Distribution network geological exploration method and system

Through the distribution network survey geological model and geological feedback regulation system, combined with abnormal detection and genetic algorithm, the stratigraphic inhomogeneity and communication challenges in tunnel excavation are solved, and efficient, intelligent and safe operations of drilling equipment are achieved.

CN119066873AActive Publication Date: 2024-12-03HAINAN POWER GRID DESIGN CO LTD

Patent Information

Application Number
CN202411202048.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-29
Publication Date
2024-12-03
Estimated Expiration
2044-08-29

AI Technical Summary

Technical Problem

In tunnel boring, uneven hardness and toughness of the formation make it difficult for drilling equipment to achieve immediate responsive adjustment, efficiency and energy consumption optimization, and at the same time, the communication challenges in equipment attitude control and multi-equipment collaborative operation are difficult to solve, resulting in operation efficiency and safety insufficient.

Method used

By establishing a distribution network survey geological model, monitoring drilling parameters and attitude data in real time, using a geological feedback regulation system and a multi-objective optimization model, optimizing drilling parameters and attitudes, combining isolated forest anomaly detection and adaptive genetic algorithms, we can achieve no-no between equipment between equipment collaboration and data transmission.

Benefits of technology

Accurate prediction and real-time adjustment of formation characteristic parameters is achieved, energy consumption and wear are reduced, attitude control stability and data transmission continuity, and drilling operations are improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119066873A_ABST
    Figure CN119066873A_ABST
Patent Text Reader

Abstract

The invention relates to a distribution network prospecting geology method, which comprises the following steps: collecting original data of a distribution network prospecting area, and establishing a distribution network prospecting geology model; drilling parameters are monitored in real time, a geological feedback regulation and control system is established, and the drilling parameters are adjusted through the geological feedback regulation and control system; collecting real-time attitude data of the drill rod and real-time attitude data of the drill bit, comparing the real-time attitude data of the drill rod and the real-time attitude data of the drill bit with a preset drilling track, and correcting the position and direction of the drill rod and the position and direction of the drill bit according to a comparison result; performing instruction distribution on the equipment through the power distribution network central data platform; an adjustment scheme is obtained through an isolated forest anomaly detection algorithm and a self-adaptive genetic algorithm, and the operation state of each operation device is optimized; predicting stratum hardness change, judging whether the working condition is abnormal or not, and adjusting the drilling process through a model prediction control algorithm; performing protection through a rule engine; performing fault prediction and diagnosis; the system comprises the method.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network survey, and in particular to a distribution network survey geological method and system. Background Art

[0002] In the context of tunnel excavation through complex geological structures, geological survey drilling operations are facing unprecedented technical challenges.

[0003] The uneven hardness and toughness of the formation not only require the drilling equipment to be highly intelligent and adaptable, but also to maintain the dual optimization of operating efficiency and quality in dynamic changes. Real-time adjustment of drill bit parameters to adapt to diverse formation conditions is a comprehensive challenge that integrates advanced sensing technology, data analysis and intelligent decision-making. Although the modern distribution network infrastructure provides the necessary energy and data flow guarantee for this process, making dynamic parameter adjustment, high-precision attitude control and multi-device collaboration possible, there are still many obstacles that need to be overcome at the practical operation level.

[0004] First, the variability of the formation requires the drilling equipment to have the ability to respond intelligently and instantly. Especially when the drill bit encounters a sudden change in hardness, how to accurately adjust parameters such as rotation speed and thrust to reduce energy consumption and wear while maximizing drilling efficiency is a complex technical problem. Even with the stable data transmission support of the distribution network, how to achieve rapid iteration and optimization of the algorithm model to accurately adapt to formation changes is still the focus of current technical exploration.

[0005] Secondly, the control of the equipment's posture is particularly critical in complex tunnel environments, which not only involves multi-dimensional precise positioning, but also requires ensuring stability and accuracy in dynamic operations. Although high-precision sensors and intelligent algorithms can provide real-time feedback, how to effectively reduce external interference in a small space and maintain high reliability of posture control is still a difficult point in technical implementation.

[0006] Furthermore, the efficient and safe execution of multi-device collaborative operations depends on a low-latency, high-stability communication network. In special operating environments such as tunnels, how to overcome unfavorable factors such as spatial layout limitations and electromagnetic interference to ensure the continuity and accuracy of data transmission is an important prerequisite for achieving seamless collaboration between devices. Although the distribution network provides infrastructure support for communications, the optimization of communication networks and the design of collaborative control strategies under complex terrain conditions are still the key bottlenecks for improving overall operational efficiency.

[0007] In summary, although the existing technology has alleviated some problems in tunnel geological survey and drilling operations to a certain extent, facing the uncertainty of geological conditions, the impact of environmental interference on posture control, and the communication challenges in collaborative operations, how to break through these technical barriers and achieve a more efficient, smarter and safer operation mode is still a comprehensive technical problem that urgently needs to be solved in this field. Summary of the invention

[0008] In order to solve the problems existing in the above-mentioned prior art, the purpose of the present disclosure is to provide a distribution network geological survey method and system, so as to break through technical barriers such as the uncertainty of formation conditions, the impact of environmental interference on posture control, and the communication challenges in collaborative operations, and achieve a more efficient, smarter and safer operation mode.

[0009] A distribution network geological survey method disclosed in the present disclosure includes the following steps: S101, collecting original data of the distribution network survey area, establishing a distribution network survey geological model, obtaining predicted stratum characteristic parameters through the distribution network survey geological model, feeding back the predicted stratum characteristic parameters to the distribution network geographic information system, and updating the geological background information of the distribution network geographic information system; S102, real-time monitoring of drilling parameters, establishment of a geological feedback control system, adjustment of drilling parameters through the geological feedback control system, and feedback of the adjusted drilling parameters to the distribution network survey geological model to correct formation characteristic parameters; establishment of a multi-objective optimization model to obtain optimal drilling parameters, establishment of a multi-criteria decision-making model to obtain the optimal construction plan; S103, collecting real-time posture data of the drill rod and the real-time posture data of the drill bit, comparing the real-time posture data of the drill rod and the real-time posture data of the drill bit with a preset drilling trajectory, and correcting the position and direction of the drill rod and the position and direction of the drill bit according to the comparison result; and establishing a drilling posture optimization knowledge base for guiding subsequent drilling operations; S104, performing real-time data exchange between the drilling equipment and the bottom scanner through the communication network, and simultaneously sharing the status between the drilling equipment and the vibration monitor, and distributing instructions to the equipment through the central data platform of the distribution network; S105, collecting the working status information of each operating device, analyzing the mutual influence between the working status information of each operating device, obtaining an adjustment plan through an isolation forest anomaly detection algorithm and an adaptive genetic algorithm, and adjusting the working parameters of each operating device through the adjustment plan, wherein the working parameters of each operating device are used to optimize the operating status of each operating device; S106, monitoring the drill bit speed, thrust, drilling depth, and borehole wall temperature of the drilling equipment; predicting changes in formation hardness, and determining whether the working condition is abnormal. When the predicted formation hardness exceeds a preset threshold, the drilling process is adjusted through a real-time optimized model prediction control algorithm; when the working condition is determined to be abnormal, protection is provided through a rule engine; S107, perform fault prediction and diagnosis, dynamically adjust the operating parameters of the drilling equipment, and optimize the distribution network survey geological model.

[0010] Preferably, the step S101 includes: Obtain the physical and chemical property parameters of rocks to form raw data; obtain the location information of distribution network equipment to perform coordinate system conversion and data format unification; pre-process the raw data to obtain standardized geological survey data; based on standardized geological survey data, establish a distribution network survey geological model for predicting stratum characteristic parameters, feed back the prediction results to the distribution network geographic information system, and update the geological background information of the distribution network geographic information system.

[0011] Preferably, the step S102 includes: The drilling parameters of the drill bit are monitored in real time by sensors, and a geological feedback control system is established. The drilling parameters include drill bit vibration frequency, drill pipe torque, drill bit pressure, drill bit temperature and drilling rig displacement. The control threshold and the corresponding drilling parameter control strategy are preset. The drilling parameters are adjusted according to the preset threshold and drilling parameter control strategy. According to the adjusted drilling parameters, the formation characteristic parameters of the current drilling formation are identified, and the formation characteristic parameters of the current drilling formation are fed back to the distribution network exploration geological model, and the formation characteristic parameters in the distribution network exploration geological model are corrected; A multi-objective optimization model for selecting the optimal drilling parameters is established based on drilling efficiency, drill bit life and construction safety; According to the formation characteristic parameters, load requirements of the distribution network and foundation type, a multi-criteria decision-making model for selecting the optimal construction scheme is established.

