Intelligent steering method for shield tunneling in volcanic ash stratum

Abstract

The application relates to the technical field of shield tunneling, and discloses an intelligent direction adjusting method for shield tunneling in a volcanic ash stratum. The method comprises the following steps: collecting construction data of a shield tunneling machine and a target region through a multi-source sensor, preprocessing the construction data, fusing the preprocessed construction data to obtain real-time monitoring data; acquiring historical tunneling error data of the target region, analyzing the historical tunneling error data and the real-time monitoring data by using a deep learning model, and generating a tunneling direction adjustment instruction; adjusting the advancing speed and the steering angle of the shield tunneling machine and the thrust distribution of a jacking oil cylinder according to the tunneling direction adjustment instruction; generating a mud cake prevention instruction according to the real-time monitoring data, adjusting the cleaning parameters of a mud cake cleaning device according to the mud cake prevention instruction, and displaying the running state and the tunneling direction of the adjusted shield tunneling machine through a man-machine interaction interface. The application solves the control problem of the tunneling direction of the shield tunneling machine, and realizes real-time regulation and control of the advancing speed and the steering angle.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of shield tunneling, and in particular to an intelligent steering method for shield tunneling in a volcanic ash stratum. BACKGROUND

[0002] With the rapid development of urban infrastructure construction, the demand for large underground projects such as subways and tunnels is increasing. Shield tunneling has become one of the main tunneling methods due to its safety, high efficiency, and minimal disturbance to the ground.

[0003] However, in complex geological conditions such as volcanic ash strata, the traditional shield tunneling method faces many challenges due to uneven stratum structure signals, fine and loose particles, and high water content. In particular, it is difficult to control the direction and easy to form mud cake, which affects the tunneling efficiency and construction safety. Current intelligent steering technology is mostly aimed at ordinary strata and lacks targeted optimization for the special properties of volcanic ash strata. SUMMARY

[0004] Therefore, the present application aims to overcome the shortcomings of the prior art and provide an intelligent steering method for shield tunneling in a volcanic ash stratum.

[0005] The present application provides the following technical solutions:

[0006] In a first aspect, the present application provides an intelligent steering method for shield tunneling in a volcanic ash stratum, comprising:

[0007] Collecting construction data of a shield tunneling machine and a target area using multiple source sensors, preprocessing the construction data, and fusing the preprocessed construction data to obtain real-time monitoring data;

[0008] Obtaining historical tunneling error data of the target area, analyzing the historical tunneling error data and the real-time monitoring data using a deep learning model, and generating a tunneling direction adjustment instruction;

[0009] Adjusting the propulsion speed and steering angle of the shield tunneling machine according to the tunneling direction adjustment instruction, and adjusting the thrust distribution of multiple jacking cylinders in the shield tunneling machine according to the tunneling direction adjustment instruction;

[0010] Generating a mud cake prevention instruction according to the real-time monitoring data, adjusting the cleaning parameters of a mud cake cleaning device according to the mud cake prevention instruction, and displaying the running state and tunneling direction of the adjusted shield tunneling machine in real time through a human-machine interaction interface.

[0011] In an optional implementation, the multi-source sensor includes a posture sensor, a geological radar, a camera, a mud parameter sensor, the construction data includes a tilt angle, an acceleration, a stratum structure signal, a tunneling face image, pressure data, flow data, viscosity data, and stratum humidity data, the construction data of the target region is collected by the multi-source sensor, and the construction data includes:

[0012] The installation position of the multi-source sensor on the shield machine is optimized by using a genetic algorithm;

[0013] The tilt angle and the acceleration of the shield machine are collected by the posture sensor, and the stratum structure signal of the target region is collected by the geological radar;

[0014] The tunneling face image of the target region is collected by the camera, and the mud parameters of the target region are collected by the mud parameter sensor, including at least one of pressure data, flow data, viscosity data, and stratum humidity data.

