Grouting drilling orifice reinforcement construction method based on intelligent control

Abstract

The grouting drilling orifice reinforcement construction method based on intelligent control comprises the five steps that S1, drilling machine parameter signals are collected through operation of a drilling machine, and meanwhile geological data signals are collected through in-hole detection equipment; s2, the drilling machine parameter signals and the geological data signals are input into an intelligent decision-making center for fusion processing, and geological sensing signals are generated; s3, based on the geological sensing signal, calling a preset decision-making model through the intelligent decision-making center for calculation, and generating a grouting control instruction signal; s4, grouting operation is executed according to the grouting control instruction signal, and meanwhile a grouting process monitoring signal is collected and fed back to the intelligent decision center in real time; and S5, after grouting is completed, hole opening stable monitoring signals are continuously collected, and when the effect evaluation signals do not reach the standard, reinforcing grouting signals are automatically triggered, and grouting operation is executed again. According to the intelligent die adjusting method for the die-casting machine, the problems that drill holes are prone to collapse, reinforcement depends on experience, and quality is difficult to control can be solved.

Description

Technical Field

[0001] This invention relates to the field of geotechnical engineering and tunnel construction technology, specifically to a method for reinforcing the borehole opening of grouting based on intelligent control. Background Technology

[0002] In tunnel excavation, mining construction, and geological disaster control projects, drilling and grouting operations are frequently required in weak, fractured, or water-rich geological strata. Existing technologies mainly employ borehole reinforcement or surface pre-grouting, but these methods have significant limitations: During drilling, the surrounding soil and rock mass is prone to collapse and necking due to stress release and drill bit disturbance, leading to difficult borehole formation and requiring repeated cleaning; traditional reinforcement parameters, such as grouting pressure, grout mix ratio, and reinforcement range, largely rely on the experience of construction personnel, lacking scientific basis. In complex and variable strata, insufficient reinforcement often results in secondary collapse, while over-reinforcement leads to material waste; grouting process control is rudimentary—excessive pressure can fracture the strata and cause grout leakage, while insufficient pressure fails to guarantee reinforcement effectiveness, and the entire process lacks effective monitoring; the geological information revealed by the borehole is disconnected from subsequent reinforcement parameter settings, making it impossible to dynamically adjust process parameters based on real-time geological conditions; construction quality assessment relies on post-construction inspection, and problems are only discovered when they have already created potential engineering hazards. The root cause of these problems lies in the fact that existing methods lack the ability to perceive geological conditions in real time, lack data-based intelligent decision-making mechanisms, and lack closed-loop control of the construction process.

[0003] Although some automated grouting equipment has emerged in recent years, these devices are mostly limited to single-parameter control and have failed to achieve full-process intelligentization from geological identification to parameter optimization and effect evaluation, especially when dealing with extremely complex geological conditions. Therefore, there is an urgent need for a comprehensive solution that can sense changes in the strata in real time, make intelligent decisions on control parameters, dynamically adjust construction techniques, and automatically evaluate the reinforcement effect, so as to fundamentally improve the reliability, safety, and economy of grouting borehole reinforcement. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the purpose of this invention is to provide an intelligent mold adjustment method for die-casting machines to solve the problems of easy collapse of borehole openings, reliance on experience for reinforcement, and difficulty in quality control. This invention generates sensing signals by real-time acquisition of borehole parameters and geological data, which are then fused and analyzed by an intelligent decision-making center to generate grouting control signals. During the grouting process, pressure, flow rate, and displacement signals are continuously monitored, and grouting parameters are dynamically adjusted through an adaptive algorithm. Finally, the reinforcement effect is evaluated based on stable monitoring signals, and supplementary grouting is automatically triggered if the target is not met, forming a closed-loop intelligent control of "perception-decision-execution-evaluation," achieving precise and adaptive borehole reinforcement.

[0005] This invention provides a method for reinforcing the borehole opening of grouting based on intelligent control, comprising: S1: Collect drilling rig parameter signals through drilling rig operation, and simultaneously collect geological data signals through in-hole detection equipment; S2: Input drilling rig parameter signals and geological data signals into the intelligent decision-making center for fusion processing to generate geological sensing signals; S3: Based on geological sensing signals, the intelligent decision-making center calls the preset decision-making model to perform calculations and generate grouting control command signals; S4: Execute grouting operation according to grouting control command signal, and collect grouting process monitoring signal and feed it back to the intelligent decision center in real time. The updated grouting control command signal is dynamically adjusted and generated through adaptive algorithm. S5: After grouting is completed, continuously collect the wellhead stability monitoring signal and input it into the intelligent decision center to generate the effect evaluation signal. When the effect evaluation signal fails to meet the standard, automatically trigger the reinforcement grouting signal and re-execute the grouting operation.

[0006] In one embodiment of the present invention, step S1 involves collecting drilling rig parameter signals through drilling rig operation, including collecting drill bit rotation speed signals, drilling thrust signals, and rotational torque signals. Simultaneously, geological data signals are collected through borehole detection equipment, including acquiring borehole wall image signals using a borehole camera and acquiring formation resistivity signals and acoustic velocity signals using a borehole geophysical instrument. In step S2, when the drilling rig parameter signals and geological data signals are input into the intelligent decision-making center for fusion processing, a multi-source information fusion algorithm is used to extract features and correlate data from the drill bit rotation speed signals, drilling thrust signals, rotational torque signals, borehole wall image signals, formation resistivity signals, and acoustic velocity signals. A preset machine learning model is used to classify and identify the lithology of the formations, an expert knowledge base is used to grade and evaluate the degree of formation fragmentation, and a neural network algorithm is used to quantitatively analyze the water-bearing capacity of the formations. Finally, a geological perception signal that comprehensively reflects the geological conditions of the borehole section is generated. This geological perception signal includes formation stability indicators and permeability indicators.

