Rotary drilling pile machine drilling method and device suitable for complex stratum

Abstract

The invention relates to the field of drilling automation, and discloses a rotary drilling pile machine drilling method and device suitable for complex stratums.The method comprises the following steps that S1, working condition feedback parameters are obtained in real time; s2, recognizing the real-time geological working condition of the drill bit; s3, optimal drilling operation parameters are determined in combination with a preset intelligent decision model; and S4, according to the optimal drilling operation parameters, an execution mechanism of the rotary drilling pile machine is automatically controlled, self-adaptive drilling is carried out, and the device comprises a data acquisition module, a working condition recognition module, a decision optimization module, a closed-loop control module and a drilling tool type selection and recommendation module. According to the method, depth data mining is carried out on key parameters such as rock-soil body specific energy, so that real-time working conditions are accurately classified into specific types such as soft clay and pebble beds, and the technical problems that in the prior art, the geological misjudgment rate is high, the recognition precision is low, and complex stratums cannot be effectively distinguished are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of drilling automation, and in particular to a rotary pile driver drilling method and device suitable for complex strata. Background Art

[0002] As a key piece of equipment in pile foundation construction, rotary pile drivers operate across a wide range of geological structures. In actual engineering projects, geological conditions are often complex and varied. The same pile, at varying depths, may penetrate a variety of strata, including soft clay, medium-dense sand, hard plastic clay, gravel, and even heavily weathered bedrock. The current rotary drilling operation model's adaptability to geology is primarily based on a combination of geological survey reports and the operator's on-site skills. Typically, a rough estimate of the ground conditions is made based on the survey report before construction begins. During construction, the operator adjusts drilling parameters in real time based on the drill's sound, vibration feedback, and operational feel.

[0003] However, this traditional approach, which relies on prior data and manual perception, is gradually becoming increasingly limited when faced with complex geology. First, geological survey reports provide geological information at discrete points and cannot accurately reveal specific changes in the stratigraphic structure between two exploration points, such as the presence of interlayers or lenses. This inherently limits the accuracy of predictions. Second, on-site operations rely heavily on the operator's personal experience. This intuitive judgment method varies, and different operators differ in their judgment and handling of the same working conditions, making it difficult to effectively ensure uniform construction quality. When encountering unexpected hard or soft interlayers, untimely or inappropriate parameter adjustments can easily lead to a sharp drop in drilling efficiency, increased energy consumption, and even safety incidents such as stuck drills and hole collapse. Furthermore, even when the operator identifies stratigraphic changes, subsequent parameter adjustments are often tentative and rough, making it difficult to quickly converge to the optimal operating point that balances efficiency, energy consumption, and safety. The entire process has a long response chain and requires extremely high operator skills, limiting the standardization and intelligentization of the construction process. Summary of the Invention

[0004] The purpose of the present invention is to provide a rotary pile drilling method and device suitable for complex strata, which solves the problems of low construction efficiency, unstable hole quality and high safety risks caused by excessive reliance on manual experience when constructing in complex and changeable strata, resulting in inaccurate geological identification, non-optimal drilling parameters, and lack of closed-loop control.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: A rotary pile drilling method suitable for complex strata comprises the following steps: S1. Data acquisition: Real-time acquisition of working condition feedback parameters of the rotary pile driver during the drilling process; S2. Working condition identification: Based on the working condition feedback parameters, the real-time geological working condition of the drill bit is identified; S3, decision optimization: determining optimal drilling operation parameters based on the identified real-time geological conditions and in combination with a preset intelligent decision model; S4. Closed-loop control: Automatically controlling the actuator of the rotary pile driver according to the optimal drilling operation parameters to perform adaptive drilling.

[0006] Preferably, in step S1, the real-time acquisition of working condition feedback parameters of the rotary pile driver during the drilling process specifically includes the following steps: S11. Directly measure and obtain drilling torque, drilling speed, pressure and drilling speed through sensors installed in the rotary drilling rig power head and main winch hydraulic system; S12, using an in-situ bottom hole monitoring assembly installed at the bottom of the drill pipe, close to the drill bit, to measure and obtain the real-time specific gravity and viscosity of the mud at the bottom of the hole, as well as the real-time verticality of the drill pipe; S13. Based on the data obtained in step S11, a built-in physical model is used to perform real-time calculation to generate the specific energy of the rock and soil.

[0007] Preferably, in step S13, the generated rock and soil mass ratio can be calculated using the following formula: Where, E s is the specific energy of the rock and soil, F is the applied pressure obtained in real time, A is the preset cross-sectional area of ​​the borehole, n is the drilling speed obtained in real time, M is the drilling torque obtained in real time, and v p is the real-time drilling speed, and 2π is the constant coefficient for converting the rotational speed and torque into power.

