Automatic deviation rectifying method for long-distance jacking of pipe jacking tunneling roadway

Abstract

The invention discloses an automatic deviation rectifying method for long-distance jacking of a pipe-jacking tunneling roadway, and relates to the technical field of automatic deviation rectifying of pipe tunneling, which comprises the following steps: collecting the compactness, water content and hardness data of a soil layer in a tunneling path in real time through a multi-point distributed sensing device, and calculating the deviation of the pipe-jacking tunneling roadway in combination with soil parameters and pipe-jacking equipment propulsion mechanical characteristics; and generating a geological condition profile map, constructing a resistance distribution model, predicting resistance change trends under different geological conditions, and establishing a correction early warning mechanism. According to the invention, through real-time stress monitoring, multi-point sensing and intelligent feedback control, the accuracy, efficiency and complex geology adaptability of pipe-jacking tunneling are improved. The equipment can quickly identify resistance abnormity, dynamically regulate and control propulsion in real time, reduce the influence of asymmetric force, and reduce personal error and construction loss. And in combination with a geological profile map, the equipment actively deals with soil layer changes through automatic deviation correction, the clamping stagnation problem is solved, safety and continuity are guaranteed, and risks and cost are reduced.

Description

Technical Field

[0001] The invention relates to the technical field of automatic deviation correction for pipe excavation, and in particular to a method for automatic deviation correction for long-distance jacking in a jacking pipe excavation tunnel. Background Art

[0002] Automatic deviation correction technology for long-distance pipe jacking tunnels is an automated correction technology developed to address directional deviation issues that may occur during pipe jacking projects. This technology automatically adjusts the thrust posture of the pipe jacking equipment by monitoring the direction, angle, and position of the pipe in real time. This technology, combined with high-precision sensors (such as gyroscopes and laser rangefinders) and intelligent control systems, ensures that the tunnel remains within the designed trajectory during long-distance advancement, thereby avoiding construction quality issues or resource waste caused by deviations. This deviation correction method has significant advantages in reducing manual intervention, improving construction efficiency, and ensuring project quality.

[0003] The existing technology has the following deficiencies:

[0004] In the existing technology for automatic deviation correction during long-distance jacking tunneling, nonlinear resistance changes caused by soil heterogeneity are a significant issue. Pipe jacking often requires traversing complex and changing geological conditions. The density, moisture content, and hardness of the soil layer can vary significantly in different areas, resulting in uneven resistance distribution around the jacking pipe. This resistance unevenness exerts asymmetric forces on the jacking equipment, increasing the difficulty of the deviation correction system's response and adjustment. If not addressed promptly, it may cause the jacking pipe to deflect, become stuck, and even lead to equipment damage or tunnel collapse, posing a serious threat to construction safety and project progress.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Summary of the Invention

[0006] The purpose of the present invention is to provide a method for automatic deviation correction of long-distance jacking in jacking tunnels. Through the synergistic effect of real-time force monitoring, multi-point distributed sensing and intelligent feedback control systems, the above scheme significantly improves the accuracy, efficiency and adaptability to complex geological conditions of jacking tunnels. The equipment can quickly identify resistance anomalies and dynamically adjust the advancement. It reduces the influence of asymmetric forces through hydraulic drive and resistance balancing devices, avoids repeated construction and losses, and reduces human errors. Combined with geological profiles and resistance models, the equipment can actively respond to changes in complex soil layers, effectively solve the problem of stagnation, ensure construction safety and continuity, reduce risks and costs, and provide technical support for high-risk scenarios to solve the problems in the above-mentioned background technology.

[0007] In order to achieve the above object, the present invention provides the following technical solution: a method for automatically correcting the deviation of a long-distance jacking tunnel, comprising the following steps:

[0008] Multi-point distributed sensing devices are used to collect data on the density, moisture content, and hardness of the soil along the excavation path in real time. Combined with soil parameters and the mechanical properties of the pipe jacking equipment, a geological profile is generated, and a resistance distribution model is constructed to predict resistance trends under different geological conditions.

[0009] Install multi-axis force sensors on the pipe jacking equipment to monitor the force status of the equipment in real time during the advancement process. Compare the measured data with the resistance distribution model to identify abnormal resistance areas.

[0010] The hydraulic drive system controls the thrust angle and direction of the pipe jacking equipment, and the equipment posture is roughly adjusted based on monitoring data to reduce the impact of asymmetric forces on the tunneling path.

[0011] When the nonlinear resistance change area is detected, the resistance balancing device is activated to achieve uniform resistance by adjusting the pressure distribution around the top pipe head, thereby reducing the impact of asymmetric force on the propulsion of the equipment;

[0012] Based on real-time monitoring data, an improved Dijkstra algorithm is used to dynamically adjust the tunneling path, avoid areas of abnormal resistance, and ensure the smoothness of the advancement path and the accuracy of the correction operation;

[0013] Through the neural network learning model, historical advancement data and current operation data are analyzed, and the correction control parameters are optimized in real time to further improve the equipment's correction efficiency and accuracy, ensuring the stability of the path during long-distance advancement.

[0014] Preferably, the density, moisture content, and hardness data of the soil layer in the excavation path are collected in real time by a multi-point distributed sensing device, and the geological condition profile is generated by combining the soil parameters and the propulsion mechanical characteristics of the jacking equipment. The resistance distribution model is constructed to predict the resistance change trend under different geological conditions. The specific steps are as follows:

[0015] Arrange sensors in key areas of the tunneling path, calibrate the equipment, and transmit data in real time via wireless networks;

[0016] Use high-frequency sampling technology to collect geological data in real time, and use filtering and correction algorithms to ensure data accuracy and consistency;

[0017] Through GIS modeling and spatial interpolation algorithms, a visual geological profile is generated to fully display the distribution of geological parameters within the path;

[0018] Based on the geological profile and pipe jacking equipment parameters, a resistance distribution model is constructed and the resistance change trend is predicted to provide support for promoting adjustments.

[0019] Preferably, a multi-axis force sensor is installed on the pipe jacking equipment to monitor the force state of the equipment in real time during the advancement process, and the measured data is compared with the resistance distribution model to identify the abnormal resistance area. The specific steps are as follows:

[0020] Select high-precision multi-axis force sensors and arrange and reinforce them in key stress-bearing areas of the pipe jacking equipment to ensure the accuracy and stability of data acquisition;

[0021] Axial force, radial force and shear force data are recorded through high-frequency sampling and transmitted synchronously with geological parameters to the central control system to form a dynamic database;

[0022] Compare and analyze real-time force data with the resistance distribution model to locate abnormal areas and trigger a correction and warning mechanism;

[0023] Identify abnormal resistance areas, adjust the propulsion posture through the correction control system and dynamically optimize the strategy to ensure the stability and effectiveness of the propulsion path.

[0024] Preferably, the specific steps of using a hydraulic drive system to control the propulsion angle and direction of the pipe jacking equipment and making rough adjustments to the equipment posture according to monitoring data to reduce the impact of asymmetric forces on the tunneling path are as follows:

[0025] Based on the geological profile and resistance distribution model, the hydraulic system was initialized and tested multiple times to ensure thrust accuracy and equipment stability.

[0026] Analyze the deviation between sensor data and resistance model in real time, generate coarse adjustment signals for propulsion angle and direction, and formulate adjustment strategies;

[0027] The hydraulic jack adjusts the thrust and direction to perform rough adjustments in real time, while monitoring the adjustment effect and inputting feedback data to ensure immediacy;

[0028] Through posture verification and data analysis, the optimization and adjustment strategy is carried out, and combined with the prediction algorithm, the stability and efficiency of the subsequent propulsion path are improved.

[0029] Preferably, the specific steps of identifying the resistance nonlinear change area through multi-sensor collaboration, judging the starting conditions and activating the resistance balancing device, and setting the initial working state according to the geological profile and the resistance model are as follows:

[0030] Collect geological and force data through sensors, compare them with the resistance distribution model, and analyze and locate areas with nonlinear resistance changes;

[0031] Combined with pattern recognition algorithms to determine the characteristics of abnormal areas, providing a basis for subsequent regulation;

[0032] Determine whether the conditions for starting the balancing device are met based on the asymmetric force and resistance distribution characteristics, and generate a start signal;

[0033] Activate the resistance balancing device and set the initial working state to ensure the accuracy and reliability of the control starting point.

