Tunneling management method and system for exploration drilling operation

Abstract

The invention discloses a tunneling management method and system for exploration drilling operation, and relates to the technical field of geological exploration and drilling engineering, and the method comprises the steps: collecting original data in real time through a sensor; data transmission, filtering and standardized format conversion; setting a judgment threshold parameter of the drilling action state; analyzing the data stream, and identifying and dividing a plurality of technical action states; generating a drilling curve and a depth checking result, and calculating an action proportion and time distribution; analyzing time distribution and depth data, and outputting an evaluation report; the system comprises a multi-source sensing data acquisition module, a data preprocessing and transmission module, a drilling action state identification module, an efficiency analysis and visualization module and a user interaction and parameter configuration module. According to the technical scheme, accurate division and time statistics of drilling action states can be achieved, a quantitative and verifiable drilling efficiency evaluation index system is provided, and the authenticity and process transparency of drilling operation are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of geological exploration and drilling engineering, and particularly relates to a tunneling management method and system for surveying and drilling drilling operations. BACKGROUND

[0002] As an indispensable basic work in the early stage of engineering construction, one of the core links of geological exploration is drilling operation. This operation obtains the physical and mechanical parameters of rock-soil layers and the information of stratum structure by drilling from the ground to the underground, and provides key basis for subsequent design and construction. With the continuous expansion of infrastructure construction scale, geological drilling has been widely used in the fields of transportation, water conservancy, energy and urban underground space development, forming a relatively mature equipment system and process flow. However, the essence of drilling operation is not a simple mechanical drilling process, but a complex dynamic system composed of drilling, pressing, lifting, drill pipe connection and separation, circulation flushing and equipment idling and other action states. Its efficiency is affected by multiple factors such as geological conditions, equipment performance and operation habits, resulting in high nonlinearity and unpredictability of the operation process. Under this background, how to objectively, quantitatively and traceably evaluate the efficiency of the whole drilling process has become a key bottleneck restricting the quality improvement and efficiency increase and digital transformation of the exploration industry.

[0003] Among them, the tunneling management of surveying and drilling operation focuses on the accurate identification and efficiency analysis of each action state of the drilling machine under real working conditions. The core goal of this direction is to break through the traditional supervision mode relying on manual observation and experience judgment, and instead build a state discrimination mechanism based on multi-source sensing data. Through real-time collection and logical correlation analysis of drilling machine operating parameters, automatic division and time proportion statistics of various technical actions in the drilling process are realized. This technical path not only requires the system to have high-precision data perception ability, but also needs to establish state judgment criteria with physical meaning and engineering interpretability, so as to convert the originally fuzzy and subjective operation process into a structured and calculable efficiency index system, laying a data foundation for subsequent operation optimization and intelligent decision-making.

[0004] The prior art has significant defects in drilling operation management: first, the drilling process lacks unified action state definition standards, cannot automatically identify key links such as drilling, lifting, connection, etc., and leads to efficiency analysis remaining at the coarse-grained level of "total footage / total time", making it difficult to reveal the specific source of efficiency loss; second, on-site supervision is highly dependent on manpower, and the skill levels of operators vary, and the supervision behavior itself is lagging, subjective and intermittent, and cannot achieve continuous monitoring and immediate intervention throughout the process; third, although drilling rig operation data can be collected, there is a lack of effective algorithm model to map raw parameters to action states with engineering semantics, resulting in massive data in a "sleeping" state and failing to be converted into management value; finally, existing systems generally lack deep integration with information management platforms, cannot support background visual analysis, and are difficult to generate standardized efficiency evaluation reports, which seriously deviates from the strategic direction of the state to promote new productivity development, realize intelligentization of the whole survey process and reduce costs and increase efficiency. SUMMARY

[0005] The purpose of the present application is to provide a kind of survey drilling drilling operation's driving management method and system, to solve the problem of strong subjectivity in drilling efficiency analysis in prior art and lack of quantitative evaluation standard.

