A high-precision while-drilling curve-guiding grouting system

Abstract

The application relates to the technical field of drilling, and discloses a high-precision while-drilling curve-guided grouting system, which generates a three-dimensional stratum coupling model based on a three-dimensional coordinate system through a stratum resistivity model and a stratum wave velocity model; an initial grouting pressure, an initial slurry ratio and an initial drilling path are obtained according to the three-dimensional stratum coupling model; a multi-target reward function is constructed to output an optimized grouting pressure, an optimized slurry ratio and an optimized drilling path; and an actuator adjusts the grouting pressure, the slurry ratio and the drilling path according to the output instruction of the reinforcement learning control kernel to realize closed-loop control of the curve grouting path. The application is favorable for solving the problems of great grouting difficulty and poor grouting technical effect in the prior art, and the curve grouting reinforcement technology combines curve drilling and grouting reinforcement technology, can avoid the obstruction of ground buildings and underground pipelines, and has the advantages of small engineering quantity and low cost.

Description

Technical Field

[0001] This invention relates to the field of drilling technology, and in particular to a high-precision grouted system for guiding curves while drilling. Background Technology

[0002] Grouting technology refers to the technique of injecting a fluid, gelling slurry (such as cement slurry, chemical slurry, water glass, etc.) into the target stratum, fissures, cavities or structural defects through drilling, grouting pipes or other equipment, so that it diffuses and solidifies, thereby changing the physical properties (such as strength and permeability) of the stratum or structure or repairing defects.

[0003] Grouting technology has a wide range of applications: in subway construction, when a tunnel passes through a water-rich sand layer, a water-stop ring is formed by grouting around the perimeter to prevent water inrush; in foundation pit engineering, grouting is applied to the soil at the bottom of the pit to prevent the surrounding settlement caused by vacuum preloading; in building construction, grouting is applied to reinforce the foundation to avoid damage to the building due to foundation settlement.

[0004] Traditional grouting techniques are divided into horizontal grouting and vertical grouting. Horizontal grouting injects grout into the formation through horizontal boreholes or horizontal grouting pipes to form a horizontal reinforcement or water-stopping curtain. Vertical grouting injects grout into the formation through vertical boreholes or grouting pipes to form a vertical reinforcement or water-stopping structure.

[0005] Horizontal and vertical grouting can only diffuse in a fixed direction. However, underground structures are often very complex, and diffusion in a fixed direction cannot bypass obstacles in the strata. Furthermore, in complex strata, the grout may not diffuse evenly due to large differences in permeability, resulting in reinforcement blind spots. Moreover, when diffusing grout in a fixed direction, it may rapidly flow away in a single direction, leading to insufficient effective reinforcement radius and inefficient grout utilization, resulting in material waste. In addition, horizontal and vertical grouting require densely arranged boreholes, resulting in low construction efficiency and high costs. Furthermore, vertical or horizontal grouting is difficult to match with the stratum interface, leading to poor reinforcement effects.

[0006] Curved grouting is a grouting technique used in tunnels, underground engineering, or geotechnical engineering. Its core characteristic is that the grouting path (such as boreholes or grouting pipes) is curved, rather than a traditional straight borehole. This technique is mainly used to solve reinforcement, seepage prevention, or ground stability problems under complex geological conditions, and is particularly suitable for scenarios that require bypassing obstacles (such as existing structures or pipelines) or for precise grouting of specific areas.

[0007] While curve grouting can effectively solve problems related to complex geology and limited space, its technical difficulty is significantly higher than that of traditional straight-line grouting. It requires real-time adjustments to the drill bit direction to maintain the preset curve path, and is prone to deviation from the target, especially in hard rock or strata with alternating soft and hard surfaces. Therefore, curve grouting relies heavily on human experience for guidance, resulting in high grouting difficulty and unsatisfactory results. Summary of the Invention

[0008] The main objective of this invention is to provide a high-precision grouted curve grouting system that aims to solve the technical problems of high grouting difficulty and poor grouting effect in existing curve grouting technologies.

[0009] To achieve the above objectives, the present invention provides a high-precision drilling directional curve grouting system, comprising:

[0010] The data fusion module is used to generate a three-dimensional coupled formation model based on a three-dimensional coordinate system by using the formation resistivity model and the formation wave velocity model.

[0011] The parameter acquisition module is used to: acquire the initial grouting pressure, initial grout mix ratio, and initial borehole path determined based on the three-dimensional formation coupling model;

[0012] The reinforcement learning control kernel is used to measure the quality of real-time grouting actions based on the formation electrical parameters in the formation resistivity model, the formation mechanical parameters in the formation wave velocity model, the real-time engineering status data of the grouting construction process, and the constructed multi-objective reward function, so as to generate optimized control actions and output optimized grouting pressure, optimized grout ratio and optimized borehole path.