[0034] Preferably, the specific steps for using the resistance balancing device to adjust the pressure distribution around the jacking head in real time, verify the control effect and optimize the strategy to ensure uniform force and path stability during the advancement process are as follows:

[0035] Adjust the pressure distribution through the resistance balancing device, dynamically equalize the resistance, and reduce the impact of asymmetric force;

[0036] Monitor the equipment's stress state and propulsion angle in real time to assess whether the control effect has achieved the expected goal;

[0037] Optimize control parameters and strategies based on feedback results to improve the device's ability to adapt to complex geological conditions;

[0038] After completing the balancing, the device is restored to the standard state and the control data is stored to provide a reference for subsequent operations.

[0039] Preferably, based on real-time monitoring data, the improved Dijkstra algorithm is used to dynamically adjust the excavation path, avoid abnormal resistance areas, and ensure the smoothness of the advancement path and the accuracy of the correction operation. The specific steps are as follows:

[0040] Using multi-point distributed sensing devices and multi-axis force sensors, we collect resistance-related parameters at each node in the tunneling path, including soil density, moisture content, and hardness, as well as the axial and radial forces acting on the equipment. The acquired data is used to calculate the node resistance value and construct a resistance distribution map. The calculation expression is as follows:

[0041] Where R i is the resistance value of the ith node, d i is the soil density of the i-th node, w i is the soil moisture content at the i-th node, h i is the soil hardness of the i-th node, α1 is the density weight factor, α2 is the water content weight factor, α3 is the hardness weight factor, F a，i is the axial force at the ith node, F r，i is the radial force at the i-th node, A is the cross-sectional area of the pipe jacking equipment, P is the circumference of the pipe jacking equipment, β1 is the weight factor of the axial force, and β2 is the weight factor of the radial force;

[0042] The tunneling path is abstracted into a graph structure consisting of nodes and edges. The edge weights are dynamically calculated using the resistance value and path smoothing factor to form a weighted graph. The calculation expression is as follows:

[0043] W ij =γ1(R i +R j )+γ2S ij , where W ij is the edge weight from node i to node j, R i is the resistance value of node i, R j is the resistance value of node j, γ1 is the weight factor of resistance, γ2 is the weight factor of smoothing factor, S ij is the path smoothing factor, which is used to measure the path turning cost between node i and node j and indicates the smoothness of the path.

[0044] Preferably, based on the dynamic weight graph, an improved Dijkstra algorithm is used to calculate the optimal path, and historical path weights and real-time adjustment factors are introduced to optimize the flexibility and accuracy of path planning. The formula is as follows:

[0045] Where, P * is the optimal path, P is the set of optional paths, H i is the historical path weight, T ij is the real-time adjustment factor from node i to node j, λ1 is the historical weight adjustment factor, and λ2 is the adjustment parameter of the real-time adjustment factor;

[0046] According to the calculated optimal path P * , drives the pipe jacking equipment to advance along the planned path, records the actual path point set and its force data in real time, and compares the actual path points with the optimal path P * , identifies path deviations and generates feedback, dynamically adjusts the weight map and resistance distribution model, and optimizes subsequent path planning. The formula is as follows:

[0047]

[0048] In the formula, ΔW ij is the dynamic adjustment of the edge weight from node i to node j, F actual，ij is the actual force value of the equipment passing through edge node i and node j, F model，ij is the force value of edge node i and node j predicted by the model, P actual is the actual path, η1 is the adjustment factor of force deviation, η2 is the adjustment factor of force change rate, η3 is the adjustment factor of path deviation, dist(P * , P actual ) is the optimal path P * and the actual path P catualDeviation distance.

[0049] Preferably, the neural network learning model is used to analyze historical propulsion data and current operation data, optimize the correction control parameters in real time, further improve the correction efficiency and accuracy of the equipment, and ensure the stability of the path during long-distance propulsion. The specific steps are as follows:

[0050] First, historical propulsion data and current operation data, including propulsion angle, propulsion speed, force state, and geological characteristic parameters, are collected. The acquired parameters are converted into the same numerical range through normalization processing. Key features are extracted using principal component analysis to generate the input feature matrix of the neural network. The generation formula is as follows:

[0051] X t =PCA(Norm([θ t , v t , F t , G t ])), where X t is the feature matrix after normalization and dimensionality reduction, F t is the force vector of the device, F t =[F ax , F rad , F shear ], where F ax is the axial force, F rad is the radial force, F shear is the shear force, G t is the geological parameter vector, G t =[d, w, h], where d is density, w is water content, h is hardness, θ t is the current propulsion angle, v t is the current propulsion speed, Norm(·) is the normalization function, and PCA(·) is the principal component analysis function;

[0052] The generated normalized and dimensionally reduced feature matrix X t Input to the multi-layer feedforward neural network, the neural network optimizes the model parameters by training historical data, predicts the correction control parameters, and in the real-time inference stage, inputs the normalized and dimension-reduced feature matrix X t , calculate the correction control parameter vector, the calculation expression is as follows:

[0053] C t =f(W·X t +b), where W is the weight matrix of the neural network, b is the bias vector of the neural network, f is the activation function, and C t It is the correction control parameter.

[0054] Preferably, the correction parameter C predicted by the neural network ist Applied to propulsion equipment, it adjusts the propulsion angle, speed, and force distribution in real time, measures the equipment response results, and calculates the control error vector by comparing the equipment response results with the target state. The error vector is used to dynamically update the parameters of the neural network through the back propagation algorithm to optimize the model performance and adapt to the propulsion needs under complex geological conditions. The control error vector calculation expression is as follows:

[0055] E t =[θ target -(θ t +Δθ), F target -(F t +ΔF), v target -(v t +Δv)], where θ target is the target propulsion angle, F target is the target force, v target is the target advancement speed, E t is the control error vector, E t =[E θ , E F , E v ], where E θ is the angular error, E F is the force error, E v is the velocity error;

[0056] Verify the stability of the adjusted propulsion path and calculate the correction efficiency index. The correction efficiency is used to evaluate the adjustment effect. If the efficiency meets the optimization goal, the model parameters are updated to adapt to the complex environment. The calculation formula of the correction efficiency is as follows:

[0057] Where η t is the correction efficiency, ||E t || 2 is the sum of squares of the error vector.

[0058] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0059] Through the synergistic effects of real-time force monitoring, multi-point distributed sensing, and intelligent feedback control systems, the present invention significantly improves the accuracy and construction efficiency of the jacking tunneling path. The equipment can quickly identify areas of abnormal resistance and automatically generate adjustment plans. It achieves dynamic regulation through the hydraulic drive system and resistance balancing device, reducing the interference of asymmetric forces on the propulsion path. Compared with the traditional method that relies on manual monitoring and manual adjustment, this intelligent solution significantly reduces human errors, avoids repeated construction and equipment loss, and improves construction quality and efficiency. At the same time, it reduces downtime and the need for manual operation, speeds up project progress, and provides technical support for long-distance jacking tunneling.

[0060] This invention significantly improves the equipment's ability to adapt to complex geological conditions. Combined with real-time analysis of geological profiles and resistance distribution models, pipe jacking equipment can predict and proactively respond to changes in resistance in heterogeneous soil layers or weak interlayers. The resistance balancing device effectively resolves equipment jams and advancement difficulties caused by sudden changes in resistance by dynamically adjusting the pressure distribution, thereby ensuring construction safety and continuity. This technical capability not only reduces construction risks and costs in complex geological environments, but also makes pipe jacking technology more widely applicable, providing reliable protection for pipe jacking construction in high-risk scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction to the drawings required for use in the embodiments will be given below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0062] Figure 1 The present invention provides a method flow chart of a long-distance automatic deviation correction method for jacking a pipe tunnel. DETAILED DESCRIPTION

[0063] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of this disclosure will be thorough and complete and will fully convey the concepts of the example embodiments to those skilled in the art.

