is filled into the cells in the two-dimensional grid coordinate system corresponding to the axial position and azimuth interval of the reflection point. Local peak points are determined by S5 on the friction feature depth curve by comparing the magnitude of the torque consumption equivalent at adjacent depth sampling points. Each local peak point carries three attribute information: the axial position of the reflection point, the torque consumption equivalent value, and the borehole section azimuth interval assigned after rotational phase transformation. The filling operation essentially maps the torque consumption equivalent after azimuth assignment in S5 to the corresponding cell in the two-dimensional grid coordinate system according to its axial position of the reflection point and its assigned azimuth interval. When multiple local peak points have their torque consumption equivalents within the same cell, the arithmetic mean of these torque consumption equivalents is taken as the fill value for that cell; when a cell has no torque consumption equivalent assigned at any depth sampling point, that cell remains empty. After filling, for each reflection point along its axial position, a two-dimensional cross-sectional mesh diagram corresponding to the corresponding depth section is obtained. Each column of this mesh diagram represents an azimuth interval on the borehole cross-section, and the value in the column represents the friction intensity of the borehole wall against the drill pipe at that depth and azimuth. This cross-sectional mesh diagram intuitively reflects the directional distribution of borehole wall friction intensity on a single depth section: if the torque consumption equivalent values ​​of each azimuth interval are similar, it indicates that there is no obvious directional bias in borehole wall friction at that depth, and the contact between the borehole wall and the drill pipe is relatively uniform; if the torque consumption equivalent of a certain azimuth interval is significantly higher than that of other azimuth intervals, it indicates that there is a bias or unilateral fracture zone at that depth, and borehole wall friction is concentrated in that azimuth.

[0050] The filled two-dimensional meshes corresponding to the axial positions of different reflection points are sequentially stacked along the vertical coordinate axis to obtain a three-dimensional spatiotemporal characteristic model of drilling parameters. As the drilling depth increases, S3 continuously locates the axial positions of new reflection points, S4 calculates the torque consumption equivalent corresponding to each reflection point, S5 completes the azimuth assignment of each local peak point, and S6 continuously adds the two-dimensional cross-sectional mesh map of the new depth section to the end of the existing stacked structure. The stacking operation integrates the two-dimensional friction feature distribution originally scattered on various depth sections into a three-dimensional data volume. This three-dimensional data volume uses the drilling depth as the vertical axis, the circumferential direction of the drilling section as the horizontal axis, and the torque consumption equivalent as the cell value, intuitively presenting the joint distribution characteristics of borehole wall friction intensity in depth and direction. In the vertical axis direction, intervals where the torque consumption equivalent value is consistently high along the depth can be identified as abnormal borehole wall friction sections. The depth boundary between intact and fractured surrounding rock can be delineated by combining the changing trend of torque consumption equivalent between adjacent depth sections. In the horizontal axis direction, the azimuth interval with the largest torque consumption equivalent within the same depth section can be located as the main friction direction. The direction of bias and the spatial distribution of asymmetric fracture zones can be determined by combining the continuous change or maintenance of the main friction direction along the depth. The establishment of this three-dimensional model expands the information dimension of traditional drilling technology, which can only output one-dimensional torque curves along the depth direction, to three dimensions. This allows for the complete recording and intuitive expression of the spatial distribution characteristics of borehole wall friction, providing a data foundation for identifying potential borehole collapse sections, eliminating false hard rock misjudgments, and designing differentiated support systems by simultaneously viewing depth and azimuth information within the same model.