[0013] The actuator is used to adjust the grouting pressure, grout ratio, and drilling path according to the output instructions of the reinforcement learning control kernel to achieve closed-loop control of the curved grouting path.

[0014] Optionally, the system further includes:

[0015] The data acquisition module includes a resistivity imaging module and a seismic wave inversion module. The resistivity imaging module is used to acquire the stratum electrical parameters of the grouting area to form a stratum resistivity model, and the seismic wave inversion module is used to acquire the stratum mechanical parameters of the grouting area to form a stratum wave velocity model.

[0016] Optionally, the resistivity imaging module includes:

[0017] The apparent resistance calculation unit is used to obtain the potential difference collected after arranging the electrode array in the grouting area at a set resolution, so as to perform apparent resistance calculation.

[0018] The first conversion unit is used to reverse reconstruct the apparent resistivity data to convert it into a formation resistivity model that represents the true resistivity distribution of the formation, so as to obtain a three-dimensional resistivity image of the grouting area.

[0019] The seismic wave inversion module includes:

[0020] The data acquisition unit is used to acquire the travel time and waveform data of seismic waves generated when the source vibrates within the grouting area.

[0021] The building unit is used to construct the initial formation wave velocity model and update the velocity distribution through travel-time inversion.

[0022] The second conversion unit is used to reverse reconstruct the initial formation wave velocity model to convert it into a formation wave velocity model that represents the true formation wave velocity distribution, so as to obtain a three-dimensional image of the formation wave velocity in the grouting area.

[0023] Optionally, the data fusion module includes:

[0024] Alignment cells are used to map the formation resistivity model and the formation wave velocity model to the same three-dimensional coordinate system and to perform resolution alignment between the formation resistivity model and the formation wave velocity model.

[0025] The first calculation unit is used to establish the joint probability distribution of resistivity and wave velocity;

[0026] The second calculation unit is used to establish the statistical relationship between resistivity and wave velocity, and to describe the spatial correlation between resistivity and wave velocity.

[0027] Sample units are used to sample from the posterior distribution to form a sample set;

[0028] The third computational unit is used to calculate the posterior mean as model parameters in the three-dimensional stratigraphic coupling model.

[0029] Optionally, the joint probability distribution of resistivity and wave velocity can be established using the following formula:

[0030] Define the prior distribution:

[0031] ;

[0032] The likelihood function is:

[0033] ;

[0034] ;

[0035] The posterior distribution is:

[0036] ;

[0037] Where p is the joint prior distribution; To represent the true formation resistivity distribution, For the longitudinal wave velocity, The transverse wave velocity; For observational data; This is a vector of resistivity observation data. For the qth resistivity observation, This represents the total amount of observation data in resistivity imaging measurements. This is the response of the resistivity forward model. This represents the variance of the resistivity data error. This is a vector of seismic wave observation data. For the first e Each seismic wave observation value This represents the total amount of observation data in seismic wave measurements. The response of the seismic wave forward model. The variance of the seismic wave rate data. It is the probability density function of a normal distribution.

[0038] Optionally, the statistical relationship between resistivity and wave velocity can be established using the following formula:

[0039] ;

[0040] Where a and b are regression coefficients, For noise; ; The standard deviation of the noise fluctuation range. For reference P-wave velocity, Reference resistivity of the formation;

[0041] The spatial correlation between resistivity and wave velocity is described by the following formula:

[0042] ;

[0043] in, This represents the spatial correlation between the resistivity of the k-th grid point and the wave velocity of the j-th grid point in a three-dimensional stratigraphic coupling model. This represents the three-dimensional coordinates of the k-th grid point. Represents the three-dimensional coordinates of the j-th grid point; This represents the actual spatial distance between the k-th grid point and the j-th grid point; L is the relevant length.

[0044] Alternatively, sampling can be performed from the posterior distribution using the following formula:

[0045] ;

[0046] Generate sample set ;

[0047] in, These are candidate samples for resistivity. Candidate samples for P-wave velocity. Candidate samples for shear wave velocity; Let be the resistivity accepted in the m-th iteration. Let be the longitudinal wave velocity received in the m-th iteration. The velocity of the transverse wave received during the m-th iteration; , Used to construct the posterior distribution; Let M represent the posterior probability density, and M be the number of iterations.

[0048] Optionally, the third calculation unit is further configured to: calculate the variance field quantization uncertainty and output the coupling parameter field;

[0049] The parameter acquisition module is used for:

[0050] Obtain preset wave velocity and resistivity thresholds, and locate weak areas in the three-dimensional stratigraphic coupling model based on the wave velocity and resistivity thresholds.

