A mine water detection and drainage drilling intelligent control method

Abstract

This invention relates to the field of mining drilling control technology, and more particularly to an intelligent control method for mining water exploration and drainage boreholes. The method includes: acquiring drilling rig advance pressure, rotation torque, drill rod vibration, trajectory measurements, and formation parameters; constructing a borehole bottom attitude state quantity and calculating an estimated borehole bottom axis value using a drill rod mechanical deformation model; establishing a four-dimensional deviation field by standardizing and registering the trajectory measurements and axis estimates; calculating the formation effect index and phase migration coefficient; multiplying the deviation field by the weighted sum of segments with the two values ​​to obtain the deviation compensation amount; calculating the transition constraint amount and adjusting the compensation amount when the rate of change exceeds a threshold; calculating the target correction attitude; and applying joint amplitude limiting modulation to generate a constrained path tracking command when the power exceeds the rated value. This invention can ensure borehole trajectory tracking accuracy and drilling rig operation stability under complex cross-strata geological conditions.

Description

Technical Field

[0001] This invention relates to the field of mining drilling control technology, and in particular to an intelligent control method for mining water exploration and drainage boreholes. Background Technology

[0002] With the continuous advancement of intelligent coal mine construction, directional drilling technology for mining has been widely applied in fields such as water exploration and drainage, gas extraction, and geological exploration. Precise control of the borehole trajectory for water exploration and drainage is directly related to the effectiveness of water hazard control and the safety of underground operations. However, underground geological conditions in coal mines are extremely complex, with frequent alternations of lithology and anisotropic characteristics. The drill rod, subjected to the complex coupling effects of propulsion pressure, rotational torque, and heterogeneous forces within the narrow borehole, is prone to bending deformation, causing the actual drilling trajectory to deviate from the preset path. Therefore, how to achieve high-precision control of the borehole trajectory under complex bottom-hole mechanical environment and geological disturbance conditions has become a technical problem that needs to be solved in the field of mine water exploration and drainage drilling.

[0003] In the prior art, Chinese patent document CN115434637B discloses a method and control system for directional drilling trajectory control based on piecewise fuzzy control. This method designs a piecewise fuzzy controller after selecting a drilling trajectory model, segments the annular value interval of the tool face angle, and designs membership functions and fuzzy rules for each segment, thereby replacing manual experience adjustments to achieve automatic tracking control of the directional drilling trajectory. This method explores membership function design and piecewise control strategies, contributing to the automation level of directional drilling construction.

[0004] However, the aforementioned method relies solely on the geometric angle deviation obtained from the inclinometer system for trajectory correction during the deviation modeling stage. It fails to incorporate multidimensional physical parameters reflecting the borehole bottom mechanical state, such as propulsion pressure, rotation torque, and drill pipe vibration, into the deviation model. Furthermore, it lacks a mechanical model of the discrete bending section of the drill pipe to estimate the actual axis position at the bottom of the borehole, resulting in incomplete deviation information and limited accuracy in compensation calculation. When formation conditions change drastically, this method lacks the ability to characterize the migration law of the bending phase difference along the borehole depth direction and lacks a transition constraint mechanism for abrupt lithological changes between adjacent borehole depths, easily leading to compensation jumps and correction oscillations at the cross-layer interface. In the control law generation stage, this method directly outputs control commands based on the current deviation, without considering the cumulative effect of historical attitude offsets on the dynamic response of the drill string. High-frequency correction commands may induce resonance and overcorrection in the servo components. Moreover, this method outputs control commands in isolation without implementing joint amplitude limiting modulation of the propulsion and rotation speeds. When encountering high-hardness rock formations, this can easily lead to hydraulic system power overload, increasing the risk of drill rig stalling.

[0005] In summary, existing borehole trajectory control methods have shortcomings in areas such as multidimensional physical parameter fusion deviation modeling, cross-layer abrupt transition constraints, suppression of attitude deviation accumulation effects, and joint limiting of propulsion and rotation power. They are unable to simultaneously meet the dual requirements of high-precision trajectory tracking and drilling rig operation safety under complex geological conditions. Summary of the Invention

[0006] To address the technical problems of insufficient deviation modeling dimensions, lack of cross-layer abrupt change constraints, and inadequate power joint limiting in existing borehole trajectory control methods, this invention provides an intelligent control method for mine water exploration and drainage boreholes, comprising: The drilling rig's feed pressure, slewing torque, drill rod vibration, trajectory measurement values, and formation parameters are obtained. Based on the preset trajectory, the hole bottom attitude state is constructed, and the hole bottom axis is estimated by combining the drill rod mechanical deformation model. A deviation field containing azimuth deviation, dip angle deviation, bending phase difference, and hole depth position is established by benchmarking and time-series registration of trajectory measurement values ​​and axis estimation values. Based on propulsion pressure, slewing torque, drill pipe vibration, and formation parameters, the formation effect index is calculated. The phase migration coefficient is calculated by combining the bending phase difference between adjacent hole depth intervals. The deviation field is then weighted and summed according to hole depth, and multiplied by the formation effect index and the phase migration coefficient to obtain the deviation compensation amount. When the deviation change rate exceeds the threshold, the transition constraint amount is calculated to adjust the compensation amount. Based on the deviation compensation amount, the bottom hole attitude state amount and the preset trajectory curvature, the target correction attitude is calculated. A dimensionless attitude accumulation factor is constructed according to the attitude offset to limit the attenuation of the target correction attitude. The propulsion speed, rotation speed and orientation angle control amounts are obtained by transforming the drilling rig dynamic mapping matrix. When the sum of propulsion power and rotation power exceeds the rated power, joint amplitude limiting modulation is applied to generate a constrained path tracking command, and the control is cyclically controlled until the set hole depth.

[0007] This invention can construct a four-dimensional deviation field by fusing multi-dimensional physical parameters, and complete the adaptive adjustment of deviation compensation by combining the formation action index, phase migration coefficient and transition constraint amount. Furthermore, it generates constrained path tracking commands by suppressing attitude accumulation factors and combining amplitude limiting modulation, thereby ensuring the tracking accuracy of the borehole trajectory and the stability of the drilling rig operation under complex cross-layer geological conditions.