[0015] In an optional implementation, the construction data is preprocessed, and the preprocessed construction data is fused to obtain real-time monitoring data, including:

[0016] The tilt angle and the acceleration of the shield machine are smoothed by using a Kalman filtering algorithm to obtain a preprocessed tilt angle and a preprocessed acceleration;

[0017] The stratum structure signal of the target region is analyzed by using a wavelet transform algorithm to obtain a preprocessed stratum structure signal;

[0018] The tunneling face image of the target region is analyzed by using a convolutional neural network to obtain a preprocessed tunneling face image;

[0019] The mud parameters of the target region are classified and predicted by using a support vector machine to obtain preprocessed mud parameters;

[0020] The preprocessed tilt angle, the preprocessed acceleration, the preprocessed stratum structure signal, the preprocessed tunneling face image, and the preprocessed mud parameters are fused by using an extended Kalman filtering algorithm to generate the real-time monitoring data.

[0021] In an optional implementation, the preprocessed stratum structure signal includes rock, pore, moisture, and stratum heterogeneity, the stratum structure signal of the target region is analyzed by using a wavelet transform algorithm to obtain a preprocessed stratum structure signal, including:

[0022] A target wavelet function is determined according to the frequency characteristics and the target resolution of the stratum structure signal.

[0023] wavelet decomposing the stratum structure signal by using the target wavelet function to obtain wavelet components, including low-frequency components and high-frequency components;

[0024] filtering out noise signals higher than a preset frequency threshold in the high-frequency components, identifying rocks, pores and water in the target region by analyzing mutation points and intensity of the denoised high-frequency components, and identifying stratum heterogeneity in the target region by analyzing fluctuations of the low-frequency components.

[0025] In an optional embodiment, the analysis of the historical tunneling error data and the real-time monitoring data by using the deep learning model to generate the tunneling direction adjustment instruction further includes:

[0026] regression analysis of the historical tunneling error data by using a support vector regression model to determine a tunneling direction deviation trend and a tunneling direction deviation factor;

[0027] comprehensive analysis of geological parameters of the stratum structure signal by using a Bayesian network to determine an influence weight of each geological layer section in the target region on the tunneling direction;

[0028] determination of thrust distribution of each jacking oil cylinder in the shield machine by using a fuzzy adaptive control algorithm and according to the historical tunneling error data and the stratum structure signal;

[0029] processing of the real-time monitoring data, the tunneling direction deviation trend, the tunneling direction deviation factor, the influence weight of each geological layer section in the target region on the tunneling direction, and the thrust distribution of each jacking oil cylinder in the shield machine by using the deep learning model to generate the tunneling direction adjustment instruction.

[0030] In an optional embodiment, the regression analysis of the historical tunneling error data by using the support vector regression model to determine the tunneling direction deviation trend and the tunneling direction deviation factor includes:

[0031] regression analysis of the historical tunneling error data by using the support vector regression model to generate an error prediction curve;

[0032] identification of a linear trend, periodic fluctuations and mutation points of tunneling displacement errors according to the error prediction curve to obtain the tunneling direction deviation trend;

[0033] The additional data collected by the additional sensors is acquired, correlation analysis is performed on the additional data and the historical tunneling error data, basic factors having a correlation with the tunneling displacement error are determined, and a support vector regression model is used to determine target factors in the basic factors having a correlation greater than a preset correlation threshold with the tunneling displacement error, so as to obtain the tunneling direction deviation factor.

[0034] In an optional implementation, the geological parameters of the stratum structure signal are comprehensively analyzed by using the Bayesian network, and an influence weight of each geological layer section in the target region on the tunneling direction is determined, including:

[0035] Nodes of the Bayesian network are defined according to the geological parameters, and conditional dependency relationships and conditional probability tables between the nodes are defined according to the historical tunneling error data, wherein the conditional probability tables are used to describe probability distribution of a node under a given condition of parent nodes of the node;

[0036] The geological parameters are input into the Bayesian network, and probabilities of tunneling direction deviation of the geological parameters under a given condition are calculated;

[0037] According to the probabilities of tunneling direction deviation of the geological parameters under a given condition and influence degrees of the tunneling direction deviations on the tunneling direction, the influence weight of each geological layer section in the target region on the tunneling direction is calculated.