[0051] Based on the weak areas, an initial drilling path is formed;

[0052] The regional geostress parameters of the weak zone are obtained by calling the geological database.

[0053] The initial grouting pressure and initial grout mix ratio are calculated based on the regional geostress parameters.

[0054] Optionally, in the reinforcement learning control kernel:

[0055] Encode formation parameters and engineering states into state vectors for reinforcement learning. :

[0056] ;

[0057] in, This is the resistivity value at the current location. The longitudinal wave velocity at the current position. The elastic modulus at the current location. The Poisson ratio at the current position. The grouting pressure of the previous time period, This refers to the slurry mix ratio from the previous time period. This is a sequence of spatial coordinates of historical borehole paths recorded up to the previous time point;

[0058] Multi-objective reward function for:

[0059] ;

[0060] ;

[0061] ;

[0062] ;

[0063] in, , and These are the weighting coefficients; As a reward for the grouting effect, As a cost penalty, As a risk penalty; This represents the change in resistivity after grouting. This represents the change in longitudinal waves after grouting. and These are the grouting effect bonus coefficients; This is the initial resistivity reference value; This is the initial P-wave velocity reference value; and These are cost penalty coefficients; For grouting pressure, For slurry preparation ratio, As the reference grouting pressure, The standard slurry mix ratio; The severity of the risk penalty; For indicator functions, The value is 1 at time. The value is 0 at that time; This represents the maximum formation pressure under the current grouting pressure. Tensile strength of the stratum;

[0064] The formula for the advantage function, which measures the superiority of an action relative to the average level, is:

[0065] ;

[0066] in, Let g be the dominance function at the current time t, g be the g-th time step in the future, and T be the upper limit of the time step; As a discount factor, The instant reward for the g-th time step in the future. For use in evaluating state vectors Value assessment network;

[0067] in, This indicates that the action was better than expected, and similar actions should be strengthened. This indicates that the action was ineffective and that similar actions should be reduced.

[0068] The formula for updating the policy gradient based on the advantage function is:

[0069] ;

[0070] in, For policy gradient, Represent the expected function; This represents the probability distribution of actions under the new strategy; This represents the action vector, which includes the increments of grouting pressure, grout mix ratio, and borehole path. ; s is the state vector; This represents the probability distribution of actions under the old strategy. This represents the probability ratio between the new strategy and the old strategy. To truncate the threshold, This represents the truncation function.

[0071] Optionally, the system further includes an adaptive filtering module, which is used to:

[0072] Magnetic field signals in the grouting pipeline are monitored using an optical fiber sensor array.

[0073] The filter cutoff frequency is adjusted in real time based on the formation inhomogeneity.

[0074] The technical solution of this invention is beneficial to solving the technical problems of high difficulty and poor technical effect in grouting technology in the prior art. The specific analysis is as follows: This invention employs a high-precision drilling-guided curve grouting system. The data fusion module is used to establish a three-dimensional coupled formation model based on a three-dimensional coordinate system by combining the formation resistivity model and the formation wave velocity model. This three-dimensional coupled formation model can evaluate the comprehensive formation performance at each location point in the grouting area. Thus, the initial grouting pressure, initial slurry ratio, and initial borehole path determined according to the three-dimensional coupled formation model can be obtained. Furthermore, this system sets up a reinforcement learning control kernel. This kernel, based on the formation electrical parameters in the formation resistivity model, the formation mechanical parameters in the formation wave velocity model, real-time engineering status data during the grouting construction process, and a constructed multi-objective reward function, measures the quality of real-time grouting actions to generate optimized control actions, thereby outputting optimized grouting pressure, optimized slurry ratio, and optimized borehole path. Therefore, according to the output instructions of the reinforcement learning kernel, the grouting pressure, slurry ratio, and borehole path can also be adjusted to achieve closed-loop control of the curve grouting path. In the technical solution of this invention, the grouting parameters and grouting path under the three-dimensional formation coupling conditions are determined and dynamically adjusted based on the three-dimensional formation coupling model and the reinforcement learning control kernel. This is beneficial to optimizing the grouting effect and solving the technical problems of high difficulty and poor effect of grouting technology in the prior art. Attached Figure Description

[0075] Figure 1 This is a schematic diagram of the method flow of the high-precision drilling directional curve grouting system in this invention;

[0076] Figure 2 This is a schematic diagram of the high-precision drilling directional curve grouting system in this invention;

[0077] Figure 3This is a schematic diagram for determining the initial drilling path based on the weak zone.

[0078] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0079] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0080] In the following description, the use of suffixes such as "unit," "component," or "element" to denote elements is solely for the purpose of illustrative purposes and has no specific meaning in itself. Therefore, "unit," "component," or "element" may be used interchangeably.