[0064] The present invention provides Figure 1 The method for automatically correcting the deviation of a long-distance jacking tunnel in a jacking tunnel shown in the figure comprises the following steps:

[0065] Multi-point distributed sensing devices collect data on the density, moisture content, and hardness of soil layers along the tunneling path in real time, generating a geological profile. Combining soil parameters with the mechanical properties of the pipe jacking equipment, a resistance distribution model is constructed to predict resistance trends under different geological conditions.

[0066] The density, moisture content, and hardness data of the soil along the tunneling path are collected in real time using multi-point distributed sensing devices to generate a geological profile. Combining soil parameters with the propulsion mechanical properties of the pipe jacking equipment, a resistance distribution model is constructed to predict resistance trends under different geological conditions. The specific steps are as follows:

[0067] In key areas of the tunneling path, sensors are rationally placed on the tunneling equipment, and the equipment is calibrated to achieve real-time data transmission via wireless networks;

[0068] Multi-point distributed sensing devices, including soil density sensors, moisture meters, and hardness meters, are deployed at key points along the tunneling path. Sensor installation locations are determined based on geological survey reports to ensure coverage of critical areas along the entire tunneling path. During initialization, calibration equipment is used to eliminate environmental noise and sensor bias, ensuring data accuracy. Furthermore, wireless communication modules are used to connect sensors and the central data processing unit, enabling real-time data transmission.

[0069] In complex geological conditions, sensor placement needs to account for soil heterogeneity. For example, sensor density should be increased in areas where rock cracks, sandy layers, or weak clay may be present. Data collected during the pilot phase verifies sensor sensitivity and reliability, establishing a complete data collection network.

[0070] Use high-frequency sampling technology to collect geological data in real time, and use filtering and correction algorithms to ensure data accuracy and consistency;

[0071] Sensors are activated to collect real-time geological parameters along the tunneling path, including soil density (reflecting soil compressibility), moisture content (affecting shear strength and friction), and hardness (determining the penetration of the jacking head). After data collection, it enters the central data processing system for preliminary processing, using filtering algorithms to eliminate occasional noise or erroneous signals. Furthermore, data deviations caused by changes in ambient temperature and humidity are corrected to ensure data stability and consistency.

[0072] The real-time data acquisition system uses high-frequency sampling technology (e.g., 10 to 50 times per second) to capture subtle changes in geological parameters, especially sudden changes in resistance that may occur during advancement. A built-in historical data comparison module identifies abnormal data points and smoothes them using data from adjacent points.

[0073] Through GIS modeling and spatial interpolation algorithms, a visual geological profile is generated to fully display the distribution of geological parameters within the path;

[0074] The collected multi-point data is input into GIS (Geographic Information System) modeling software and combined with the geographic coordinates of the tunneling path to generate a visual geological profile. This profile displays the distribution of soil density, moisture content, and hardness in different areas using color zoning or three-dimensional format, providing a visual basis for subsequent resistance prediction. Furthermore, using the geological model's spatial interpolation algorithm, areas without sensors are inferred and filled in, generating continuous geological profile data.

[0075] The generation of profiles requires the use of various visualization methods, such as 2D profiles to display the cross-sectional distribution of soil parameters and 3D profiles to present the entire excavation path. The model's integrity is enhanced by overlaying historical survey data with real-time data. Warning markers are placed on the profiles for high-risk areas (such as weak interlayers).

[0076] Based on the geological profile and pipe jacking equipment parameters, a resistance distribution model is constructed and resistance change trends are predicted to provide support for adjustments.

[0077] Based on the generated geological profiles and combined with the propulsion mechanics parameters of the pipe jacking equipment (such as cutterhead speed, propulsion torque, and ambient pressure), a mathematical model of resistance distribution was established. This model employed finite element analysis to simulate soil deformation and resistance distribution after load application. A time series prediction algorithm was used to simulate resistance trends under different geological conditions, providing data support for the propulsion equipment's posture adjustment and correction operations.

[0078] The resistance distribution model requires sub-models for different soil types, such as sandy soil, clay, and gravel. Model parameters (such as friction coefficient and cohesion) are adjusted to optimize and verify the model. Furthermore, the model's predicted resistance trends are used to generate an early warning system. When resistance exceeds a set threshold, it prompts corrections or parameter adjustments to ensure continuous and stable excavation.

[0079] Multi-axis force sensors are installed on the pipe jacking equipment to monitor the force status of the equipment in real time during the advancement process, including axial force, radial force, and shear force. The measured data is compared with the resistance distribution model to identify abnormal resistance areas;

[0080] Install a multi-axis force sensor on the pipe jacking equipment to monitor the force status of the equipment in real time during the advancement process, including axial force, radial force, and shear force. Compare the measured data with the resistance distribution model to identify abnormal resistance areas. The specific steps are as follows:

[0081] Reasonably arrange and reinforce high-precision multi-axis force sensors at key stress-bearing locations of pipe jacking equipment to ensure the accuracy and stability of data acquisition;

[0082] Based on the pipe jacking equipment's structure and construction requirements, high-precision multi-axis force sensors are selected to ensure real-time monitoring of multi-dimensional stress states, including axial, radial, and shear forces. Sensors are strategically placed at key stress-bearing locations on the pipe jacking equipment (such as the cutterhead center, sidewalls, and propulsion mechanism) to ensure data acquisition covers the equipment's primary stress-bearing surfaces. The sensors are secured to the equipment with reinforced brackets to prevent data deviations caused by vibration or impact, while ensuring good coupling between the sensors and the propulsion system.

[0083] The sensor must meet the requirements of complex environments, such as being resistant to moisture, dust, and high temperatures, and capable of real-time sampling and wireless data transmission. Before installation, simulated pipe jacking equipment operating scenarios are tested to verify the sensor's stability and response speed, laying the foundation for subsequent data collection.

[0084] Axial force, radial force and shear force data are recorded through high-frequency sampling and transmitted synchronously with geological parameters to the central control system to form a dynamic database;

[0085] During operation, the multi-axis force sensor is activated to collect force data, including axial force (propulsion), radial force (sidewall friction), and shear force (soil cutting resistance). Force changes are recorded through high-frequency sampling (e.g., 10 to 50 times per second). The collected data is transmitted in real time to the central control system and integrated with geological data obtained by multi-point distributed sensing devices to form a dynamic monitoring database.

[0086] Real-time data acquisition utilizes synchronized clock technology to ensure timeline consistency between force data and geological parameters. Multidimensional data collected by sensors is filtered to remove noise, while a self-calibration mechanism addresses deviations caused by environmental changes, ensuring data accuracy.

[0087] Compare and analyze real-time force data with the resistance distribution model to locate abnormal areas and trigger a correction and warning mechanism;

[0088] Real-time force data collected by sensors is compared with a resistance distribution model constructed based on geological profiles to analyze the degree of match between equipment force and soil resistance. Abnormal areas, such as sudden increases in local force, are located by calculating deviation values (e.g., the vector difference between the force and resistance directions). These areas are marked as potential risk points, triggering a correction and early warning mechanism.

[0089] The comparative analysis utilizes a force simulation model based on the finite element method, comparing sensor data with the model's predictions point by point to generate a force distribution map. For areas where deviations exceed a set threshold, the correlation between force and geological parameters is analyzed to determine whether the propulsion attitude needs to be adjusted or the equalization device needs to be activated.

[0090] Identify abnormal resistance areas, adjust propulsion posture through the correction control system and dynamically optimize strategies to ensure the stability and effectiveness of the propulsion path;

[0091] Based on the comparison and analysis results, abnormal resistance areas are identified and feedback signals are generated. These signals are fed into the pipe jacking equipment's correction control system, which optimizes the equipment's propulsion posture by adjusting the thrust angle, distributing hydraulic thrust, or activating the resistance equalization device. Simultaneously, based on trend prediction data from the resistance distribution model, the adjustment strategy is dynamically updated to ensure the continued effectiveness of correction measures.