[0007] In one embodiment of the present invention, when calculating based on the geological sensing signal by calling the preset decision model through the intelligent decision center in step S3, the preset decision model includes a grouting parameter decision module based on fuzzy logic and an adaptive optimization module based on reinforcement learning. The fuzzy logic-based grouting parameter decision module receives the formation stability index and permeability index from the geological sensing signal and outputs an initial grouting parameter set through a fuzzy inference system. The reinforcement learning-based adaptive optimization module constructs a reward function based on historical grouting effect data and continuously optimizes the grouting parameter decision strategy through strategy iteration. The outputs of the two modules are weighted and fused to generate a grouting control command signal containing a grouting pressure benchmark value, a grouting flow rate benchmark value, a grout water-cement ratio benchmark value, and a grouting duration benchmark value. The grouting control command signal also includes the allowable adjustment range and priority setting of each parameter.

[0008] In one embodiment of the present invention, when performing grouting operation according to the grouting control command signal in step S4, an intelligent grouting device is used to receive the grouting control command signal and perform grouting operation according to the grouting pressure reference value, grouting flow rate reference value, grout water-cement ratio reference value, and grouting duration reference value in the command. At the same time, real-time grouting pressure signals are collected by pressure sensors arranged in the grouting pipeline, real-time grouting flow rate signals are collected by electromagnetic flowmeters, real-time grout consistency signals are collected by grout consistency meters, and orifice deformation signals are collected by orifice displacement monitoring devices. These signals together constitute the grouting process monitoring signal. The process of real-time feedback to the intelligent decision center is achieved by using an industrial Internet of Things gateway to realize real-time data transmission. The adaptive algorithm in the intelligent decision center dynamically adjusts and generates updated grouting control command signals based on the deviation between the grouting process monitoring signals and expected parameters, combined with the formation response characteristics.

[0009] In one embodiment of the present invention, the adaptive algorithm in step S4 employs a model predictive control algorithm. This algorithm establishes a multivariate control system model including grouting pressure, grouting flow rate, grout consistency, and orifice displacement, and calculates the optimal control quantity through a rolling optimization strategy. Simultaneously, the adaptive algorithm also includes abnormal condition handling logic. When an abnormal increase in grouting pressure is detected, a pressure protection mode is automatically activated; when an abnormal grout loss is detected, a loss compensation mode is automatically activated; and when the orifice displacement exceeds the limit, a displacement control mode is automatically activated. Each handling mode includes specific parameter adjustment strategies and priority settings to ensure the stability and safety of the grouting process under complex conditions. Finally, an updated grouting control command signal is generated through multi-objective optimization.

[0010] In one embodiment of the present invention, step S5, after grouting is completed, continuously collecting wellhead stability monitoring signals includes collecting three-dimensional displacement signals of the wellhead using a high-precision displacement sensor, collecting grouting pressure dissipation signals using a pressure sensor, and collecting formation density signals around the wellhead using an acoustic detection device. After these signals are input into the intelligent decision-making center, an effect evaluation signal is generated using a deep learning-based evaluation model. This evaluation model is trained on a large amount of historical data and can predict the long-term stability of the wellhead by integrating multiple monitoring parameters. When the effect evaluation signal does not meet the standard, the intelligent decision-making center automatically generates a targeted reinforcement grouting signal based on the specific indicators that do not meet the standard. This signal includes the location information of the reinforcement grouting, grouting parameter adjustment suggestions, and reinforcement priority settings, and the grouting operation is re-executed until the effect evaluation signal reaches the predetermined standard.

[0011] In one embodiment of the present invention, the intelligent decision-making center adopts a distributed computing architecture, including edge computing nodes and a cloud computing platform. The edge computing nodes are deployed at the construction site and are responsible for processing high-frequency acquired drilling rig parameter signals and geological data signals in real time, performing preliminary data filtering, feature extraction, and anomaly detection. The cloud computing platform receives the data processed by the edge computing nodes, runs complex machine learning models and optimization algorithms, and generates grouting control command signals and effect evaluation signals. The two computing layers synchronize data through an encrypted communication protocol to ensure the security and real-time performance of data processing. This distributed architecture ensures both the low latency requirements of key controls and the high computing resource requirements of complex algorithms.

[0012] In one embodiment of the present invention, the method further includes a digital management step for the construction process. A construction data acquisition module automatically records the entire construction data for each borehole, including geological sensing signals, grouting control command signals, grouting process monitoring signals, and effect evaluation signals. This data is stored in a construction database and deeply analyzed by a data analysis module to generate construction quality reports and process optimization suggestions. Simultaneously, a quality traceability system based on blockchain technology is established, generating tamper-proof digital fingerprints for key construction parameters and decision-making processes, providing a reliable basis for project quality acceptance and accountability. Construction personnel can view the construction progress and quality status in real time via mobile terminals, achieving transparent management of the construction process.

[0013] In one embodiment of the present invention, the generation of grouting control command signals also considers the influence of environmental factors. Temperature, humidity, and vibration signals at the construction site are collected by environmental monitoring equipment. The intelligent decision-making center couples and analyzes these environmental signals with geological sensing signals to establish a correlation model between environmental factors and grouting effect. When generating grouting control command signals, grouting parameters are dynamically adjusted according to real-time environmental conditions. For example, the initial setting time of grout is appropriately reduced when the temperature is high, and the grouting pressure benchmark value is appropriately increased when the vibration is large. This environmentally adaptive control strategy can effectively compensate for the impact of environmental changes on grouting quality and improve the stability of construction quality.

[0014] In one embodiment of the present invention, the method employs a progressive grouting control strategy, dividing the borehole reinforcement into multiple grouting stages. At the beginning of each grouting stage, the grouting control command signal for the current stage is dynamically adjusted based on the grouting effect evaluation signal of the previous stage and the current geological sensing signal. In the initial stage, a lower grouting pressure and a smaller grout diffusion radius are used to focus on reinforcing the weakest area. In subsequent stages, the grouting parameters are gradually increased to expand the reinforcement range. This progressive control, through multi-stage optimization achieved by an intelligent decision center, avoids formation damage that may be caused by a single grouting and ensures the uniformity and reliability of the final reinforcement effect, making it particularly suitable for borehole reinforcement under extremely fractured formation conditions.