[0081] Please see Figures 1 to 2 This invention provides a high-precision drilling directional curve grouting system, comprising:

[0082] The data fusion module is used to generate a three-dimensional coupled formation model based on a three-dimensional coordinate system by using the formation resistivity model and the formation wave velocity model.

[0083] The parameter acquisition module is used to: acquire the initial grouting pressure, initial grout mix ratio, and initial borehole path determined based on the three-dimensional formation coupling model;

[0084] The reinforcement learning control kernel is used to measure the quality of real-time grouting actions based on the formation electrical parameters in the formation resistivity model, the formation mechanical parameters in the formation wave velocity model, the real-time engineering status data of the grouting construction process, and the constructed multi-objective reward function, so as to generate optimized control actions and output optimized grouting pressure, optimized grout ratio and optimized borehole path.

[0085] The actuator is used to adjust the grouting pressure, grout ratio, and drilling path according to the output instructions of the reinforcement learning control kernel to achieve closed-loop control of the curved grouting path.

[0086] The technical solution of this invention is beneficial to solving the technical problems of high difficulty and poor technical effect in grouting technology in the prior art. The specific analysis is as follows: This invention employs a high-precision drilling-guided curve grouting system. The data fusion module is used to establish a three-dimensional coupled formation model based on a three-dimensional coordinate system by combining the formation resistivity model and the formation wave velocity model. This three-dimensional coupled formation model can evaluate the comprehensive formation performance at each location point in the grouting area. Thus, the initial grouting pressure, initial slurry ratio, and initial borehole path determined according to the three-dimensional coupled formation model can be obtained. Furthermore, this system sets up a reinforcement learning control kernel. This kernel, based on the formation electrical parameters in the formation resistivity model, the formation mechanical parameters in the formation wave velocity model, real-time engineering status data during the grouting construction process, and a constructed multi-objective reward function, measures the quality of real-time grouting actions to generate optimized control actions, thereby outputting optimized grouting pressure, optimized slurry ratio, and optimized borehole path. Therefore, according to the output instructions of the reinforcement learning kernel, the grouting pressure, slurry ratio, and borehole path can also be adjusted to achieve closed-loop control of the curve grouting path. In the technical solution of this invention, the grouting parameters and grouting path under the three-dimensional formation coupling conditions are determined and dynamically adjusted based on the three-dimensional formation coupling model and the reinforcement learning control kernel. This is beneficial to optimizing the grouting effect and solving the technical problems of high difficulty and poor effect of grouting technology in the prior art.

[0087] Specifically, the system also includes:

[0088] The data acquisition module includes a resistivity imaging module and a seismic wave inversion module. The resistivity imaging module is used to acquire the stratum electrical parameters of the grouting area to form a stratum resistivity model, and the seismic wave inversion module is used to acquire the stratum mechanical parameters of the grouting area to form a stratum wave velocity model.

[0089] Optionally: the resistivity imaging module includes:

[0090] The apparent resistance calculation unit is used to obtain the potential difference collected after arranging the electrode array in the grouting area at a set resolution, so as to perform apparent resistance calculation.

[0091] The first conversion unit is used to reverse reconstruct the apparent resistivity data to convert it into a formation resistivity model that represents the true resistivity distribution of the formation, so as to obtain a three-dimensional resistivity image of the grouting area.

[0092] The formula for calculating the apparent resistivity is:

[0093] ;

[0094] in, Let k be the apparent resistivity and k be the device coefficient. I is the potential difference, and I is the injection current;

[0095] The objective function for reverse reconstruction is:

[0096] ;

[0097] in, This represents the true resistivity distribution of the formation. This is the measured apparent resistivity data vector; To be based on the current resistivity model The theoretical apparent resistivity calculated through forward modeling; This is the data weight matrix; This is the regularization parameter for resistivity inversion; is the regularization matrix for resistivity inversion.

[0098] Among them, the three-dimensional resistivity image of the grouting area is marked with the actual resistivity corresponding to each three-dimensional coordinate point in the grouting area obtained by reverse reconstruction;

[0099] The seismic wave inversion module includes:

[0100] The data acquisition unit is used to acquire the travel time and waveform data of seismic waves generated when the source vibrates within the grouting area.

[0101] The building unit is used to construct the initial formation wave velocity model and update the velocity distribution through travel-time inversion.

[0102] The second conversion unit is used to reverse reconstruct the initial formation wave velocity model to convert it into a formation wave velocity model that represents the true formation wave velocity distribution, so as to obtain a three-dimensional image of the formation wave velocity in the grouting area.