[0012] Preferably, the step S103 includes: Preset the threshold range of the drill rod posture data and the threshold range of the drill bit posture data; obtain the real-time posture data of the drill rod and the real-time posture data of the drill bit through the sensor. Calculate the deviation between the real-time posture data of the drill rod and the threshold of the preset drill rod posture data, and calculate the deviation between the real-time posture data of the drill bit and the threshold of the preset drill bit posture data; establish a fuzzy rule base between the deviation value and the control amount, and obtain the control amount according to the size and direction of the deviation value, the control amount includes the drilling rig thrust and rotation speed; and according to the size and direction of the deviation value, correct the position and direction of the drill rod, as well as the position and direction of the drill bit; establish a drilling posture optimization knowledge base for guiding subsequent drilling construction operations according to the real-time posture data, deviation value and adjustment instructions generated during the drilling process.

[0013] Preferably, the step S104 includes: The MQTT protocol is used to enable real-time data interaction between the drilling equipment and the formation scanner, and the MQTT protocol is used to enable status sharing between the drilling equipment and the vibration monitor; the real-time status data of the drilling equipment is uploaded to the central data platform of the distribution network through the OPCUA protocol. The real-time status data of the drilling equipment includes: drilling position, drilling speed, thrust, rotation speed and vibration level.

[0014] Preferably, the step S105 includes: collecting the working status information of each device through the central data platform of the distribution network, the working status information of each device including the real-time data of the drilling speed, drill bit speed, propulsion force, geological radar and soil analyzer of the drilling equipment; using algorithms such as the Pearson correlation coefficient to calculate the correlation matrix between different equipment parameters, and quantitatively analyze the mutual influence of each parameter; using the isolated forest anomaly detection algorithm to analyze the abnormal behavior of the equipment real-time data, and selecting the anomaly threshold by statistical methods according to the distribution characteristics of historical data; designing a fitness function according to the historical data of equipment parameters and the anomaly detection results, and obtaining the optimal equipment working parameters according to the fitness function; and distributing the optimal equipment working parameters to each device through the intelligent operation and maintenance system of the distribution network.

[0015] Preferably, the step S106 includes: collecting parameter data such as drill bit speed, thrust, drilling depth, and borehole wall temperature, and performing streaming processing and real-time analysis; predicting and analyzing the change trend of formation hardness through the distribution network survey geological model, and performing autocorrelation analysis and stationarity test on the formation hardness data. When the predicted formation hardness meets the preset requirements, a geological condition mutation warning is triggered, and a drilling system dynamic model is established online to predict the system output in the future and optimize the control amount; collecting drill pipe vibration and drill bit temperature data, performing unsupervised learning through the isolation forest anomaly detection algorithm, selecting an anomaly threshold through cross-validation, and judging that an abnormal working condition such as drill bit overheating or rock crushing occurs when the anomaly score exceeds the threshold; processing the abnormal working condition through the rule engine, and when the rule triggers the distribution network protection action, subscribing to relevant messages, parsing the abnormal state of the drilling equipment, and automatically deciding whether to implement protection actions on the affected lines according to the distribution network protection control strategy; pre-processing the real-time working condition parameters, anomaly detection results, and protection measure logs of the drilling equipment to establish a multi-dimensional and multi-scenario distribution network risk assessment model.

[0016] Preferably, the step S107 includes: collecting operating parameter data of each drilling equipment and performing preprocessing; obtaining geological information such as the stratigraphic structure and rock and soil properties of the drilling location, and performing correlation analysis with the operating parameter data of the drilling equipment, establishing a multi-dimensional knowledge graph of equipment-geology-process and establishing an equipment failure prediction model, and when the failure probability exceeds a preset threshold, triggering a fault warning, generating a diagnosis report and pushing it to relevant personnel; obtaining fuzzy control rules between fault types and control operations based on historical data, and calculating control quantities through fuzzy reasoning and defuzzification processes; collecting distribution network operation data, the distribution network operation data includes parameters such as voltage, current, and power, and performing correlation analysis on the fault diagnosis results and the distribution network operation data to evaluate the impact of the drilling equipment failure on the safe operation of the distribution network; if the failure may cause a power outage in the distribution network or equipment damage, generating a distribution network dispatching plan, and reducing the scope of the fault impact through load dispatching, line switching and other measures to ensure the power supply reliability of the distribution network; correcting the model state by introducing observation data; and optimizing the distribution network survey geological model.

[0017] The present disclosure also provides a distribution network geological survey system, including the distribution network geological survey method as described above.

[0018] The advantages of the distribution network geological survey method and system disclosed in the present invention are: First, the present invention establishes a distribution network survey geological model, and based on the formation characteristic parameters predicted by the distribution network survey geological model, uses a geological feedback control system to adjust the drilling parameters, accurately adjusts parameters such as rotation speed and propulsion force to reduce energy consumption and wear, and maximize drilling efficiency.

[0019] Secondly, by collecting the real-time posture data of the drill rod and the real-time posture data of the drill bit, the real-time posture data of the drill rod and the real-time posture data of the drill bit are compared with the preset drilling trajectory. According to the comparison results, the position and direction of the drill rod, as well as the position and direction of the drill bit are corrected; and a drilling posture optimization knowledge base is established to guide subsequent drilling construction operations to ensure stability and precision in dynamic operations.

[0020] In addition, real-time data exchange between the drilling equipment and the underlying scanner is carried out through the communication network, and the status of the drilling equipment and the vibration monitor is shared at the same time. In addition, instructions are distributed to the equipment through the central data platform of the distribution network, realizing seamless collaboration between devices and ensuring the continuity and accuracy of data transmission. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a structural schematic diagram of a distribution network geological survey method and system described in the present invention. DETAILED DESCRIPTION

[0022] like Figure 1 As shown, a distribution network geological survey method and system described in the present disclosure may specifically include: Step S101, collect original data of the distribution network survey area, establish a distribution network survey geological model, obtain predicted stratum characteristic parameters through the distribution network survey geological model, feed back the predicted stratum characteristic parameters to the distribution network geographic information system, and update the geological background information of the distribution network geographic information system.

[0023] Specifically, the physical and chemical property parameters of rocks are obtained to form raw data; the location information of distribution network equipment is obtained for coordinate system conversion and data format unification; the raw data is preprocessed to obtain standardized geological survey data; based on the standardized geological survey data, a distribution network survey geological model for predicting stratum characteristic parameters is established, the prediction results are fed back to the distribution network geographic information system, and the geological background information of the distribution network geographic information system is updated.

[0024] More specifically, geological survey instruments are used to collect data on the physical and chemical properties of rocks in the distribution network survey area, including density, porosity, mineral composition and chemical composition, remove noise and outliers in the data, unify the dimensions of different data sources, scale and normalize the data, and input the processed data into the distribution network geographic information system to provide geological background information for the layout and installation of distribution network equipment, calculate the mean, variance, kurtosis and skewness, extract spectral features, and calculate and predict formation characteristic parameters, including hardness, toughness and formation thickness.

[0025] Use geological survey instruments to collect rock samples in the distribution network survey area, obtain the physical property parameters of the rock, including density, porosity, etc., and chemical property parameters, such as mineral composition, chemical composition, etc., to form the original geological survey data set. At the same time, the location information of the distribution network equipment is obtained through on-site mapping, GPS positioning, etc., and the coordinate system conversion and data format unification are performed. The collected original geological survey data is preprocessed, and data cleaning algorithms such as filtering and smoothing are used to remove noise and outliers. The data from different sources and different dimensions are unified in dimension, data scaling and normalization are performed to make the data set more standardized and normalized. Based on the standardized geological survey data, the statistical characteristics of the physical and chemical property parameters of each sampling point, such as mean, variance, kurtosis and skewness, are calculated. At the same time, Fourier transform (such as fast Fourier transform FFT) is used to convert the time domain signal into a frequency domain signal, and frequency domain feature parameters such as amplitude spectrum and phase spectrum are extracted to comprehensively characterize the characteristics of the geological sample. The statistical characteristics and frequency domain characteristics of the sampling points are used to construct feature vectors as the input of the machine learning model. The predicted formation characteristic parameters are compared and verified with the measured values ​​of known sampling points. The prediction performance of the model is evaluated by cross-validation and other methods. The model parameters are optimized through grid search and other technologies. The prediction model is continuously iterated and optimized to improve the prediction accuracy. At the same time, the prediction results are fed back to the distribution network geographic information system to dynamically update the geological background information.