[0006] To solve the above technical problems, the present application provides the following technical solutions: On the one hand, a driving management method for drilling operation of survey drilling, the specific steps of which are: Step S110, through the multi-source sensor deployed on the drilling rig, real-time collection of working lifting oil pressure, down pressure, drill pipe speed, drill pipe steering, real-time drilling position of drill bit and circulating water flow rate raw data throughout the whole process of geological drilling operation; Step S120, the collected raw data is transmitted to the server, and the embedded software is used to perform data import, preliminary filtering and standardized format conversion operation on the host MCU, to generate a data stream suitable for algorithm processing; Step S130, based on the converted standard data, using manual user intervention or fully automatic mode, setting specific node parameter threshold values for drilling action state determination, providing prior information input for subsequent algorithm model; Step S140, applying drilling action state differentiation criteria and algorithms, real-time analysis of data stream, automatic identification and division of drilling, down pressure, lifting, drill pipe connection and separation, circulating flushing and equipment idle, and recording the start time stamp and duration of each state; Step S150, according to the identified action state sequence, generating a complete drilling curve and drilling depth checking result, and synchronously calculating the graphical proportion and time proportion distribution of each key action; Step S160, based on the time distribution of the action state and the drilling depth data, the working time statistics, action decomposition, distribution rule analysis and drilling efficiency comprehensive evaluation are completed, and a standardized evaluation report is output.

[0007] The drilling action state distinguishing algorithm in step S140 adopts a real-time classification method based on multi-feature fusion, specifically including three sub-steps of feature extraction, feature fusion and state classification: in the feature extraction stage, time domain features (mean, variance, peak) and frequency domain features (main frequency, energy spectrum entropy) are extracted from the preprocessed data stream; in the feature fusion stage, the extracted multiple features are combined into a high-dimensional feature vector, and dimension reduction processing is performed through principal component analysis, and the principal components with cumulative variance contribution rate exceeding 95% are retained; in the state classification stage, the dimension-reduced feature vector is input into the trained ensemble learning model, and the action state label corresponding to the current time window is output.

[0008] The drilling efficiency comprehensive evaluation in step S160 adopts a multi-level index system, including time utilization rate index, footage efficiency index and energy consumption efficiency index; the time utilization rate index is defined as the ratio of pure drilling time to total operation time; the footage efficiency index is defined as the average drilling footage per unit time, and the calculation formula is:

[0009] Among them, is the total drilling footage, is the pure drilling time; the energy consumption efficiency index is defined as the hydraulic energy consumption per unit footage, which is obtained by integrating the product of oil pressure and flow rate with respect to time.

[0010] On the other hand, a tunneling management system for survey drilling drilling operation, the system comprises the following components: A multi-source sensor data acquisition module is used to acquire the working lifting oil pressure, the down pressure, the drill pipe rotating speed, the drill pipe steering, the real-time drilling position of the drill bit and the circulating water flow rate parameters in the drilling operation process in real time; A data preprocessing and transmission module is connected to the multi-source sensor data acquisition module, used for preliminary filtering and format standardization processing of the collected raw data, and uploading the processed data to the cloud database through the communication interface; A drilling action state recognition module is deployed on the server side, receiving data stream from the data preprocessing and transmission module, based on the preset drilling action state distinguishing criteria and algorithm model, automatically distinguishing and time stamping the drilling, down pressure, lifting, drill pipe connection and separation, circulating flushing and equipment idle of the drilling rig; An efficiency analysis and visualization module connected to the drilling action state recognition module, configured to generate a complete drilling curve, drilling depth checking results, time proportion analysis graphics of each key action, and drilling efficiency summary report according to the identified action state sequence; A user interaction and parameter configuration module integrated in the embedded software and running on the main control MCU, configured to provide device calibration, historical data query, result data packet file generation functions, and allow the user to set specific node parameters through a manual intervention mode, to provide prior information for the algorithm to improve the state recognition accuracy.

[0011] The multi-source sensing data acquisition module includes a pressure sensor, a rotating speed encoder, a displacement sensor, and a flow meter. The pressure sensor is installed on the working lifting oil path and the down pressure oil path of the drilling rig hydraulic system, with a measurement range of 0-40 MPa and a sampling frequency of not less than 100 Hz. The rotating speed encoder is coupled to the drill rod driving shaft to monitor the drill rod rotating speed and steering signal in real time, with a rotating speed measurement accuracy of ±0.5 rpm. The displacement sensor is a magnetostrictive linear displacement sensor installed on the drill bit advancing mechanism to feed back the drill bit drilling position in real time, with a resolution of better than 0.1 mm. The flow meter is integrated in the circulating water system to monitor the circulating water flow rate, with a measurement range of 0-200 L / min.