[0103] Among them, the three-dimensional image of the formation wave velocity in the grouting area is marked with the actual formation wave velocity corresponding to each three-dimensional coordinate point in the grouting area;

[0104] The initial formation wave velocity model is as follows:

[0105] ;

[0106] ;

[0107] in, For the longitudinal wave velocity, Where E is the transverse wave velocity; E is the elastic modulus. Poisson's ratio, The volume density of the target formation for grouting;

[0108] The objective function for time-inversion is:

[0109] ;

[0110] in, For the formation wave velocity model to be inverted, for or ; Let r be the travel time observation value of the seismic wave measured for the rth time. R represents the total number of seismic wave travel time observations; Based on the current wave speed model The current timeline is in progress; For seismic wave inversion, the regularization parameter is used. This is the regularization matrix for seismic wave inversion;

[0111] The formula for calculating the forward travel time is:

[0112] ;

[0113] in, For the theory to be outdated, Based on three-dimensional coordinates of spatial location Stratified wave velocity distribution; This represents the minute displacement increment along the path of the seismic wave rays;

[0114] Among them, the three-dimensional image of the formation wave velocity in the grouting area is marked with the formation wave velocity data corresponding to each three-dimensional coordinate point in the area.

[0115] Optionally: the data fusion module includes:

[0116] Alignment cells are used to map the formation resistivity model and the formation wave velocity model to the same three-dimensional coordinate system and to perform resolution alignment between the formation resistivity model and the formation wave velocity model.

[0117] The first calculation unit is used to establish the joint probability distribution of resistivity and wave velocity;

[0118] The second calculation unit is used to establish the statistical relationship between resistivity and wave velocity, and to describe the spatial correlation between resistivity and wave velocity.

[0119] Sample units are used to sample from the posterior distribution to form a sample set;

[0120] The third computational unit is used to calculate the posterior mean as model parameters in the three-dimensional stratigraphic coupling model.

[0121] Optionally: The joint probability distribution of resistivity and wave velocity can be established using the following formula:

[0122] Define the prior distribution:

[0123] ;

[0124] The likelihood function is:

[0125] ;

[0126] ;

[0127] The posterior distribution is:

[0128] ;

[0129] Where p is the joint prior distribution; This represents the true resistivity distribution of the formation. For observational data; This is a vector of resistivity observation data. For the qth resistivity observation, This represents the total amount of observation data in resistivity imaging measurements. This is the response of the resistivity forward model. This represents the variance of the resistivity data error. This is a vector of seismic wave observation data. For the first e Each seismic wave observation value This represents the total amount of observation data in seismic wave measurements. The response of the seismic wave forward model. The variance of the seismic wave rate data. It is the probability density function of a normal distribution.

[0130] Optionally: Establish the statistical relationship between resistivity and wave velocity using the following formula:

[0131] ;

[0132] Where a and b are regression coefficients, For noise; ; The standard deviation of the noise fluctuation range. For reference P-wave velocity, Reference resistivity of the formation;

[0133] The spatial correlation between resistivity and wave velocity is described by the following formula:

[0134] ;

[0135] in, This represents the spatial correlation between the resistivity of the k-th grid point and the wave velocity of the j-th grid point in a three-dimensional stratigraphic coupling model. This represents the three-dimensional coordinates of the k-th grid point. Represents the three-dimensional coordinates of the j-th grid point; This represents the actual spatial distance between the k-th grid point and the j-th grid point; L is the relevant length.

[0136] Optionally: Sample from the posterior distribution using the following formula:

[0137] ;

[0138] Generate sample set ;

[0139] in, These are candidate samples for resistivity. Candidate samples for P-wave velocity. Candidate samples for shear wave velocity; Let be the resistivity accepted in the m-th iteration. Let be the longitudinal wave velocity received in the m-th iteration. The velocity of the transverse wave received during the m-th iteration; , and Used to construct the posterior distribution; This represents the posterior probability density; M is the number of iterations;

[0140] The posterior mean is calculated as the optimal coupling model using the following formula:

[0141] ;

[0142] ;

[0143] in, This is the posterior mean of resistivity, i.e., the optimal estimate of resistivity. This is the posterior mean of the P-wave velocity, i.e., the optimal estimate of the P-wave velocity;

[0144] Optionally, the third calculation unit is further configured to: calculate the quantization uncertainty of the variance field and output the coupling parameter field;

[0145] The variance field uncertainty is calculated using the following formula:

[0146] ;

[0147] The output coupling parameter field is: ;

[0148] in, ;

[0149] in, For resistivity at The variance field of a point For the m-th iteration sample in The resistivity value at the point The resistivity for optimal estimation is The value of the point; The optimal longitudinal wave velocity is in The value of the point, The optimal transverse wave velocity is at The value of the point, This represents the ratio of spatial variations.