[0026] For example, in the distribution network survey area, geological survey instruments such as geological radar and resistivity imager are used to set sampling points along the survey line, and sampling is carried out every 50 meters to obtain physical parameters such as rock density and porosity, as well as chemical parameters such as mineral composition and chemical composition. At the same time, surveying and mapping equipment such as total station and differential GPS are used to accurately locate the position of each sampling point, with a plane coordinate accuracy of better than 5 cm and an elevation accuracy of better than 10 cm. The acquired raw data is preprocessed, and the median filter algorithm is used to remove random noise, and the 3σ criterion is used to identify and eliminate outliers. Then, the data is normalized by the minimum-maximum normalization method, so that the value range of each attribute parameter is unified to [0,1]. The sampling point data is expanded to the entire survey area using a spatial interpolation algorithm (such as Kriging interpolation) to generate geological attribute raster data covering the entire area, with a grid resolution of 10 meters × 10 meters. The statistical characteristics of the physical and chemical property parameters such as mean, variance, kurtosis and skewness are calculated for each grid, and the frequency domain characteristic parameters such as amplitude spectrum and phase spectrum are extracted by fast Fourier transform (FFT). The statistical characteristics and frequency domain characteristics of the grid are constructed as feature vectors as the input of the support vector machine (SVM) model. The feature vectors are mapped to high-dimensional space through radial basis kernel function, and the optimal classification hyperplane is found in the high-dimensional space. The hyperparameters of the SVM model (such as penalty coefficient C, kernel function parameter γ, etc.) are optimized using grid search and 5-fold cross-validation methods, and the optimal parameter combination is C=10, γ=0.01. The optimized SVM model is used to predict the formation characteristic parameters of the unknown area, and the prediction results are compared with the drilling verification data. The average relative error is less than 15%. Taking into account the types of distribution network equipment, load demand, geological conditions and other multi-source heterogeneous data, the analytic hierarchy process (AHP) is used to construct an evaluation index system for distribution network planning and design, and the weight of each index is determined by pairwise comparison; the TOPSIS algorithm is used to comprehensively evaluate and rank the candidate schemes, and the scheme with the highest comprehensive score is selected as the optimal distribution network planning and design scheme under the premise of meeting the constraints of safe and stable operation of the equipment, taking into account both the project cost and construction efficiency. The predicted formation characteristic parameters are comprehensively analyzed with the data such as the location of distribution network equipment and load demand, and the analytic hierarchy process (AHP) evaluation index system is constructed. The weight of each index is determined by expert scoring, and the TOPSIS algorithm is used to comprehensively evaluate the candidate schemes to obtain the optimal distribution network planning and design scheme. This scheme reduces the project cost by 8% and improves the construction efficiency under the premise of ensuring that the equipment safety margin meets 1.25 times.

[0027] Step S102: monitor the drilling parameters in real time, establish a geological feedback control system, adjust the drilling parameters through the geological feedback control system, and feed back the adjusted drilling parameters to the distribution network survey geological model to correct the formation characteristic parameters; establish a multi-objective optimization model to obtain the optimal drilling parameters, establish a multi-criteria decision-making model to obtain the optimal construction plan.

[0028] Specifically, the drilling parameters of the drill bit are monitored in real time through sensors, and a geological feedback control system is established. The drilling parameters include drill bit vibration frequency, drill pipe torque, drill bit pressure, drill bit temperature and drilling rig displacement. The control threshold and the corresponding drilling parameter control strategy are preset. The drilling parameters are adjusted according to the preset threshold and drilling parameter control strategy. According to the adjusted drilling parameters, the formation characteristic parameters of the current drilling formation are identified, and the formation characteristic parameters of the current drilling formation are fed back to the distribution network exploration geological model, and the formation characteristic parameters in the distribution network exploration geological model are corrected; A multi-objective optimization model for selecting the optimal drilling parameters is established based on drilling efficiency, drill bit life and construction safety; According to the formation characteristic parameters, load requirements of the distribution network and foundation type, a multi-criteria decision-making model for selecting the optimal construction scheme is established.

[0029] More specifically, based on the real-time monitoring data of vibration sensors, torque sensors, pressure sensors, temperature sensors and displacement sensors, a geological feedback control system is established to adjust the dynamic parameters of the drill bit. If hard formations are detected, the geological feedback control system increases the propulsion force; if tough formations are detected, the rotation speed is reduced. The drilling parameters are also fed back to the distribution network survey geological model, the formation characteristic parameters are dynamically corrected, and the distribution network foundation construction plan is adjusted to cope with different geological conditions.

[0030] Before drilling, an initial distribution network exploration geological model is established based on existing geological exploration data and empirical knowledge. By statistically analyzing historical drilling data and geological sample data, the distribution law of different formation characteristic parameters, such as hardness and toughness, is obtained, and a reasonable threshold range is preset as the judgment basis of the geological feedback control system. Various types of sensors such as vibration sensors, torque sensors, pressure sensors, temperature sensors and displacement sensors are installed on the drilling rig to collect various parameter data during the drilling process in real time, including drill bit vibration frequency, drill pipe torque, drill bit pressure, drill bit temperature and rig displacement, etc. The collected real-time monitoring data is transmitted to the geological feedback control system. The geological feedback control system analyzes and judges the monitoring data in real time based on the preset threshold and drilling parameter control strategy, identifies the hardness and toughness characteristics of the current drilling formation, and dynamically adjusts the parameters such as the propulsion force and speed of the drill bit according to the analysis results. In the actual drilling process, the threshold setting can also be dynamically adjusted according to the real-time monitoring data to adapt to changes in geological conditions. When the monitoring data shows that the drill bit encounters a hard formation, the geological feedback control system automatically increases the thrust to improve the drilling depth and rock breaking efficiency of the drill bit; when the monitoring data shows that the drill bit enters a tough formation, the system automatically reduces the speed to reduce the torque and wear of the drill bit and extend the service life of the drill bit. The drilling parameter adjustment command output by the geological feedback control system is transmitted to the drilling rig control unit. The control unit accurately controls the thrust and speed of the drilling rig through hydraulic and electronic control methods according to the command signal, so that the drilling process is always kept within the optimal operating range. The optimal operating range refers to the parameter setting range that comprehensively considers drilling efficiency, drill bit life, construction safety and other factors to achieve the best balance state for the entire drilling process. By establishing a multi-objective optimization model, balancing multiple indicators such as drilling speed, drill bit wear, and vibration level, the optimal combination of drill bit thrust and speed parameters can be solved. According to the dynamically updated distribution network survey geological model, comprehensive consideration is given to parameters such as stratum hardness, toughness, burial depth, as well as factors such as load requirements and foundation type of distribution network equipment, to establish a multi-criteria decision-making model for the distribution network foundation construction plan. The weight of each evaluation index is determined through weighted scoring, hierarchical analysis and other methods, and the optimization algorithm is used to search for the optimal combination of construction plan parameters to achieve dynamic optimization and adjustment of the construction plan, and to reasonably select the tower foundation burial depth, cable trench excavation method and protective measures.

[0031] For example, when establishing the initial distribution network exploration geological model, by statistically analyzing the existing 500 sets of drilling data and 200 geological sample data, the mean of the formation hardness is 120MPa and the standard deviation is 30MPa; the mean of the formation toughness is 8MPa·m1 / 2 and the standard deviation is 2MPa·m1 / 2. According to the 3σ principle of normal distribution, the preset hardness threshold range is 30210MPa and the toughness threshold range is 214MPa·m1 / 2. In the real-time monitoring data transmission process, the local weighted scatter point smoothing (LOWESS) algorithm is used to smooth the data, and the smoothing span is set to 0.2 to reduce noise interference. The geological feedback control system judges the monitoring data every 0.1s. When the drill bit thrust force exceeds 150kN for three consecutive times, it is judged to encounter a hard formation, and the system will increase the thrust by 20%; when the drill rod torque is monitored to be less than 500N·m for three consecutive times, it is judged to enter a tough formation, and the system will reduce the speed by 15%. The drilling parameter optimization model uses the constraints of 0.2m / min drilling speed, 10h drill bit service life and 0.8 comprehensive index weight. The genetic algorithm is used to search for the optimal parameter combination. The initial population size is 50, the crossover probability is 0.6, and the mutation probability is 0.1. After 50 iterations, the optimal drilling parameters are: thrust 120kN, speed 60r / min. In the dynamic update process of the distribution network survey geological model, the number of ensemble samples of the EnKF algorithm is set to 100, the observation error covariance matrix is ​​determined by statistical analysis of 10 groups of drilling monitoring data, and the state variable update cycle is 10min. The dynamic optimization model of the construction plan determines the weight of each evaluation index through the hierarchical analysis method, and optimizes the search through the particle swarm algorithm. The number of particles is 30, the acceleration constant c1=c2=1.5, and the maximum number of iterations is 200. In the data mining process, the Apriori association rule algorithm was used with a minimum support threshold of 0.05 and a minimum confidence threshold of 0.5. A strong association rule was obtained that the greater the formation hardness, the slower the drilling speed, and a medium-strength association rule that the higher the formation toughness, the faster the drill bit wear, providing a basis for drilling parameter optimization.