[0008] Preferably, in step S2, identifying the real-time geological conditions of the drill bit specifically includes the following steps: S21. Establish a baseline geological profile: Based on the current drilling depth, extract the expected rock and soil layer type and its physical and mechanical property indicators corresponding to the depth from the engineering geological survey report database loaded during the construction preparation phase, which will serve as a baseline reference for subsequent geological identification and judgment; S22. Extracting real-time formation response characteristics: performing time domain analysis based on the working condition feedback parameters, and quantifying the real-time response characteristics of the formation by calculating the average value, change trend, and fluctuation amplitude within the current time window; S23, performing feature matching and operating condition classification: performing high-speed pattern matching on the real-time response features and the geological and operating condition feature database built into the intelligent control center, automatically classifying the current operating condition into the most suitable geological type through a support vector machine algorithm, and outputting the recognition result; The geological and working condition characteristic database is pre-calibrated through experiments and historical data, and stores typical parameter characteristics corresponding to different geological working conditions; The geological conditions include soft clay, medium-dense sand layer, hard plastic clay, pebble layer and strongly weathered bedrock; The typical parameter characteristics include the range of values ​​of the rock and soil specific energy and the fluctuation mode of the torque.

[0009] Preferably, in step S22, performing time domain analysis based on the operating condition feedback parameters specifically includes the following steps: S221. Establishing a sliding data window: For the operating condition feedback parameters, starting from the current time point and tracing back a fixed time length, a dynamic, continuously sliding forward time series data subset is established; S222, data preprocessing and filtering: cleaning the data subset to eliminate the influence of sensor noise and transient disturbance on the analysis results, and obtaining a smoothed data subset; The calculation of the average value, change trend and fluctuation range within the current time window includes the following steps: S223, calculating the statistical mean: calculating the arithmetic mean of the smoothed data subset; S224, calculating the slope of the change trend: performing linear regression analysis on the processed data subset to fit a best-fit straight line; S225. Calculate the fluctuation range: calculate the dispersion of the processed data subset relative to its average value, and use the standard deviation as a quantitative indicator of the fluctuation range.

[0010] Preferably, in step S23, the current working condition is automatically classified into the most suitable geological type by the support vector machine algorithm, and the identification result is output, which specifically includes the following steps: S231, constructing a real-time feature vector: combining the average value, change trend, and fluctuation amplitude into a single, ordered mathematical vector; S232, performing feature vector normalization processing: inputting the mathematical vector into a pre-set normalization model for processing to obtain a normalized feature vector; S233, calling the SVM classifier for prediction: feeding the normalized feature vector as input to an existing pre-trained offline SVM classifier; S234, mapping labels and outputting final recognition results: classifying the integers output by the SVM classifier, searching and converting them to obtain human-readable geological condition names, and calculating the distance between the feature vector and the classification hyperplane, and converting it into a confidence score; S235, final result output: the confidence score is formally output as the final result of this recognition.

[0011] Preferably, in step S3, determining the optimal drilling operation parameters specifically includes the following steps: S31. Setting optimization goals and constraints: Based on the geological conditions and in combination with the construction goals manually set before drilling, the corresponding geological strategy knowledge base is called from the preset intelligent decision-making model; S32. Constructing a quantitative evaluation function: Based on the construction goal, establishing a comprehensive performance evaluation function capable of scoring any set of drilling parameters; S33, executing an optimization search algorithm: based on the recommended range provided by the geological strategy knowledge base, starting a genetic algorithm to generate and continuously iterate multiple sets of candidate parameter combinations within the parameter range by simulating group collaboration, and scoring each generation of parameters using the evaluation function; S34. Convergence and output of the optimal solution: After multiple iterations, the genetic algorithm automatically searches for and converges to a unique set of parameter combinations that can give the evaluation function the highest comprehensive performance score, defines it as the optimal drilling operation parameters, and outputs it; the geological strategy knowledge base provides an initial, safe recommended range for various drilling operation parameters under the current geological conditions, and this recommended range will serve as the boundary constraint condition for subsequent optimization calculations.

[0012] Preferably, in step S4, automatically controlling the actuator of the rotary pile driver according to the optimal drilling operation parameters specifically includes the following steps: S41, command analysis and target value setting: converting the optimal drilling operation parameters into specific command setting points sent to the controllers of each actuator; S42, performing real-time error feedback and correction: continuously comparing the actual operating parameters with the command set point, calculating the real-time error, and dynamically fine-tuning the control signal through the PID control algorithm.