[0092] Abnormal areas are identified using a multi-dimensional assessment, including the intensity, duration, and spatial distribution of abnormal forces. A feedback system responds to abnormal signals in real time through a closed-loop control algorithm, tightly integrating deviation correction with force monitoring to achieve efficient correction of the tunneling path. Combined with forecast trends, high-risk areas can be identified in advance, effectively avoiding potential construction obstacles.

[0093] The hydraulic drive system controls the thrust angle and direction of the pipe jacking equipment, and the equipment posture is roughly adjusted based on monitoring data to reduce the impact of asymmetric forces on the tunneling path.

[0094] The specific steps for using the hydraulic drive system to control the thrust angle and direction of the pipe jacking equipment and making rough adjustments to the equipment posture based on monitoring data to reduce the impact of asymmetric forces on the tunneling path are as follows:

[0095] Based on the geological profile and resistance distribution model, the hydraulic system was initialized and tested multiple times to ensure thrust accuracy and equipment stability.

[0096] Initial settings for the hydraulic drive system are performed based on the pipe jacking equipment's structural characteristics and construction requirements, including the thrust range, movement speed, and response sensitivity. Based on the geological profile and resistance distribution model, the initial hydraulic system parameters are adjusted to accommodate potential resistance variations along the tunneling path. Multiple rounds of testing are conducted before the equipment is put into operation to ensure the accuracy of the hydraulic jacking force and overall stability.

[0097] Hydraulic system initialization must consider the complexity of geological conditions. For example, in areas with soft soil, thrust must be reduced to avoid pipeline damage, while in dense soil, thrust must be increased to maintain propulsion efficiency. Simulated stress testing is also performed to verify the hydraulic system's response speed and adjustment accuracy under different conditions, ensuring the reliability of subsequent operations.

[0098] Analyze the deviation between sensor data and resistance model in real time, generate coarse adjustment signals for propulsion angle and direction, and formulate adjustment strategies;

[0099] Data from multi-point distributed sensing devices and multi-axis force sensors is collected to calculate in real time the deviation between the device's current posture and the intended propulsion path, including changes in propulsion angle and shifts in force direction. This analysis is fed into the control system, generating posture adjustment signals. Based on the resistance characteristics of different soil layers, a coarse adjustment strategy is developed, such as adjusting the propulsion angle or reducing lateral pressure.

[0100] Data input utilizes a high-frequency acquisition and processing mode to ensure rapid response in complex geological conditions. By comparing real-time force data with the resistance distribution model, the cause of deviation, such as excessive lateral friction or insufficient propulsion, can be accurately determined, resulting in targeted adjustment plans and improved correction efficiency.

[0101] The hydraulic jack adjusts thrust and direction in real time to perform rough adjustments. Adjustment results are monitored and feedback data is input to ensure immediacy. Based on the adjustment signals, the hydraulic drive system is activated to make preliminary adjustments to the thrust angle and direction. By adjusting the thrust and sequence of the hydraulic jack, the pipe jacking equipment's posture is controlled. For example, increasing the thrust on one side of the hydraulic jack can correct the equipment's offset angle. This process requires real-time monitoring of the adjustment results, with feedback data input into the control system for reanalysis to ensure accurate and timely adjustments.

[0102] Hydraulic jacks are usually arranged around the pipe jacking equipment, forming a multi-point distributed support structure around the equipment to ensure that the thrust direction can be flexibly adjusted during the advancement process. According to construction requirements, hydraulic jacks are generally installed at the front end or side wall of the pipe section so that the posture and direction of the equipment can be adjusted during advancement. The thrust directions that can be adjusted by the hydraulic jack include: axial thrust (along the pipeline advancement direction), which is used to provide the main thrust to advance the equipment; radial thrust (perpendicular to the direction of the pipeline axis), which is used to adjust the offset of the equipment, such as correcting lateral offset by increasing the unilateral thrust; pitch and yaw angle adjustment, that is, by unevenly applying hydraulic pressure, adjusting the tilt angle of the equipment on the vertical or horizontal plane to correct the trajectory deviation. By real-time monitoring of the changes in the hydraulic jacking thrust and combining feedback data, the control system can dynamically optimize the thrust distribution to achieve precise correction and stable advancement.

[0103] Hydraulic jacking requires precise coordination of movements. For example, if uneven radial force is detected, the thrust and direction of multiple hydraulic jacks must be adjusted synchronously to avoid additional deviation. A real-time feedback mechanism quickly checks whether adjustments are achieving the desired results. If the deviation is not corrected, further adjustments are automatically triggered.

[0104] Through posture verification and data analysis, we optimize and adjust strategies, and combine them with prediction algorithms to improve the stability and efficiency of subsequent propulsion paths;

[0105] After completing the rough adjustment, the adjustment results are verified by comparing the multi-axis force sensor with the geological profile to determine whether the equipment has returned to the expected propulsion path. If posture deviation persists, the next adjustment strategy is optimized based on the adjustment results, such as modifying the hydraulic jack's action sequence or dynamically changing the thrust distribution. Furthermore, incorporating trend prediction algorithms, a more precise posture control plan is developed for the subsequent propulsion process, reducing the number of subsequent adjustments.

[0106] The verification phase is crucial. By comparing attitude data before and after adjustments, the effectiveness of the adjustment strategy is analyzed. After multiple verifications, the adjustment algorithm is gradually optimized, enabling the hydraulic system to independently and efficiently perform correction operations under similar geological conditions, thereby improving overall propulsion stability and efficiency.

[0107] When a nonlinear resistance change area is detected, the resistance balancing device is activated to achieve uniform resistance by adjusting the pressure distribution around the jacking head, thereby reducing the impact of asymmetric force on equipment propulsion. The nonlinear resistance change area is identified through multi-sensor collaboration, the starting conditions are determined, and the resistance balancing device is activated. The specific steps for setting the initial working state based on the geological profile and resistance model are as follows:

[0108] Sensors collect geological and force data, compare it with the resistance distribution model, and analyze and locate areas of nonlinear resistance variation. Multi-point distributed sensing devices and multi-axis force sensors work together to collect real-time geological parameters and equipment stress conditions along the tunneling path, including soil density, moisture content, and hardness, as well as the axial, radial, and shear forces of the equipment. Real-time data is compared with the resistance distribution model to analyze whether there are areas of significant nonlinear resistance variation along the tunneling path, such as sudden increases in localized resistance or uneven distribution.

[0109] High-frequency sampling and filtering algorithms are used during the data acquisition phase to ensure the captured parameters are highly accurate and stable. Anomaly analysis uses model-based multi-dimensional thresholds to accurately identify areas of abnormal resistance, such as comparing force gradients and resistance prediction errors at different points.

[0110] Combined with pattern recognition algorithms, the characteristics of abnormal areas are determined to provide a basis for subsequent regulation.

[0111] Identified areas of abnormal resistance are marked on a geological profile, and pattern recognition algorithms are used to determine their characteristics, such as their extent, shape, and soil composition. For areas with complex resistance variations, the need for resistance equalization is determined by comparing historical data with trend predictions.

[0112] Pattern recognition combined with machine learning algorithms, such as support vector machines or neural networks, automatically classifies soil types and resistance patterns based on regional characteristics, identifying the root cause of the problem. For example, is it due to a significant increase in resistance caused by abnormal local moisture content, or a sudden increase in soil density causing asymmetric loading?

[0113] Based on the asymmetric force and resistance distribution characteristics, it is determined whether the conditions for starting the equalizing device are met and a start signal is generated.

[0114] Based on the regional characteristic analysis results, the uniformity index of the resistance distribution model is compared to determine whether the conditions for activating the equalizer are met. If activation is determined to be necessary, a start signal is generated and a command is issued to the resistance equalizer to prepare for control operations.

[0115] The activation conditions are determined through a multi-level judgment process. For example, when the radial force difference exceeds a set threshold or the duration of the asymmetric force exceeds a specified range, the system triggers the activation command. A redundant judgment mechanism is also implemented to minimize false triggering caused by short-term fluctuations or occasional noise.