[0032] Step S103, collect the real-time posture data of the drill rod and the real-time posture data of the drill bit, compare the real-time posture data of the drill rod and the real-time posture data of the drill bit with the preset drilling trajectory, and correct the position and direction of the drill rod and the position and direction of the drill bit according to the comparison results; and establish a drilling posture optimization knowledge base for guiding subsequent drilling operations.

[0033] Specifically, the threshold range of the drill rod posture data and the threshold range of the drill bit posture data are preset; the real-time posture data of the drill rod and the real-time posture data of the drill bit are obtained through sensors. The deviation between the real-time posture data of the drill rod and the threshold of the preset drill rod posture data is calculated, and the deviation between the real-time posture data of the drill bit and the threshold of the preset drill bit posture data is calculated; a fuzzy rule base between the deviation value and the control amount is established, and the control amount is obtained according to the size and direction of the deviation value, and the control amount includes the drilling rig thrust and rotation speed; and according to the size and direction of the deviation value, the position and direction of the drill rod, as well as the position and direction of the drill bit are corrected; according to the real-time posture data, deviation value and adjustment instructions generated during the drilling process, a drilling posture optimization knowledge base for guiding subsequent drilling construction operations is established.

[0034] More specifically, sensors are used to monitor the inclination angle and direction of the drill rod and drill bit in real time, and these data are transmitted to the data processing unit for real-time analysis. The real-time attitude data is compared with the preset drilling trajectory, and the suspension height and related control parameters are adjusted according to the comparison results. The position and direction of the drill rod and drill bit are corrected, and the drilling trajectory is corrected according to the construction requirements of the distribution network tower foundation. The attitude is continuously monitored and adjusted to ensure that the construction meets the standards.

[0035] According to the standard requirements for the construction of distribution network tower foundations, the threshold range of the drill rod and drill bit posture is preset; high-precision inclination sensors and azimuth sensors are installed at key positions of the drill rod and drill bit, and the data is sent to the data processing unit on the ground. After the data processing unit receives the real-time posture data of the drill rod and drill bit, a complementary filtering algorithm is used to filter the data, and the filtering frequency and time constant are dynamically adjusted by setting the weight coefficients of the accelerometer and gyroscope; the filtered posture data is compared and analyzed with the preset drilling trajectory, and the deviation value between the actual position of the drill rod and drill bit and the preset trajectory is calculated. According to the size and direction of the deviation value, the data processing unit uses the fuzzy PID control algorithm to automatically generate the attitude adjustment instructions of the drill rod and the drill bit. By establishing a fuzzy rule base between the deviation value and the control quantity, the fuzzy reasoning and defuzzification process are used to obtain the adjustment amount of the drilling rig propulsion force, rotation speed, and lifting force parameters, and the drilling rig hydraulic system and the electronic control system are controlled to adjust the suspension height, forward angle, and azimuth of the drill rod and the drill bit accordingly to keep them consistent with the preset drilling trajectory. During the drilling process, the data processing unit continuously monitors the real-time attitude data of the drill rod and the drill bit, and according to the requirements of the distribution network tower foundation construction, the iterative optimization algorithm is used to correct the preset drilling trajectory in real time, generate dynamically updated drill bit attitude control instructions, and guide the drilling rig to dynamically adjust the drilling trajectory to ensure that the drilling position, depth, and angle parameters meet the construction standards. During the attitude adjustment process, if the deviation value of the drill rod and the drill bit is continuously monitored to exceed the preset threshold range for multiple times, the data processing unit triggers an alarm signal and automatically controls the drilling rig to stop drilling. At the same time, a warning message is sent to the construction personnel, indicating that there may be abnormal conditions such as drilling deviation or drilling tool failure, which require manual inspection and processing. The real-time posture data, deviation values, adjustment instructions and other information generated during the drilling process are recorded by the data processing unit in time series to form a complete drilling trajectory database. In combination with data such as drilling angle, drill bit depth, and mechanical parameters of drilling tools, three-dimensional intuitive drilling trajectory diagrams, deviation distribution diagrams and stress cloud diagrams are drawn to intuitively evaluate the drilling construction quality. At the same time, the drilling trajectory characteristics of different batches and different regions are compared and analyzed to optimize the drilling parameter design and construction organization plan. The analysis results are transmitted to the distribution network engineering management system in real time and correlated with other construction parameter data. The drilling trajectory database is analyzed through association rule mining algorithms such as Apriori or FP-growth to mine the intrinsic connection between the changes in the posture of the drill rod and drill bit and the formation characteristics and drilling parameters. Clustering algorithms such as K-means are used to classify the optimized drilling parameters under different geological conditions to form a drilling posture optimization knowledge base to guide subsequent drilling construction operations.

[0036] For example, according to the distribution network tower foundation construction specifications, the preset drilling verticality deviation does not exceed 1°, and the drilling position deviation does not exceed the threshold range of 50mm. Install a MEMS tilt sensor with a range of ±30°, an accuracy of 0.1°, a resolution of 0.01°, and a sampling frequency of 10Hz and an electronic compass with a range of 0~360°, an accuracy of 1°, a resolution of 0.1°, and a sampling frequency of 5Hz, and transmit data through a ZigBee module at a frequency of 2.4GHz and a rate of 250kbps. Use a complementary filtering algorithm, set the accelerometer weight to 0.6, the gyroscope weight to 0.4, the filter frequency to 20Hz, and the time constant to 0.5s. Compare the filtered data with the preset trajectory, calculate the deviation value, map the deviation value to the fuzzy rule base, input the domain [-5°, 5°] and [-100mm, 100mm], output the domain [-50%, 50%], use the triangle membership function, Mamdani reasoning method, and center of gravity defuzzification to obtain the adjustment of the drilling rig's thrust, speed, and lifting force. Control the drilling rig hydraulic system pressure ±2MPa, motor speed ±10r / min, and lifting height ±0.5m to achieve a deviation of the drill rod and drill bit posture from the preset trajectory of <0.5° and 20mm. Iterate and optimize the preset trajectory. If the deviation exceeds 2° or 100mm for 5 consecutive times, it will automatically stop and alarm. Record the drilling parameter time series data, draw a three-dimensional trajectory diagram, deviation distribution diagram, and stress cloud diagram. Use the Apriori algorithm to mine association rules with a support degree of 0.2, a confidence degree of 0.8, and a lifting degree of 1.5. It is found that the drilling speed decreases by 5% for every 10MPa increase in formation hardness. K-means algorithm clustering analysis optimizes drilling parameters and forms a posture optimization knowledge base.

[0037] Step S104: Real-time data exchange is performed between the drilling equipment and the bottom scanner through the communication network, and status sharing is performed between the drilling equipment and the vibration monitor, and instructions are distributed to the equipment through the central data platform of the distribution network.

[0038] Specifically, a drilling equipment communication network will be established, including Wi-Fi, wireless Bluetooth, Ethernet and wired optical fiber. Through standardized communication protocols, including MQTT, Modbus, and OPCUA, real-time data interaction and status sharing between drilling equipment and formation scanners and vibration monitors will be achieved. The data will be integrated into the central data platform of the distribution network to share equipment status and distribute instructions, and to share drilling locations, formation characteristic data and drilling parameter information in real time.

[0039] More specifically, the real-time data between the drilling equipment and the formation scanner are interacted through the MQTT protocol, and the status between the drilling equipment and the vibration monitor is shared through the MQTT protocol; the real-time status data of the drilling equipment is uploaded to the central data platform of the distribution network through the OPCUA protocol. The real-time status data of the drilling equipment includes: drilling position, drilling speed, thrust, rotation speed and vibration level.