[0013] A rotary pile drilling device suitable for complex strata, comprising: A data acquisition module is used to obtain in real time the working condition feedback parameters of the rotary pile driver during the drilling process, wherein the working condition feedback parameters include at least drilling mechanical parameters, in-hole environment and posture parameters, and rock and soil specific energy; a working condition identification module connected to the data acquisition module, for identifying the real-time geological working conditions of the drill bit based on the working condition feedback parameters by matching the real-time response characteristics of the parameters with a built-in geological-working condition characteristic database; a decision optimization module connected to the working condition identification module, for determining the optimal drilling operation parameters that can maximize the overall efficiency based on the real-time geological working conditions and in combination with a preset intelligent decision model through an optimization search algorithm; a closed-loop control module, connected to the decision optimization module, for parsing the optimal drilling operation parameters into command set points, automatically controlling the actuators of the rotary pile driver accordingly, and performing real-time error correction based on the actual operating parameters fed back by the data acquisition module to perform adaptive drilling; The drilling tool selection and recommendation module is connected to the working condition identification module and is used to match and output the corresponding optimal detachable drill bit model from the database after receiving the real-time geological working conditions output by the working condition identification module. Preferably, in the drilling tool selection and recommendation module, the geological drill bit matching database includes at least: When the geological conditions are soft clay and hard plastic clay, it is recommended to use a double-bottom sand bucket and a rotary drilling bucket; When the geological conditions are pebble layers, it is recommended to use a rotary drill bit inlaid with cemented carbide; When the geological conditions are strongly weathered bedrock, it is recommended to use roller drill bits and combined rock drill bits.

[0014] In summary, the present invention includes at least one of the following beneficial technical effects: 1. This invention uses deep data mining of key parameters such as rock and soil specific energy to accurately classify real-time working conditions into specific types, such as weak clay and gravel layers. Compared to traditional construction methods that rely on operator experience or a single parameter threshold for rough judgment, this method solves the technical problems of high geological misjudgment rate, low recognition accuracy, and inability to effectively distinguish complex strata.

[0015] 2. This invention uses a genetic algorithm to perform a global optimization search for drilling parameters. This allows the system to consistently find the optimal parameter combination that balances efficiency, risk, and cost. Existing technologies typically use fixed parameter tables or rely on manual judgment and adjustment. This invention addresses the inherent drawbacks of low drilling efficiency in variable formations, high energy consumption, and the inability to dynamically optimize according to construction strategies.

[0016] 3. This invention continuously compares actual drilling parameters with optimal target values ​​and performs real-time error correction. Compared to existing technologies that rely primarily on manual monitoring and intervention, this method addresses the issues of delayed response, unstable processes, heavy reliance on operator skill, and difficulty ensuring uniform hole quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A flow chart for implementing the method of the present invention; Figure 2 It is a data acquisition flow chart of the present invention; Figure 3 This is a working condition identification flow chart of the present invention; Figure 4 This is the optimization and control flow chart of the present invention; Figure 5 Schematic diagram of the device module architecture of the present invention. DETAILED DESCRIPTION

[0018] The following is combined with Figure 1 -Attached Figure 5 , the present invention is described in further detail.

[0019] like Figures 1-4 As shown, a rotary pile drilling method suitable for complex strata includes the following steps: S1. Data acquisition: Real-time acquisition of working condition feedback parameters of the rotary pile driver during the drilling process; In step S1, the working condition feedback parameters of the rotary pile driver during the drilling process are obtained in real time, which specifically includes the following steps: S11. Directly measure and obtain drilling torque, drilling speed, pressure and drilling speed through sensors installed in the rotary drilling rig power head and main winch hydraulic system; S12, using an in-situ bottom hole monitoring assembly installed at the bottom of the drill pipe, close to the drill bit, to measure and obtain the real-time specific gravity and viscosity of the mud at the bottom of the hole, as well as the real-time verticality of the drill pipe; S13. Based on the data obtained in step S11, the built-in physical model is used to perform real-time calculation to generate the specific energy of the rock and soil mass. In step S13, the specific energy of the rock and soil mass is calculated using the following formula: Where, E s is the specific energy of the rock and soil, F is the applied pressure obtained in real time, A is the preset cross-sectional area of ​​the borehole, n is the drilling speed obtained in real time, M is the drilling torque obtained in real time, and v p is the real-time drilling speed, 2π is the constant coefficient for converting speed and torque into power; Specifically, step S1 aims to comprehensively and in real time capture the working condition feedback parameters of the rotary drilling pile driver that reflect its own status and external formation response during the entire drilling operation through a multi-dimensional sensing and computing system.

[0020] Specifically, the data acquisition step is accomplished through a distributed sensor system in collaboration with a central data processing module. The sensor system exemplarily includes a surface sensor unit mounted on the drilling rig body and a bottom hole in-situ monitoring assembly mounted on the bottom end of the drill bit.

[0021] The ground sensor unit further includes a power head monitoring unit and a main winch monitoring unit. The power head monitoring unit directly measures and obtains real-time data of drilling torque and drilling speed through sensors installed in the hydraulic system of the rotary drilling rig power head.

[0022] At the same time, the main winch monitoring unit uses pressure and displacement sensors installed in the main winch hydraulic system to directly measure and convert the real-time pressure and drilling speed reflecting the drilling process. The above four parameters together constitute the core mechanical parameters of the drilling rig.