[0008] Preferably, the step of constructing the hole bottom attitude state quantity based on a preset trajectory and calculating the estimated value of the hole bottom axis in combination with the drill pipe mechanical deformation model includes: Extract the three-dimensional coordinates, azimuth angle, and inclination angle of the current actual hole bottom position as hole bottom attitude state variables; The drill pipe is divided into multiple discrete bending segments. Based on the obtained pushing pressure and rotation torque, the axial force and lateral force on each discrete bending segment in the hole are calculated. Based on the axial force, lateral force, and bending stiffness of the drill pipe, the deflection and rotation angle changes of each discrete bending segment are calculated using the force-deformation equation, so that the estimated value of the bottom hole axis can incorporate the influence of the mechanical deformation of the drill pipe.

[0009] Preferably, the calculated estimated value of the hole bottom axis further includes: The relative attitude transformation matrix between adjacent discrete bending segments is calculated based on the change in the rotation angle of each discrete bending segment. The coordinate transformation and spatial recursion accumulation are performed along the hole depth direction in combination with the flexural deformation. Then, the estimated value of the hole bottom axis is calculated by combining the hole bottom attitude state quantity. Kalman filtering is used to fuse the estimated bottom axis of the hole obtained by spatial recursion and accumulation with the measured attitude state of the bottom of the hole, generating a continuous coordinate set of estimated bottom axis of the hole and a set of tangential vectors.

[0010] By combining segmented coordinate transformation with spatial recursive accumulation and Kalman filtering data fusion, the estimated hole bottom axis value can simultaneously take into account the continuity of mechanical model calculations and the authenticity of sensor measured data, thereby improving the accuracy of full-length axis reconstruction.

[0011] Preferably, the method for obtaining each deviation component in the deviation field includes: aligning the trajectory measurement value and the estimated value of the borehole bottom axis to the same borehole depth node through spatial interpolation to eliminate the temporal difference between the sensor acquisition frequency and the model calculation step size; at each borehole depth node, calculating the azimuth difference and tilt difference between the trajectory measurement value and the estimated value of the borehole bottom axis, which are respectively used as the azimuth deviation and tilt deviation.

[0012] Preferably, establishing the deviation field further includes: Based on the spatial geometric characteristics of the borehole trajectory, the cross product of the tangential vector of the trajectory measurement value and the tangential vector of the estimated borehole bottom axis value is calculated. The direction of the cross product is used to characterize the direction of the bending phase difference, and the magnitude of the bending phase difference is used to characterize the angle between the two vectors. A four-dimensional deviation field is constructed by vector combining the azimuth deviation, tilt deviation, and bending phase difference with the corresponding hole depth nodes.

[0013] By calculating the bending phase difference through the tangential vector cross product and combining it with the azimuth deviation, tilt deviation, and hole depth nodes to construct a four-dimensional deviation field, the deviation description is extended from two-dimensional angular difference to a four-dimensional spatial representation that includes bending phase information, thus enriching the dimensions of deviation information that can be used for deviation correction decisions.

[0014] Preferably, the step of multiplying the weighted sum of the deviation field segmented by borehole depth with the formation interaction index and the phase migration coefficient to obtain the deviation compensation amount includes: performing dimensionless processing on the amplitudes of the propulsion pressure, slewing torque, and drill pipe vibration, and calculating the formation interaction index through a multivariate regression equation in combination with formation hardness and abrasiveness parameters; performing differential operation on the bending phase difference between adjacent borehole depth intervals in the four-dimensional deviation field, using the ratio of the differential operation result to the borehole depth step size as the continuity gradient, and transforming it into a dimensionless phase migration coefficient after normalization; dividing the four-dimensional deviation field into multiple historical compensation segments according to borehole depth, assigning a dimensionless distance weight factor that increases with increasing borehole depth to each segment, multiplying the distance weight factor with the azimuth deviation and dip angle deviation, performing segmented weighted summation, and multiplying the summation result with the formation interaction index and the phase migration coefficient to obtain the bottom borehole deviation compensation amount.

[0015] Preferably, the adjustment compensation amount includes: calculating the rate of change of the Euclidean distance between the azimuth deviation and the inclination deviation in adjacent hole depth intervals with respect to hole depth; when the rate of change is greater than the abrupt change threshold, extracting the bending phase difference of the previous interval and the spatial trajectory vector increment of the current interval; taking the bending phase difference as a zero vector in the initial stage; normalizing the cross product of the estimated tangential vector of the hole bottom axis and the tangential vector of the trajectory measurement to obtain the phase direction vector; dividing the absolute value of the inner product of the phase direction vector and the trajectory vector increment by the magnitude of the trajectory vector increment to obtain the dimensionless transition constraint amount; using the result of subtracting the transition constraint amount from 1 to perform exponential decay adjustment on the phase migration coefficient; and multiplying the transition constraint amount by the original deviation compensation amount to obtain the adjusted deviation compensation amount.

[0016] When the rate of change of deviation between adjacent borehole depths exceeds the abrupt change threshold, the transition constraint is generated by normalizing the inner product of the bending phase direction vector and the spatial trajectory vector increment. The phase migration coefficient is adjusted by exponential decay, and the compensation amplitude is reduced proportionally to suppress the abrupt change of compensation amount and the correction oscillation at the lithological interface.

[0017] Preferably, the step of constructing a dimensionless attitude accumulation factor based on the attitude offset to limit the attenuation of the target correction attitude includes: constructing a spatial curve Frenet frame based on the curvature, tangential vector, and normal vector of the preset trajectory to generate an attitude guidance reference vector; converting the hole bottom attitude state quantity into a two-dimensional attitude vector; mapping the adjusted hole bottom deviation compensation quantity into an azimuth correction quantity and an inclination correction quantity; and fusing the attitude guidance reference vector to obtain the initial target correction attitude; calculating the azimuth difference and inclination difference between the initial target correction attitude and the hole bottom attitude state quantity of the previous control cycle as a two-dimensional attitude offset; constructing an attitude accumulation factor based on the rate of change of attitude offsets in multiple historical control cycles, the duration of the servo control cycle, and the maximum allowable deflection angle of the drilling rig; multiplying the command adjustment increment of the initial target correction attitude by the attenuation ratio obtained by transforming the attitude accumulation factor through a nonlinear function to perform amplitude limiting suppression; and outputting the suppressed and adjusted target correction attitude.