[0015] This invention provides a method for reinforcing borehole openings through grouting based on intelligent control. It generates sensing signals by real-time acquisition of borehole parameters and geological data, which are then fused and analyzed by an intelligent decision-making center to produce grouting control signals. During the grouting process, pressure, flow rate, and displacement signals are continuously monitored, and grouting parameters are dynamically adjusted using an adaptive algorithm. Finally, the reinforcement effect is evaluated based on stable monitoring signals. If the effect is not met, supplementary grouting is automatically triggered, forming a closed-loop intelligent control system of "perception-decision-execution-evaluation," achieving precise and adaptive borehole reinforcement. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of the intelligent mold adjustment method for die-casting machines. Detailed Implementation

[0018] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.

[0019] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0020] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0021] Please see Figure 1The diagram illustrates the intelligent mold adjustment method for die-casting machines according to the present invention. This method comprises five steps: S1: Collecting drilling rig parameter signals during drilling operation and simultaneously collecting geological data signals through in-hole detection equipment; S2: Inputting the drilling rig parameter signals and geological data signals into an intelligent decision-making center for fusion processing to generate a geological sensing signal; S3: Based on the geological sensing signal, calculating using a preset decision model through the intelligent decision-making center to generate a grouting control command signal; S4: Executing the grouting operation according to the grouting control command signal, while simultaneously collecting grouting process monitoring signals and feeding them back to the intelligent decision-making center in real time, dynamically adjusting and generating an updated grouting control command signal through an adaptive algorithm; S5: After grouting is completed, continuously collecting borehole stability monitoring signals and inputting them into the intelligent decision-making center to generate an effect evaluation signal. When the effect evaluation signal fails to meet the standard, automatically triggering a reinforcement grouting signal and re-executing the grouting operation.

[0022] like Figure 1As shown, its core lies in constructing a complete technological chain from geological information perception to intelligent decision-making, dynamic execution, and closed-loop optimization. The method is characterized by five interconnected, data-driven basic steps. The first step is to collect drilling rig parameter signals through drilling rig operation and simultaneously collect geological data signals through in-hole detection equipment. This step is the fundamental data source for the entire method and the sensing source for achieving intelligent control. Specifically, the acquisition of drilling rig parameter signals relies on a high-precision sensor network integrated into the drilling rig's power head and propulsion system. These sensors can capture and transmit continuous signals reflecting the mechanical state during drilling in real time, mainly including drill bit rotation speed signals, drilling thrust signals, and rotational torque signals. These signals are direct physical quantities for judging the lithological changes and degree of fracturing in the strata where the drill bit is located. Meanwhile, the acquisition of geological data signals is accomplished through a series of advanced in-hole detection devices. These devices can be inserted into the borehole along with the drill pipe to conduct in-situ exploration of the borehole wall and surrounding strata during drilling operations or intervals. Commonly used equipment includes in-hole camera devices and in-hole geophysical instruments. In-hole camera devices can acquire high-resolution images of the borehole wall, visually displaying the joints, fissures, collapses, and spalling phenomena in the rock and soil. In-hole geophysical instruments can acquire geophysical parameters such as formation resistivity and acoustic wave velocity signals. These parameters are strongly correlated with the integrity, density, and water-bearing capacity of the formation. The second step involves inputting the drilling rig parameter signals and geological data signals acquired in the first step into a central processing unit called the intelligent decision center for fusion processing. The aim is to generate a comprehensive signal that can fully and accurately quantify the geological conditions of the borehole opening section, namely, a geological sensing signal. This processing is not a simple signal superposition, but rather a deep processing using a multi-source information fusion algorithm. This algorithm first preprocesses the various input signals, including filtering and denoising, time series alignment, and feature extraction. Then, through data association technology, it maps signals from different sensors and with different physical meanings onto a unified geological interpretation model. During this process, a pre-set machine learning model is activated, performing pattern recognition on continuous drilling parameter sequences and intermittent geophysical parameters to automatically classify and identify lithology, distinguishing different types such as clay, sandstone, and fault fracture zones. Simultaneously, an expert knowledge base storing numerous typical geological cases and expert experience is invoked. This knowledge base combines visualized fracture features from borehole wall images with abrupt changes in drilling torque to scientifically assess the degree of formation fracturing. Furthermore, neural network algorithms are applied to analyze the combined relationship between formation resistivity and acoustic wave velocity, thereby quantifying the formation's water-bearing capacity and predicting potential water inrush risks. Ultimately, all these analytical results are integrated into a comprehensive geological sensing signal that reflects the geological conditions of the borehole section. This signal typically includes two key engineering parameters: formation stability index and permeability index, providing a solid data foundation for subsequent intelligent decision-making.The third step, based on the geological sensing signals generated in the second step, uses the intelligent decision center to call its internal preset decision model for calculation, generating grouting control command signals with clear guiding significance. This preset decision model is the intelligent core; it receives formation stability and permeability indicators from the geological sensing signals as input, and outputs a set of executable grouting process parameters, namely the grouting control command signals, through built-in optimization algorithms and rule bases. This set of signals includes at least the benchmark values ​​for grouting pressure, grouting flow rate, grout water-cement ratio, and grouting duration set for the current geological conditions. These benchmark values ​​constitute the direct basis for the initial grouting operation. The fourth step, based on the grouting control command signals generated in the third step, drives the intelligent grouting equipment to perform specific grouting operations. This is a dynamic process full of feedback adjustment. While the grouting equipment injects grout into the formation around the borehole according to the reference parameters in the command signal, an independent monitoring system starts to work. It collects real-time grouting pressure signals through pressure sensors arranged in the grouting pipeline, real-time grouting flow signals through electromagnetic flow meters, real-time grout consistency signals through grout consistency meters, and borehole deformation signals through displacement gauges or inclinometers installed near the borehole. All these signals together constitute the grouting process monitoring signal reflecting the current grouting conditions. These monitoring signals are fed back to the intelligent decision-making center in real time and at high speed. The adaptive algorithm inside the center is immediately activated. This algorithm continuously compares the monitored actual parameters with the benchmark parameters set in the third step, analyzes the degree of deviation and the trend of change, and combines it with the understanding of the formation response characteristics, such as the change in the formation grout absorption rate or the rate of wellhead deformation, to dynamically fine-tune or reconstruct the grouting parameters. This generates updated grouting control command signals that are more in line with the current actual working conditions, and immediately issues them to the grouting equipment for execution, forming a rapid closed loop of "execution-monitoring-feedback-adjustment" to ensure that the grouting process is always in an optimal state. The fifth step is that after the grouting operation is completed according to the instructions, the system does not immediately stop working, but enters an effect verification and assurance stage. In this stage, it continuously collects post-stage signals that can reflect the stability of the wellhead and the surrounding formation, namely wellhead stability monitoring signals. These signals usually include the three-dimensional displacement signal of the wellhead collected by high-precision sensors, the natural dissipation signal of the grouting pressure after the injection stops, etc. These post-monitoring signals are input into the intelligent decision-making center and compared with preset reinforcement success standards. A specific evaluation algorithm generates an objective reinforcement effect evaluation signal. When the evaluation signal indicates that the reinforcement effect has not met the preset safety and stability standards, the intelligent decision-making center will not wait for manual intervention but will automatically trigger a reinforcement grouting signal. This signal contains information such as the specific location that needs reinforcement and the recommended adjustment of grouting parameters, and restarts the fourth step of the grouting operation cycle until the effect evaluation signal finally confirms that the borehole has reached a reliable stable state, thereby realizing intelligent closed-loop management of the entire process from construction to acceptance.