[0051] In another preferred embodiment of the invention, the method further includes sequentially extracting the entire row of cross-sectional data corresponding to the axial position of each reflection point from top to bottom along the vertical coordinate axis. This row of data contains the torque consumption equivalent value at each direction on the depth cross-section. For each axial position of the reflection point, the torque consumption equivalent values ​​at each direction on the depth cross-section are summed and divided by the total number of azimuth intervals to obtain the cross-sectional mean value of the torque consumption equivalent at the axial position of the reflection point. The physical meaning of the cross-sectional mean value is the average level of the friction intensity between the borehole wall and the drill pipe on the depth cross-section. It eliminates the local fluctuations caused by the uneven distribution of friction between different directions on a single depth cross-section and retains the overall trend information of the friction intensity changing along the depth direction. All axial positions of the reflection points are arranged from shallow to deep, and the cross-sectional mean values ​​corresponding to the axial positions of each reflection point are arranged sequentially to form a sequence in which the cross-sectional mean value changes with the axial position of the reflection point.

[0052] The cross-sectional mean value sequence is time-aligned with the drill bit depth variation sequence recorded during drilling. The drill bit depth variation sequence is continuously recorded by the drilling rig displacement encoder during drilling. The recording format is that each sampling moment corresponds to a depth value of the drill bit. This sequence completely records the entire time history of the drill bit advancing from the borehole opening to the bottom. The specific time alignment method is as follows: for each axial position of the reflection point in the cross-sectional mean value sequence, according to the time mapping relationship established by the pulse sequence number label in S1, the sampling moment when the cross-sectional mean value first appears at the axial position of the reflection point is determined, that is, the time when the borehole wall friction signal is first detected at this depth position; at the same time, the moment when the drill bit first arrives at the axial position of the reflection point is found in the drill bit depth variation sequence, that is, the time when the drill bit advances to this depth. The time difference at the axial position of the reflection point is obtained by subtracting the moment when the cross-sectional mean value first appears from the moment when the drill bit first arrives at the axial position of the reflection point. A time difference greater than zero indicates that the hole wall friction signal appears after the drill bit has passed through that depth, meaning that the hole wall has not yet experienced significant friction when the drill bit passes through that position, and the friction signal appears delayed after the drill bit leaves. A time difference equal to or close to zero indicates that the hole wall friction signal occurs synchronously with the drill bit reaching that depth, meaning that the hole wall immediately comes into frictional contact with the drill rod when the drill bit reaches that position.

[0053] Among the axial positions of reflection points with a time difference greater than zero, the interval of axial positions of reflection points continuously distributed along the vertical coordinate axis is defined as the potential borehole collapse section. It can be understood that drill bit entry is the only active leading event in the borehole. If the borehole wall remains stable when the drill bit passes, the frictional contact between the borehole wall and the drill pipe should occur synchronously with drilling, with a time difference approaching zero. If the borehole wall experiences delayed instability due to drill bit disturbance after the drill bit passes, the collapsed rock fragments need a certain amount of time to accumulate under gravity and encase the drill pipe before continuous borehole wall friction occurs, and the time difference must be greater than zero. Therefore, a time difference greater than zero is an objective temporal indicator of delayed borehole wall instability, independent of any empirical thresholds related to lithology, borehole diameter, or mud concentration. Multiple adjacent axial positions of reflection points along the vertical coordinate axis all satisfy the condition of a time difference greater than zero. These positions spatially constitute a continuous depth interval, indicating that the borehole wall within this interval has experienced overall delayed instability, rather than isolated local rockfalls.

[0054] The axial position range of reflection points that meet the above continuous distribution conditions is defined as the potential borehole collapse section. The starting depth of the potential borehole collapse section is the axial position of the reflection point with the smallest depth within the continuous range, and the ending depth is the axial position of the reflection point with the largest depth within the continuous range. This potential borehole collapse section is the depth range within which the drill bit will experience delayed collapse after passing through the borehole wall, and it is the direct spatial object for subsequent torque anomaly attribution diagnosis and construction early warning.