[0040] Install a variety of communication modules on the drilling equipment to achieve diversified communication connections between the drilling equipment and peripheral equipment, meet the data transmission requirements of different distances and different bandwidths, and ensure the diversity and reliability of the communication link. The MQTT protocol is used to realize real-time data interaction between the drilling equipment and the formation scanner, and to share the status between the drilling equipment and the vibration monitor; the OPCUA protocol is used to upload the real-time status data of the drilling equipment to the central data platform of the distribution network, including parameters such as drilling position, drilling speed, thrust, rotation speed, vibration level, etc. The central data platform of the distribution network deploys the OPCUA server, defines the information model of the drilling equipment, including nodes such as equipment ID, location coordinates, working status, sensor parameters, and encodes the data in JSON format. The drilling equipment writes real-time data to the corresponding node of the server through the OPCUA client; the OPCUA client of the central data platform reads the server node data and stores the data in the time series database InfluxDB for equipment status monitoring and historical data analysis. The central data platform of the distribution network sends drilling instructions and control parameters to the drilling equipment, and dynamically adjusts the drilling tasks and working parameters of the drilling equipment according to the construction plan and on-site operation requirements of the distribution network project. The OPC UA client of the central data platform writes the instructions and parameter data into the corresponding node of the OPC UA server. The OPC UA client of the drilling equipment reads the server node data, parses the instructions and parameters, and controls the drilling rig to perform the corresponding drilling operations. The symbolic clustering algorithm is used to extract features and mine patterns of time series data to discover frequent patterns and association rules of equipment conditions and drilling parameters. The optimization strategy is encapsulated as an executable rule script, which is sent to the drilling equipment through edge computing technology. The lightweight rule engine is used to infer and execute the optimization strategy in real time locally on the drilling equipment to achieve adaptive optimization control of drilling parameters and improve drilling quality and efficiency.

[0041] For example, the drilling equipment is equipped with a wireless communication module that supports the Wi-Fi 6 protocol and has a bandwidth of 1.2 Gbps, a Bluetooth module that supports the BLE 5.1 ​​protocol and has a transmission distance of 100 m, an Ethernet module that supports the 10GBASE-T standard and has a transmission rate of 10 Gbps, and a fiber optic transceiver that supports the OC-192 standard and has a transmission rate of 9.6 Gbps. The formation scanner publishes formation image data and geological parameter data to the "drill / scanner / data" topic in JSON format through the MQTT protocol every 0.1 s. The message size does not exceed 10 KB, and the QoS level is 2. The drilling equipment subscribes to the topic, receives formation data in real time, performs semantic segmentation of formation images through the convolutional neural network algorithm, identifies formation types, and optimizes drilling parameters in combination with geological parameters. When the formation hardness exceeds 150 MPa, the drilling speed is reduced by 10%. The vibration monitor reports vibration data to the drilling equipment every 50ms through the ModbusRTU protocol. The data length is 4 bytes. After the drilling equipment reads the data, it performs spectrum analysis on the vibration signal through the fast Fourier transform algorithm. When the vibration frequency exceeds 80Hz and the amplitude exceeds 2mm, the drill rod speed is reduced by 20%. The drilling equipment uploads real-time status data to the central data platform through the OPCUA protocol every 1s, including 18 parameters such as drilling position, drilling speed, thrust, speed, vibration level, etc., each parameter is represented by a 64-bit floating point number. The central data platform uses the InfluxDB time series database to store equipment status data. The data retention period is 1 year, and Grafana is used for real-time data visualization and historical data query and analysis. According to the distribution network construction plan and on-site operation requirements, the central data platform sends drilling instructions and control parameters to the drilling equipment through the OPCUA protocol every 10min. The instructions and parameters are encapsulated in JSON format, and the data size does not exceed 1KB. The drilling equipment stores the received instructions and parameters through the SQLite database, and uses the PID control algorithm to realize automatic control and optimization of the drilling process, controlling the drilling trajectory deviation within ±5cm. The drilling equipment publishes the drilling location data to the "drill / location" topic in JSON format through the MQTT protocol every 1s, and the data size does not exceed 100 bytes. The Web-side GIS platform receives the drilling location data in real time through the WebSocket protocol, uses the Baidu Map API to display the location of the drilling equipment on the electronic map of the construction site, and uses the AntVL7 visualization library to draw the drilling trajectory to achieve real-time positioning and trajectory tracking of the drilling equipment. The central data platform uses a symbolic clustering algorithm to mine the equipment status data, converts the time series data into a symbolic sequence, and calculates the frequency and mutual information of the symbolic subsequence to find the frequent patterns and association rules of the equipment working conditions. For example, when the drilling speed is lower than 0.5m / min and the thrust is higher than 10kN, the drill bit wears faster.The FP-Growth algorithm is used to mine the association rules between formation characteristics and drilling efficiency, with a support threshold of 0.2 and a confidence threshold of 0.8. It is found that when the formation lithology is sandstone and fractures are developed, the drilling speed increases by more than 20%. The association rules are converted into Drools rule scripts, which are sent to the drilling equipment through the MQTT protocol. The Drools rule engine is used to match the equipment status data in real time and automatically adjust the drilling parameters.

[0042] Step S105: collect the working status information of each operating equipment, analyze the mutual influence between the working status information of each operating equipment, obtain an adjustment plan through the isolation forest anomaly detection algorithm and the adaptive genetic algorithm, and adjust the working parameters of each operating equipment through the adjustment plan. The working parameters of each operating equipment are used to optimize the operating status of each operating equipment.

[0043] Specifically, the working status information of each device is collected through the central data platform of the distribution network, and the working status information of each device includes the real-time data of the drilling speed, drill bit speed, propulsion force, geological radar and soil analyzer of the drilling equipment; the correlation matrix between different equipment parameters is calculated using algorithms such as the Pearson correlation coefficient, and the mutual influence of each parameter is quantitatively analyzed; the isolated forest anomaly detection algorithm is used to analyze the abnormal behavior of the equipment real-time data, and the anomaly threshold is selected by statistical methods according to the distribution characteristics of historical data; the fitness function is designed according to the historical data of equipment parameters and the anomaly detection results, and the optimal equipment working parameters are obtained according to the fitness function; the optimal equipment working parameters are distributed to each device through the intelligent operation and maintenance system of the distribution network.

[0044] More specifically, the central data platform of the distribution network collects working status information fed back by collaborative equipment, including the drilling speed, drill bit speed, propulsion force of drilling equipment, and real-time data of geological radar and soil analyzer, and analyzes the mutual influence of real-time data. The adjustment plan is calculated through the isolated forest anomaly detection algorithm and adaptive genetic algorithm, and the working parameters of each operating equipment are dynamically adjusted. The adjusted parameters are applied to the intelligent operation and maintenance system of the distribution network to optimize the operating status of the equipment.

[0045] The data platform caches and distributes real-time data streams through Kafka, and then uses Storm or Spark Streaming for real-time processing, and writes the processed data to the InfluxDB time series database; the Pearson correlation coefficient and other algorithms are used to calculate the correlation matrix between different equipment parameters, quantitatively analyze the mutual influence of each parameter, and discover key influencing factors. The isolated forest anomaly detection algorithm is used to analyze the abnormal behavior of the equipment's real-time data, and multiple isolated trees are constructed by randomly selecting feature subsets. The abnormal threshold is selected through statistical methods based on the distribution characteristics of historical data. If the threshold is exceeded, it is judged as abnormal, and the alarm mechanism is triggered to notify the operation and maintenance personnel to check and repair. The historical data of equipment parameters and anomaly detection results are input into the adaptive genetic algorithm, which uses floating-point encoding. Each gene represents a equipment parameter, and the fitness function is designed in a weighted sum manner. The multiple optimization objectives such as equipment energy consumption, output, and failure rate are comprehensively considered. The weights are adjusted to dynamically adapt to changes in demand. The maximum number of iterations or the fitness threshold is set as the termination condition, and the optimal combination of equipment operating parameters is searched and used as the new equipment operating parameters. The optimized equipment operating parameters are passed to the distribution network intelligent operation and maintenance system, and sent to each operating equipment in real time through the equipment control interface to dynamically adjust its operating status.

[0046] For example, the data platform uses the Kafka cluster to cache and distribute real-time data streams, and can process 100,000 data per second. The Storm cluster consists of 20 nodes. Through parallel computing, the data processing delay is controlled within 100 milliseconds. The processed data is written to the InfluxDB time series database, which stores historical data for the past year, and the data compression rate reaches 70%. By calculating the Pearson correlation coefficient between device parameters, it is found that the correlation coefficient between device energy consumption and ambient temperature is 8, and the correlation coefficient with load is 6. Energy consumption can be effectively reduced by controlling ambient temperature and optimizing load distribution. The isolated forest anomaly detection algorithm randomly selects 100 feature subsets from historical data, and constructs an isolated tree for each subset. Through statistics, it is found that the anomaly threshold is 3 standard deviations, that is, 97% of historical data is distributed within this threshold. Once the real-time data exceeds the threshold, an alarm is automatically triggered, notifying the operation and maintenance personnel to arrive at the site within 30 minutes to troubleshoot the fault. The adaptive genetic algorithm uses 20-bit floating point encoding, each bit represents a device parameter. The fitness function comprehensively considers the three optimization goals of equipment energy consumption, output, and failure rate. The initial weight values ​​are 5, 3, and 2, respectively. The weight is dynamically adjusted every 50 iterations to adapt to changes in demand. The maximum number of iterations of the algorithm is 500 times, and the convergence speed is increased by 30%. The optimal parameter combination searched can reduce equipment energy consumption by 15% and failure rate by 20% while meeting production requirements. The optimized equipment operating parameters are sent to the operating equipment in real time through the distribution network intelligent operation and maintenance system, and the average sending delay is less than 10 seconds. Through data visualization technology, the trend of key indicators such as equipment energy consumption, output, and failure rate over time is displayed in a curve chart, and the distribution of various equipment failures is displayed in a pie chart, providing intuitive data support and scientific decision-making basis for distribution network operation and maintenance management, and improving operation and maintenance efficiency.