[0023] Furthermore, its further optimization and concretization are reflected in the in-depth expansion of the signal acquisition and fusion processing in steps S1 and S2. In step S1, the step of acquiring drilling rig parameter signals through drilling rig operation is explicitly refined to include acquiring three key mechanical parameters: drill bit rotation speed signal, drilling thrust signal, and rotational torque signal. These together constitute the core dataset for judging drilling resistance and formation strength. The drill bit rotation speed signal reflects the efficiency of the drill bit in cutting the formation and the frictional resistance encountered. The drilling thrust signal directly reflects the work done by the propulsion system to overcome the formation resistance, while the rotational torque signal sensitively indicates the shear strength of the interaction between the drill bit and the formation. The synchronous acquisition and comprehensive analysis of these three signals can accurately capture the location of the soft-hard transition interface and potential fracture zones in the formation. At the same time, the step of acquiring geological data signals through borehole exploration equipment is also concretized to include acquiring borehole wall image signals using borehole camera devices and acquiring formation resistivity signals and acoustic wave velocity signals using borehole geophysical instruments. Borehole wall images provide direct visual evidence of rock mass structure, fracture aperture, infill material properties, and groundwater seepage points, serving as valuable data for geological logging and qualitative assessment. Formation resistivity signals are highly sensitive to water content and mineralization within the formation, while acoustic wave velocity signals are directly correlated with the rock mass's elastic modulus and integrity index. The combination of these two geophysical signals can effectively and quantitatively evaluate the formation's water-bearing capacity and degree of fragmentation. In step S2, when all these multi-source, heterogeneous signals are input into the intelligent decision-making center for fusion processing, the multi-source information fusion algorithm employed is a complex and sophisticated data processing flow. This algorithm first preprocesses the various input raw signals, including applying moving average filtering to the drilling rig parameter signals to eliminate random fluctuations, performing environmental correction on the geophysical signals to eliminate interference, and enhancing and segmenting the image signals to highlight key features. The algorithm then proceeds to the feature extraction and data association stage. It extracts representative feature values ​​from the preprocessed signal, such as the mean and variance of drilling torque, the attenuation gradient of acoustic wave velocity, and the density and direction of fractures in the image. These data, which are not directly aligned temporally and spatially, are associated using common borehole depth coordinates to establish a multi-parameter profile indexed by depth. Based on this, a pre-defined machine learning model, such as a support vector machine or random forest model, is loaded and run. Utilizing patterns learned from a large amount of historical engineering data, it quickly and accurately classifies and identifies the lithology of the formation at the current borehole depth, outputting classification results such as "intact bedrock," "strongly weathered layer," and "fault gouge." In parallel, an expert knowledge base accumulating knowledge and successful cases from numerous domain experts is activated. Based on the associated feature data, especially combining image fracture features and torque abrupt change information, it grades and evaluates the degree of formation fragmentation according to predefined logical rules, outputting levels such as "intact," "relatively fragmented," and "extremely fragmented."Furthermore, specially designed neural network algorithms, such as feedforward neural networks or convolutional neural networks, are used to process the complex nonlinear relationship between formation resistivity signals and acoustic wave velocity signals. Through their powerful mapping capabilities, they quantify the water-bearing capacity of the formation, providing a specific water-bearing index or estimated inflow. Finally, all these analytical results from different algorithm modules—lithological classification, fracturing grade, and water-bearing index—are weighted, fused, and normalized to generate a comprehensive geological sensing signal that includes formation stability and permeability indices. The stability index may be a composite index integrating lithological strength, fracturing grade, and geostress state, while the permeability index reflects the formation's ability to allow grout flow. Together, they provide accurate and reliable quantitative input for subsequent grouting decisions.