[0055] In another preferred embodiment of the present invention, step S6 further includes a drilling rig torque monitoring system that synchronously samples the torque value during drilling using the same trigger pulse source as the torsional wave sensor. Each sampling moment records a drill rod torque value, and the torque values ​​are arranged in chronological order to form a torque value sequence. Based on the time mapping relationship established by the pulse sequence number labels in step S1, the torque value at each sampling moment in the torque value sequence is paired with the drill bit depth value at the same moment in the drill bit depth versus time sequence. The drill bit depth value is used as the abscissa, and the torque value as the ordinate, to plot a torque versus depth curve point by point. This curve reflects the change in the total torque required for drill rod rotation along the borehole depth. A rise in the curve indicates that the total torque in a certain depth range is significantly higher than in adjacent depth ranges. This rise may be caused by increased rock-breaking resistance due to the drill bit encountering high-strength surrounding rock, or it may be caused by strong friction between the drill rod and the borehole wall in that depth range.

[0056] The depth locations where torque increases on the torque-depth curve are compared with potential borehole collapse sections. The depth location of the torque increase refers to the continuous depth range between the initial depth where the torque value transitions from a low to a high value region on the torque-depth curve and the final depth where the torque value falls back to a low value region. Potential borehole collapse sections are determined by the time difference between the cross-sectional mean sequence and the drill bit depth-time change sequence. These are the depth ranges where the borehole wall experiences delayed instability after the drill bit passes through. The borehole wall friction signals within this range have been confirmed to be later than the drill bit's arrival in time; therefore, any friction phenomena occurring within this range can only be causally attributed to delayed borehole wall collapse, not drill bit rock breaking. The depth ranges of torque increases are compared one by one with the depth ranges of potential borehole collapse sections to determine whether the depth location of the torque increase falls completely or partially within the potential borehole collapse section range.

[0057] It is understandable that, within the potential collapse section, time-series verification has confirmed that the borehole wall friction signal occurs after the drill bit passes through. The torque consumption equivalent at this location is generated by the friction between the delayed-collapsed rock fragments and the drill pipe. The torque rise at the same depth location on the torque-depth curve coincides with this delayed friction event both temporally and spatially. Therefore, the physical source of this rise should be attributed to the frictional contact between the drill pipe and the borehole wall, rather than changes in the surrounding rock strength. Depth locations not falling within the potential collapse section are identified as torque anomalies contributing to changes in surrounding rock strength. It is understandable that when the depth location of the torque rise does not spatially coincide with all known delayed friction sources, the rise cannot be caused by delayed friction between the drill pipe and the borehole wall. Its physical source can only be the change in rock-breaking resistance when the drill bit contacts the surrounding rock at that depth location, i.e., changes in surrounding rock strength contribute to this torque rise. Through this comparison and judgment, each abnormal uplift segment on the torque-depth curve is given a clear physical cause. The previously indistinguishable drill bit rock-breaking torque response and drill rod friction torque response are separated, thus severing the inherent erroneous correspondence between torque increase and hard rock in traditional methods. This provides torque information that removes the interference of borehole wall friction for the identification of surrounding rock strength in tunnel advanced geological prediction.

[0058] Please see Figure 2 As shown, the present invention also includes a spatiotemporal feature modeling system for drilling parameters inside tunnel geological exploration boreholes, used to implement the above-mentioned spatiotemporal feature modeling method for drilling parameters inside tunnel geological exploration boreholes, including: The parameter acquisition module is used to pre-install a torsional wave sensor on the top of the drill pipe, continuously collect the instantaneous torsional angle of the drill pipe surface and record it in time sequence to establish a torsional wave curve; The interval acquisition module is used to take a fixed depth of drilling by the drill bit as a drilling segment, extract the wave curve segment corresponding to the drilling segment from the torsional wave curve, identify the wave peaks and calculate the time interval between adjacent wave peaks to form a time interval set. The reflection recognition module is used to extract the interval value with the maximum energy accumulation from the time interval set as the feature period, determine the torsional wave reflection position that generates the feature period as the reflection point, calculate the distance from the reflection point to the torsional wave sensor based on the feature period, and determine the axial position of the reflection point based on the distance. The torque calculation module is used to perform exponential decay fitting based on the amplitude of each peak corresponding to the characteristic period to obtain the decay ratio, and at the same time obtain the average drill rod torque corresponding to the drilling section. Based on the average torque and the decay ratio, the torque consumption equivalent at the reflection point is calculated. The position allocation module is used to establish a friction feature depth curve with the axial position of the reflection point as the abscissa and the torque consumption equivalent as the ordinate, determine the local peak points on the curve and obtain the drill pipe rotation phase corresponding to each local peak point, and allocate the torque consumption equivalent of each local peak point to the borehole section orientation according to the rotation phase. The results generation module is used to superimpose the orientation of torque consumption equivalents at different depths and rotation phases on the borehole cross section, and generate a three-dimensional spatiotemporal feature model of drilling parameters with the borehole depth as the vertical axis and the circumferential direction of the borehole cross section as the horizontal axis.