[0018] The attitude guidance reference vector is generated by Frenet frame and the initial target correction attitude is obtained by fusing the deviation compensation amount. Then, the attitude accumulation factor is constructed based on the historical attitude offset change rate and the maximum allowable deflection angle to apply nonlinear limiting, thereby suppressing the high-frequency resonance and command oscillation of the servo component.

[0019] Preferably, the generation of constrained path tracking instructions includes: converting the suppressed and adjusted target correction attitude into the desired propulsion speed, slewing speed and orientation angle control values ​​through the drilling rig dynamic mapping matrix; When the sum of propulsion power and slewing power exceeds the system's rated power, the ratio of the system's rated power to the sum of propulsion power and slewing power is calculated as the derating ratio. The propulsion power is calculated based on the product of propulsion pressure, the effective working area of ​​the hydraulic cylinder, and the propulsion speed control quantity. The slewing power is calculated based on the product of slewing torque and slewing speed control quantity. According to the derating ratio, the propulsion speed control quantity and the slewing speed control quantity are reduced by the same proportion. The reduced propulsion speed, slewing speed, and orientation angle control quantity are packaged to generate a constrained path tracking command.

[0020] By mapping the target correction attitude into a control variable through the drilling rig's power mapping matrix, when the sum of the propulsion power and the slewing power exceeds the rated power, the propulsion speed and slewing speed are reduced proportionally according to the derating ratio, so that the total output power of the drilling rig's dual hydraulic circuits is constrained within the rated range, avoiding the risk of power overload and stall during hard rock drilling.

[0021] Preferably, the acquisition of drilling rig advance pressure, slewing torque, drill rod vibration, trajectory measurement values, and formation parameters further includes acquiring the current advance speed, slewing speed, and orientation angle of the drilling rig; the constrained path tracking command is converted into an analog voltage signal by a programmable logic controller and sent to the drilling rig actuator; the borehole depth change is monitored in real time, and cyclic control is executed until the set target borehole depth is reached. This forms a complete control loop from perception and decision-making to execution.

[0022] The technical solution of the present invention has the following beneficial technical effects: This invention incorporates multi-dimensional physical parameters such as propulsion pressure, slewing torque, drill pipe vibration, and formation parameters into the deviation modeling system, constructing a four-dimensional deviation field containing bending phase information as the basis for correction decisions. This allows the compensation calculation to respond to changes in formation lithology and the characteristics of drill pipe bending phase migration. For abrupt lithological changes between adjacent hole depths during cross-strata drilling, a transition constraint mechanism adaptively compresses the compensation amplitude according to the Euclidean distance change rate, suppressing compensation jumps and correction oscillations at lithological boundaries. An attitude accumulation factor applies nonlinear amplitude limiting to the target correction attitude, blocking the transmission of high-frequency jitter to the servo components. Joint amplitude limiting modulation of propulsion speed and slewing speed constrains the system power within the rated safe range, balancing trajectory tracking accuracy, correction process stability, and drilling rig operation safety under complex cross-strata geological conditions.

[0023] Furthermore, in cross-strata drilling operations involving frequent alternating interfaces between sandstone and mudstone, the complete solution of this invention can control the target deviation within the centimeter range, reduce the number of high-frequency oscillation overshoots per 100 meters to 0, reduce the number of overload shutdowns to 0, and effectively improve the average drilling efficiency compared to the traditional proportional-integral-derivative control scheme. This verifies the effectiveness of the coupling of deviation compensation, anti-shaking mechanism and power joint limiting for high-precision and stable drilling under complex strata conditions. Attached Figure Description

[0024] Figure 1 This is a flowchart of an intelligent control method for mine water exploration and drainage boreholes according to the present invention; Figure 2 This is a schematic diagram of the evolution and compensation convergence curve of the four-dimensional deviation field with hole depth; Figure 3 This is a schematic diagram comparing the overall effectiveness of cross-layer drilling under different control strategies. Detailed Implementation

[0025] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0026] Reference Figure 1 A method for intelligent control of mine water exploration and drainage boreholes includes steps S1 to S3, which are described in detail below.

[0027] S1. Collect drilling parameters and construct hole bottom attitude and axis estimation values.

[0028] The drilling rig collects propulsion pressure via pressure sensors deployed on its hydraulic system pipelines, rotational torque via torque sensors installed at the output of the rotary motor, and triaxial vibration signals via piezoelectric accelerometers fixed to the drill pipe surface. Measurement-while-drilling (MWD) instruments acquire trajectory measurements including the actual dip and azimuth angles. Formation parameters, including rock hardness and abrasiveness, are imported from a geological exploration database. The current propulsion speed, rotational speed, and orientation angle are read from the drilling rig control system and transmitted to the host computer controller via a serial communication protocol.

[0029] Servo control cycle of the host computer controller The time interval is set to 100ms. The sensor collects raw data at a frequency of 50Hz, and the latest frame is used as the input for the current cycle in each control cycle. The correspondence between the spatial interpolation step size of 0.1m and the control cycle depends on the actual propulsion speed. When the propulsion speed is 0.6m / min, it takes 10s to advance 0.1m. During this period, the controller executes 100 control cycles. Deviation compensation and attitude correction in each cycle are updated based on the latest sensor data.

[0030] The borehole trajectory is discretized according to a set spatial step size. A cubic spline interpolation algorithm is used to calculate the preset dip angle and azimuth angle at each discrete node, which are then combined into a three-dimensional vector as the desired attitude reference value. The current actual borehole bottom three-dimensional coordinates, azimuth angle, and dip angle in the northeast-southeast coordinate system, calculated by the measurement-while-drilling system, are extracted at a frequency of 50Hz. The azimuth angle ranges from 0° to 360°, and the dip angle ranges from -90° to 90°.