[0024] In one embodiment of the present invention, another aspect of its deepening lies in the detailed definition and description of the specific implementation mechanism of how the intelligent decision-making center generates grouting control command signals in step S3. The core of this step is that, based on the geological sensing signals generated in the aforementioned steps, the intelligent decision-making center calls its internally preset decision model for calculation. This preset decision model is not a single algorithm, but a complex system composed of multiple intelligent modules working collaboratively. Specifically, this system mainly includes a grouting parameter decision module based on fuzzy logic and an adaptive optimization module based on reinforcement learning. The fuzzy logic-based grouting parameter decision module is specifically responsible for handling the fuzziness and uncertainty of the input information in the geological sensing signals. It receives key indicators from the geological sensing signals, namely, formation stability indicators and permeability indicators, as input variables. This module internally defines a complete fuzzy inference system, including a fuzzified interface that converts precise input values ​​into fuzzy linguistic values ​​such as "high stability" and "low permeability," a rule base containing a large number of "if-then" fuzzy rules, and an inference engine for fuzzy inference. For example, a typical fuzzy rule might be "If formation stability is low and permeability is high, then the grouting pressure should be medium-high, and the grout water-cement ratio should be low." This system maps quantitative geological indicators to qualitative decisions that align with human expert thinking, outputting an initial set of grouting parameters. This set includes initial suggested values ​​for grouting pressure, flow rate, water-cement ratio, and duration considered reasonable under current geological conditions. Simultaneously, the reinforcement learning-based adaptive optimization module plays a continuous learning and improvement role. It doesn't rely directly on instantaneous manually set rules but constructs a reward function based on historical grouting effect data. This function uses grouting effect evaluation signals and grouting material consumption as evaluation criteria. Through continuous interaction with the environment—repeatedly simulating or trying different parameter strategies during actual grouting—and evaluating the effectiveness of each strategy through feedback from the reward function, the module continuously optimizes its grouting parameter decision-making strategy using iterative algorithms such as Q-learning or policy gradient methods. The goal is to maximize long-term accumulated rewards, i.e., finding the most economical and efficient grouting scheme while ensuring reinforcement effectiveness. In each decision-making cycle, the initial parameter set output by the fuzzy logic-based grouting parameter decision module and the optimal strategy recommended by the reinforcement learning-based adaptive optimization module are simultaneously submitted to the intelligent decision center. The center has an internal weight allocation and fusion mechanism that assigns appropriate real-time weight coefficients to the outputs of the two modules based on factors such as the confidence level of the current working condition and the completeness of historical data, and then fuses the weighted results.Finally, a comprehensive and optimized grouting control command signal is generated. This signal not only includes core operation commands such as the grouting pressure benchmark value, grouting flow rate benchmark value, grout water-cement ratio benchmark value, and grouting duration benchmark value determined by fusion calculation, but also includes the allowable adjustment range of each parameter to increase the flexibility and fault tolerance of execution. For example, the grouting pressure can fluctuate within a specific range above and below the benchmark value, and the priority setting when multiple parameters need to be adjusted. For example, the grout water-cement ratio is adjusted first while ensuring that the grouting pressure does not exceed the limit. This design enables the subsequent execution system to have stronger environmental adaptability while ensuring safety.

[0025] Furthermore, its key advancement lies in the specific architecture and operational mechanism of the adaptive algorithm in step S4, which is the core technological support for achieving precise dynamic control of the grouting process. This adaptive algorithm employs an advanced control strategy called model predictive control, which differs from traditional feedback control in that it performs forward-looking rolling optimization based on a model. Specifically, the algorithm first needs to establish a multivariable control system model that describes the dynamic characteristics of the grouting system. This model correlates key variables such as grouting pressure, grouting flow rate, grout consistency, and orifice displacement, expressing their coupling relationships and how they respond to control commands and external disturbances through mathematical equations. In each control cycle, the algorithm does not simply calculate the control quantity based on the deviation between the current monitored value and the target value. Instead, it uses the established multivariable model, starting from the current system state, to predict the system's behavior over a future period. Subsequently, it calculates a series of future control action sequences by solving an optimization problem in the rolling time domain. The goal is to make the predicted system output as close as possible to the desired trajectory while satisfying various operational constraints, such as the upper limit of grouting pressure and the capacity limitations of the grouting equipment. However, the model predictive control algorithm does not mechanically execute the entire calculated control sequence. Instead, it only implements the first control action in the sequence. In the next sampling period, the algorithm acquires the latest measured data, refreshes the system state, and then repeats the prediction and optimization process. This "rolling optimization and feedback correction" mechanism enables it to effectively handle the system's hysteresis and nonlinearity, and possesses strong robustness to unknown disturbances. In addition to this core optimization framework, the adaptive algorithm also incorporates a highly intelligent abnormal condition handling logic. This logic acts like an experienced safety officer, constantly alerting to potential risks during grouting. When the system detects an abnormal increase in grouting pressure, with its rate or absolute value exceeding the safety threshold, indicating a potential risk of grout channel blockage or impending formation fracturing, the algorithm automatically activates the pressure protection mode. In this mode, the system prioritizes and rapidly reduces the grouting flow rate or even temporarily stops the pump to ensure the pressure returns to a safe range, preventing irreversible damage to the formation structure. When the grouting flow rate is consistently high while the pressure rises slowly, indicating that the grout may be flowing into large fissures or cavities leading to significant loss, the algorithm will activate a loss compensation mode. This mode may trigger automatic adjustments to the grout mix ratio, such as instantly increasing the grout consistency or switching to a special grout containing plugging materials, to quickly seal the loss channels. When a high-precision displacement sensor detects displacement or deformation exceeding the allowable range in the formation at or around the wellhead, the algorithm will immediately activate a displacement control mode. In this mode, the system will comprehensively analyze the relationship between displacement and grouting pressure, and may adopt various strategies such as reducing grouting pressure, changing the grouting rhythm, or adjusting the grouting point location to stabilize the formation.Importantly, these processing modes do not operate in isolation. Each mode includes parameter adjustment strategies and explicit priority settings for specific anomalies. For example, in extreme cases where both pressure and displacement are excessively high, the system coordinates control actions according to preset priorities (perhaps displacement safety takes precedence over grouting progress). Ultimately, all these functions—model prediction, rolling optimization, anomaly monitoring, and multi-mode processing—are integrated into a multi-objective optimization framework. This framework must simultaneously balance multiple, sometimes conflicting, objectives such as reinforcement effectiveness, construction efficiency, material consumption, and process safety. Through complex online calculations, it generates updated, safe, and efficient grouting control command signals in real time, ensuring optimal performance throughout the reinforcement process in complex and variable geological environments.