[0023] To obtain the real-world environmental parameters inside the borehole, especially at the drill bit's location, this embodiment further includes a bottom hole in-situ monitoring assembly. This assembly is preferably installed at the bottom of the drill pipe, as close as possible to the drill bit, to ensure real-time and representative measurement data.

[0024] The bottomhole in-situ monitoring assembly integrates multiple micro-sensors to measure and obtain real-time specific gravity and viscosity of the bottomhole mud, key environmental parameters that reflect the effectiveness of wall protection and deslagging efficiency. It also measures the real-time verticality of the drill pipe, a posture parameter that ensures hole quality.

[0025] In order to generate comprehensive formation response parameters that can more sensitively reflect changes in formation lithology, this embodiment also includes real-time calculation of rock and soil specific energy. This calculation is performed by a built-in physical model that receives the mechanical parameters obtained in step S11 as input.

[0026] The rock and soil mass specific energy can be calculated using the following formula: Where, E s is the specific energy of the rock and soil, F is the applied pressure obtained in real time, A is the preset cross-sectional area of ​​the borehole, n is the drilling speed obtained in real time, M is the drilling torque obtained in real time, and v p is the real-time drilling speed, and 2π is the constant coefficient for converting the rotational speed and torque into power.

[0027] By combining the above steps, the present invention can obtain a complete working condition feedback parameter set including the mechanical parameters of the drilling rig body, the in-situ environment and posture parameters of the bottom hole, and the comprehensive formation response parameters calculated through depth.

[0028] S2. Working condition identification: Based on working condition feedback parameters, identify the real-time geological working conditions of the drill bit; In step S2, the real-time geological conditions of the drill bit are identified, which specifically includes the following steps: S21. Establish a baseline geological profile: Based on the current drilling depth, extract the expected rock and soil layer type and its physical and mechanical property indicators corresponding to the depth from the engineering geological survey report database loaded during the construction preparation phase, which will serve as a baseline reference for subsequent geological identification and judgment; S22. Extracting real-time formation response characteristics: Performing time domain analysis based on the working condition feedback parameters to quantify the real-time formation response characteristics by calculating their average value, change trend, and fluctuation amplitude within the current time window; S23. Feature matching and working condition classification: The real-time response features are matched with the geological and working condition feature database built into the intelligent control center at high speed. The current working condition is automatically classified into the most suitable geological type through the support vector machine algorithm, and the identification result is output; The geological and working condition characteristic database is pre-calibrated through experiments and historical data, and stores typical parameter characteristics corresponding to different geological conditions; The geological conditions include weak clay, medium-dense sand, hard plastic clay, pebble layer and strongly weathered bedrock; Typical parameter characteristics include the range of values ​​of rock mass specific energy and the fluctuation pattern of torque; In step S22, time domain analysis is performed based on the operating condition feedback parameters, specifically including the following steps: S221. Establishing a sliding data window: For the operating condition feedback parameters, starting from the current time point and tracing back a fixed length of time, a dynamic, continuously sliding time series data subset is established; S222, data preprocessing and filtering: clean the data subset to eliminate the influence of sensor noise and transient disturbance on the analysis results, and obtain a smooth data subset; Calculating its average value, change trend, and fluctuation range in the current time window includes the following steps: S223, calculating the statistical mean: calculating the arithmetic mean of the smoothed data subset; S224. Calculate the slope of the change trend: perform linear regression analysis on the processed data subset to fit a best-fit straight line; S225. Calculate the fluctuation range: Calculate the dispersion of the processed data subset relative to its mean value, and use the standard deviation as a quantitative indicator of the fluctuation range; In step S23, the current working condition is automatically classified into the most suitable geological type through the support vector machine algorithm, and the identification result is output. Specifically, the following steps are included: S231. Constructing a real-time feature vector: combining the average value, trend, and fluctuation amplitude into a single, ordered mathematical vector; S232, performing feature vector normalization processing: inputting the mathematical vector into a pre-set normalization model for processing to obtain a normalized feature vector; S233, calling the SVM classifier for prediction: feeding the normalized feature vector as input to an existing pre-trained offline SVM classifier; S234, mapping labels and outputting final recognition results: classifying the integers output by the SVM classifier, searching and converting them to obtain human-readable geological condition names, and calculating the distance between the feature vector and the classification hyperplane, and converting it into a confidence score; S235, final result output: the confidence score is officially output as the final result of this recognition; Specifically, step S2 aims to convert the raw data stream obtained in step S1 into an accurate and automated judgment of the real-time geological conditions of the drill bit through a series of data processing and pattern recognition algorithms.

[0029] The condition identification step first establishes a geological reference. Based on the current drilling depth reported by the drill rig in real time, the expected rock and soil layer type corresponding to that depth is automatically extracted from a database of engineering geological survey reports pre-loaded into the system during the construction preparation phase.