[0012] The data filtering operation performed by the data preprocessing and transmission module adopts a sliding window mean filtering algorithm, with a window length of 50 sampling points and a standard deviation multiple threshold of 2.0, to eliminate abnormal data points caused by sensor noise or transient interference. The data standardization processing includes dimension unification and normalization operation, which normalizes the oil pressure data to the interval [0, 1], converts the rotating speed data to standard unit rpm, converts the displacement data to absolute depth value, and converts the flow data to L / min unit. The communication interface supports 4G / 5G wireless transmission protocol and Ethernet wired transmission protocol, and the data transmission interval can be configured to 1-10 seconds.

[0013] The drilling action state recognition module is preconfigured with a drilling action state distinguishing criterion based on multi-parameter joint decision logic, which is specifically defined as: when the real-time drilling position of the drill bit continuously increases and the down pressure oil pressure is greater than a set threshold , and the drill rod rotating speed is greater than a set threshold , it is determined as a drilling state; when the working lifting oil pressure is greater than a set threshold and the drill bit position continuously rises, it is determined as a lifting state; when the drill rod rotating speed is lower than a set threshold and the circulating water flow rate is lower than a set threshold , it is determined as an idle equipment state; the drill rod connection and separation state is recognized by the mutation of the drill rod steering signal and the rotating speed zeroing feature; the circulating flushing state is recognized by the circulating water flow rate being greater than a set threshold And the drill pipe rotation speed is lower than Make a judgment.

[0014] The drilling action state recognition module employs an ensemble learning model based on time series classification. This model integrates decision trees, support vector machines, and long short-term memory networks. The long short-term memory network has 128 hidden layer units, an input time window of 60 seconds, and uses cross-entropy loss as the loss function for model training. The optimizer is Adam, and the initial learning rate is set to 0.001. The final output of the ensemble learning model is determined through a weighted voting mechanism, with the decision tree weight at 0.3, the support vector machine weight at 0.3, and the long short-term memory network weight at 0.4.

[0015] The node parameter setting function provided by the user interaction and parameter configuration module includes a drilling status determination threshold. Increase the state determination threshold Idle state determination threshold and 1. Threshold for determining flushing status The adjustable interface allows users to dynamically adjust these threshold parameters within a preset range based on actual geological conditions and drilling rig model. The module also provides a historical data query function, supporting queries based on various conditions such as time range, borehole number, and action type. The query results are displayed in the form of tables and trend charts. The equipment calibration function includes sensor zero-point calibration, range calibration, and communication self-test. Calibration data is stored in non-volatile memory.

[0016] The efficiency analysis and visualization module generates a complete drilling curve, which includes a synchronous overlay display of drill bit depth-time curve, oil pressure-time curve, rotational speed-time curve, and flow rate-time curve. Drilling depth verification is calculated by comparing the drill bit position measured by sensors with the theoretical drilling footage. The verification error calculation formula is as follows:

[0017] in, For the sensor to measure depth, For the theoretical calculation depth, when A depth anomaly alarm is triggered when the depth exceeds a set threshold of 3%. The time percentage analysis of each key action is visualized using a combination of pie charts and bar charts, and the percentage of pure drilling time, auxiliary operation time, and idle time are automatically calculated.

[0018] Compared with the prior art, the beneficial technical effects of the present invention are as follows: This invention establishes an automatic identification system for drilling action status based on multi-source sensor data, which realizes accurate classification and time statistics of various technical actions such as drilling, hoisting, connection, flushing, and idleness, and solves the problems of subjectivity and lag caused by the reliance on manual observation in traditional methods.

[0019] This invention integrates embedded software, cloud data processing, and visualization analysis modules to construct an intelligent management platform covering the entire chain from data acquisition, transmission, processing to result output. It supports remote real-time monitoring and historical data traceability, significantly improving the ability of exploration units to prove the authenticity of drilling operations and enhancing process transparency.

[0020] This invention introduces a user-configurable parameter threshold and algorithm prior information input mechanism, which enhances the system's adaptability to different geological conditions and equipment models, and improves the accuracy of state recognition and engineering practicality by dynamically adjusting the judgment criteria.