[0031] The drill pipe is divided into multiple discrete bending segments, each with a length between 0.5m and 1m in this embodiment. Based on the obtained propulsion pressure and rotation torque, combined with the drill string's self-weight determined by the drill pipe material density and cross-sectional dimensions, the buoyancy coefficient calculated based on the borehole flushing fluid density and drill pipe outer diameter, and the borehole wall friction coefficient obtained from a table of rock mechanics parameters, the axial and lateral forces borne by the two end faces of each discrete bending segment are calculated using the infinitesimal element method force balance equation. After obtaining the force parameters, the bending stiffness of the drill pipe of a specific specification is input. Substituting the differential force-deformation equations based on the Euler-Bernoulli beam theory, the deflection and rotation changes of each discrete bending segment in the local coordinate system are solved by the fourth-order Runge-Kutta numerical integration method.

[0032] Based on the principles of spatial kinematics, a relative attitude transformation matrix between adjacent discrete bending segments is constructed using the changes in rotation angles of each segment. The deflection deformation of each segment is represented as a local translation vector, which, combined with the corresponding relative attitude transformation matrix, performs a segment-by-segment homogeneous coordinate transformation. Spatial recursive accumulation is then performed along the hole depth direction starting from the borehole reference point. Kalman filtering is used to fuse and adjust the estimated bottom axis value obtained from the spatial recursive accumulation with the measured bottom attitude state quantities. During the filtering process, the three-dimensional coordinates and attitude angles output segment by segment from the mechanical model are used as the predictors of the state transition equation, while the three-dimensional coordinates and attitude angles measured by the measurement-while-drilling instrument are used as the observations of the observation equation. The process noise covariance matrix is ​​calibrated according to the modeling error of the drill pipe mechanical model, and the observation noise covariance matrix is ​​set according to the factory accuracy specifications of the measurement-while-drilling instrument. After filtering and fusion, a continuous set of coordinates and tangential vectors for the estimated bottom axis value is generated.

[0033] S2. Establish the deviation field and calculate the deviation compensation amount and transition constraint adjustment.

[0034] A spatial interpolation algorithm based on piecewise cubic Hermite polynomials is employed to resample asynchronously acquired trajectory measurements and the estimated borehole axis values ​​calculated based on model accumulation. These are then strictly aligned to standard borehole depth nodes with a spacing of 0.1m, eliminating the temporal difference between the sensor acquisition frequency and the model calculation step size. A time warping algorithm is used to eliminate the phase delay between the two in the data acquisition time.

[0035] At each aligned standard hole depth node, the difference between the actual measured azimuth and the azimuth corresponding to the model estimate is used to obtain the azimuth deviation; the difference between the actual measured dip and the dip corresponding to the model estimate is used to obtain the dip deviation. The accuracy of the azimuth and dip deviations is on the order of ±0.01°, and the positive and negative signs are retained to indicate the direction of deviation.

[0036] Extract the tangential vectors of the measured trajectory and the estimated trajectory at the current borehole depth node, calculate the cross product of the two tangential vectors, use the direction of the cross product to represent the direction of the bending phase difference, and use the numerical value of the angle between the two vectors to represent the magnitude of the bending phase difference, together constituting the bending phase difference. The azimuth deviation, tilt deviation, bending phase difference, and their corresponding one-dimensional borehole depth scalars are then combined into the following form: The vector combination is continuously stored and updated in the database in matrix form to obtain a four-dimensional deviation field representing the real-time evolution of the comprehensive deviation of the borehole in three-dimensional space with drilling depth. This is due to orientation deviation. For tilt angle deviation, For the curved phase difference, This indicates the depth of the hole.

[0037] The Min-Max normalization method was used to proportionally map the real-time acquired propulsion pressure, rotational torque, and drill pipe triaxial acceleration vibration amplitude to the [0, 1] interval. Simultaneously, abrasive parameters such as the Protodyakonov coefficient and quartz content, representing formation hardness, were extracted from the current borehole logging data. Since different lithologies result in varying pressure resistance and cutting resistance to the drill bit, the normalized propulsion pressure... Rotational torque Drill pipe vibration amplitude Formation hardness and grinding parameters Substitute into the multivariate linear regression equation Calculation of the formation effect index In the formula to The regression coefficients were obtained using the least squares method with a real drilling dataset. The formation effect index ranges from [0.5, 2.5] and is used to characterize the combined effect of the formation on the drilling process under different lithological conditions. The larger the formation effect index, the stronger the resistance of the current rock formation to the drill string, and a larger correction range must be allocated during deviation compensation.

[0038] Taking the calibration results of a real drilling dataset of alternating sandstone and mudstone formations in a coal mine as an example, a set of regression coefficients obtained by fitting using the least squares method are: It is 0.12. It is 0.35. It is 0.28. It is 0.15. It is 0.42. The value is 0.18. The following working conditions were used as a sampling point: normalized propulsion pressure of 0.6, slewing torque of 0.45, drill pipe vibration amplitude of 0.3, formation hardness of 0.7, and abrasiveness parameter of 0.5. These conditions were then substituted into the above regression equation for verification: The calculated result falls within the range of [0.5, 2.5], indicating that the overall intensity of the rock strata's effect on the drilling tool at the current sampling time is at a moderately low level, and the correction range allocated during deviation compensation is appropriate. Regression coefficients must be refitted for different mining areas based on their specific geological conditions and drilling rig models.

[0039] The bending phase difference data of the continuous hole depth range in the deviation field is retrieved and subjected to a first-order difference operation. The resulting phase difference change is divided by a hole depth step of 0.1m to obtain the continuous gradient per unit length. The absolute value of the continuous gradient is then taken. Then an exponential decay function was adopted. Map the gradient sequence to the interval (0, 1], where A preset attenuation coefficient is used to map and generate a dimensionless phase migration coefficient. The more drastic the change in the bending phase difference along the hole depth direction, the larger the absolute value of the continuity gradient. After exponential attenuation, the phase migration coefficient approaches zero, meaning the contribution of the bending phase to the compensation is suppressed. When the bending phase difference transitions smoothly along the hole depth direction, the phase migration coefficient approaches 1, and the compensation fully inherits the bending phase information. Attenuation coefficient The average lithological variation frequency of the drilled strata was obtained through calibration experiments in this embodiment. It is 2. When the value is too low (below 1), the attenuation force is insufficient, and compensating oscillations are prone to occur at the interlayer interface; Within the range of [1, 4], the suppression effect in the region of dramatic changes in bending phase difference can be balanced with the compensation sensitivity in the normal drilling region. In other embodiments, the range of [1, 4] can be adjusted according to the alternation frequency of the formation in the target mining area.