[0030] The expected rock and soil layer types and their related physical and mechanical property indicators constitute the prior knowledge and benchmark reference for subsequent real-time geological judgment, which can significantly improve the accuracy of identification.

[0031] Subsequently, in order to extract key indicators that can characterize the core characteristics of the formation from the continuous operating condition feedback parameters, this embodiment performs time domain analysis on the operating condition feedback parameters, especially on the time series data of the rock and soil specific energy.

[0032] Time-domain analysis begins by establishing a sliding data window. This step targets the time series data of the operating condition feedback parameters, starting at the current time point and moving forward for a preset length of time or number of data points, thereby creating a dynamic, continuously sliding subset of the time series data.

[0033] To eliminate the effects of sensor noise and transient mechanical disturbances on the analysis results, this embodiment then performs data preprocessing and filtering on the time series data subset. This processing includes removing outliers and applying a smoothing filter, ultimately obtaining a smoothed data subset that reflects the true trend of change.

[0034] After obtaining the smoothed data subset, this embodiment quantifies the multi-dimensional response characteristics of the formation through calculation. Specifically, by calculating the statistical average value, change trend and fluctuation amplitude of the smoothed data subset, the real-time response characteristics of the formation are quantified.

[0035] Statistical mean, exemplarily by calculating the arithmetic mean of a subset of smoothed data It is obtained, which represents the average hardness of the formation in the current time window. Its calculation formula is: Where x i is the i-th data point in the smoothed data subset, and N is the total number of data points in the subset.

[0036] For example, the trend of change is determined by performing linear regression analysis on a subset of smoothed data and fitting a best-fit line. The slope of this line is then used as a quantitative indicator. The magnitude and sign of this slope can reflect whether the formation is gradually hardening, softening, or remaining stable.

[0037] The fluctuation amplitude is obtained by calculating the dispersion of the smoothed data subset relative to its average value, and preferably the standard deviation is used as a quantitative indicator of the fluctuation amplitude, which characterizes the homogeneity of the formation. The calculation formula is: Where, is the arithmetic mean of the smoothed data subset, N is the total number of data points in the subset, x i is the i-th data point in the smoothed data subset, and σ is the standard deviation of the fluctuation amplitude.

[0038] After extracting the above-mentioned real-time response features, this embodiment then performs feature matching and working condition classification. This step first calculates the average value The slope of the changing trend and the standard deviation of the fluctuation amplitude σ are combined into a single, ordered real-time feature vector.

[0039] To eliminate differences in dimensionality and numerical range between eigenvalues, the real-time eigenvector is then fed into a pre-set normalization model to produce a normalized eigenvector. This step is crucial for ensuring the accuracy of the subsequent classification algorithm.

[0040] Next, the normalized feature vector is fed as input to a pre-trained offline support vector machine classifier for prediction. The SVM classifier is trained based on a database of geological and working condition characteristics.

[0041] The geological and working condition database is calibrated using extensive field testing and historical construction data. It stores typical parameter characteristics corresponding to different geological conditions. For example, these conditions include at least soft clay, medium-dense sand, hard plastic clay, gravel, and heavily weathered bedrock.

[0042] Accordingly, the typical parameter characteristics at least include the value range of the rock and soil specific energy and the fluctuation pattern of the drilling torque under different geological conditions.

[0043] After receiving the normalized feature vector, the SVM classifier outputs an integer classification label representing the geological category. This embodiment further searches and converts the integer classification label through a preset mapping table to obtain a human-readable geological condition name.

[0044] At the same time, in order to evaluate the credibility of this recognition result, this embodiment further calculates the distance between the feature vector and the SVM classification hyperplane and converts it into a quantitative confidence score.

[0045] Finally, the geological condition name and confidence score are officially output as the final result of this condition identification.

[0046] S3, Decision Optimization: Determine the optimal drilling parameters based on the identified real-time geological conditions and the preset intelligent decision-making model; In step S3, the optimal drilling operation parameters are determined, which specifically includes the following steps: S31. Setting optimization goals and constraints: Based on the geological conditions and the construction goals set manually before drilling, the corresponding geological strategy knowledge base is called from the preset intelligent decision-making model; S32. Constructing a quantitative evaluation function: Based on the construction objectives, establish a comprehensive performance evaluation function that can score any set of drilling parameters; S33. Execute the optimization search algorithm: Based on the recommended range provided by the geological strategy knowledge base, start a genetic algorithm to simulate group collaboration to generate and iterate multiple sets of candidate parameter combinations within the parameter range, and score each generation of parameters using an evaluation function; S34, convergence and output of optimal solution: After multiple iterations, the genetic algorithm automatically searches for and converges to a unique set of parameter combinations that can maximize the comprehensive performance score of the evaluation function, defines this set as the optimal drilling operation parameters, and outputs the set; The geological strategy knowledge base provides an initial, safe recommended range for various drilling operation parameters under the current geological conditions. This recommended range will serve as the boundary constraint for subsequent optimization calculations. Specifically, step S3 aims to automatically determine a set of optimal drilling operation parameter combinations based on the identified geological conditions through an intelligent decision-making and optimization process.