[0021] This invention generates drilling curves, depth verification, time percentage analysis, and efficiency reports based on action state sequences, providing a quantitative and verifiable drilling efficiency evaluation index system. It provides a data foundation for operation optimization and decision support, and promotes the transformation of geological drilling operations towards scientific, refined, and intelligent management. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the tunneling management system architecture for exploration drilling operations proposed in this invention; Figure 2 This is a schematic diagram of a tunneling management method for exploration drilling operations proposed in this invention. Detailed Implementation

[0023] To further illustrate the technical means and effects adopted by the present invention to achieve the intended purpose, the specific embodiments according to the present invention will be described in detail below with reference to the accompanying drawings and preferred embodiments.

[0024] Example 1 In complex geological exploration drilling projects, multi-source sensor data acquisition modules comprehensively collect data throughout the entire drilling operation. Pressure sensors are installed in the working lift and lower pressure oil circuits of the drilling rig's hydraulic system, monitoring the working lift and lower pressure parameters in real time. The measurement range covers 0-40MPa, and the sampling frequency is set to 100Hz to ensure the capture of instantaneous pressure fluctuations in the hydraulic system. A speed encoder is rigidly connected to the drill rod drive shaft via a coupling, acquiring drill rod speed and direction signals in real time. The speed measurement accuracy reaches ±0.5rpm, and the direction signal is determined through orthogonal encoded pulse sequences. A magnetostrictive linear displacement sensor is installed on the guide device of the drill bit propulsion mechanism, providing real-time feedback on the drill bit's drilling position with a resolution better than 0.1mm. Absolute value encoding output ensures the continuity of position data. A flow meter is integrated into the outlet pipe of the circulating water system, using electromagnetic flow measurement principles to monitor the circulating water flow rate, covering a range of 0-200L / min, with a measurement error controlled within ±1%. All sensors are connected to the data acquisition terminal via shielded cables, and the signals are digitized using a 24-bit analog-to-digital converter, with a sampling synchronization error of less than 1 millisecond.

[0025] The data preprocessing and transmission module receives the raw data stream from the multi-source sensor data acquisition module. Data filtering employs a sliding window mean filtering algorithm with a window length of 50 sampling points and a standard deviation threshold of 2.0. By calculating the statistical characteristics of the data within the window, it automatically identifies and removes abnormal data points caused by sensor noise or transient interference. Data standardization processing includes dimension unification and normalization operations. Oil pressure data is normalized to the 0-1 range through linear transformation; speed data is converted to standard units (rpm); displacement data is converted to absolute depth values ​​through coordinate transformation; and flow rate data is converted to units (L / min). The communication interface supports 4G and 5G wireless transmission protocols and is also compatible with Ethernet wired transmission protocols. The data transmission interval is configurable from 1 to 10 seconds depending on network conditions. Data compression and encryption technologies ensure data transmission security and real-time performance. See also... Figure 1 This system includes a multi-source sensor data acquisition module, a data preprocessing and transmission module, a drilling action status identification module, an efficiency analysis and visualization module, and a user interaction and parameter configuration module.

[0026] The drilling action status recognition module is deployed on a cloud server cluster, receiving standard data streams from the data preprocessing and transmission module. This module constructs drilling action status differentiation criteria based on multi-parameter joint judgment logic, specifically defined as follows: when the drill bit's real-time drilling position continuously increases and the downpressure oil pressure exceeds a set threshold of 5 MPa, and the drill rod rotation speed exceeds a set threshold of 30 rpm, it is determined to be in drilling status; when the working lifting oil pressure exceeds a set threshold of 8 MPa and the drill bit position continuously rises, it is determined to be in lifting status; when the drill rod rotation speed is below a set threshold of 5 rpm and the circulating water flow rate is below a set threshold of 10 L / min, it is determined to be in idle status; the drill rod connection and disconnection status is identified through the abrupt change characteristics of the drill rod steering signal and the zero-speed characteristic. The rapid switching of the steering signal between forward and reverse rotation, combined with the speed dropping to zero, indicates a drill rod connection operation; the circulating flushing status is determined by the circulating water flow rate exceeding a set threshold of 50 L / min and the drill rod rotation speed being below 30 rpm.

[0027] The drilling action state recognition module includes a feature extraction unit, a feature fusion unit, and a state classification unit.