[0040] The four-dimensional deviation field is divided into multiple historical compensation segments according to the historical borehole depth direction. In this embodiment, it is divided into 5 compensation segments, each with a length of 0.5m. The number of compensation segments is determined through calibration experiments based on the alternation spacing of the strata in the mining area. In other embodiments, it can be adjusted between 3 and 8 segments according to the actual working conditions. A dimensionless distance weighting factor is assigned to each compensation segment based on an exponential law, which increases exponentially with the absolute value of the borehole depth, calculated using the following formula: ; In the formula For the first The distance weighting factor for each compensation segment. The value is increased to compensate for the segment number, and the closer it is to the current drilling face, the larger the value. The growth rate coefficient is obtained through calibration experiments. Compensation sections closer to the current drilling face receive higher weights, especially under conditions of frequent alternation between sandstone and mudstone during cross-strata drilling. The value range is [0.3, 1.5]. In this embodiment... Take 0.8. When the weights are too low, the differences between the compensation segments tend to flatten out, the sensitivity of the compensation amount to recent deviation changes decreases, and the correction action becomes lagging. When the weight is set to 0.8, a reasonable weight gradient is formed between the near-end compensation segment and the far-end compensation segment, which takes into account both the response speed of recent deviations and the reference value of historical deviations.

[0041] The distance weighting factor of each compensation segment is multiplied by the azimuth deviation and dip deviation within the corresponding depth interval, and the base compensation amount is obtained by performing piecewise weighted summation, calculated according to the following formula: , ; In the formula This is the azimuth base compensation amount. This is the amount of compensation for the dip angle base. and The first Azimuth and dip deviations within each compensation segment The total number of compensation segments is given. The calculated basement compensation is multiplied by the formation action index and the phase migration coefficient, respectively, to output the bottom hole deviation compensation that simultaneously includes the horizontal azimuth control component and the vertical dip control component, with a calculation accuracy of 0.01°.

[0042] Within each 0.1m borehole depth advancement update cycle, the rate of change of the Euclidean distance between the azimuth and dip deviations of the current borehole depth interval and the previous borehole depth interval is calculated in real time. When this rate of change exceeds a pre-calibrated abrupt change threshold, it indicates that abnormal conditions such as lithological abrupt changes or borehole wall collapse have occurred in adjacent borehole depth intervals. At this time, the bending phase difference of the previous interval is extracted from the memory stack, and the difference between the three-dimensional coordinates of the start and end points of the current borehole depth interval is extracted as the spatial trajectory vector increment. When there is no historical data of the previous interval at the beginning of the drilling stage, the initial value of the bending phase difference is set to zero vector, and the transition constraint is set to 0 accordingly. The system runs in the conventional compensation mode until the deviation data of more than two borehole depth intervals are accumulated. The abrupt change threshold is calibrated through statistical analysis based on the geological conditions of the mining area: the rate of change of the Euclidean distance of the deviation is calculated for each borehole depth interval in the test borehole, and the mean and standard deviation of the rate of change of the normal drilling interval are statistically analyzed. The mean plus twice the standard deviation is taken as the initial value of the abrupt change threshold, and then fine-tuned in combination with the actual response effect of the cross-layer interface position. In this embodiment, the mutation threshold is 0.5° / m. A value within the range of [0.2, 1]° / m can balance the tolerance for normal drilling fluctuations with the timeliness of the response to lithological mutations. In other embodiments, the threshold can be adjusted within the range of [0.2, 1]° / m according to the frequency of strata crossing.

[0043] Based on the cross product of the estimated tangential vector of the current hole bottom axis and the measured tangential vector of the current trajectory, a three-dimensional bending phase direction vector is obtained after normalization. The three-dimensional bending phase direction vector is then multiplied by the spatial trajectory vector increment. The absolute value of the multiplication result is divided by the geometric modulus of the spatial trajectory increment to complete the normalization operation, generating a dimensionless and non-negative transition constraint quantity constrained within the interval [0, 1]. When the geometric modulus of the spatial trajectory increment is less than the preset minimum displacement threshold, the transition constraint is set to 0, and the constraint adjustment calculation for the current interval is skipped. From a physical perspective, the transition constraint represents the degree of consistency between the bending phase direction and the drilling direction: when the two directions tend to be consistent, the transition constraint is close to 1, the phase migration coefficient attenuation is minimal, and the compensation amplitude remains at a relatively large proportion; when the two directions tend to be orthogonal, the transition constraint is close to 0, the phase migration coefficient attenuation is maximum, and the compensation amplitude is significantly reduced, avoiding the generation of erroneous compensation commands in the lithological abrupt change interval.

[0044] Using natural constants The exponential function model with base is used to multiply the original phase shift coefficient by the attenuation penalty factor. The attenuation control coefficient Taking a value of 1.2, the adjusted phase shift coefficient is calculated. Attenuation control coefficient. Taking 1.2 as a compromise in engineering, a value within the range of [0.8, 2] can balance the penalty intensity of the abrupt change interval with the compensation sensitivity of the normal drilling interval. In other embodiments, the value can be adjusted within the range of [0.8, 2] according to the intensity of the formation abrupt change.

[0045] Transition constraint Multiplying the original hole bottom deviation compensation amount by a proportional coefficient, the amplitudes of the horizontal azimuth compensation component and the vertical inclination compensation component are adjusted proportionally to generate the adjusted hole bottom deviation compensation amount.

[0046] S3. Calculate the target's corrected attitude and generate constrained tracking commands.

[0047] The host computer controller uses the differential geometric difference method to differentiate the tangent and normal of the preset three-dimensional spiral trajectory at the current actual hole depth, obtaining the tangential and normal vectors of the local preset trajectory. Based on the tangential vector, normal vector, and the binormal vector obtained by their cross product, a Frenet frame of spatial curves is constructed. When the curvature of the preset trajectory is zero, the principal normal is replaced by the tangential vector of the previous normal node combined with the local normal obtained by least-squares fitting, to ensure the continuous generation of the attitude guidance reference vector. The tangential vector is then normalized into an attitude guidance reference vector of length 1.