[0059] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A method for modeling the spatiotemporal characteristics of drilling parameters in tunnel geological exploration boreholes, characterized in that, Includes the following steps: S1, a torsional wave sensor is pre-installed on the top of the drill pipe to continuously collect the instantaneous torsional angle of the drill pipe surface and record it in chronological order to establish a torsional wave curve; S2, take each fixed depth drilled by the drill bit as a drilling segment, extract the wave curve segment corresponding to the drilling segment from the torsional wave curve, identify the wave peaks and calculate the time interval between adjacent wave peaks to form a time interval set. S3, extract the interval value with the largest energy accumulation from the time interval set as the characteristic period, determine the torsional wave reflection position that generates the characteristic period as the reflection point, calculate the distance from the reflection point to the torsional wave sensor based on the characteristic period, and determine the axial position of the reflection point as the distance; S4. Based on the amplitude of each wave peak corresponding to the characteristic period, perform exponential decay fitting to obtain the decay ratio. At the same time, obtain the average drill rod torque corresponding to the drilling section. Based on the average torque and the decay ratio, calculate the torque consumption equivalent at the reflection point. S5. Establish a friction feature depth curve with the axial position of the reflection point as the abscissa and the torque consumption equivalent as the ordinate. Determine the local peak points on the curve and obtain the drill pipe rotation phase corresponding to each local peak point. Distribute the torque consumption equivalent of each local peak point to the borehole section orientation according to the rotation phase. S6 superimposes the orientation of the torque consumption equivalent at different depths and rotation phases on the borehole cross section to generate a three-dimensional spatiotemporal feature model of drilling parameters with the borehole depth as the vertical axis and the circumferential direction of the borehole cross section as the horizontal axis.

2. The method for modeling the spatiotemporal characteristics of drilling parameters in tunnel geological exploration boreholes according to claim 1, characterized in that, In S1, the specific construction process of the torsional fluctuation curve is as follows: A torsional wave sensor is pre-installed on the top of the drill rod. The torsional wave sensor rotates synchronously with the drill bit and continuously collects the instantaneous torsional angle value of the drill rod surface at a preset sampling frequency during the drilling process. The instantaneous torsional angle value corresponding to each sampling moment is recorded in chronological order to form a torsional wave curve with time as the horizontal axis and instantaneous torsional angle value as the vertical axis.

3. The method for modeling the spatiotemporal characteristics of drilling parameters in tunnel geological exploration boreholes according to claim 1, characterized in that, In S3, the specific process for determining the reflection point is as follows: For each time interval in the time interval set, starting from each peak in the wave curve segment corresponding to the drilling section, the peaks with adjacent peak time intervals equal to the time interval are selected along the time axis in the positive direction. The amplitude values ​​of the peaks selected from each starting point are squared and summed, and the maximum value is taken as the energy accumulation value corresponding to the time interval, forming a set of energy accumulation values. The set of energy accumulation values ​​is subjected to binary clustering. The set of energy accumulation values ​​is divided into high-energy class and low-energy class based on the maximum inter-class variance. The minimum energy accumulation value in the high-energy class is taken as the adaptive energy threshold. The time interval in which the energy accumulation value exceeds the adaptive energy threshold is determined as the candidate period. Based on the candidate period and the pre-calibrated propagation speed of the torsional wave in the drill pipe, the distance from the reflection point to the torsional wave sensor corresponding to each candidate period is calculated. Reflection points whose distance is greater than the current drilling depth are removed. The remaining reflection points are determined as reflection points within the drilling section, and the corresponding distance of each remaining reflection point is determined as the axial position of the reflection point.