[0116] Activate the resistance balancing device and set the initial operating state to ensure the accuracy and reliability of the control starting point. After the resistance balancing device is activated, the initial operating state is set based on regional characteristics, such as determining the hydraulic pressure distribution, grouting pressure, or the position and angle of the support rods. The initial state setting should refer to the geological profile and resistance distribution model to ensure the accuracy and reliability of the control starting point of the balancing device.

[0117] The initial state is set by an automated control system. For example, it assigns maximum pressure to the point with the most concentrated resistance in the cross-sectional diagram, while adjusting the gradient distribution of surrounding pressure to gradually achieve equilibrium. Once the setting is complete, the device enters the control phase.

[0118] The specific steps for using the resistance balancing device to adjust the pressure distribution around the jacking head in real time, verify the control effect and optimize the strategy to ensure uniform force and path stability during the advancement process are as follows:

[0119] The resistance equalization device adjusts pressure distribution, dynamically equalizing resistance and reducing the impact of asymmetric forces. The resistance equalization device adjusts pressure distribution around the jacking pipe head through hydraulic drive or controllable grouting. In areas where resistance varies nonlinearly, pressure is applied or released point by point, dynamically adjusting the soil stress state to make resistance uniform and reduce asymmetric forces acting on the equipment.

[0120] Pressure regulation is precisely adjusted based on real-time feedback data. For example, the hydraulic system automatically increases pressure to a specified range, while the grouting system selects the optimal slurry density and injection rate based on soil characteristics. Pressure changes must be continuously monitored during the regulation process to ensure that adjustments are immediate and precise.

[0121] Real-time monitoring of the equipment's stress state and thrust angle assesses whether the desired control effect is being achieved. During pressure regulation, real-time data on the equipment's stress state and thrust angle is collected to determine whether the adjustment is effectively improving the asymmetric stress state. This data is input into the control system and compared with the resistance distribution model to generate a balance evaluation report to guide subsequent optimization operations.

[0122] Feedback verification uses a closed-loop control algorithm, such as comparing the uniformity index of force distribution and the deviation value of the propulsion path. If the expected target is achieved, the adjustment is ended and the equipment status is updated; if the target is not achieved, the control parameters are continuously optimized.

[0123] Optimize control parameters and strategies based on feedback results to enhance the device's ability to adapt to complex geological conditions.

[0124] Based on this feedback, the resistance balancing device's operating strategy is optimized, such as adjusting the hydraulic thrust distribution ratio or the grouting point selection rules. During the optimization process, geological profiles and trend forecast data are combined to develop a more efficient balancing control plan, enhancing the device's ability to adapt to complex geological conditions.

[0125] The optimization process incorporates machine learning models, training algorithms based on historical control data and current feedback, gradually improving the accuracy and robustness of control strategies. For example, the hydraulic jacking sequence can be dynamically adjusted to suit specific soil characteristics.

[0126] After balancing is complete, the device is restored to its standard state, and the control data is stored for reference in subsequent operations. After the balancing control is completed, the resistance balancing device parameters are restored to their standard state, preparing for subsequent tunneling operations. At the same time, all data from the control period is stored in the system database for subsequent analysis and decision-making optimization in similar scenarios.

[0127] The recovery process ensures that the equipment restarts safely. This includes verifying the initial position of the hydraulic jack, the draining of the grouting system, and the balance of the overall pressure system. Simultaneously, empirical models are generated using stored data to provide a basis for rapid response during subsequent excavation.

[0128] Based on real-time monitoring data, the improved Dijkstra algorithm is used to dynamically adjust the excavation path, avoid areas of abnormal resistance, and ensure the smoothness of the advancement path and the accuracy of the correction operation.

[0129] Based on real-time monitoring data, the improved Dijkstra algorithm is used to dynamically adjust the tunneling path to avoid areas of abnormal resistance and ensure the smoothness of the advancement path and the accuracy of the correction operation. The specific steps are as follows:

[0130] Using multi-point distributed sensing devices and multi-axis force sensors, we collect resistance-related parameters at each node in the tunneling path, including soil density, moisture content, and hardness, as well as the axial and radial forces acting on the equipment. The acquired data is used to calculate the node resistance value and construct a resistance distribution map. The calculation expression is as follows:

[0131]

[0132] Where R i is the resistance value of the ith node, d i is the soil density of the i-th node, w i is the soil moisture content at the i-th node, h i is the soil hardness of the i-th node, α1 is the density weight factor, which measures the soil density d iThe relative importance of the soil water content in the comprehensive resistance value is α2, which is the water content weight factor, measuring the soil water content w i The relative importance of the comprehensive resistance value, α3 is the hardness weight factor, which measures the hardness of the soil layer h i The relative importance of F in the overall resistance value a，i is the axial force at the i-th node, which represents the force applied by the equipment along the excavation direction during the advancement of the i-th node. r，i is the radial force at the i-th node, the sidewall friction and reaction force that the equipment is subjected to during the advancement of the i-th node, A is the cross-sectional area of the pipe jacking equipment, P is the circumference of the pipe jacking equipment, β1 is the weight factor of the axial force, which measures the axial force F of the equipment a，i The comprehensive resistance value R i The relative importance of β2 is the weight factor of the radial force, which measures the radial force F of the device. r，i The comprehensive resistance value R i the relative importance of

[0133] Calculate R by the above steps i Display the numerical characteristics of resistance distribution, which is used to construct the resistance distribution map of the tunneling path. i , reflecting resistance changes in real time, ensuring that subsequent path planning is based on the latest data.

[0134] The tunneling path is abstracted into a graph structure consisting of nodes and edges. The edge weights are dynamically calculated using the resistance value and path smoothing factor to form a weighted graph. The calculation expression is as follows:

[0135] W ij =γ1(R i +R j )+γ2S ij , where W ij is the edge weight from node i to node j, R i is the resistance value of node i, R j is the resistance value of node j, γ1 is the weight factor of the resistance, which is used to control the resistance value R i and R j The proportion of edge weight, γ2 is the weight factor of the smoothing factor, which is used to control the path smoothness S ij The importance of S ij is the path smoothing factor, which is used to measure the path turning cost between node i and node j and indicates the smoothness of the path;

[0136] A graph is a mathematical abstraction consisting of nodes (vertices) and edges (connections). In this scenario, a graph is used to represent the excavation path, where nodes represent physical locations or working condition measurement points, and edges represent the connections between these nodes. This structure clearly describes the path's topology and facilitates path optimization and calculation.

[0137] A weighted graph is a numerically weighted graph structure, where each edge is assigned a weight that represents a characteristic or cost of the path. In excavation route optimization, weights are typically related to resistance and path smoothness, reflecting the difficulty of different paths. By calculating a weighted graph, data support can be provided for path planning, optimized scheduling, and other tasks, ensuring that the selected path better meets project requirements.

[0138] This step balances resistance and path smoothness in path planning, ensuring that the planned path avoids high-resistance areas while maintaining a reasonable propulsion angle.

[0139] Based on the dynamic weight graph, the improved Dijkstra algorithm is used to calculate the optimal path. The historical path weight and real-time adjustment factor are introduced to optimize the flexibility and accuracy of path planning. The formula is as follows:

[0140] Where, P * is the optimal path, P is a set of optional paths, each path includes all nodes from the starting point to the end point, H i is the historical path weight, which represents the average historical weight of node i, T ij is the real-time adjustment factor from node i to node j, which represents the rate of change of edge weight over time. λ1 is the historical weight adjustment factor, which controls the influence of historical paths on current path planning. λ2 is the adjustment parameter of the real-time adjustment factor, which is used to balance the influence of real-time data on path planning.

[0141] The algorithm combines historical experience and real-time data to enable path planning to dynamically adapt to complex geological conditions.