[0047] Step S106, monitor the drill bit speed, thrust, drilling depth, and borehole wall temperature of the drilling equipment; predict changes in formation hardness, and determine whether the working condition is abnormal. When the predicted formation hardness exceeds a preset threshold, adjust the drilling process through a real-time optimized model predictive control algorithm; when the working condition is determined to be abnormal, perform protection through a rule engine.

[0048] Specifically, the data of parameters such as drill bit speed, thrust, drilling depth, and borehole wall temperature are collected for streaming processing and real-time analysis. The geological model of the distribution network survey is used to predict and analyze the changing trend of the formation hardness. By performing autocorrelation analysis and stationarity test on the formation hardness data, when the predicted formation hardness meets the preset requirements, a geological condition mutation warning is triggered, and a dynamic model of the drilling system is established online to predict the system output in the future and optimize the control quantity. The drill pipe vibration and drill bit temperature data are collected, and unsupervised learning is performed through the isolated forest anomaly detection algorithm. The anomaly threshold is selected through cross-validation. When the anomaly score exceeds the threshold, it is judged that abnormal conditions such as drill bit overheating or rock crushing occur. The abnormal conditions are processed through the rule engine. When the rule triggers the distribution network protection action, the relevant message is subscribed to parse the abnormal state of the drilling equipment. According to the distribution network protection control strategy, it is automatically decided whether it is necessary to implement protection actions on the affected lines. The real-time operating parameters of the drilling equipment, the anomaly detection results, and the protection measures log are pre-processed to establish a multi-dimensional and multi-scenario distribution network risk assessment model.

[0049] More specifically, the real-time response mechanism of the monitoring system is used to detect sudden changes in geological conditions. When the formation hardness predicted by the distribution network survey geological model increases by a predetermined amount within a predetermined time period, the drill speed and thrust are adjusted within milliseconds. When rock breakage or drill overheating occurs, corresponding protective measures are taken immediately through the isolated forest anomaly detection algorithm and the Drools rule engine algorithm, including suspending drilling, replacing the drill bit, and adjusting the cooling system.

[0050] The monitoring system is used to monitor the drilling parameters and formation status of the drilling equipment in real time. Through the sensor network deployed at the drilling site, the drill bit speed, thrust, drilling depth, borehole wall temperature and other parameter data are collected, and the data is transmitted to the time series database of the monitoring system in real time through industrial Ethernet for streaming processing and real-time analysis. Based on the geological model of the distribution network survey, the monitoring system predicts and analyzes the change trend of the formation hardness. By performing autocorrelation analysis and stationarity test on the formation hardness data, when the predicted formation hardness meets the preset requirements, the geological condition mutation warning is triggered and the warning information is sent to the drilling equipment control system. After receiving the warning, the drilling equipment control system uses the model predictive control algorithm (MPC) based on real-time optimization to establish the dynamic model of the drilling system online, predict the system output in the future, and optimize the control quantity; during the drilling process, the monitoring system collects the drill rod vibration and drill bit temperature data in real time through the deployed vibration sensors and infrared thermal imagers, uses the isolated forest anomaly detection algorithm for unsupervised learning, selects the anomaly threshold through cross-validation, and judges that the drill bit is overheated or the rock is broken when the anomaly score exceeds the threshold. The monitoring system passes the detected abnormal working conditions to the Drools rule engine through the Kafka message queue. The rule engine pre-configures the exception handling rules through the rule language and sets the rule priority through the salience attribute of the rule. At the same time, the association rules between the abnormal state and the distribution network protection action are defined in the rule engine, and the full chain from abnormal detection to protection measures to distribution network intelligent protection is automatically triggered through the rule chain. When the rule triggers the distribution network protection action, it subscribes to the relevant messages, parses the abnormal state of the drilling equipment, and automatically decides whether it is necessary to implement protection actions on the affected lines according to the distribution network protection control strategy, such as tripping to isolate the faulty equipment, switching to the backup power supply, etc., to ensure the safe and stable operation of the distribution network. In addition, the monitoring system can also clean and convert the real-time working parameters, abnormal detection results, protection measures logs and other data of the drilling equipment through ETL tools, and load them into the distribution network geographic information system to explore the association rules and early warning indicators between the equipment status and the distribution network risk, and establish a multi-dimensional and multi-scenario distribution network risk assessment model.

[0051] Exemplarily, the monitoring system collects drilling parameters and formation status data, with a sampling frequency of 100Hz and a data transmission delay of less than 10ms. By performing an ADF test on the formation hardness data, the p value is less than 0.01, proving that the data is stable and suitable for prediction using the ARIMA model. The ratio of the training set to the test set is 8:2, and the model order is determined to be ARIMA (2,1,1) by the AIC criterion. The prediction error MAPE on the test set is 6.3%. When the predicted formation hardness increase exceeds 20%, the MPC controller optimizes the drilling parameters with a control cycle of 128ms. The objective function is to minimize drilling costs and energy consumption. The constraints are that the drilling speed does not exceed 1m / min, the drill bit temperature does not exceed 90°C, and the thrust does not exceed 10kN. The DDPG algorithm adopts the Actor-Critic architecture. The input state includes the current drilling parameters and the formation state. The output action is the adjustment value of the drilling parameters. The reward function is the weighted sum of drilling efficiency and equipment wear. The learning rate is 0.001 and the discount factor is 0.95. After training 1000 episodes, the algorithm converged and the average reward on the test set increased by 23%. The isolation forest algorithm constructs 500 isolated trees with a subsampling size of 256. The anomaly threshold is determined to be 0.6 through cross-validation. The anomaly detection accuracy on the test set is 95.2% and the F1 value is 0.87. 20 exception handling rules are defined in the Drools rule engine. The LHS part of each rule contains the anomaly detection results and equipment status conditions, and the RHS part contains the corresponding protection measures. The rule priority is set by the salience attribute, with a value range of 1~10. The larger the value, the higher the priority. When the rule engine triggers the drilling suspension rule, the system response time is less than 200ms; when the drill bit replacement rule is triggered, the system response time is less than 1s; when the cooling system adjustment rule is triggered, the system response time is less than 500ms. In the rule engine, 10 distribution network protection rules are also defined to map the abnormal state of drilling equipment with the fault type of the distribution network, such as line tripping for drilling suspension, overheating of the drill bit corresponding to overcurrent protection, etc. The rule chain execution delay is less than 100ms. The distribution network risk assessment model adopts a three-layer BP neural network. The input features include 20 dimensions such as equipment status, environmental conditions, and operating parameters. The output feature is the risk level of the distribution network. The number of hidden layer neurons is 50, the activation function is ReLU, and the learning rate is 0.01. After 5000 trainings, the model converges, and the risk discrimination accuracy on the test set is 93.6%. Risk perception data is pushed to the GIS platform in real time through the WebSocket protocol. The data update frequency is 1s. The visualization adopts the ECharts framework and is presented through risk heat maps, equipment status tree maps, etc. The interactive response delay is less than.

[0052] Step S107, perform fault prediction and diagnosis, dynamically adjust the operating parameters of the drilling equipment, and optimize the distribution network survey geological model.

[0053] Specifically, the operating parameter data of each drilling equipment is collected and preprocessed; the geological information such as the stratigraphic structure and rock and soil properties of the drilling location is obtained, and correlated with the operating parameter data of the drilling equipment to establish a multi-dimensional knowledge graph of equipment-geology-process and an equipment failure prediction model. When the failure probability exceeds the preset threshold, the fault warning is triggered, and a diagnosis report is generated and pushed to relevant personnel; based on historical data, the fuzzy control rules between the fault type and the control operation are obtained, and the control quantity is calculated through fuzzy reasoning and defuzzification process; the distribution network operation data is collected, and the distribution network operation data includes parameters such as voltage, current, and power. The fault diagnosis results are correlated with the distribution network operation data to evaluate the impact of the drilling equipment failure on the safe operation of the distribution network; if the fault may cause a power outage in the distribution network or equipment damage, a distribution network dispatching plan is generated, and the fault impact range is reduced through load dispatching, line switching and other measures to ensure the reliability of the distribution network power supply; the model state is corrected by introducing observation data; and the distribution network survey geological model is optimized.