[0038] In an optional implementation, the thrust distribution of each jacking oil cylinder in the shield machine is determined by using a fuzzy adaptive control algorithm and according to the historical tunneling error data and the stratum structure signal, including:

[0039] Geological parameters of the stratum structure signal, the historical tunneling error data and real-time tunneling error data are converted into fuzzy sets, and ranges of the fuzzy sets are defined by using a preset membership function;

[0040] Fuzzy rules are determined according to the historical tunneling error data, the fuzzy adaptive control algorithm is used to apply the fuzzy rules to the fuzzy sets, and fuzzy thrust distributions of the jacking oil cylinders are calculated;

[0041] The fuzzy thrust distributions of the jacking oil cylinders are defuzzified by using a defuzzification algorithm, so as to obtain the thrust distributions of the jacking oil cylinders.

[0042] In an optional embodiment, the deep learning model comprises a convolutional neural network module, a long short-term memory network module, and a comprehensive output module, the real-time monitoring data, the tunneling direction deviation trend, the tunneling direction deviation factors, the influence weight of each geological section in the target area on the tunneling direction, and the thrust distribution of each jacking oil cylinder in the shield machine are processed by the deep learning model to generate the tunneling direction adjustment instruction, comprising:

[0043] The spatial feature data of the real-time monitoring data is extracted by the convolutional neural network module;

[0044] The time correlation of the historical tunneling error data and the real-time monitoring data is analyzed by the long short-term memory network module to obtain time series data;

[0045] The prediction result is generated by combining the spatial feature data and the time series data by the comprehensive output module, and the prediction result is used to represent the prediction of the stratum change trend and the deviation risk;

[0046] According to the prediction result, the target tunneling direction adjustment amount is calculated, and the PID control algorithm is adopted to combine the target tunneling direction adjustment amount, the tunneling direction deviation trend, the tunneling direction deviation factors, the influence weight of each geological section in the target area on the tunneling direction, and the thrust distribution of each jacking oil cylinder in the shield machine to generate the tunneling direction adjustment instruction.

[0047] In an optional embodiment, the mud cake prevention instruction is generated according to the real-time monitoring data, and the removal parameters of the mud cake removal device are adjusted according to the mud cake prevention instruction, comprising:

[0048] According to the tunneling face image, the mud cake position, mud cake thickness, mud cake distribution range, and mud cake physical properties are identified;

[0049] According to the mud cake position, mud cake thickness, mud cake distribution range, mud cake physical properties, and mud parameters, the mud cake prevention instruction is generated;

[0050] According to the mud cake prevention instruction, the removal strength and frequency of the mud cake removal device are adjusted.

[0051] The beneficial effects of the present application are:

[0052] The volcano ash stratum shield tunneling intelligent direction adjusting method provided by the embodiment of the present application comprehensively utilizes multi-source sensor data, historical tunneling error analysis, uneven geological distribution in the volcano ash stratum, thrust distribution of different jacking cylinders and advanced intelligent algorithms, solves the shield tunneling direction control problem, realizes real-time regulation and control of the advancing speed and the turning angle, and ensures accurate tunneling and construction safety of the shield tunneling machine under complex geological conditions.

[0053] In order to make the above object, features and advantages of the present application more obvious and easy to understand, the following preferred embodiments are specifically described below, and the accompanying drawings are referred to for detailed description. BRIEF DESCRIPTION OF DRAWINGS

[0054] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can also be obtained without creative labor. In each drawing, similar components are marked with similar reference numerals.

[0055] Figure 1 A flow chart of a volcano ash stratum shield tunneling intelligent direction adjusting method provided by the embodiment of the present application is shown;

[0056] Figure 2 A flow chart of another volcano ash stratum shield tunneling intelligent direction adjusting method provided by the embodiment of the present application is shown;

[0057] Figure 3 A flow chart of still another volcano ash stratum shield tunneling intelligent direction adjusting method provided by the embodiment of the present application is shown. DETAILED DESCRIPTION

[0058] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the drawings, wherein the same or similar reference numerals represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary and are only used to explain the present application, and cannot be understood as a limitation on the present application.

[0059] The terms "first", "second" are only used for description purpose, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "multiple" is two or more, unless otherwise specifically limited.

[0060] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein in the template description is for the purpose of describing particular embodiments only and is not intended to limit the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0061] Example 1

[0062] like Figure 1 The diagram shown is a flowchart of a smart tunneling method for volcanic ash formations according to an embodiment of this application. The smart tunneling method for volcanic ash formations provided in this embodiment includes the following steps:

[0063] Step S110: Collect construction data of the tunnel boring machine and the target area through multi-source sensors, preprocess the construction data, and fuse the preprocessed construction data to obtain real-time monitoring data.