[0142] According to the calculated optimal path P * , drives the pipe jacking equipment to advance along the planned path, records the actual path point set and its force data in real time, and compares the actual path points with the optimal path P * , identifies path deviations and generates feedback, dynamically adjusts the weight map and resistance distribution model, and optimizes subsequent path planning. The formula is as follows:

[0143]

[0144] In the formula, ΔW ij is the dynamic adjustment of the edge weight from node i to node j, Factual，ij is the actual force value of the equipment passing through edge node i and node j, F model，ij is the force value of edge node i and node j predicted by the model, P actual is the actual path, η1 is the adjustment factor of the force deviation, which adjusts the degree of influence of the deviation between the actual force value and the model predicted force value on the edge weight adjustment, η2 is the adjustment factor of the force change rate, which adjusts the actual force change rate The degree of influence on the edge weight adjustment, η3 is the adjustment factor of path deviation, which adjusts the optimal path P * and the actual path P actual The influence of the deviation distance on the edge weight adjustment, dist(P * , P actual ) is the optimal path P * and the actual path P actual Deviation distance.

[0145] Actual path P actual Is the device along P * The result of advancement, but it may deviate due to uneven force or geological mutation. * and P actual , calculate the bias and dynamically adjust the weight W ij , optimize the resistance model, so that the subsequent path planning is more in line with the actual situation. The introduction of the path deviation indicator ensures that the algorithm can quickly adapt to the actual driving situation.

[0146] Through the neural network learning model, historical advancement data and current operation data are analyzed, and the correction control parameters are optimized in real time to further improve the equipment's correction efficiency and accuracy, ensuring the stability of the path during long-distance advancement.

[0147] The neural network learning model analyzes historical propulsion data and current operation data, optimizes the correction control parameters in real time, further improves the equipment's correction efficiency and accuracy, and ensures path stability during long-distance propulsion. The specific steps are as follows:

[0148] First, historical propulsion data and current operation data, including propulsion angle, propulsion speed, force state, and geological characteristic parameters, are collected. The obtained parameters are converted into the same numerical range (0 to 1) through normalization processing (Norm) to facilitate subsequent processing. Principal component analysis (PCA) is used to extract key features, remove redundant data, and reduce the dimension to generate the input feature matrix of the neural network. The generation formula is as follows:

[0149] X t =PCA(Norm([θ t , v t , F t , G t])), where X t It is the normalized and dimensionally reduced feature matrix used for subsequent neural network model input, F t is the force vector of the device, F t =[F ax , F rad , F shear ], where F ax是 Axial force, that is, the propulsion force of the equipment in the propulsion direction, F rad is the radial force, that is, the lateral friction force on both sides of the equipment, F shear is the shear force, i.e. the cutting resistance between the equipment and the soil, G t is the geological parameter vector, G t =[d, w, h], where d is the density, which reflects the compactness of the soil; w is the water content, which indicates the percentage of water in the soil to the total weight, affecting the friction and shear properties of the soil; h is the hardness, which indicates the soil's ability to resist damage; θ t is the current propulsion angle, v t is the current propulsion speed. Norm(·) is a normalization function that adjusts the numerical ranges of different parameters to a uniform range (for example, 0 to 1) to eliminate the influence of different parameter dimensions and amplitudes. PCA(·) is a principal component analysis function that performs dimensionality reduction on the normalized data to extract the main features that best explain the data changes and reduce data redundancy.

[0150] The generated normalized and dimensionally reduced feature matrix X t Input to the multi-layer feedforward neural network (MLP, Multi-Layer Perceptron), the neural network optimizes the model parameters by training historical data, predicts the correction control parameters, and in the real-time inference stage, inputs the normalized and reduced dimension feature matrix X t , calculate the correction control parameter vector, the calculation expression is as follows:

[0151] C t =f(W·X t +b), where W is the weight matrix of the neural network, b is the bias vector of the neural network, f is the activation function, and C t is the deviation correction control parameter, C t =[Δθ, ΔF, Δv], where Δθ is the propulsion angle adjustment, which is used to correct the direction of equipment deviation so that the equipment is close to the ideal excavation path; ΔF is the force adjustment, which is used to optimize the force distribution of the equipment and reduce the impact of asymmetric forces; and Δv is the speed adjustment, which is used to control the propulsion speed to adapt to complex geological conditions.

[0152] The correction parameter c predicted by the neural network tApplied to propulsion equipment, it adjusts the propulsion angle, speed, and force distribution in real time, and measures the equipment response results (actual propulsion state after adjustment). By comparing the equipment response results with the target state, the control error vector is calculated. The error vector is used to dynamically update the parameters of the neural network through the back propagation algorithm to optimize the model performance and adapt to the propulsion needs under complex geological conditions. The control error vector calculation expression is as follows:

[0153] E t =[θ target -(θ t +Δθ), F target -(F t +ΔF), v target -(v t +Δv)], where θ target is the target propulsion angle, F target is the target force, v target is the target advancement speed, E t is the control error vector, E t =[E θ , E F , E v ], where E θ is the angular error, E F is the force error, E v is the velocity error, which quantifies the deviation between the actual adjustment state and the target state;

[0154] Verify the stability of the adjusted propulsion path and calculate the correction efficiency index. The correction efficiency is used to evaluate the adjustment effect. If the efficiency meets the optimization goal, the model parameters are updated (to adapt to complex environments). The calculation formula for the correction efficiency is as follows:

[0155] Where η t Is the correction efficiency. The larger the value, the better the adjustment effect. t || 2 is the sum of squares of the error vector.

[0156] Implementation method one: This implementation method is based on real-time force monitoring, and realizes dynamic adjustment and precise deviation correction during the advancement of the equipment through multi-axis force sensors and multi-point distributed sensing devices installed on the pipe jacking equipment. First, multi-axis force sensors are installed at key parts of the equipment, such as the center of the cutter head, the side wall and the hydraulic jacking mechanism. These sensors can collect the axial force (propulsion force), radial force (lateral friction force) and shear force (cutting soil resistance) during the advancement of the equipment in real time. The sensor data is transmitted to the central control system via a wireless network, and is synchronously integrated with the geological parameters collected by the multi-point distributed sensing device to form a dynamic monitoring database.

[0157] Combining resistance distribution models with real-time data, the equipment's stress state is compared and analyzed with the soil resistance distribution, quickly identifying areas of asymmetric stress. Based on these deviations, the control system generates adjustment signals, such as the thrust angle adjustment range, the hydraulic jacking force distribution ratio, and the dynamic adjustment sequence. Upon receiving these signals, the hydraulic drive system precisely adjusts the hydraulic jacking force and direction to coarsely adjust the equipment's thrust angle and direction, thereby minimizing equipment deflection caused by soil resistance variations.

[0158] At the same time, when a significant anomaly in the resistance distribution is identified (such as a sudden increase in local resistance), the resistance balancing device is activated to further optimize the pressure distribution. For example, by injecting reinforcement slurry around the equipment or adjusting the pressure balance of the hydraulic top, the soil resistance is homogenized and the impact of asymmetric forces is reduced. The entire process is managed in a closed loop by the feedback control system, and the adjustment effect is evaluated in real time to ensure that the equipment always remains within the expected path. This implementation method is applicable to various geological conditions, especially in areas with significant soil heterogeneity or high geological complexity, and can significantly improve propulsion stability and construction efficiency.

[0159] Implementation Method 2: This method utilizes geological profiles and resistance distribution models for active pressure control, ensuring the stability of pipe jacking equipment in areas of nonlinear resistance. A multi-point distributed sensing system, deployed along the tunneling path, collects real-time data on soil density, moisture content, and hardness, generating an accurate geological profile. Combining soil mechanical properties with pipe jacking equipment propulsion parameters, a resistance distribution model is constructed to predict resistance trends in different soil layers, providing a basis for pressure equalization and control.

[0160] During excavation, if the equipment enters an area with significant soil resistance variations, the system activates the resistance equalization device for real-time control. This device dynamically adjusts the pressure distribution around the equipment using a hydraulic jack and a controlled grouting system. For example, it increases the thrust of the hydraulic jack in areas of high resistance while applying relatively low pressure in areas of low resistance, thereby evening out the soil resistance. Furthermore, the controlled grouting system automatically selects the appropriate slurry density and injection rate based on real-time data, further optimizing the resistance distribution by altering the soil structure.