[0028] The drilling action state recognition module employs an ensemble learning model based on time series classification, which integrates decision trees, support vector machines, and long short-term memory (LSTM) networks. The feature extraction stage extracts time-domain features, including mean, variance, and peak values, and frequency-domain features, including dominant frequency and energy spectral entropy parameters, from the preprocessed data stream. The feature fusion stage combines the extracted features into a high-dimensional feature vector, performs dimensionality reduction using principal component analysis (PCA), and retains principal components with a cumulative variance contribution exceeding 95%. The state classification stage inputs the dimensionality-reduced feature vector into the trained ensemble learning model. The LSM network has 128 hidden layer units, an input time window of 60 seconds, and uses cross-entropy loss as the loss function. The optimizer is Adam, and the initial learning rate is set to 0.001. The final output of the ensemble learning model is determined through a weighted voting mechanism: the decision tree weight is set to 0.3, the support vector machine weight to 0.3, and the LSM network weight to 0.4.

[0029] The efficiency analysis and visualization module connects to the drilling action status recognition module, generating a complete drilling curve based on the recognized action status sequence. The complete drilling curve includes a synchronized overlay display of drill bit depth-time, oil pressure-time, rotational speed-time, and flow rate-time curves, using different colors and line types to distinguish each parameter curve, and supporting zooming and panning operations on the time axis. Drilling depth verification is calculated by comparing the drill bit position measured by sensors with the theoretical drilling footage. The verification error calculation formula is the absolute value of the difference between the measured depth and the theoretical depth divided by the theoretical depth and then multiplied by 100%. When the calculated result exceeds a set threshold of 3%, a depth anomaly alarm is triggered, and the alarm information is notified to the operator via audible and visual signals and message push notifications. The time percentage analysis of each key action is visualized using a combination of pie charts and bar charts, automatically calculating the percentage of pure drilling time, auxiliary operation time, and idle time, supporting statistical analysis by shift, borehole, and date.

[0030] The user interaction and parameter configuration modules are integrated into the embedded software and run on the main control MCU. The node parameter setting function provides an adjustable interface for drilling status judgment thresholds, lifting status judgment thresholds, idle status judgment thresholds, and flushing status judgment thresholds. Users can dynamically adjust these threshold parameters within preset ranges according to actual geological conditions and drilling rig models. The historical data query function supports queries based on multiple conditions such as time range, borehole number, and action type. Query results are displayed in table and trend chart formats, and data can be exported to Excel and PDF formats. The equipment calibration function includes sensor zero-point calibration, range calibration, and communication self-test. Calibration data is stored in non-volatile memory to ensure that calibration parameters are not lost after power failure. The manual intervention mode allows users to set specific node parameters through a touchscreen interface, providing prior information to the algorithm to improve status recognition accuracy.

[0031] Step S110 uses multi-source sensors deployed on the drilling rig to collect real-time raw data on the working oil pressure, downpressure, drill rod speed, drill rod rotation, drill bit real-time drilling position, and circulating water flow rate throughout the geological drilling operation. The pressure sensor uses a piezoresistive sensing principle with a built-in temperature compensation circuit to ensure measurement accuracy remains within ±0.5%FS in ambient temperatures ranging from -20℃ to 80℃. The speed encoder uses a photoelectric incremental encoder, outputting 1024 pulses per revolution, and the resolution is increased to 4096 pulses / revolution through quadruple frequency multiplication. The displacement sensor is installed parallel to the drill rod axis and uses an absolute position measurement mode to avoid position information loss due to power outages. The flow meter is installed in the straight pipe section at the outlet of the circulating water pump, ensuring sufficient straight pipe length before and after to guarantee a stable flow field.

[0032] Step S120 involves transmitting the collected raw data to the server. Embedded software on the main control MCU performs data import, preliminary filtering, and standardized format conversion operations. During data import, data integrity is checked, verifying the start bit, checksum, and end bit of the data frame, discarding data frames that fail verification. Preliminary filtering uses a sliding window algorithm with a window size of 50 sampling points. The mean and standard deviation of the data within the window are calculated, and data points deviating from the mean by more than twice the standard deviation are removed. Standardized format conversion converts the raw oil pressure value by dividing it by the 40MPa range to a dimensionless value between 0 and 1. Rotational speed data retains the rpm unit but is uniformly formatted as a floating-point number. Displacement data is converted to an absolute depth value based on the orifice. Flow rate data is uniformly formatted as L / min and retained to one decimal place.