[0048] Under the established unified pose space reference system of Northeast Sky, the azimuth and tilt angles in the hole bottom attitude state variables are converted into two-dimensional attitude vectors. The horizontal azimuth compensation component in the adjusted hole bottom deviation compensation variable is mapped to the azimuth correction variable, and the vertical tilt compensation component is mapped to the tilt correction variable. Combined with the attitude guidance reference vector, vector fusion is performed on the above three vectors in the unified two-dimensional attitude space to obtain the initial target corrected attitude.

[0049] The azimuth and tilt differences between the initial target corrected attitude and the actual hole bottom attitude state in the previous control cycle are calculated and combined to form a two-dimensional attitude offset. A scrolling window is used to read the attitude offset sequence data from multiple historical servo control cycles. In this embodiment, the window length is... Take 10 control cycles, The value range is [5, 20]. When the skewness is small, the statistical sample is insufficient, the mean square value of the rate of change of the skewness fluctuates greatly, and the accumulation factor is overly sensitive to instantaneous disturbances. When the window coverage time is too long, the accumulation factor lags in response to recent attitude changes. Therefore, the offset change rate for each cycle is calculated using a first-order difference method, dividing the difference in attitude offset between adjacent control cycles by the servo control cycle duration. The maximum permissible deflection angle limit of the system, calibrated at the drilling rig's factory, is retrieved, and the mean square value of the offset change rate within the window period is first calculated. : ; Then the mean square value After taking the root mean square (RMS) value, divide it by the maximum permissible deflection rate to generate a dimensionless attitude accumulation factor with a value in the interval [0, 1]. : ; In the formula For the first The rate of change of attitude deviation per control cycle, in degrees per second. The number of control cycles contained in the scrolling window. This is the maximum permissible deflection angle of the drilling rig system. This refers to the duration of the servo control cycle. The physical meaning is the maximum allowable deflection angular velocity of the system within a single control cycle, and the molecular... Dimensions are the same (° / s), after division It is dimensionless. When the calculation result is greater than 1, it will be... The clamp is set to 1 to ensure that the attitude accumulation factor is always constrained within the range of [0, 1]. The attitude accumulation factor is specifically used to characterize the high-frequency resonance degree and command oscillation state of the system's servo components: the closer the attitude accumulation factor is to 1, the higher the frequency and amplitude of recent attitude adjustments, and the greater the risk of high-frequency oscillation; when the attitude accumulation factor is close to 0, the system is in a stable tracking state.

[0050] Upon entering the final output stage, the controller utilizes a nonlinear saturation function to construct a limiting filter. This saturation function possesses smooth soft-inflection point characteristics, multiplying the incremental portion of the command adjustment within the initial target attitude correction by... The mapped attenuation ratio is used for command suppression processing. After amplitude limiting suppression, the output is the target attitude correction after suppression adjustment.

[0051] The azimuth correction component and the tilt correction component in the suppressed and adjusted target attitude are combined into a two-dimensional input vector, which is then multiplied by a pre-calibrated drilling rig dynamic mapping matrix. The output consists of three-dimensional control vectors, corresponding to the desired propulsion speed, desired yaw speed, and orientation angle control values, respectively. Mapping matrix. It is a 3x2 matrix. The matrix elements are obtained by least squares regression through multi-condition drilling calibration tests in the target mining area, using different attitude command inputs and corresponding drilling rig execution response outputs as samples.

[0052] Taking the calibration results of a certain mining area as an example, the mapping matrix The values ​​are: the first row has elements of 1.25 and -0.38, the second row has elements of 0.42 and 1.15, and the third row has elements of 0.85 and 0.62. Substituting the azimuth correction component of 0.8° and the dip correction component of -0.5° into a two-dimensional input vector, the desired propulsion speed is calculated to be 1.25×0.8+(-0.38)×(-0.5)=1.19m / min, and the desired slewing speed is 0.42×0.8+1.15×(-0.5)=-0.239r / min. The negative sign indicates that the slewing direction is opposite to the positive direction. In actual control, the absolute value of 0.239r / min is taken and the direction indicated by the negative sign is followed. The orientation angle control value is 0.85×0.8+0.62×(-0.5)=0.37°. Different mining areas require recalibration based on the drilling rig response characteristics of that area.

[0053] Because the inverse kinematic mapping does not take into account the capacity limitations of the power source, a power coordination check must be performed before outputting electrical commands. The measured feedback pressure of the current propulsion hydraulic cylinder (in MPa) and the measured cutting torque of the rotary motor (in N·m) are read in real time. The measured pressure is multiplied by the effective working area of ​​the cylinder, then multiplied by the desired propulsion speed (in m / min), and the propulsion power (in kW) is obtained after unit conversion. Multiply the measured torque (in N·m) by the desired rotational speed (in r / min), and then divide by the power conversion constant 9550 to obtain the rotational power (in kW). In the formula, 9550 is the engineering coefficient for converting N·m and r / min to kW, and the total power demand of the system is obtained by summing the propulsion power and the slewing power.

[0054] The total power demand of the system is compared with the rated safe power of the drilling rig. If the total power demand exceeds the rated safe power, the ratio of the rated safe power to the total power demand is calculated, and a derating ratio of less than 1 is output. In this embodiment, the rated safe power is set to 30kW.

[0055] Based on the derating ratio, a synchronous proportional reduction is applied to the desired propulsion speed control and the desired rotation speed control. Taking the desired propulsion speed of 1.19 m / min and the desired rotation speed of 0.239 r / min calculated by the above matrix as an example, assuming the current measured feedback pressure is 12 MPa, the effective working area of ​​the hydraulic cylinder is 0.02 m², and the measured cutting torque is 1200 N·m, then the propulsion power... kW, rotational power The total system power requirement is 4.79kW, which does not exceed the rated safe power of 30kW, so no limiting is required. When drilling encounters high-hardness rock formations, causing the measured feedback pressure to rise to 25MPa, the measured cutting torque to rise to 2500N·m, the desired advance speed to be 1.5m / min, and the desired rotation speed to be 150r / min, the advance power... kW, rotational power The total power demand is 51.77kW, exceeding the rated safe power by 30kW. The derating ratio is [value missing]. The expected propulsion speed after amplitude limiting is m / min, the desired rotational speed after limiting is The total output power of the dual hydraulic circuits is constrained within the rated range of 30kW (r / min). The directional angle control signal is a pure hydraulic fine-tuning signal, which is not involved in power calculation and does not require reduction.