[0150] Among them, the three-dimensional stratigraphic coupling model uses the coupling parameter field of each coordinate point. To represent;

[0151] The parameter acquisition module is used for:

[0152] Obtain preset wave velocity and resistivity thresholds, and locate weak areas in the three-dimensional stratigraphic coupling model based on the wave velocity and resistivity thresholds.

[0153] Specifically, when the P-wave velocity is below the P-wave velocity threshold, the S-wave velocity is below the S-wave velocity threshold, and the resistivity is below the resistivity threshold, this point is identified as a weak point. The distribution area of ​​these weak points determines the weak zones in the three-dimensional formation coupling model. In this embodiment, weak points are located simultaneously by combining P-wave velocity, S-wave velocity, and resistivity.

[0154] Based on the weak areas, an initial drilling path is formed (see...). Figure 3 , Figure 3 This is a schematic diagram of a cross-section of the weak area, based on Figure 3 The initial drilling path is formed in the light-colored area (indicated by the direction of the arrow).

[0155] The geological database is called to obtain the regional geostress parameters of the weak area, including the geological type, the average unit weight of the overlying strata, and the average depth of the weak area.

[0156] Calculate the initial grouting pressure and initial grout mix ratio based on the regional geostress parameters;

[0157] Among them, the initial grouting pressure The calculation formula is as follows:

[0158] ;

[0159] Where c is the geological type coefficient. H represents the average unit weight of the overlying strata, and H represents the average depth of the weak zone. For correction factor, For the P-wave velocity in the weak region, For reference stress, For reference wave speed;

[0160] in, This reflects the contribution of the overlying strata's own weight to the grouting pressure. Corrections reflecting formation mechanical properties (wave velocity ratio) and reference stress; wave velocity ratio The larger the area, the softer the weak zone, and the greater the grouting pressure needs to be adjusted.

[0161] The initial slurry mix ratio was determined by the mapping relationship table between the longitudinal wave velocity and resistivity of the weak zone and the slurry mix ratio.

[0162] Optionally: In the reinforcement learning control kernel:

[0163] Encode formation parameters and engineering states into state vectors for reinforcement learning. :

[0164] ;

[0165] in, This is the resistivity value at the current location. The longitudinal wave velocity at the current position. The elastic modulus at the current location. The Poisson ratio at the current position. The grouting pressure of the previous time period, This refers to the slurry mix ratio from the previous time period. This is a sequence of spatial coordinates of historical borehole paths recorded up to the previous time point;

[0166] Multi-objective reward function for:

[0167] ;

[0168] ;

[0169] ;

[0170] ;

[0171] in, , and These are the weighting coefficients; As a reward for the grouting effect, As a cost penalty, As a risk penalty; This represents the change in resistivity after grouting. This represents the change in longitudinal waves after grouting. and These are the grouting effect bonus coefficients; This is the initial resistivity reference value; This is the initial P-wave velocity reference value; and These are cost penalty coefficients; For grouting pressure, For slurry preparation ratio, As the reference grouting pressure, The standard slurry mix ratio; The severity of the risk penalty; For indicator functions, The value is 1 at time. The value is 0 at that time; This represents the maximum formation pressure under the current grouting pressure. It represents the tensile strength of the formation.

[0172] The formula for the advantage function, which measures the superiority of an action relative to the average level, is:

[0173] ;

[0174] in, Let g be the dominance function at the current time t, g be the g-th time step in the future, and T be the upper limit of the time step; As a discount factor, The instant reward for the g-th time step in the future. For use in evaluating state vectors Value assessment network;

[0175] in, This indicates that the action was better than expected, and similar actions should be strengthened. This indicates that the action was ineffective and that similar actions should be reduced.

[0176] The formula for updating the policy gradient based on the advantage function is:

[0177] ;

[0178] in, For policy gradient, Represent the expected function; This represents the probability distribution of actions under the new strategy; This represents the action vector, which includes the increments of grouting pressure, grout mix ratio, and borehole path. ; s is the state vector; This represents the probability distribution of actions under the old strategy. This represents the probability ratio between the new strategy and the old strategy. To truncate the threshold, This represents the truncation function;

[0179] Therefore, the strategy gradient is updated based on the dominance function, which involves determining the increments of grouting pressure, grout mix ratio, and borehole path. Therefore, based on the increments of grouting pressure, grout ratio, and borehole path, Based on the actions at the current moment, determine the optimal actions for the next moment, namely, optimize the grouting pressure, optimize the grout mix ratio, and optimize the drilling path. The actions at the current moment are the grouting pressure, grout mix ratio, and drilling path at the current moment.

[0180] Optionally: the system further includes an adaptive filtering module, the adaptive filtering module being used for:

[0181] Magnetic field signals in the grouting pipeline are monitored using an optical fiber sensor array.