[0033] Step S130, based on the converted standard data, sets specific node parameter thresholds for determining the drilling action status using either manual user intervention or fully automatic mode. In manual user intervention mode, the operator inputs the threshold values ​​for each status through a touchscreen interface. The system updates the algorithm parameters after verifying that the input values ​​are within a reasonable range. In fully automatic mode, the system automatically calculates the optimal threshold parameters for each status determination based on statistical analysis of historical operation data. These threshold parameters include a drilling status determination threshold of 5 MPa, a hoisting status determination threshold of 8 MPa, an idle state rotation speed threshold of 5 rpm, an idle state flow rate threshold of 10 L / min, and a flushing state flow rate threshold of 50 L / min. These parameters are stored in a configuration file and support batch modification and import / export.

[0034] Step S140 applies drilling action state differentiation criteria and algorithms to perform real-time analysis of the data stream, automatically identifying and classifying various technical action states such as drilling, pressing down, hoisting, drill pipe connection and separation, circulating flushing, and equipment idleness. In the feature extraction stage, 12 time-domain features and 8 frequency-domain features are extracted from the data stream within a 60-second time window. Time-domain features include mean, variance, peak factor, impulse factor, waveform factor, skewness, kurtosis, root mean square value, absolute mean, variance, standard deviation, and average amplitude. Frequency-domain features are calculated using Fast Fourier Transform and include dominant frequency, centroid frequency, mean square frequency, frequency variance, spectral kurtosis, spectral skewness, spectral entropy, and energy spectral entropy. In the feature fusion stage, the 20-dimensional feature vector is reduced to 8 dimensions using principal component analysis, ensuring an information loss rate of less than 5%. In the state classification stage, the dimensionality-reduced feature vector is simultaneously input into a decision tree, support vector machine, and long short-term memory network. The final state label is determined based on the weighted voting results, and the start timestamp and duration of each state are recorded with millisecond-level accuracy.

[0035] Step S150 generates a complete drilling curve and drilling depth verification results based on the identified action state sequence. The complete drilling curve is displayed in a four-quadrant format: the first quadrant displays the depth-time curve, the second quadrant displays the oil pressure-time curve, the third quadrant displays the rotational speed-time curve, and the fourth quadrant displays the flow rate-time curve. All curves use a unified time axis and support cursor tracking to display parameter values ​​at specific moments. The drilling depth verification compares the drill bit position measured by the displacement sensor with the theoretical depth calculated based on drilling time and drilling rate. If the verification error exceeds 3%, the system automatically marks the abnormal section and prompts the operator to check the sensors or drilling parameters. The graphical proportions and time percentages of each key action are visually displayed using a pie chart. The pie chart is divided into seven sector areas: drilling state, hoisting state, pressing state, drill pipe connection state, drill pipe separation state, circulation flushing state, and equipment idle state. The area proportion of each sector corresponds to the time percentage of each state.

[0036] Step S160, based on the time distribution of action states and drilling depth data, completes the statistics of working time, action decomposition, distribution pattern analysis, and comprehensive evaluation of drilling efficiency. Working time statistics are compiled by shift, summarizing pure drilling time, auxiliary operation time, and equipment idle time, and calculating the time utilization rate index, i.e., the ratio of pure drilling time to total working time. Action decomposition breaks down each work cycle into sub-actions such as drilling, hoisting, connecting drill pipe, unloading drill pipe, flushing the bottom of the hole, and equipment adjustment, and counts the number of times and average time for each sub-action. Distribution pattern analysis uses statistical methods to study the changing patterns of the time proportion of each action in different formations, establishing a correlation model between formation characteristics and working efficiency. The comprehensive evaluation of drilling efficiency adopts a multi-level index system, including time utilization rate index, footage efficiency index, and energy consumption efficiency index. The footage efficiency index is defined as the average drilling footage per unit time, calculated by dividing the total drilling footage by the pure drilling time. The energy consumption efficiency index is defined as the hydraulic energy consumed per unit footage, obtained by integrating the product of oil pressure and flow rate over time. The final output is a standardized evaluation report, which includes four parts: work overview, efficiency analysis, anomaly record, and improvement suggestions.