4. The method for modeling the spatiotemporal characteristics of drilling parameters in tunnel geological exploration boreholes according to claim 1, characterized in that, In S4, the specific calculation process for the attenuation ratio is as follows: Extract the peak sequence of the characteristic period from the wave curve segment corresponding to the drilling section, and take the amplitude value of each peak in turn to form an amplitude sequence; establish an exponential decay model for the amplitude sequence, use the least squares method to fit the amplitude sequence, determine the decay constant by finding the minimum value of the sum of squared residuals, and determine the decay ratio by taking the inverse power of the natural constant.

5. The method for modeling the spatiotemporal characteristics of drilling parameters in tunnel geological exploration boreholes according to claim 1, characterized in that, In step S4, the specific calculation process for the torque consumption equivalent is as follows: Obtain the torque value sequence of the drill rod torque as a function of time, which is synchronously recorded by the drill rig torque monitoring system corresponding to the drilling section. The torque value sequence corresponds one-to-one with the fluctuation curve segment in time. The arithmetic mean of all torque values ​​in the torque value sequence is calculated by summing all torque values ​​in the torque value sequence and dividing by the total number of torque values ​​in the torque value sequence to obtain the average torque value corresponding to the drilling section. The attenuation ratio is multiplied by the average torque value, and the resulting product is determined as the torque consumption equivalent of the surrounding rock on the drill pipe at the reflection point.

6. The method for modeling the spatiotemporal characteristics of drilling parameters in tunnel geological exploration boreholes according to claim 1, characterized in that, In step S5, the specific process of distributing the torque consumption equivalent of each local peak point to the borehole cross-section orientation according to the rotation phase is as follows: On the friction feature depth curve, the sampling point where the torque consumption equivalent is greater than the torque consumption equivalent of the two adjacent depth sampling points is determined as a local peak point, and the axial position of the reflection point and the torque consumption equivalent corresponding to each local peak point are recorded. The drilling rig monitoring system records the drill rod rotation phase at the sampling time of each local peak point, and establishes a one-to-one correspondence between each local peak point and the drill rod rotation phase. The drill rod rotation phase recorded before drilling begins is obtained as the initial phase, and the geographical orientation of the initial phase measured before drilling begins on the borehole cross section is obtained as the reference orientation. For each local peak point, the drill rod rotation phase corresponding to the local peak point is subtracted from the initial phase to obtain the phase offset of the local peak point relative to the initial phase. Add the phase offset to the reference azimuth to obtain the geographical azimuth of the local peak point on the borehole cross section; With the center of the borehole section as the origin and the magnetic north direction as the zero-degree reference direction, the circumference of the borehole section is divided into a predetermined number of azimuth intervals in a clockwise direction; For each local peak point, the azimuth range that the geographical azimuth points to on the borehole section is determined according to the geographical azimuth of the local peak point. The torque consumption equivalent of the local peak point is allocated to the determined azimuth range, and the axial position of the reflection point of the local peak point is recorded as the depth position corresponding to the torque consumption equivalent within the azimuth range.