[0026] like Figure 1As shown, its ultimate embodiment of quality assurance and self-improvement is concentrated in step S5, which constructs a continuous intelligent evaluation and reinforcement system that surpasses traditional construction acceptance standards. After the grouting operation is completed according to the instructions, the system does not declare the task over, but immediately enters a stage of effect verification and capability maintenance. In this stage, through a precision monitoring network deployed at and around the borehole, a series of signals that can comprehensively reflect the subsequent behavior of the reinforced body and the surrounding soil and rock are continuously and frequently collected, collectively referred to as borehole stability monitoring signals. These signals are not single parameters, but a collection of multi-dimensional data. They primarily include: three-dimensional displacement signals at the borehole opening, acquired through high-precision displacement sensors. These signals accurately capture minute movements of the borehole opening in three spatial directions, thus determining its overall stability; grouting pressure dissipation signals, acquired through pressure sensors retained in the grouting pipeline. These signals reflect the attenuation pattern of grout pressure in the formation after grouting, and their attenuation curves can indirectly reveal the grout solidification, the density of the reinforced body, and the presence of unfilled fractures; and formation density signals around the borehole opening, acquired through acoustic detection devices or other geophysical exploration equipment. These signals non-destructively assess the grout diffusion range and the strength formation of the solidified body by measuring changes in the propagation velocity of elastic waves in the reinforced area. All these information-rich post-monitoring signals are transmitted in real time and input into the intelligent decision-making center. Inside the center, a mature, deep learning-based evaluation model, pre-trained with extensive historical engineering data, is activated. This deep learning model, employing deep neural networks or long short-term memory network architectures, possesses the ability to process complex time-series data and discover deep nonlinear relationships. It uses continuously input signals such as three-dimensional displacement sequences, pressure dissipation curves, and formation density changes as input features. Through its multi-layer nonlinear transformation, it ultimately generates a highly condensed and reliable reinforcement effect evaluation signal. This evaluation signal is no longer just a list of a few independent parameters, but a comprehensive criterion. It may be a stability probability value or an index related to the long-term safety factor, capable of proactively predicting the stability state of the borehole in the future. When the effect evaluation signal calculated by this deep learning model is lower than the system's preset qualified standard threshold, it indicates that the current reinforcement status has potential risks or has not met design expectations. The intelligent decision center will not wait for manual judgment and instructions, but will immediately initiate an automated remedial process. It will first diagnose the main reasons for the failure to meet the standard based on the specific numerical pattern of the effect evaluation signal and similar cases in the historical database, such as insufficient local reinforcement, grout solidification shrinkage, or the existence of undiscovered hidden channels.Based on this diagnosis, the center automatically generates a highly targeted reinforcement grouting signal. This signal is not a simple repetition of previous grouting commands, but includes a detailed reinforcement strategy, such as the three-dimensional spatial location information of areas requiring key reinforcement, suggested grouting parameters adjusted according to the characteristics of the reinforcement area, and reinforcement priority settings determined based on risk levels when multiple potential reinforcement points exist. This intelligently generated reinforcement grouting signal is then sent to the execution system, automatically restarting the grouting operation cycle starting from step S4 in claim 1 for precise reinforcement. This process can be iterated until subsequent monitoring signals are input into the evaluation model again, and the generated new effect evaluation signal stably indicates that the borehole has reached and maintained above the predetermined safety and reliability standards, thus achieving a truly data-driven and self-learning quality closed-loop control.

[0027] Specifically, its powerful computing capabilities and real-time response characteristics are rooted in the advanced computing architecture adopted by the intelligent decision-making center, namely, a distributed computing architecture. This architecture does not centralize all computing tasks in one place, but intelligently distributes them to two different but closely collaborative computing units based on the characteristics and requirements of the tasks: edge computing nodes deployed at the construction site and a cloud computing platform at the back end. This design fully considers the dual high requirements of real-time performance and computational complexity in grouting control. The edge computing nodes typically consist of high-performance industrial computers or dedicated embedded systems deployed at the construction site, directly connected to data sources such as drilling rig sensors, borehole detection equipment, and grouting monitoring instruments. Their main responsibility is to handle tasks with extremely high response speed requirements and large data throughput, and to process high-frequency acquired raw drilling rig parameter signals and geological data signals in real time. Specifically, the edge nodes perform preliminary but crucial data processing tasks, including filtering raw signals to eliminate noise from electromagnetic interference and mechanical vibrations, extracting features to calculate engineering-significant feature values ​​from massive data streams, and performing anomaly detection to quickly identify emergency situations such as sensor malfunctions or sudden changes in operating conditions. Because of its proximity to the data source, edge computing achieves extremely low latency, ensuring timely judgment and response to the instantaneous state of the drilling and grouting processes, providing a foundation for rapid closed-loop control. On the other hand, the cloud computing platform, relying on its near-limitless storage resources and powerful parallel computing capabilities, undertakes complex model calculations and long-term learning tasks requiring enormous computing power. It receives pre-processed and compressed data packets from one or more edge computing nodes at the construction site via a secure network connection. In the cloud, more complex machine learning models run, such as deep neural networks for geological sensing, reinforcement learning algorithms for parameter optimization, and deep learning models for performance evaluation. The training and iterative optimization of these models are typically completed in the cloud, as this consumes significant computing resources and time. The cloud platform can also integrate massive amounts of data from different projects and geological conditions, performing cross-project in-depth mining and knowledge discovery, continuously optimizing and updating its models and knowledge base. These two computing layers, edge and cloud, do not operate in isolation; they maintain continuous and secure data synchronization and command transmission through data channels employing advanced encrypted communication protocols. Edge nodes upload refined feature data and event reports to the cloud for model updates and macro-analysis; the cloud, in turn, distributes the latest trained model, optimized decision-making strategies, and updated knowledge base to the edge nodes, enhancing their local intelligence. This distributed architecture cleverly balances computational load, ensuring low latency requirements for critical control loops, enabling field devices to respond quickly to changes, while also meeting the high computational resource demands of complex intelligent algorithms. This allows the system to become increasingly intelligent, forming a crucial foundation for achieving high-performance intelligent control in harsh industrial environments.