[0047] The decision-making optimization step begins by setting optimization objectives and constraints. Based on the real-time geological conditions output in step S2 and the construction objectives manually set by the operator before drilling, the system calls the corresponding geological strategy knowledge base from a pre-set intelligent decision-making model.

[0048] Construction objectives include different strategic modes such as "efficiency first," "minimum risk," or "cost balance." The geological strategy knowledge base is pre-calibrated and provides an initial, safe range of recommended drilling parameters (such as rotation speed and pressure) for the current geological conditions (such as gravel).

[0049] This recommended range will serve as the boundary constraint for subsequent optimization calculations, ensuring that the final parameters are within the range allowed by engineering safety, and has the technical effect of limiting the optimization range and avoiding dangerous solutions.

[0050] Subsequently, this embodiment constructs a quantitative evaluation function. Based on the construction target set in S31, the function establishes a mathematical model capable of comprehensively scoring the performance of any set of drilling parameters.

[0051] The comprehensive performance evaluation function can quantitatively predict: when a certain candidate parameter combination is used for drilling, the expected construction efficiency index, the possible comprehensive risk index (such as the weighted probability of risks such as hole collapse and drill sticking), and the resource consumption index.

[0052] After determining the parameter boundary constraints and evaluation criteria, this embodiment then executes the optimization search algorithm. The system starts a genetic algorithm within the parameter recommendation range determined in S31.

[0053] By simulating the collaborative and evolutionary process of a population, the genetic algorithm generates and iterates a population of multiple candidate parameter combinations within parameter boundaries. During each iteration, each candidate parameter combination is scored using a comprehensive performance evaluation function to assess its fitness.

[0054] Finally, after multiple iterations, the genetic algorithm population will automatically converge to a unique parameter combination that maximizes the overall performance evaluation function score. This parameter combination is defined as the optimal drilling parameters because it has the best overall performance under the current constraints and objectives.

[0055] S4, closed-loop control: Automatically control the actuator of the rotary pile driver according to the optimal drilling operation parameters to perform adaptive drilling; In step S4, the actuator of the rotary pile driver is automatically controlled according to the optimal drilling operation parameters, which specifically includes the following steps: S41, command analysis and target value setting: converting the optimal drilling operation parameters into specific command set points sent to the controllers of each actuator; S42, perform real-time error feedback and correction: continuously compare the actual operating parameters with the command set point, calculate the real-time error, and dynamically fine-tune the control signal through the PID control algorithm; Specifically, step S4 aims to accurately convert the optimization decision output by S3 into the physical action of the rotary pile driver actuator and achieve stable and adaptive control of the action process.

[0056] The closed-loop control process begins with command analysis and target value setting. The system analyzes the optimal drilling parameter combination (such as optimal speed and pressure) output by S3 and converts it into specific command set points sent to the controllers of each actuator.

[0057] For example, the speed and torque set points are sent to the hydraulic controller of the power head, the pressure set point is sent to the main winch controller, and the mud density and viscosity set points are sent to the automatic mud mixing system. This step establishes the instruction bridge from optimization decision to physical execution.

[0058] The system then sends precise control signals, via a programmable logic controller, to the electro-hydraulic proportional servo valves that manage the power unit and main winch. Based on these signals, the servo valves adjust the flow and pressure of oil entering the hydraulic motor or cylinder in real time, precisely driving the drill bit to achieve the target value set by S41.

[0059] In addition, the system also issues coordinated control instructions to auxiliary operation execution agencies such as the mud circulation and deployment system to ensure that the auxiliary operation conditions in the hole (such as mud performance) can also reach and maintain the target value set in S41.

[0060] To accurately track and maintain optimal operating parameters, the core of this embodiment is to perform real-time error feedback and correction. This process forms a high-frequency closed-loop control circuit, which is the key to achieving adaptive drilling.

[0061] In this closed loop, the system continuously compares the actual operating parameters (such as actual speed, actual pressure, etc.) obtained in step S1 with the command set point issued in step S41 at high frequency, thereby calculating the real-time error between the two.

[0062] The real-time error is then fed into a built-in PID (Proportional-Integral-Derivative) control algorithm. Based on the magnitude, cumulative amount, and trend of the real-time error, the algorithm dynamically and fine-tunes the control signals sent to various actuators (such as electro-hydraulic proportional servo valves).

[0063] The rotary pile driver drilling device applicable to complex strata described below and the rotary pile driver drilling method applicable to complex strata described above may refer to each other.