[0054] More specifically, a multi-dimensional control strategy is implemented, including real-time monitoring of drilling equipment, fault prediction and diagnosis, and energy consumption reduction. The comprehensive information processing module is called to integrate the data of the drilling equipment with the data of the geological radar and soil analyzer, and interact with the data of the distribution network. The random forest machine learning model and principal component analysis statistical analysis method are used to identify patterns and anomalies, perform fault prediction and diagnosis, and provide real-time feedback of analysis results. The operating parameters of the drilling equipment are dynamically adjusted to optimize the distribution network survey geological model.

[0055] Through various types of sensors installed on the drill pipe, drill rack, mud pump and other key components of the drilling equipment, such as vibration sensors, pressure sensors, temperature sensors, etc., various parameter data during the operation of the equipment are collected in real time, and the data is transmitted to the comprehensive information processing module through the MQTT Internet of Things communication protocol. The transmission delay is controlled within milliseconds. The comprehensive information processing module performs preprocessing operations such as cleaning, normalization, and feature extraction on the collected data; through industrial communication protocols such as OPCUA, the comprehensive information processing module interacts with exploration equipment such as geological radars and soil analyzers to obtain geological information such as the stratigraphic structure and rock and soil properties of the drilling location, and performs correlation analysis with the operating parameter data of the drilling equipment. The FP-growth algorithm is used to mine frequent association rules to discover the intrinsic connection between equipment operating conditions and geological conditions; the PrefixSpan algorithm is used to mine time series patterns to discover the laws of equipment status changes over time and construct a multi-dimensional knowledge graph of equipment-geology-process; the random forest machine learning model is used to train the historical operating data of the drilling equipment, and the importance of features is evaluated through indicators such as information gain and Gini coefficient. The feature combination with the greatest contribution to fault prediction is selected, and the feature variables of multiple dimensions such as equipment parameters, geological conditions, and process parameters are comprehensively considered to establish an equipment fault prediction model. The model evaluation uses indicators such as precision, recall, and F1 value to measure performance. The current status data of the equipment is compared with the output of the prediction model in real time, the fault probability is calculated and dynamically updated. When the fault probability exceeds the preset threshold, the fault warning is triggered, and a diagnostic report is generated and pushed to relevant personnel. For the equipment failures warned, the comprehensive information processing module automatically calls the fault diagnosis knowledge base to make a preliminary judgment on the fault type based on multiple dimensions such as fault mode, components, phenomena, and causes. Through statistical analysis methods, the covariance matrix of the characteristic parameters is calculated, and the eigenvalue decomposition is performed. The eigenvectors corresponding to the first k largest eigenvalues ​​are selected as the principal components. The k value is determined according to the cumulative variance contribution rate; by analyzing the correlation between the principal components and the original characteristic parameters, the physical meaning represented by each principal component is determined, the key characteristic parameters causing the fault are identified, and the fault diagnosis results and disposal suggestions are generated and fed back to the drilling equipment control system. According to the fault diagnosis results and disposal suggestions, the drilling equipment control system uses fuzzy control and adaptive control methods to dynamically adjust the drilling parameters, including drilling pressure, speed, pump volume, etc., to optimize the drilling process and reduce the impact of the fault on the equipment. Based on historical data, the fuzzy control rules between fault types and control operations are summarized, and the fuzzy set is represented by a triangular or trapezoidal membership function, and the control quantity is calculated through fuzzy reasoning and defuzzification. The self-correcting PID control algorithm is used for adaptive control, and the PID parameters are adjusted in real time according to the equipment status and fault diagnosis results, so that the control system can adapt to changes in equipment status.The distribution network operation data, including voltage, current, power and other parameters, are collected through distribution automation systems, smart meters and other equipment. The data is transmitted to the comprehensive information processing module through the communication protocol. The comprehensive information processing module associates the fault diagnosis results with the distribution network operation data. The causal relationship and time dependence between the fault diagnosis results and the distribution network operation parameters are explored by time series analysis, Granger causality test and other methods, and the impact of drilling equipment failure on the safe operation of the distribution network is evaluated. If the fault may cause power outage or equipment damage in the distribution network, the distribution network dispatch plan is automatically generated. Through load dispatching, line switching and other measures, the scope of fault impact is reduced to ensure the reliability of distribution network power supply. Data assimilation algorithms such as ensemble Kalman filtering are used to correct the model state by continuously introducing observation data; optimization methods such as gradient descent and genetic algorithm are used for parameter inversion to minimize the error between model output and observation data, solve the optimal model parameter combination, realize real-time correction of geological models, and improve the accuracy and timeliness of models.

[0056] For example, the vibration sensor, pressure sensor and temperature sensor installed on the drilling equipment have a sampling frequency of 100Hz, and the data is uploaded to the comprehensive information processing module in JSON format through the MQTT protocol, and the transmission delay is less than 10ms. The comprehensive information processing module uses the K-Means clustering algorithm to clean the data, and the proportion of outliers is controlled within 2%; the Min-Max normalization algorithm is used to normalize the data and map the data to the [0,1] interval. The stratigraphic structure and geotechnical property data are obtained from the geological radar and soil analyzer through the OPCUA protocol, and the data sampling interval is 1m. The FP-growth algorithm is used for association rule mining, with a minimum support of 0.05 and a minimum confidence of 0.8. 500 strong association rules are mined from 100,000 equipment working conditions and geological conditions data. The PrefixSpan algorithm is used for time series pattern mining, with a time window of 10s and a minimum support of 0.1, and 200 frequent time series patterns are mined from 10,000 equipment status time series data. Five features, namely drilling pressure, rotation speed, pump pressure, vibration and temperature, were selected to construct a random forest fault prediction model. The ratio of training set to test set was 8:2, the number of trees was 100, and the top three features in terms of importance were drilling pressure, rotation speed and vibration. The model had a precision of 0.95, a recall of 0.92 and an F1 value of 0.93 on the test set. The model predicted the probability of equipment failure in real time, and triggered an early warning when the probability of failure exceeded 0.8. The fault diagnosis knowledge base contained 1,000 fault cases, each of which contained four dimensions: fault mode, component, phenomenon and cause. The KNN algorithm was used for case reasoning, with a K value of 5 and a similarity threshold of 0.9. Principal component analysis was performed on 100 equipment failure data, and the first three principal components were selected, with a cumulative variance contribution rate of 85%. The first principal component represented the vibration characteristics of the equipment, the second principal component represented the temperature characteristics of the equipment, and the third principal component represented the pressure characteristics of the equipment. According to the fault diagnosis results, the fuzzy PID control algorithm is used to dynamically adjust the drilling parameters. The input variables are vibration and temperature, and the output variables are drilling pressure and speed. A 5×5 control rule table is constructed, using the triangular membership function, Mamdani reasoning method, and centroid defuzzification method. The initial parameters of the self-correcting PID control algorithm are Kp=1.5, Ki=0.2, and Kd=0.1, and the parameters are automatically adjusted every 5 minutes. The distribution network operation data is obtained from the distribution automation system through the IEC61850MMS protocol. The data sampling interval is 10ms. The Granger causality test is used to analyze the causal relationship between the fault diagnosis results and the distribution network operation parameters. The causal probability threshold is 0.9. It is found that some serious faults will cause voltage fluctuations and load mutations in the distribution network. The ensemble Kalman filter algorithm is used for geological model data assimilation. The state vector dimension is 50, the observation vector dimension is 20, the initial state covariance matrix is ​​the unit matrix, and the process noise covariance matrix and the observation noise covariance matrix are set according to experience.The genetic algorithm was used to invert the geological model parameters, with a chromosome length of 20, a population size of 50, a crossover probability of 0.8, a mutation probability of 0.1, and convergence after 500 iterations. The consistency between the inversion results and the drilling data was improved.

[0057] The present disclosure also provides a distribution network geological survey system, including the distribution network geological survey method as described above.

[0058] In the description of the present disclosure, it is necessary to understand that the directions or positional relationships indicated by directional words such as "front, back, up, down, left, right", "lateral, vertical, horizontal" and "top, bottom" are usually based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present disclosure and simplifying the description. Unless otherwise specified, these directional words do not indicate or imply that the device or element referred to must have a specific orientation or be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the scope of protection of the present disclosure.

[0059] For those skilled in the art, various other corresponding changes and deformations can be made according to the technical solutions and concepts described above, and all of these changes and deformations should fall within the protection scope of the claims of the present disclosure.