[0161] The entire pressure equalization and control process is monitored by the control system, providing real-time feedback on the adjustment results. For example, multi-axis force sensors verify whether the soil resistance has reached the target uniformity and whether the propulsion path has returned to normal. If the control effect is unsatisfactory, the system will dynamically optimize the control parameters based on the feedback, such as redistributing the hydraulic jacking force or adjusting the grouting points. The advantage of this implementation is that it proactively identifies and addresses resistance changes, making it particularly suitable for construction scenarios involving long-distance jacking and complex geological conditions.

[0162] Implementation Method 3: This implementation method is based on an intelligent feedback control system, which uses machine learning algorithms such as neural networks to achieve adaptive control during the propulsion process. The system uses real-time force data, geological parameters, and resistance distribution models as input. By learning the propulsion behavior of the equipment under different geological conditions, it continuously optimizes the correction and control strategies. First, the data collected by the sensors (including axial force, radial force, and shear force) is compared with the geological profile and resistance model to identify asymmetric force areas and abnormal resistance changes.

[0163] The intelligent control system generates a multi-level correction strategy based on feedback data. For example, when the deviation is small, the hydraulic jack is used to fine-tune the propulsion angle. When the deviation is even smaller, the system uses the resistance equalizer to fine-tune the pressure distribution to ensure balanced propulsion force and prevent small deviations from accumulating into large errors. At the same time, the hydraulic jack can be used to fine-tune the propulsion angle for precise correction.

[0164] When deviations are large or resistance distribution is complex, the system prioritizes large-scale adjustments, such as adjusting the overall propulsion path or redistributing propulsion force, to quickly reduce the deviation and restore the device to normal propulsion. Furthermore, the system uses trend prediction algorithms to proactively identify potential high-resistance areas and optimize propulsion path planning, thus avoiding significant deviations caused by the device entering abnormal areas.

[0165] In addition, the system can also identify potential high-resistance areas in advance based on trend prediction algorithms and optimize propulsion path planning, thereby avoiding major deviations caused by equipment entering abnormal areas.

[0166] Neural network models play a key role in this process, gradually improving control accuracy through training and analysis of historical and real-time data. For example, the system can automatically adjust hydraulic thrust distribution or grouting parameters for similar soil characteristics, significantly improving the equipment's propulsion efficiency under similar conditions. The system's closed-loop control design also ensures that each adjustment receives immediate feedback and optimizes the next step, achieving dynamic self-adaptation throughout the process.

[0167] The highlight of this implementation method lies in its efficient learning ability and control flexibility, especially in long-distance excavation and complex geological conditions. By continuously accumulating empirical data, the propulsion performance of the jacking equipment is gradually optimized, providing smarter and more reliable technical support for construction.

[0168] Through the synergistic effects of real-time force monitoring, multi-point distributed sensing, and intelligent feedback control systems, the present invention significantly improves the accuracy and construction efficiency of the jacking tunneling path. The equipment can quickly identify areas of abnormal resistance and automatically generate adjustment plans. It achieves dynamic regulation through the hydraulic drive system and resistance balancing device, reducing the interference of asymmetric forces on the propulsion path. Compared with the traditional method that relies on manual monitoring and manual adjustment, this intelligent solution significantly reduces human errors, avoids repeated construction and equipment loss, and improves construction quality and efficiency. At the same time, it reduces downtime and the need for manual operation, speeds up project progress, and provides technical support for long-distance jacking tunneling.

[0169] This invention significantly improves the equipment's ability to adapt to complex geological conditions. Combined with real-time analysis of geological profiles and resistance distribution models, pipe jacking equipment can predict and proactively respond to changes in resistance in heterogeneous soil layers or weak interlayers. The resistance balancing device effectively resolves equipment jams and advancement difficulties caused by sudden changes in resistance by dynamically adjusting the pressure distribution, thereby ensuring construction safety and continuity. This technical capability not only reduces construction risks and costs in complex geological environments, but also makes pipe jacking technology more widely applicable, providing reliable protection for pipe jacking construction in high-risk scenarios.

[0170] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0171] The above description is merely illustrative of certain exemplary embodiments of the present invention. It goes without saying that those skilled in the art will be able to modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the scope of protection of the claims.

[0172] It should be noted that, in this document, if there are relational terms such as first and second, etc., they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprises", "comprising" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or device that includes the element.

[0173] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0174] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0175] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0176] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0177] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0178] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0179] The above description is merely illustrative of certain exemplary embodiments of the present invention. Those skilled in the art will appreciate that the described embodiments may be modified in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and description are illustrative in nature and should not be construed as limiting the scope of the claims.

Claims

1. A method for automatically correcting the deviation of a long-distance jacking tunnel, characterized in that: The following steps are involved: Multi-point distributed sensing devices are used to collect data on the density, moisture content, and hardness of the soil along the excavation path in real time. Combined with soil parameters and the mechanical properties of the pipe jacking equipment, a geological profile is generated, a resistance distribution model is constructed, and resistance change trends under different geological conditions are predicted, establishing a correction and early warning mechanism. Install multi-axis force sensors on the pipe jacking equipment to monitor the force status of the equipment in real time during the advancement process. Compare the measured data with the resistance distribution model to identify abnormal resistance areas. The hydraulic drive system controls the thrust angle and direction of the pipe jacking equipment, and the equipment posture is roughly adjusted based on monitoring data to reduce the impact of asymmetric forces on the tunneling path. When the nonlinear resistance change area is detected, the resistance balancing device is activated to achieve uniform resistance by adjusting the pressure distribution around the top pipe head, thereby reducing the impact of asymmetric force on the propulsion of the equipment; Based on real-time monitoring data, an improved Dijkstra algorithm is used to dynamically adjust the tunneling path, avoid areas of abnormal resistance, and ensure the smoothness of the advancement path and the accuracy of the correction operation; Through the neural network learning model, historical advancement data and current operation data are analyzed, and the correction control parameters are optimized in real time to further improve the equipment's correction efficiency and accuracy, ensuring the stability of the path during long-distance advancement.

2. The method for automatically correcting the long-distance jacking of a jacking tunnel according to claim 1 is characterized in that: The density, moisture content, and hardness data of the soil along the tunneling path are collected in real time using multi-point distributed sensing devices. Combined with soil parameters and the propulsion mechanical properties of the pipe jacking equipment, a resistance distribution model is constructed to generate a geological condition profile. The specific steps for predicting resistance change trends under different geological conditions are as follows: In key areas of the tunneling path, sensors are rationally placed on the tunneling equipment, and the equipment is calibrated to achieve real-time data transmission via wireless networks; Use high-frequency sampling technology to collect geological data in real time, and use filtering and correction algorithms to ensure data accuracy and consistency; Through GIS modeling and spatial interpolation algorithms, a visual geological profile is generated to fully display the distribution of geological parameters within the path; Based on the geological profile and pipe jacking equipment parameters, a resistance distribution model is constructed and the resistance change trend is predicted, and a correction and early warning mechanism is established to provide support for promoting adjustments.

3. The method for automatically correcting the long-distance jacking of a jacking tunnel according to claim 1 is characterized in that: Install a multi-axis force sensor on the pipe jacking equipment to monitor the force status of the equipment in real time during the advancement process. Compare the measured data with the resistance distribution model to identify abnormal resistance areas. The specific steps are as follows: Select high-precision multi-axis force sensors and arrange them reasonably at the key stress-bearing parts of the pipe jacking equipment. If necessary, reinforce the installation to ensure the accuracy and stability of data acquisition; Axial force, radial force and shear force data are recorded through high-frequency sampling and transmitted synchronously with geological parameters to the central control system to form a dynamic database; Compare and analyze real-time force data with the resistance distribution model to locate abnormal areas and trigger correction warnings; Identify abnormal resistance areas, adjust the propulsion posture through the correction control system and dynamically optimize the strategy to ensure the stability and effectiveness of the propulsion path.