[0182] Adjusting the filter cutoff frequency in real time based on formation inhomogeneity:

[0183] ;

[0184] in, The adjusted cutoff frequency. The initial cutoff frequency, NI These are normalized index parameters obtained through joint calculation of resistivity and seismic wave inversion data. NI max yes NI The upper limit of the reference value.

[0185] The adaptive filtering module uses a fiber optic sensor array to eliminate magnetic field interference caused by formation inhomogeneity.

[0186] Curved grouting reinforcement technology combines the concepts of curved drilling and grouting reinforcement. It avoids obstructions from surface buildings and underground pipelines, achieving long-distance trenchless grouting through the placement of curved grouting boreholes. Long-distance trenchless drilling eliminates the need for large-scale excavation, reducing the project's scope and cost, and minimizing disruption to surrounding residents. The drilling path can be arbitrary, making it suitable for densely populated urban areas and complex underground tunnel and pipeline networks. Furthermore, by modifying the grouting equipment, it can achieve various grouting processes such as high-pressure jet grouting, permeation grouting, and fracturing grouting for different construction scenarios.

[0187] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms, or by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a readable storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to enter the methods described in the various embodiments of the present invention.

[0188] In the description of this specification, references to terms such as "one embodiment," "another embodiment," "other embodiments," or "first embodiment to Xth embodiment," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, method steps, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0189] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0190] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0191] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A high-precision while-drilling curve-guided grouting system, characterized in that, The system comprises: a data fusion module configured to generate a three-dimensional stratum coupling model based on a three-dimensional coordinate system by using a stratum resistivity model and a stratum wave velocity model; a parameter acquisition module configured to acquire an initial grouting pressure, an initial slurry ratio and an initial drilling path determined according to the three-dimensional stratum coupling model; a reinforcement learning control kernel configured to measure the advantages and disadvantages of real-time grouting actions based on stratum electrical parameters in the stratum resistivity model, stratum mechanical parameters in the stratum wave velocity model, real-time engineering state data of the grouting construction process and a multi-objective reward function constructed to generate an optimized control action, thereby outputting an optimized grouting pressure, an optimized slurry ratio and an optimized drilling path; an actuator configured to adjust the grouting pressure, the slurry ratio and the drilling path according to the output instruction of the reinforcement learning control kernel to realize closed-loop control of the curved grouting path. In the reinforcement learning control kernel: Encoding formation parameters and engineering conditions into a state vector for reinforcement learning : ； wherein, is a resistivity value for the current location, is a P-wave velocity for the current location, is a modulus of elasticity for the current location, is a Poisson's ratio for the current location, is a grouting pressure at a previous time, is a slurry mix at a previous time, is a sequence of spatial coordinates of a historical borehole path recorded before the previous time; Multi-objective reward function is: ； ； ； ； wherein, , and are weight coefficients, respectively; is a grouting effect reward, is a cost penalty, is a risk penalty; is a post-grouting resistivity change amount, is a post-grouting P-wave change amount, and are grouting effect reward coefficients, respectively; is an initial resistivity reference value; is an initial P-wave velocity reference value; and are cost penalty coefficients, respectively; is a grouting pressure, is a grout proportion, is a reference grouting pressure, is a reference grout proportion; is a risk penalty intensity; is an indicator function, is 1 when is 0 when is a maximum formation pressure under the current grouting pressure; is a formation tensile strength.

2. The high-precision while-drilling curve-guided grouting system according to claim 1, characterized in that: The system further comprises: a data acquisition module comprising a resistivity imaging module and a seismic wave inversion module, the resistivity imaging module being configured to acquire stratum electrical parameters of a grouting area to form a stratum resistivity model, and the seismic wave inversion module being configured to acquire stratum mechanical parameters of the grouting area to form a stratum wave velocity model.

3. The high-precision while-drilling curve-guided grouting system according to claim 2, characterized in that: The resistivity imaging module comprises: a apparent resistivity calculation unit configured to acquire a potential difference collected after arranging an electrode array in the grouting area at a set resolution to perform apparent resistivity calculation; a first conversion unit configured to inversely reconstruct the apparent resistivity data to convert into a stratum resistivity model for representing a real resistivity distribution of the stratum to obtain a resistivity three-dimensional image of the grouting area; The seismic wave inversion module comprises: a data acquisition unit configured to acquire seismic wave travel time and waveform data generated when a seismic source vibrates in the grouting area; a construction unit configured to construct an initial stratum wave velocity model and update the velocity distribution through travel time inversion; a second conversion unit configured to inversely reconstruct the initial stratum wave velocity model to convert into a stratum wave velocity model for representing a real wave velocity distribution of the stratum to obtain a stratum wave velocity three-dimensional image of the grouting area.