[0037] Example 2 At deep mineral resource exploration drilling sites, multi-source sensor data acquisition modules are optimized for the specific needs of ultra-deep hole drilling operations. Pressure sensors utilize high-pressure resistant products, extending the measurement range to 0-60 MPa, ensuring accurate monitoring of hydraulic system conditions even at drilling depths exceeding 1000 meters. The speed encoder's waterproof and dustproof rating is upgraded to IP68, adapting to the high-temperature and high-humidity environment downhole. Displacement sensors employ a multi-stage cascade installation method, using data fusion from multiple sensors to address the challenges of ultra-long stroke measurements. Flow meters feature a self-cleaning function to prevent rock cuttings from affecting measurement accuracy. The data preprocessing and transmission module adopts an edge computing architecture, with a high-performance processor built into the data acquisition terminal, enabling localized execution of data filtering and standardization, reducing data transmission volume. The transmission interval is adjusted to 5 seconds, ensuring stable transmission even in weak network environments.

[0038] The drilling action status recognition module optimizes the status judgment criteria for deep hole drilling. The drilling status judgment threshold is dynamically adjusted according to the hole depth, using a 5MPa threshold for shallow holes and gradually increasing to 8MPa for deeper holes to adapt to changes in formation resistance. For lifting status, the elastic deformation of ultra-long drill strings is considered, and an acceleration parameter is introduced into the judgment logic to avoid misjudgments. Torque monitoring parameters are added to the drill pipe connection status recognition, improving the accuracy of connection operation recognition by analyzing torque change curves. The algorithm model incorporates a transfer learning mechanism, using a basic model trained on existing borehole data, which can be fine-tuned with a small amount of new borehole data to adapt to new formation conditions. The efficiency analysis and visualization module adds a formation lithology identification function, automatically identifying the encountered formation lithology by analyzing drilling parameters and vibration signals, and marking the lithology interface on the drilling curve. The user interaction and parameter configuration module adds an adaptive parameter tuning function; the system automatically optimizes the status judgment threshold based on real-time operation data, reducing the need for manual intervention.

[0039] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for managing tunneling operations during exploration drilling, characterized in that, The method includes the following steps: Step S110: Through multi-source sensors deployed on the drilling rig, real-time data of working oil pressure, downpressure, drill rod speed, drill rod rotation, real-time drilling position of the drill bit, and circulating water flow rate are collected throughout the geological drilling operation. Step S120: The collected raw data is transmitted to the server. Embedded software is used on the main control MCU to perform data import, preliminary filtering and standardized format conversion operations to generate a data stream suitable for algorithm processing. Step S130: Based on the converted standard data, set the threshold values ​​for specific node parameters for determining the drilling action status using either manual user intervention or fully automatic mode. Step S140: Apply drilling action state differentiation criteria and algorithms to perform real-time analysis of the data stream, automatically identify and classify drilling, pressing, hoisting, drill pipe connection and separation, circulation flushing and equipment idle technical action states, and record the start timestamp and duration of each state. Step S150: Based on the identified action state sequence, generate a complete drilling curve and drilling depth verification results, and simultaneously calculate the graphical proportion and time percentage distribution of each key action. Step S160: Based on the time distribution of action status and drilling depth data, complete the statistics of working time, action decomposition, distribution pattern analysis and comprehensive evaluation of drilling efficiency, and output a standardized evaluation report.

2. The tunneling management method for exploration drilling operations according to claim 1, characterized in that: In step S140, the drilling action state differentiation algorithm adopts a real-time classification method based on multi-feature fusion, which specifically includes three sub-steps: feature extraction, feature fusion, and state classification. The feature extraction stage extracts 12 time-domain features and 8 frequency-domain features from the data stream within a 60-second time window. The time-domain features include mean, peak factor, impulse factor, waveform factor, skewness, kurtosis, root mean square value, absolute mean, variance, standard deviation, and average amplitude. The frequency-domain features are calculated using Fast Fourier Transform and include dominant frequency, centroid frequency, mean square frequency, frequency variance, spectral kurtosis, spectral skewness, spectral entropy, and energy spectral entropy.