[0064] Please see the attached Figure 5 The present invention also provides a rotary pile drilling device suitable for complex strata, comprising: A data acquisition module is used to obtain in real time the working condition feedback parameters of the rotary pile driver during the drilling process. The working condition feedback parameters include at least drilling mechanical parameters, in-hole environment and posture parameters, and rock and soil specific energy; The working condition identification module is connected to the data acquisition module and is used to identify the real-time geological working conditions of the drill bit by matching the real-time response characteristics of the parameters with the built-in geological-working condition characteristic database based on the working condition feedback parameters; The decision optimization module is connected to the working condition identification module and is used to determine the optimal drilling operation parameters that can maximize the overall efficiency based on the real-time geological conditions and the preset intelligent decision model through the optimization search algorithm; The closed-loop control module, connected to the decision-making optimization module, is used to interpret the optimal drilling operation parameters into command set points and automatically control the actuators of the rotary pile driver accordingly. At the same time, it performs real-time error correction based on the actual working condition parameters fed back by the data acquisition module to achieve adaptive drilling. The drilling tool selection and recommendation module is connected to the working condition identification module and is used to match and output the corresponding optimal detachable drill bit model from the database after receiving the real-time geological conditions output by the working condition identification module; In the drilling tool selection and recommendation module, the geological drill bit matching database includes at least: When the geological conditions are soft clay and hard plastic clay, it is recommended to use a double-bottom sand bucket and a rotary drilling bucket; When the geological conditions are pebble layers, it is recommended to use a rotary drill bit inlaid with cemented carbide; When the geological conditions are strongly weathered bedrock, it is recommended to use roller drill bits and combined rock drill bits.

[0065] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A rotary pile drilling method suitable for complex strata, characterized in that: The following steps are involved: S1. Data acquisition: Real-time acquisition of working condition feedback parameters of the rotary pile driver during the drilling process; S2. Working condition identification: Based on the working condition feedback parameters, the real-time geological working condition of the drill bit is identified; S3, decision optimization: determining optimal drilling operation parameters based on the identified real-time geological conditions and in combination with a preset intelligent decision model; S4. Closed-loop control: Automatically controlling the actuator of the rotary pile driver according to the optimal drilling operation parameters to perform adaptive drilling.

2. The rotary pile drilling method suitable for complex strata according to claim 1, characterized in that: In step S1, the real-time acquisition of working condition feedback parameters of the rotary pile driver during the drilling process specifically includes the following steps: S11. Directly measure and obtain drilling torque, drilling speed, pressure and drilling speed through sensors installed in the rotary drilling rig power head and main winch hydraulic system; S12, using an in-situ bottom hole monitoring assembly installed at the bottom of the drill pipe, close to the drill bit, to measure and obtain the real-time specific gravity and viscosity of the mud at the bottom of the hole, as well as the real-time verticality of the drill pipe; S13. Based on the data obtained in step S11, a built-in physical model is used to perform real-time calculation to generate the specific energy of the rock and soil.

3. The rotary pile drilling method suitable for complex strata according to claim 2, characterized in that: In step S13, the generated rock and soil mass specific energy is calculated using the following formula: Where, E s is the specific energy of the rock and soil, F is the applied pressure obtained in real time, A is the preset cross-sectional area of ​​the borehole, n is the drilling speed obtained in real time, M is the drilling torque obtained in real time, and v p is the real-time drilling speed, and 2π is the constant coefficient for converting the rotational speed and torque into power.

4. The rotary pile drilling method suitable for complex strata according to claim 1, characterized in that: In step S2, the identification of the real-time geological conditions of the drill bit specifically includes the following steps: S21. Establish a baseline geological profile: Based on the current drilling depth, extract the expected rock and soil layer type and its physical and mechanical property indicators corresponding to the depth from the engineering geological survey report database loaded during the construction preparation phase, which will serve as a baseline reference for subsequent geological identification and judgment; S22. Extracting real-time formation response characteristics: performing time domain analysis based on the working condition feedback parameters, and quantifying the real-time response characteristics of the formation by calculating the average value, change trend, and fluctuation amplitude within the current time window; S23, performing feature matching and operating condition classification: performing high-speed pattern matching on the real-time response features and the geological and operating condition feature database built into the intelligent control center, automatically classifying the current operating condition into the most suitable geological type through a support vector machine algorithm, and outputting the recognition result; The geological and working condition characteristic database is pre-calibrated through experiments and historical data, and stores typical parameter characteristics corresponding to different geological working conditions; The geological conditions include soft clay, medium-dense sand layer, hard plastic clay, pebble layer and strongly weathered bedrock; The typical parameter characteristics include the range of values ​​of the rock and soil specific energy and the fluctuation mode of the torque.