4. The high-precision while-drilling curve-guided grouting system according to claim 2, characterized in that: The data fusion module comprises: an alignment unit configured to map the stratum resistivity model and the stratum wave velocity model to the same three-dimensional coordinate system and perform resolution alignment on the stratum resistivity model and the stratum wave velocity model; a first calculation unit configured to establish a joint probability distribution of resistivity and wave velocity; a second calculation unit configured to establish a statistical relationship of resistivity and wave velocity and describe a spatial correlation of resistivity and wave velocity; a sample unit configured to sample from a posterior distribution to form a sample set; a third calculation unit configured to calculate a posterior mean as a model parameter in the three-dimensional stratum coupling model.

5. The high-precision while-drilling curve-guided grouting system according to claim 4, characterized in that, The joint probability distribution of resistivity and wave velocity is established by the following formula: The prior distribution is defined as: ； The likelihood function is: ； ； The posterior distribution is: ； Where p is the joint prior distribution; To represent the true formation resistivity distribution, For the longitudinal wave velocity, The transverse wave velocity; For observational data; This is a vector of resistivity observation data. For the q-th resistivity observation, This represents the total amount of observation data in resistivity imaging measurements. This is the response of the resistivity forward model. This represents the variance of the resistivity data error. This is a vector of seismic wave observation data. For the first e Each seismic wave observation value This represents the total amount of observation data in seismic wave measurements. The response of the seismic wave forward model. The variance of the seismic wave rate data. It is the probability density function of a normal distribution.

6. The high-precision while-drilling curve-guided grouting system according to claim 5, characterized in that, The statistical relationship of resistivity and wave velocity is established by the following formula: ； wherein a and b are regression coefficients, respectively, is noise; ; is the standard deviation of the fluctuation range of the noise, is the reference P-wave velocity, is the reference resistivity of the formation; The spatial correlation of resistivity and wave velocity is described by the following formula: ； wherein, represents the spatial correlation of the resistivity of the kth grid point and the wave velocity of the jth grid point in the three-dimensional formation coupling model, represents the three-dimensional coordinates of the kth grid point, represents the three-dimensional coordinates of the jth grid point; represents the actual spatial distance between the kth grid point and the jth grid point; L is the correlation length.

7. The high-precision while-drilling curve-guided grouting system according to claim 6, characterized in that, The sample is taken from the posterior distribution by the following formula: ； Generating a sample set ; wherein, is a candidate sample for resistivity, is a candidate sample for P-wave velocity, is a candidate sample for S-wave velocity; is the accepted resistivity at the mth iteration, is the accepted P-wave velocity at the mth iteration, is the accepted S-wave velocity at the mth iteration; , , for constructing the posterior distribution; denotes the posterior probability density, and M is the number of iterations.

8. The high-precision while-drilling curve-guided grouting system according to claim 7, characterized in that, The third calculation unit is further configured to calculate a variance field to quantify uncertainty and output a coupling parameter field; The parameter acquisition module is configured to: Obtaining preset wave velocity threshold and resistivity threshold, positioning the weak zone in the three-dimensional stratum coupling model according to the wave velocity threshold and the resistivity threshold; Forming an initial drilling path according to the weak zone; Calling a geological database to obtain regional geo-stress parameters of the weak zone; Calculating an initial grouting pressure and an initial slurry ratio according to the regional geo-stress parameters.

9. The high-precision while-drilling curve-guided grouting system according to claim 8, characterized in that, In the reinforcement learning control kernel: The formula of the advantage function for measuring the pros and cons of the action relative to the average level is: ； wherein, is the advantage function for the current time t, g is the gth time step in the future, and T is the upper limit on the time steps; is a discount factor, is the immediate reward for the gth time step in the future, is a value evaluation network for evaluating the value of the state vector V(s). wherein, indicates that the action is better than expected and that similar actions should be reinforced; indicates that the action is less effective than expected and that similar actions should be reduced; The formula for policy gradient update according to the advantage function is: ； wherein, is a policy gradient, represents an expected function; represents an action probability distribution of a new policy; represents an action vector, including a grouting pressure increment, a slurry ratio increment, and a borehole path increment ; s is a state vector; represents an action probability distribution of an old policy, represents a probability ratio of the new policy to the old policy; is a truncation threshold, represents a truncation function.

10. The high-precision while-drilling guided curve grouting system according to any one of claims 1 to 9, characterized in that, The system further comprises an adaptive filtering module, which is used to: Monitoring the grouting pipeline magnetic field signal through the optical fiber sensing array; Adjusting the filtering cutoff frequency in real time through the stratum heterogeneity.

Images