3. The tunneling management method for exploration drilling operations according to claim 1, characterized in that: In step S150, the complete drilling curve is displayed in a four-quadrant mode. The first quadrant displays the depth-time curve, the second quadrant displays the oil pressure-time curve, the third quadrant displays the rotational speed-time curve, and the fourth quadrant displays the flow rate-time curve. All curves use a unified time axis and support cursor tracking to display parameter values ​​at specific moments.

4. The tunneling management method for exploration drilling operations according to claim 1, characterized in that: In step S160, the working time statistics are summarized by shift, including pure drilling time, auxiliary operation time, and equipment idle time, and the time utilization rate index is calculated; the action decomposition subdivides each work cycle into drilling, drill lifting, drill rod connection, drill rod unloading, bottom hole flushing, and equipment adjustment sub-actions, and the number of times and average time of each sub-action are counted.

5. The tunneling management method for exploration drilling operations according to claim 1, characterized in that, In step S160, the energy efficiency index for the comprehensive evaluation of drilling efficiency is defined as the hydraulic energy consumed per unit of drilling footage, which is obtained by integrating the product of oil pressure and flow rate over time.

6. A tunneling management system for exploration drilling operations, characterized in that, The system includes the following components: The multi-source sensor data acquisition module is used to collect parameters such as working lifting oil pressure, lowering oil pressure, drill rod speed, drill rod rotation direction, real-time drilling position of drill bit, and circulating water flow rate during drilling rig operation in real time. The data preprocessing and transmission module is connected to the multi-source sensor data acquisition module. It is used to perform preliminary filtering and format standardization on the acquired raw data, and upload the processed data to the cloud database through the communication interface. The drilling action status recognition module is deployed on the server side. It receives data streams from the data preprocessing and transmission module. Based on the preset drilling action status differentiation criteria and algorithm model, it automatically identifies and timestamps the drilling, pressing, lifting, drill rod connection and separation, circulation flushing and equipment idle action status of the drilling rig. The efficiency analysis and visualization module is connected to the drilling action status recognition module. It is used to generate a complete drilling curve, drilling depth verification results, time percentage analysis graphs of each key action, and drilling efficiency summary report based on the recognized action status sequence. The user interaction and parameter configuration module is integrated into the embedded software and runs on the main control MCU. It provides functions such as device calibration, historical data query, and result data package file generation. It also allows users to set specific node parameters through manual intervention mode, providing prior information for the algorithm.

7. The tunneling management system for exploration drilling operations according to claim 6, characterized in that: In the multi-source sensor data acquisition module, the pressure sensor is installed in the working lifting oil circuit and the lowering oil circuit of the drilling rig hydraulic system, and adopts the piezoresistive sensing principle with a built-in temperature compensation circuit; the speed encoder adopts a photoelectric incremental encoder, coupled to the drill rod drive shaft, to monitor the drill rod speed and steering signal in real time; the displacement sensor is installed in the drill bit propulsion mechanism, and adopts a magnetostrictive linear displacement sensor to provide real-time feedback on the drill bit drilling position; the flow meter is integrated into the circulating water system to monitor the circulating water flow rate.

8. The tunneling management system for exploration drilling operations according to claim 6, characterized in that: In the data preprocessing and transmission module, the data import stage checks data integrity, verifies the start bit, checksum, and end bit of the data frame, discards data frames that fail to pass verification, and uses data compression and encryption transmission technology for data transmission.

9. The tunneling management system for exploration drilling operations according to claim 6, characterized in that: In the drilling action state recognition module, the feature extraction stage involves extracting time-domain and frequency-domain features from the preprocessed data stream. The time-domain features include mean, variance, and peak value; the frequency-domain features include dominant frequency and energy spectral entropy. In the feature fusion stage, the extracted time-domain features and frequency-domain features are combined into a high-dimensional feature vector, and then dimensionality reduction is performed through principal component analysis to retain principal components with a cumulative variance contribution rate of over 95%.

10. The tunneling management system for exploration drilling operations according to claim 6, characterized in that: In the efficiency analysis and visualization module, the comprehensive evaluation of drilling efficiency adopts a multi-level index system, including time utilization rate index, footage efficiency index and energy consumption efficiency index; the time utilization rate index is defined as the ratio of pure drilling time to total operating time; the footage efficiency index is defined as the average drilling footage per unit time.

Images