Claims

1. A method for geological survey of a distribution network, characterized in that: The following steps are involved: Collect original data of the distribution network survey area, establish a distribution network survey geological model, obtain predicted stratum characteristic parameters through the distribution network survey geological model, feed the predicted stratum characteristic parameters back to the distribution network geographic information system, and update the geological background information of the distribution network geographic information system; S102. Real-time monitoring of drilling parameters, establishment of a geological feedback control system, adjustment of drilling parameters through the geological feedback control system, and feedback of the adjusted drilling parameters to the distribution network geological survey model to correct formation characteristic parameters; establishment of a multi-objective optimization model to obtain optimal drilling parameters, and establishment of a multi-criteria decision-making model to obtain the optimal construction plan; S103, collecting real-time posture data of the drill rod and the real-time posture data of the drill bit, comparing the real-time posture data of the drill rod and the real-time posture data of the drill bit with a preset drilling trajectory, and correcting the position and direction of the drill rod and the position and direction of the drill bit based on the comparison results; and establishing a drilling posture optimization knowledge base for guiding subsequent drilling operations; S104, exchanging real-time data between the drilling equipment and the underlying scanner via the communication network, sharing status between the drilling equipment and the vibration monitor, and distributing instructions to the equipment via the distribution network central data platform; S105: Collecting operating status information of each operating device, analyzing the mutual influence between the operating status information of each operating device, obtaining an adjustment plan using an isolation forest anomaly detection algorithm and an adaptive genetic algorithm, and adjusting the operating parameters of each operating device using the adjustment plan. The operating parameters of each operating device are used to optimize the operating status of each operating device; S106. Monitor the drill bit speed, thrust, drilling depth, and borehole wall temperature of the drilling equipment; predict changes in formation hardness and determine whether the operating conditions are abnormal. When the predicted formation hardness exceeds a preset threshold, adjust the drilling process using a real-time optimized model predictive control algorithm. When the working condition is judged to be abnormal, protection is carried out through the rule engine; S107. Perform fault prediction and diagnosis, dynamically adjust drilling equipment operating parameters, and optimize the distribution network survey geological model.

2. The distribution network geological survey method according to claim 1, characterized in that: The step S101 includes: Obtain the physical and chemical property parameters of rocks to form raw data; obtain the location information of distribution network equipment to perform coordinate system conversion and data format unification; preprocess the raw data to obtain standardized geological exploration data; based on the standardized geological exploration data, establish a distribution network exploration geological model for predicting stratum characteristic parameters, feed the prediction results back to the distribution network geographic information system, and update the geological background information of the distribution network geographic information system.

3. The method for geological survey of distribution network according to claim 1, characterized in that: The step S102 includes: The drilling parameters of the drill bit are monitored in real time through sensors, and a geological feedback control system is established. The drilling parameters include drill bit vibration frequency, drill pipe torque, drill bit pressure, drill bit temperature, and drilling rig displacement. The control thresholds and corresponding drilling parameter control strategies are preset. The drilling parameters are adjusted according to the preset thresholds and drilling parameter control strategies. Based on the adjusted drilling parameters, the formation characteristic parameters of the current drilling formation are identified. The formation characteristic parameters of the current drilling formation are fed back to the distribution network exploration geological model, and the formation characteristic parameters in the distribution network exploration geological model are corrected; A multi-objective optimization model for selecting optimal drilling parameters is established based on drilling efficiency, drill bit life, and construction safety. According to the formation characteristic parameters, load requirements of the distribution network and foundation type, a multi-criteria decision-making model for selecting the optimal construction plan is established.

4. The method for geological survey of distribution network according to claim 1, characterized in that: The step S103 includes: A threshold range for the drill rod posture data and a threshold range for the drill bit posture data are preset; real-time drill rod posture data and real-time drill bit posture data are obtained via a sensor. The deviation between the real-time drill rod posture data and the preset drill rod posture data threshold is calculated, as is the deviation between the real-time drill bit posture data and the preset drill bit posture data threshold. A fuzzy rule base is established between the deviation value and the control variable, and the control variable, including the drilling rig thrust and rotational speed, is obtained based on the magnitude and direction of the deviation value. The position and direction of the drill rod, as well as the position and direction of the drill bit, are corrected based on the magnitude and direction of the deviation value. A drilling posture optimization knowledge base is established to guide subsequent drilling operations based on the real-time posture data, deviation value, and adjustment instructions generated during the drilling process.

5. The distribution network geological survey method according to claim 1, characterized in that: The step S104 includes: The MQTT protocol enables real-time data exchange between the drilling equipment and the formation scanner, and the MQTT protocol enables status sharing between the drilling equipment and the vibration monitor. The OPCUA protocol is used to upload the real-time status data of the drilling equipment to the central data platform of the distribution network. The real-time status data of the drilling equipment includes: drilling position, drilling speed, thrust, rotation speed and vibration level.

6. The method for geological survey of distribution network according to claim 1, characterized in that: The step S105 includes: collecting the working status information of each device through the central data platform of the distribution network, the working status information of each device including the drilling speed, drill bit speed, propulsion force, geological radar and soil analyzer of the drilling equipment; using algorithms such as the Pearson correlation coefficient to calculate the correlation matrix between different device parameters, and quantitatively analyzing the mutual influence of each parameter; using the isolation forest anomaly detection algorithm to analyze the abnormal behavior of the device real-time data, and selecting the anomaly threshold through statistical methods based on the distribution characteristics of historical data; designing a fitness function based on the historical data of device parameters and the anomaly detection results, and obtaining the optimal device working parameters based on the fitness function; and distributing the optimal device working parameters to each device through the intelligent operation and maintenance system of the distribution network.

7. The method for geological survey of distribution network according to claim 1, characterized in that: Step S106 includes: collecting parameter data such as drill bit speed, thrust, drilling depth, and borehole wall temperature, and performing streaming processing and real-time analysis; predicting and analyzing the changing trend of formation hardness using a distribution network survey geological model; performing autocorrelation analysis and stationarity testing on the formation hardness data; triggering a geological condition mutation warning when the predicted formation hardness meets preset requirements; establishing an online dynamic model of the drilling system, predicting system output for a period of time in the future, and optimizing control variables; collecting drill pipe vibration and drill bit temperature data, performing unsupervised learning using an isolation forest anomaly detection algorithm, selecting an anomaly threshold through cross-validation, and determining the presence of abnormal operating conditions such as drill bit overheating or rock breakage when the anomaly score exceeds the threshold; processing abnormal operating conditions using a rule engine; subscribing to relevant messages when a rule triggers a distribution network protection action, analyzing the abnormal status of the drilling equipment, and automatically determining whether to implement a protection action on the affected line based on the distribution network protection control strategy; and preprocessing the real-time operating parameters of the drilling equipment, anomaly detection results, and protection measure logs to establish a multi-dimensional, multi-scenario distribution network risk assessment model.

8. The method for geological survey of distribution network according to claim 1, characterized in that: The step S107 includes: collecting operating parameter data of each drilling equipment and performing preprocessing; obtaining geological information such as the stratum structure and rock and soil properties of the drilling location, and performing correlation analysis with the operating parameter data of the drilling equipment, establishing a multi-dimensional knowledge graph of equipment-geology-process and establishing an equipment failure prediction model, and when the failure probability exceeds a preset threshold, triggering a fault warning, generating a diagnosis report and pushing it to relevant personnel; obtaining fuzzy control rules between fault types and control operations based on historical data, and calculating control quantities through fuzzy reasoning and defuzzification processes; collecting distribution network operation data, the distribution network operation data including parameters such as voltage, current, and power, correlating the fault diagnosis results with the distribution network operation data, and evaluating the impact of the drilling equipment failure on the safe operation of the distribution network; if the failure may cause power outage in the distribution network or equipment damage, generating a distribution network dispatch plan, and reducing the scope of the fault impact through load dispatching, line switching and other measures to ensure the power supply reliability of the distribution network; correcting the model state by introducing observation data; and optimizing the distribution network survey geological model.

9. A distribution network geological survey system, characterized in that: The method comprises the distribution network geological survey method as described in any one of claims 1 to 8.

Citation Information

Patent Citations

  • Distribution network line unmanned aerial vehicle autonomous inspection integrated management system

    CN112668847A

  • Hydraulic ring geological survey system and method based on GPS positioning

    CN118428094A

  • Optimal trajectory control for directional drilling

    US20150330209A1

Cited By

  • Remote operation control method and system for drilling machine

    CN119393118A

  • Self-adaptive control method and control system for drilling anchor of airborne drilling arm

    CN119825330A

  • While-drilling intelligent rock drill and method for self-calibrating tunnel blast hole drilling track

    CN120139779A

  • Exploration drilling parameter control method and system

    CN120575835A

  • Anti-deviation hole drilling method and system based on geological feedback information

    CN120701308A