[0028] like Figure 1As shown, the key to its value extension and modernization of engineering management lies in the fact that the method also systematically integrates a comprehensive digital management link for the construction process. This link aims to transform the traditional construction management model, which relies on paper records and manual experience, into a modern engineering management paradigm that is full-process, traceable, and data-driven. At the core of this digital management link is a continuously operating construction data acquisition module. This module acts like a tireless recorder, automatically and comprehensively recording all construction data for each borehole from drilling commencement to final acceptance confirmation. These recorded data are comprehensive, forming a complete digital archive of the borehole, including but not limited to: geological sensing signals generated during the borehole perception phase, which record how the formation was "diagnosed"; grouting control command signals generated during the decision-making phase, which record the "prescription" given by the system based on the diagnostic results; grouting process monitoring signals collected during the grouting execution phase, which record every detail of the "treatment" process; and the final effect evaluation signal generated during the effect verification phase, which records the evaluation result of the "treatment effect." All these timestamped, multi-dimensional data are systematically stored in a centralized construction database. More profoundly, a dedicated data analysis module periodically or irregularly performs in-depth mining of this ever-growing database, employing methods such as statistical analysis, trend prediction, and machine learning to automatically generate detailed construction quality reports. These reports not only reflect the quality of individual boreholes but also reveal the quality distribution patterns across the entire work area. Furthermore, based on historical data comparison and effect retrospection, it can propose process optimization suggestions, such as determining the optimal grout mix ratio and grouting rate for specific geological conditions to achieve the best cost-effectiveness. To ensure the authenticity and immutability of this critical data, and to provide legally valid and credible evidence for project quality acceptance and potential accountability, this invention also establishes a quality traceability system based on blockchain technology. This system uses a hash algorithm to generate unique, irreversible digital fingerprints from key construction parameters, such as the final geological sensing signals, core grouting control commands, important process monitoring inflection point data, and final effect evaluation conclusions, along with their decision-making process logs, and stores these fingerprints in a distributed manner within a blockchain network. Any slight alteration to the original data will result in a significant change in the hash value, thus being immediately identified. This fundamentally eliminates the possibility of data tampering afterward, greatly enhancing the credibility of quality records. Furthermore, to achieve transparency and collaborative management of the construction process, authorized personnel, such as project managers, supervising engineers, or owner representatives, can use mobile devices such as smartphones and tablets, after identity verification, to view the real-time construction progress of the entire work area, the real-time quality status of each borehole, and automatically generated early warning information. This allows management decisions to be based on real-time, accurate on-site data, significantly improving the efficiency and precision of project management.This digital management process transforms one-off construction operations into analyzable, traceable, and optimizable digital assets, greatly enhancing the overall management value of the project.

[0029] This invention discloses a method for reinforcing borehole openings using intelligent control. The method involves real-time acquisition of borehole parameters and geological data to form a sensing signal, which is then fused and analyzed by an intelligent decision-making center to generate a grouting control signal. During the grouting process, pressure, flow rate, and displacement signals are continuously monitored, and grouting parameters are dynamically adjusted using an adaptive algorithm. Finally, the reinforcement effect is evaluated based on stable monitoring signals. If the effect is not met, supplementary grouting is automatically triggered, forming a closed-loop intelligent control system of "perception-decision-execution-evaluation," achieving precise and adaptive borehole reinforcement.

[0030] Therefore, the intelligent mold adjustment method for die casting machines of the present invention can solve the problems of easy collapse of the drill hole opening, reliance on experience for reinforcement, and difficulty in quality control.

[0031] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A method for reinforcing the borehole opening of grouting based on intelligent control, characterized in that, include: S1: Collect drilling rig parameter signals through drilling rig operation, and simultaneously collect geological data signals through in-hole detection equipment; S2: Input the drilling rig parameter signals and geological data signals into the intelligent decision-making center for fusion processing to generate geological sensing signals; S3: Based on the geological sensing signal, the intelligent decision-making center calls the preset decision-making model to perform calculations and generate grouting control command signals; S4: Execute grouting operation according to the grouting control command signal, and at the same time collect grouting process monitoring signals and feed them back to the intelligent decision center in real time. Dynamically adjust and generate updated grouting control command signals through adaptive algorithms. S5: After grouting is completed, continuously collect the wellhead stability monitoring signal and input it into the intelligent decision center to generate an effect evaluation signal. When the effect evaluation signal fails to meet the standard, automatically trigger the reinforcement grouting signal and re-execute the grouting operation.

2. The method for reinforcing grouting borehole openings based on intelligent control according to claim 1, characterized in that, In step S1, drilling rig parameter signals are collected during drilling operation, including drill bit speed, drilling thrust, and rotational torque signals. Simultaneously, geological data signals are collected using borehole detection equipment, including borehole wall image signals obtained using a borehole camera and formation resistivity and acoustic velocity signals obtained using a borehole geophysical instrument. In step S2, when the drilling rig parameter signals and geological data signals are input into the intelligent decision-making center for fusion processing, a multi-source information fusion algorithm is used to extract features and correlate data from the drill bit speed, drilling thrust, rotational torque, borehole wall image, formation resistivity, and acoustic velocity signals. A preset machine learning model is used to classify and identify formation lithology, an expert knowledge base is used to grade and assess the degree of formation fragmentation, and a neural network algorithm is used to quantitatively analyze the formation's water-bearing capacity. Finally, a comprehensive geological sensing signal reflecting the geological conditions of the borehole section is generated, which includes formation stability and permeability indicators.

3. The method for reinforcing grouting borehole openings based on intelligent control according to claim 1, characterized in that, In step S3, when the preset decision model is invoked by the intelligent decision center based on the geological sensing signal for calculation, the preset decision model includes a grouting parameter decision module based on fuzzy logic and an adaptive optimization module based on reinforcement learning. The fuzzy logic-based grouting parameter decision module receives the formation stability index and permeability index from the geological sensing signal and outputs an initial grouting parameter set through the fuzzy inference system. The reinforcement learning-based adaptive optimization module constructs a reward function based on historical grouting effect data and continuously optimizes the grouting parameter decision strategy through strategy iteration. The outputs of the two modules are weighted and fused to generate a grouting control command signal that includes the grouting pressure reference value, grouting flow rate reference value, grout water-cement ratio reference value and grouting duration reference value. The grouting control command signal also includes the allowable adjustment range and priority settings of each parameter.