5. The rotary pile drilling method suitable for complex strata according to claim 4, characterized in that: In step S22, the time domain analysis is performed based on the operating condition feedback parameters, specifically including the following steps: S221. Establishing a sliding data window: For the operating condition feedback parameters, starting from the current time point and tracing back a fixed time length, a dynamic, continuously sliding forward time series data subset is established; S222, data preprocessing and filtering: cleaning the data subset to eliminate the influence of sensor noise and transient disturbance on the analysis results, and obtaining a smoothed data subset; The calculation of the average value, change trend and fluctuation range within the current time window includes the following steps: S223, calculating the statistical mean: calculating the arithmetic mean of the smoothed data subset; S224, calculating the slope of the change trend: performing linear regression analysis on the processed data subset to fit a best-fit straight line; S225. Calculate the fluctuation range: calculate the dispersion of the processed data subset relative to its average value, and use the standard deviation as a quantitative indicator of the fluctuation range.

6. The rotary pile drilling method suitable for complex strata according to claim 4, characterized in that: In step S23, the current working condition is automatically classified into the most suitable geological type through the support vector machine algorithm, and the identification result is output, which specifically includes the following steps: S231, constructing a real-time feature vector: combining the average value, change trend, and fluctuation amplitude into a single, ordered mathematical vector; S232, performing feature vector normalization processing: inputting the mathematical vector into a pre-set normalization model for processing to obtain a normalized feature vector; S233, calling the SVM classifier for prediction: feeding the normalized feature vector as input to an existing pre-trained offline SVM classifier; S234, mapping labels and outputting final recognition results: classifying the integers output by the SVM classifier, searching and converting them to obtain human-readable geological condition names, and calculating the distance between the feature vector and the classification hyperplane, and converting it into a confidence score; S235, final result output: the confidence score is formally output as the final result of this recognition.

7. The rotary pile drilling method suitable for complex strata according to claim 1, characterized in that: In step S3, determining the optimal drilling operation parameters specifically includes the following steps: S31. Setting optimization goals and constraints: Based on the geological conditions and in combination with the construction goals manually set before drilling, the corresponding geological strategy knowledge base is called from the preset intelligent decision-making model; S32. Constructing a quantitative evaluation function: Based on the construction goal, establishing a comprehensive performance evaluation function capable of scoring any set of drilling parameters; S33, executing an optimization search algorithm: based on the recommended range provided by the geological strategy knowledge base, starting a genetic algorithm to generate and continuously iterate multiple sets of candidate parameter combinations within the parameter range by simulating group collaboration, and scoring each generation of parameters using the evaluation function; S34. Convergence and output of the optimal solution: After multiple iterations, the genetic algorithm automatically searches for and converges to a unique set of parameter combinations that can give the evaluation function the highest comprehensive performance score, defines it as the optimal drilling operation parameters, and outputs it; the geological strategy knowledge base provides an initial, safe recommended range for various drilling operation parameters under the current geological conditions, and this recommended range will serve as the boundary constraint condition for subsequent optimization calculations.

8. The rotary pile drilling method suitable for complex strata according to claim 1, characterized in that: In step S4, the automatic control of the actuator of the rotary pile driver according to the optimal drilling operation parameters specifically includes the following steps: S41, command analysis and target value setting: converting the optimal drilling operation parameters into specific command setting points sent to the controllers of each actuator; S42, performing real-time error feedback and correction: continuously comparing the actual operating parameters with the command set point, calculating the real-time error, and dynamically fine-tuning the control signal through the PID control algorithm.

9. A rotary pile driver drilling device suitable for complex strata, a rotary pile driver drilling method suitable for complex strata according to any one of claims 1 to 8, characterized in that: include: A data acquisition module is used to obtain in real time the working condition feedback parameters of the rotary pile driver during the drilling process, wherein the working condition feedback parameters include at least drilling mechanical parameters, in-hole environment and posture parameters, and rock and soil specific energy; a working condition identification module connected to the data acquisition module, for identifying the real-time geological working conditions of the drill bit based on the working condition feedback parameters by matching the real-time response characteristics of the parameters with a built-in geological-working condition characteristic database; a decision optimization module connected to the working condition identification module, for determining the optimal drilling operation parameters that can maximize the overall efficiency based on the real-time geological working conditions and in combination with a preset intelligent decision model through an optimization search algorithm; a closed-loop control module, connected to the decision optimization module, for parsing the optimal drilling operation parameters into command set points, automatically controlling the actuators of the rotary pile driver accordingly, and performing real-time error correction based on the actual operating parameters fed back by the data acquisition module to perform adaptive drilling; The drilling tool selection and recommendation module is connected to the working condition identification module and is used to match and output the corresponding optimal detachable drill bit model recommendation from the database after receiving the real-time geological working conditions output by the working condition identification module.

10. The rotary pile drilling device suitable for complex strata according to claim 9, characterized in that: In the drilling tool selection and recommendation module, the geological drill bit matching database includes at least: When the geological conditions are soft clay and hard plastic clay, it is recommended to use a double-bottom sand bucket and a rotary drilling bucket; When the geological conditions are pebble layers, it is recommended to use a rotary drill bit inlaid with cemented carbide; When the geological conditions are strongly weathered bedrock, it is recommended to use roller drill bits and combined rock drill bits.

Images