4. The method for automatically correcting the deviation of a long-distance jacking tunnel according to claim 1 is characterized in that: The specific steps for using the hydraulic drive system to control the thrust angle and direction of the pipe jacking equipment and making rough adjustments to the equipment posture based on monitoring data to reduce the impact of asymmetric forces on the tunneling path are as follows: Based on the geological profile and resistance distribution model, the hydraulic system was initialized and tested multiple times to ensure thrust accuracy and equipment stability. Analyze the deviation between sensor data and resistance model in real time, generate coarse adjustment signals for propulsion angle and direction, and formulate adjustment strategies; The hydraulic jack adjusts the thrust and direction to perform rough adjustments in real time, while monitoring the adjustment effect and inputting feedback data to ensure immediacy; Through posture verification and data analysis, the optimization and adjustment strategy is carried out, and combined with the prediction algorithm, the stability and efficiency of the subsequent propulsion path are improved.

5. The method for automatically correcting the long-distance jacking of a jacking tunnel according to claim 1 is characterized in that: The specific steps for identifying the nonlinear resistance change area through multi-sensor collaboration, determining the starting conditions and activating the resistance balancing device, and setting the initial working state based on the geological profile and resistance model are as follows: Collect geological and force data through sensors, compare them with the resistance distribution model, and analyze and locate areas with nonlinear resistance changes; Combined with pattern recognition algorithms to determine the characteristics of abnormal areas, providing a basis for subsequent regulation; Determine whether the conditions for starting the balancing device are met based on the asymmetric force and resistance distribution characteristics, and generate a start signal; Activate the resistance balancing device and set the initial working state to ensure the accuracy and reliability of the control starting point.

6. The method for automatically correcting the long-distance jacking of a jacking tunnel according to claim 1 is characterized in that: The specific steps for using the resistance balancing device to adjust the pressure distribution around the jacking head in real time, verify the control effect and optimize the strategy to ensure uniform force and path stability during the advancement process are as follows: Adjust the pressure distribution through the resistance balancing device, dynamically equalize the resistance, and reduce the impact of asymmetric force; Monitor the equipment's stress state and propulsion angle in real time to assess whether the control effect has achieved the expected goal; Optimize control parameters and strategies based on feedback results to improve the device's ability to adapt to complex geological conditions; After completing the balancing, the device is restored to the standard state and the control data is stored to provide a reference for subsequent operations.

7. The method for automatically correcting the deviation of a long-distance jacking tunnel according to claim 1 is characterized in that: Based on real-time monitoring data, the improved Dijkstra algorithm is used to dynamically adjust the tunneling path to avoid areas of abnormal resistance and ensure the smoothness of the advancement path and the accuracy of the correction operation. The specific steps are as follows: Using multi-point distributed sensing devices and multi-axis force sensors, we collect resistance-related parameters at each node in the tunneling path, including soil density, moisture content, and hardness, as well as the axial and radial forces acting on the equipment. The acquired data is used to calculate the node resistance value and construct a resistance distribution map. The calculation expression is as follows: Where R i is the resistance value of the ith node, d i is the soil density of the i-th node, w i is the soil moisture content at the i-th node, h i is the soil hardness of the i-th node, α1 is the density weight factor, α2 is the water content weight factor, α3 is the hardness weight factor, F a，i is the axial force at the ith node, F r，i is the radial force at the i-th node, A is the cross-sectional area of the pipe jacking equipment, P is the circumference of the pipe jacking equipment, β1 is the weight factor of the axial force, and β2 is the weight factor of the radial force; The tunneling path is abstracted into a graph structure consisting of nodes and edges. The edge weights are dynamically calculated using the resistance value and path smoothing factor to form a weighted graph. The calculation expression is as follows: W ij =γ1(R i +R j )+γ2S ij , where W ij is the edge weight from node i to node j, R i is the resistance value of node i, R j is the resistance value of node j, γ1 is the weight factor of resistance, γ2 is the weight factor of smoothing factor, S ij is the path smoothing factor, which is used to measure the path turning cost between node i and node j and indicates the smoothness of the path.

8. The method for automatically correcting the deviation of a long-distance jacking tunnel according to claim 7, characterized in that: Based on the dynamic weight graph, the improved Dijkstra algorithm is used to calculate the optimal path. The historical path weight and real-time adjustment factor are introduced to optimize the flexibility and accuracy of path planning. The formula is as follows: Where, P * is the optimal path, P is the set of optional paths, H i is the historical path weight, T ij is the real-time adjustment factor from node i to node j, λ1 is the historical weight adjustment factor, and λ2 is the adjustment parameter of the real-time adjustment factor; According to the calculated optimal path P * , drives the pipe jacking equipment to advance along the planned path, records the actual path point set and its force data in real time, and compares the actual path points with the optimal path P * , identifies path deviations and generates feedback, dynamically adjusts the weight map and resistance distribution model, and optimizes subsequent path planning. The formula is as follows: In the formula, ΔW ij is the dynamic adjustment of the edge weight from node i to node j, F actual，ij is the actual force value of the equipment passing through edge node i and node j, F model，ij is the force value of edge node i and node j predicted by the model, P actual is the actual path, η1 is the adjustment factor of force deviation, η2 is the adjustment factor of force change rate, η3 is the adjustment factor of path deviation, dist(P * , P actual ) is the optimal path P * and the actual path P actual Deviation distance.

9. The method for automatically correcting the deviation of a long-distance jacking tunnel according to claim 1, characterized in that: The neural network learning model analyzes historical propulsion data and current operation data, optimizes the correction control parameters in real time, further improves the equipment's correction efficiency and accuracy, and ensures path stability during long-distance propulsion. The specific steps are as follows: First, historical propulsion data and current operation data, including propulsion angle, propulsion speed, force state, and geological characteristic parameters, are collected. The acquired parameters are converted into the same numerical range through normalization processing. Key features are extracted using principal component analysis to generate the input feature matrix of the neural network. The generation formula is as follows: X t =PCA(Norm([θ t , v t , F t , G t ])), where X t is the feature matrix after normalization and dimensionality reduction, F t is the force vector of the device, F t =[F ax , F rad , F shear ], where F ax is the axial force, F rad is the radial force, F shear is the shear force, G t is the geological parameter vector, G t =[d, w, h], where d is density, w is water content, h is hardness, θ t is the current propulsion angle, v t is the current propulsion speed, Norm(·) is the normalization function, and PCA(·) is the principal component analysis function; The generated normalized and dimensionally reduced feature matrix X t Input to the multi-layer feedforward neural network, the neural network optimizes the model parameters by training historical data, predicts the correction control parameters, and in the real-time inference stage, inputs the normalized and dimension-reduced feature matrix X t , calculate the correction control parameter vector, the calculation expression is as follows: C t =f(W·X t +b), where W is the weight matrix of the neural network, b is the bias vector of the neural network, f is the activation function, and C t It is the correction control parameter.

10. The method for automatically correcting the deviation of a long-distance jacking tunnel according to claim 9, characterized in that: The correction parameter C predicted by the neural network t Applied to propulsion equipment, it adjusts the propulsion angle, speed, and force distribution in real time, measures the equipment response results, and calculates the control error vector by comparing the equipment response results with the target state. The error vector is used to dynamically update the parameters of the neural network through the back propagation algorithm to optimize the model performance and adapt to the propulsion needs under complex geological conditions. The control error vector calculation expression is as follows: E t =[θ target -(θ t +Δθ), F target -(F t +ΔF), v target -(v t +Δv)], where θ target is the target propulsion angle, F target is the target force, v target is the target advancement speed, E t is the control error vector, E t =[E θ , E F , E v ], where E θ is the angular error, E F is the force error, E v is the velocity error; Verify the stability of the adjusted propulsion path and calculate the correction efficiency index. The correction efficiency is used to evaluate the adjustment effect. If the efficiency meets the optimization goal, the model parameters are updated to adapt to the complex environment. The calculation formula of the correction efficiency is as follows: Where η t is the correction efficiency, ||E t || 2 is the sum of squares of the error vector.

Images