4. The method for reinforcing the borehole opening of grouting based on intelligent control according to claim 1, characterized in that, In step S4, when performing grouting operations according to the grouting control command signal, an intelligent grouting device receives the grouting control command signal and performs grouting operations according to the grouting pressure reference value, grouting flow rate reference value, grout water-cement ratio reference value, and grouting duration reference value in the command. At the same time, real-time grouting pressure signals are collected by pressure sensors arranged in the grouting pipeline, real-time grouting flow rate signals are collected by electromagnetic flowmeters, real-time grout consistency signals are collected by grout consistency meters, and orifice deformation signals are collected by orifice displacement monitoring devices. These signals together constitute the grouting process monitoring signal. The real-time feedback to the intelligent decision center is achieved by using an industrial Internet of Things gateway to realize real-time data transmission. The adaptive algorithm in the intelligent decision center dynamically adjusts and generates updated grouting control command signals based on the deviation between the grouting process monitoring signals and expected parameters, combined with the formation response characteristics.

5. The method for reinforcing the borehole opening of grouting based on intelligent control according to claim 1, characterized in that, The adaptive algorithm in step S4 employs a model predictive control algorithm. This algorithm establishes a multivariate control system model that includes grouting pressure, grouting flow rate, grout consistency, and orifice displacement, and calculates the optimal control quantity through a rolling optimization strategy. Simultaneously, the adaptive algorithm also includes abnormal condition handling logic. When an abnormal increase in grouting pressure is detected, a pressure protection mode is automatically activated; when abnormal grout loss is detected, a loss compensation mode is automatically activated; and when orifice displacement exceeds the limit, a displacement control mode is automatically activated. Each handling mode includes specific parameter adjustment strategies and priority settings to ensure the stability and safety of the grouting process under complex conditions. Finally, an updated grouting control command signal is generated through multi-objective optimization.

6. The method for reinforcing the borehole opening of grouting based on intelligent control according to claim 1, characterized in that, In step S5, after grouting is completed, the continuous acquisition of wellhead stability monitoring signals includes acquiring three-dimensional displacement signals of the wellhead through a high-precision displacement sensor, acquiring grouting pressure dissipation signals through a pressure sensor, and acquiring formation density signals around the wellhead through an acoustic detection device. After these signals are input into the intelligent decision-making center, an effect evaluation signal is generated using a deep learning-based evaluation model. This evaluation model is trained on a large amount of historical data and can predict the long-term stability of the wellhead by integrating multiple monitoring parameters. When the effect evaluation signal does not meet the standard, the intelligent decision-making center automatically generates a targeted reinforcement grouting signal based on the specific indicators that do not meet the standard. This signal includes the location information of the reinforcement grouting, suggestions for adjusting grouting parameters, and reinforcement priority settings. The grouting operation is then re-executed until the effect evaluation signal reaches the predetermined standard.

7. The method for reinforcing grouting borehole openings based on intelligent control according to claim 1, characterized in that, The intelligent decision-making center adopts a distributed computing architecture, including edge computing nodes and a cloud computing platform. The edge computing nodes are deployed at the construction site and are responsible for processing high-frequency acquired drilling rig parameter signals and geological data signals in real time, performing preliminary data filtering, feature extraction, and anomaly detection. The cloud computing platform receives the data processed by the edge computing nodes, runs complex machine learning models and optimization algorithms, and generates grouting control command signals and effect evaluation signals. The two computing layers synchronize data through an encrypted communication protocol to ensure the security and real-time performance of data processing. This distributed architecture ensures both the low latency requirements of key controls and the high computing resource requirements of complex algorithms.

8. The method for reinforcing grouting borehole openings based on intelligent control according to claim 1, characterized in that, The method also includes a digital management component for the construction process. A construction data acquisition module automatically records all construction data for each borehole, including geological sensing signals, grouting control command signals, grouting process monitoring signals, and effect evaluation signals. This data is stored in a construction database and deeply analyzed by a data analysis module to generate construction quality reports and process optimization suggestions. Simultaneously, a blockchain-based quality traceability system is established, generating tamper-proof digital fingerprints for key construction parameters and decision-making processes, providing a reliable basis for project quality acceptance and accountability. Construction personnel can view construction progress and quality status in real time via mobile terminals, achieving transparent management of the construction process.

9. The method for reinforcing the borehole opening of grouting based on intelligent control according to claim 1, characterized in that, The generation of the grouting control command signal also considers the influence of environmental factors. Temperature, humidity, and vibration signals from the construction site are collected through environmental monitoring equipment. The intelligent decision-making center couples and analyzes these environmental signals with geological sensing signals to establish a correlation model between environmental factors and grouting effect. When generating the grouting control command signal, the grouting parameters are dynamically adjusted according to real-time environmental conditions. For example, the initial setting time of the grout is appropriately reduced when the temperature is high, and the grouting pressure benchmark value is appropriately increased when the vibration is large. This environmentally adaptive control strategy can effectively compensate for the impact of environmental changes on grouting quality and improve the stability of construction quality.

10. The method for reinforcing the borehole opening of grouting based on intelligent control according to claim 1, characterized in that, The method employs a progressive grouting control strategy, dividing the borehole reinforcement into multiple grouting stages. At the beginning of each grouting stage, the grouting control command signal for that stage is dynamically adjusted based on the grouting effect evaluation signal of the previous stage and the current geological sensing signal. In the initial stage, a lower grouting pressure and a smaller grout diffusion radius are used to focus on reinforcing the weakest area. In subsequent stages, the grouting parameters are gradually increased to expand the reinforcement range. This progressive control, through multi-stage optimization achieved by an intelligent decision center, avoids formation damage that may be caused by a single grouting and ensures the uniformity and reliability of the final reinforcement effect, making it particularly suitable for borehole reinforcement under extremely fractured formation conditions.

Images