7. The method for modeling the spatiotemporal characteristics of drilling parameters in tunnel geological exploration boreholes according to claim 1, characterized in that, In step S6, the specific generation process of the three-dimensional drilling parameter spatiotemporal characteristic model is as follows: A two-dimensional grid coordinate system is established with the axial position of the reflection point as the vertical coordinate axis and the directional intervals divided by the circumferential direction of the borehole section as the horizontal coordinate axis. The unit length of the vertical coordinate axis corresponds to the axial position of a reflection point, and the unit length of the horizontal coordinate axis corresponds to a directional interval. Fill the torque consumption equivalent of each local peak point, which has been allocated to each azimuth interval, into the cell in the two-dimensional grid coordinate system corresponding to the axial position and azimuth interval of the reflection point; The filled two-dimensional meshes corresponding to the axial positions of different reflection points are stacked sequentially along the vertical coordinate axis to obtain a three-dimensional spatiotemporal characteristic model of drilling parameters.

8. The method for modeling the spatiotemporal characteristics of drilling parameters in tunnel geological exploration boreholes according to claim 7, characterized in that, It also includes extracting the cross-sectional torque consumption equivalent distribution corresponding to the axial position of each reflection point along the vertical coordinate axis from the three-dimensional drilling parameter spatiotemporal feature model, calculating the cross-sectional mean value of torque consumption equivalent at the axial position of each reflection point, and forming a sequence of cross-sectional mean value changing with the axial position of the reflection point; The cross-sectional mean value sequence is time-aligned with the drill bit depth change sequence recorded during drilling. For each reflection point axial position, the time difference between the moment when the cross-sectional mean value first appears at that reflection point axial position and the moment when the drill bit first arrives at that reflection point axial position is determined. The depth intervals that are continuously distributed along the vertical coordinate axis among the axial positions of reflection points with a time difference greater than zero are defined as potential hole collapse sections.

9. The method for modeling the spatiotemporal characteristics of drilling parameters in tunnel geological exploration boreholes according to claim 8, characterized in that, S6 further includes generating a torque-depth curve from the drill rod torque value sequence synchronously recorded by the drilling rig torque monitoring system; comparing the depth positions where torque increases appear on the torque-depth curve with potential hole collapse sections, determining the depth positions falling within the potential hole collapse section range as torque anomalies caused by friction between the drill rod and the hole wall, and determining the depth positions not falling within the potential hole collapse section range as torque anomalies including contributions from changes in surrounding rock strength.

10. A system for modeling the spatiotemporal characteristics of drilling parameters inside a tunnel geological exploration borehole, used to implement the method for modeling the spatiotemporal characteristics of drilling parameters inside a tunnel geological exploration borehole as described in any one of claims 1-9, characterized in that, include: The parameter acquisition module is used to pre-install a torsional wave sensor on the top of the drill pipe, continuously collect the instantaneous torsional angle of the drill pipe surface and record it in time sequence to establish a torsional wave curve; The interval acquisition module is used to take a fixed depth of drilling by the drill bit as a drilling segment, extract the wave curve segment corresponding to the drilling segment from the torsional wave curve, identify the wave peaks and calculate the time interval between adjacent wave peaks to form a time interval set. The reflection recognition module is used to extract the interval value with the maximum energy accumulation from the time interval set as the feature period, determine the torsional wave reflection position that generates the feature period as the reflection point, calculate the distance from the reflection point to the torsional wave sensor based on the feature period, and determine the axial position of the reflection point based on the distance. The torque calculation module is used to perform exponential decay fitting based on the amplitude of each peak corresponding to the characteristic period to obtain the decay ratio, and at the same time obtain the average drill rod torque corresponding to the drilling section. Based on the average torque and the decay ratio, the torque consumption equivalent at the reflection point is calculated. The position allocation module is used to establish a friction feature depth curve with the axial position of the reflection point as the abscissa and the torque consumption equivalent as the ordinate, determine the local peak points on the curve and obtain the drill pipe rotation phase corresponding to each local peak point, and allocate the torque consumption equivalent of each local peak point to the borehole section orientation according to the rotation phase. The results generation module is used to superimpose the orientation of torque consumption equivalents at different depths and rotation phases on the borehole cross section, and generate a three-dimensional spatiotemporal feature model of drilling parameters with the borehole depth as the vertical axis and the circumferential direction of the borehole cross section as the horizontal axis.

Images