[0056] The limited advance speed, rotation speed, and orientation angle control values ​​are packaged into a constrained path tracking command that satisfies the physical constraints of the drilling rig. The constrained path tracking command is converted into an analog voltage signal by a programmable logic controller (PLC) and sent to the drilling rig's actuators. Electromagnetic proportional valves control the hydraulic cylinders and rotary motors to perform the corresponding actions. The controller monitors the borehole depth changes in real time, continuously repeating the above calculation and control steps to perform cyclic control until the current depth value recorded by the drilling measurement instrument reaches the set target borehole depth value.

[0057] Reference Figure 2 The schematic diagram of the evolution and compensation convergence curve of the four-dimensional deviation field with hole depth includes the trajectory deviation curve after correction using the control method of this invention and the original trajectory deviation curve without correction. As the hole depth increases, this invention can continuously constrain the azimuth deviation and dip angle deviation within a small range, and the deviation does not diverge significantly during the drilling process.

[0058] To verify the comprehensive effectiveness of this invention in borehole trajectory control and drilling rig operation protection, a control experiment was conducted under complex cross-strata geological conditions. The experimental conditions were set as cross-strata drilling operations with a target depth of 300m, involving frequent alternating abrupt interfaces between sandstone and mudstone along the route, and the drilling rig system's rated safe power was 30kW. Four comparison groups were set up: the first group was the basic group using traditional proportional-integral-differential trajectory tracking; the second group retained the basic framework but removed the abrupt change constraints and high-frequency jitter suppression module; the third group removed the power joint amplitude limiting control module; and the fourth group was the complete scheme using all features of this invention.

[0059] Independent drilling tests were conducted under identical operating conditions, and key performance indicators were recorded. The target deviation in the basic group was 56.4 cm, with 18 instances of high-frequency oscillation overshoot per 100 meters and 5 instances of overload shutdown. In contrast, the target deviation in Group 1 decreased to 23.5 cm, and no overload shutdowns occurred, but the high-frequency oscillation overshoot per 100 meters still reached 14 times. This was due to the lack of abrupt change constraints and high-frequency suppression mechanisms, leading to frequent attitude overshoot and vibration of the drill string at the lithological interface. In contrast, the target deviation in Group 2 was 15.8 cm, and the high-frequency oscillation overshoot per 100 meters decreased to 2 times, but overload shutdowns occurred 4 times, frequently exceeding power limits when encountering hard rock without applying combined amplitude limiting. The target deviation in the complete solution group was only 8.2 cm, the high-frequency oscillation overshoot per 100 meters decreased to 0 times, and the number of overload shutdowns was also 0, with an average drilling efficiency improvement of 27% compared to the basic group.

[0060] Reference Figure 3 The horizontal axis of the comparative experimental results of cross-strata drilling under different control strategies represents, in order: traditional PID control group, group with removal of abrupt change constraints and high-frequency suppression group, group with removal of power joint limiting group, and the complete scheme group of this invention. The vertical axis corresponds to the target deviation, the number of high-frequency oscillation overshoots per 100 meters, and the number of overload shutdowns, respectively. The complete scheme group, through the triple coupling of deviation compensation, abrupt change anti-shaking mechanism based on Euclidean distance change rate, and power joint limiting, absorbs the disturbances caused by abrupt changes in the strata, eliminates mechanical high-frequency resonance damage, and curbs the risk of hydraulic system overload from the source. Under complex cross-strata geological conditions, the target deviation is controlled within the centimeter range, and the drilling process is stable and continuous.

[0061] It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept, and these all fall within the scope of protection of this invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for intelligent control of mine water exploration and drainage boreholes, characterized in that, include: S1. Obtain drilling rig propulsion pressure, rotation torque, drill rod vibration, trajectory measurement values ​​and formation parameters. Construct hole bottom attitude state quantities based on preset trajectory, and calculate the hole bottom axis estimate value by combining drill rod mechanical deformation model. S2. Establish a deviation field containing azimuth deviation, dip angle deviation, bending phase difference and hole depth position by benchmarking and time-series registration of trajectory measurement values ​​and axis estimation values. Calculate the formation effect index based on propulsion pressure, slewing torque, drill pipe vibration and formation parameters. Calculate the phase migration coefficient by combining the bending phase difference between adjacent hole depth intervals. Multiply the deviation field by the formation effect index and phase migration coefficient after weighted summation according to hole depth to obtain the deviation compensation amount. When the deviation change rate exceeds the threshold, calculate the transition constraint amount to adjust the compensation amount. S3. Calculate the target correction attitude based on the deviation compensation amount, the hole bottom attitude state amount and the preset trajectory curvature. Construct a dimensionless attitude accumulation factor to limit the attenuation of the target correction attitude based on the attitude offset amount. Obtain the propulsion speed, rotation speed and orientation angle control amount through the drilling rig dynamic mapping matrix transformation. When the sum of propulsion power and rotation power exceeds the rated power, apply joint amplitude limiting modulation to generate a constrained path tracking command, and cyclically control until the set hole depth.

2. The intelligent control method for mine water exploration and drainage boreholes according to claim 1, characterized in that, The process of constructing the hole bottom attitude state variables based on a preset trajectory and calculating the estimated value of the hole bottom axis using a drill pipe mechanical deformation model includes: Extract the three-dimensional coordinates, azimuth angle, and inclination angle of the current actual hole bottom position as hole bottom attitude state variables; The drill pipe is divided into multiple discrete bending segments. Based on the obtained pushing pressure and rotation torque, the axial force and lateral force on each discrete bending segment in the hole are calculated. Based on the axial force, lateral force, and bending stiffness of the drill pipe, the deflection and rotation angle changes of each discrete bending segment are calculated using the force-deformation equation.

3. The intelligent control method for mine water exploration and drainage boreholes according to claim 2, characterized in that, The estimated value of the bottom axis of the calculated hole also includes: The relative attitude transformation matrix between adjacent discrete bending segments is calculated based on the change in the rotation angle of each discrete bending segment. The coordinate transformation and spatial recursion accumulation are performed along the hole depth direction in combination with the flexural deformation. Then, the estimated value of the hole bottom axis is calculated by combining the hole bottom attitude state quantity. Kalman filtering is used to fuse the estimated bottom axis of the hole obtained by spatial recursion and accumulation with the measured attitude state of the bottom of the hole, generating a continuous coordinate set of estimated bottom axis of the hole and a set of tangential vectors.

4. The intelligent control method for mine water exploration and drainage boreholes according to claim 1, characterized in that, The method for obtaining each deviation component in the deviation field includes: aligning the trajectory measurement value and the estimated value of the borehole bottom axis to the same borehole depth node through spatial interpolation to eliminate the temporal difference between the sensor acquisition frequency and the model calculation step size; at each borehole depth node, calculating the azimuth angle difference and the tilt angle difference between the trajectory measurement value and the estimated value of the borehole bottom axis, which are respectively used as the azimuth deviation and the tilt angle deviation.

5. The intelligent control method for mine water exploration and drainage boreholes according to claim 4, characterized in that, Establishing the deviation field also includes: Based on the spatial geometric characteristics of the borehole trajectory, the cross product of the tangential vector of the trajectory measurement value and the tangential vector of the estimated borehole bottom axis value is calculated. The direction of the cross product is used to characterize the direction of the bending phase difference, and the magnitude of the bending phase difference is used to characterize the angle between the two vectors. A four-dimensional deviation field is constructed by vector combining the azimuth deviation, tilt deviation, and bending phase difference with the corresponding hole depth nodes.

6. The intelligent control method for mine water exploration and drainage boreholes according to claim 1, characterized in that, The step of obtaining the deviation compensation amount by multiplying the weighted sum of the deviation field segmented by borehole depth with the formation interaction index and the phase migration coefficient includes: dimensionlessly processing the amplitudes of the propulsion pressure, slewing torque, and drill pipe vibration, and calculating the formation interaction index through a multivariate regression equation in conjunction with formation hardness and abrasiveness parameters; performing differential calculation on the bending phase difference between adjacent borehole depth intervals in the four-dimensional deviation field, using the ratio of the differential calculation result to the borehole depth step size as the continuity gradient, and transforming it into a dimensionless phase migration coefficient after normalization; dividing the four-dimensional deviation field into multiple historical compensation segments according to borehole depth, assigning a dimensionless distance weight factor that increases with increasing borehole depth to each segment, multiplying the distance weight factor with the azimuth deviation and dip angle deviation, performing segmented weighted summation, and multiplying the summation result with the formation interaction index and the phase migration coefficient to obtain the bottom borehole deviation compensation amount.

7. The intelligent control method for mine water exploration and drainage boreholes according to claim 1, characterized in that, The adjustment compensation amount includes: The rate of change of the Euclidean distance between the azimuth and inclination deviations within adjacent borehole depth intervals with respect to borehole depth is calculated. When the rate of change exceeds the abrupt change threshold, the bending phase difference of the previous interval and the spatial trajectory vector increment of the current interval are extracted. In the initial stage, the bending phase difference is taken as a zero vector. The cross product of the estimated tangential vector of the borehole bottom axis and the measured tangential vector of the trajectory is normalized to obtain the phase direction vector. The absolute value of the inner product of the phase direction vector and the trajectory vector increment is divided by the magnitude of the trajectory vector increment to obtain the dimensionless transition constraint. The phase migration coefficient is adjusted by exponential decay using the result of subtracting the transition constraint from 1. The adjusted deviation compensation is obtained by multiplying the transition constraint by the original deviation compensation.

8. The intelligent control method for mine water exploration and drainage boreholes according to claim 1, characterized in that, The step of constructing a dimensionless attitude accumulation factor based on the attitude offset to limit the attenuation of the target correction attitude includes: constructing a spatial curve Frenet frame based on the curvature, tangential vector, and normal vector of the preset trajectory to generate an attitude guidance reference vector; converting the bottom hole attitude state quantity into a two-dimensional attitude vector; mapping the adjusted bottom hole deviation compensation quantity into azimuth correction quantity and inclination correction quantity; and fusing the attitude guidance reference vector to obtain the initial target correction attitude; calculating the azimuth and inclination differences between the initial target correction attitude and the bottom hole attitude state quantity of the previous control cycle as a two-dimensional attitude offset; constructing an attitude accumulation factor based on the rate of change of attitude offsets in multiple historical control cycles, the duration of the servo control cycle, and the maximum allowable deflection angle of the drilling rig; multiplying the command adjustment increment of the initial target correction attitude by the attenuation ratio obtained by nonlinear function transformation of the attitude accumulation factor to perform amplitude limiting suppression; and outputting the suppressed and adjusted target correction attitude.

9. The intelligent control method for mine water exploration and drainage boreholes according to claim 1, characterized in that, The generation of constrained path tracking instructions includes: converting the suppressed and adjusted target correction attitude into the desired propulsion speed, slewing speed and orientation angle control values ​​through the drilling rig dynamic mapping matrix; When the sum of propulsion power and slewing power exceeds the system's rated power, the ratio of the system's rated power to the sum of propulsion power and slewing power is calculated as the derating ratio. The propulsion power is calculated based on the product of propulsion pressure, the effective working area of ​​the hydraulic cylinder, and the propulsion speed control quantity. The slewing power is calculated based on the product of slewing torque and slewing speed control quantity. According to the derating ratio, the propulsion speed control quantity and the slewing speed control quantity are reduced by the same proportion. The reduced propulsion speed, slewing speed, and orientation angle control quantity are packaged to generate a constrained path tracking command.

10. The intelligent control method for mine water exploration and drainage boreholes according to claim 1, characterized in that, The acquisition of drilling rig propulsion pressure, slewing torque, drill rod vibration, trajectory measurement values, and formation parameters also includes acquiring the current propulsion speed, slewing speed, and orientation angle of the drilling rig; the constrained path tracking command is converted into an analog voltage signal by a programmable logic controller and sent to the drilling rig actuator; the borehole depth change is monitored in real time and cyclic control is executed until the set target borehole depth is reached.

Images