[0064] In this embodiment, multi-source sensors are installed at the front end and key parts of the tunnel boring machine (TBM) to monitor the construction data of the TBM and the target area in real time. These multi-source sensors include, but are not limited to, attitude sensors, ground-penetrating radar, cameras, and mud parameter sensors. The construction data includes, but is not limited to, tilt angle, acceleration, ground structure signals, tunnel face images, pressure data, flow rate data, viscosity data, and ground moisture data.

[0065] Preferably, a genetic algorithm (GA) can be used to optimize the arrangement of multi-source sensors at the front end and key parts of the tunnel boring machine, ensuring the comprehensiveness and real-time nature of sensor data and improving data acquisition efficiency.

[0066] Specifically, attitude sensors (gyroscopes, accelerometers) are used to collect the tilt angle and acceleration of the tunnel boring machine to ensure stability during the tunneling process; ground-penetrating radar is used to collect geological structure signals of the target area, including rock, pores, moisture, and geological inhomogeneity, providing accurate geological basis for intelligent orientation adjustment; cameras are used to collect real-time images of the tunnel face of the target area to help identify geological changes, potential obstacles, and mud cake formation, improving the accuracy and efficiency of image recognition; mud parameter sensors are used to collect mud parameters of the target area, including at least one of pressure data, flow data, viscosity data, and ground moisture data, to detect signs of mud cake formation in a timely manner and ensure the normal operation of the slurry shield tunneling system.

[0067] Alternatively, in one implementation, the sensors can be calibrated periodically to ensure the accuracy and reliability of the data. The calibration process may include static calibration and dynamic calibration to eliminate errors and drift inherent in the sensors themselves.

[0068] The above process optimizes the sensor arrangement through a genetic algorithm, ensures the comprehensiveness and real-time nature of data acquisition, reduces monitoring blind spots, and improves data acquisition efficiency. The construction data covers multiple dimensions of information such as shield machine posture, geological structure, tunneling face image, and mud parameters, providing a high-quality data foundation for subsequent analysis.

[0069] Further, the Kalman Filter algorithm is used to smooth the inclination angle and acceleration of the shield machine to obtain preprocessed inclination angle and preprocessed acceleration, eliminate noise interference, and improve the accuracy of shield machine stability monitoring. The Wavelet Transform algorithm is used to perform multi-resolution analysis on the stratum structure signal of the target area to obtain preprocessed stratum structure signal. The CNN is used to perform image analysis on the tunneling face image of the target area to obtain preprocessed tunneling face image, improving the accuracy and efficiency of image recognition. The SVM is used to classify and predict the mud parameters of the target area to obtain preprocessed mud parameters, ensuring the normal operation of the mud system.

[0070] In an alternative embodiment, as shown in Figure 2 The Wavelet Transform algorithm is used to perform multi-resolution analysis on the stratum structure signal of the target area to obtain preprocessed stratum structure signal, including steps S111-S113:

[0071] Step S111, determine the target wavelet function according to the frequency characteristics of the stratum structure signal and the target resolution;

[0072] Step S112, use the target wavelet function to perform wavelet decomposition on the stratum structure signal to obtain wavelet components, including low-frequency components and high-frequency components;

[0073] Step S113, filter out noise signals above the preset frequency threshold in the high-frequency components, identify the rock, pore, and water in the target area by analyzing the mutation points and intensity of the denoised high-frequency components, and identify the stratum heterogeneity in the target area by analyzing the fluctuations of the low-frequency components.

[0074] Understandably, first, the target wavelet function (such as Daubechies wavelet, Haar wavelet, etc.) is determined according to the frequency characteristics of the stratum structure signal and the target resolution, then the stratum structure signal is decomposed by the target wavelet function, the stratum structure signal is decomposed into wavelet components of different scales through continuous or discrete wavelet transform, including low-frequency components and high-frequency components, the low-frequency components usually contain the overall trend information of the signal, which can be used to identify large-scale stratum heterogeneity, and the high-frequency components reflect local changes, which are suitable for identifying the presence of rocks, pores or water.

[0075] Then, a threshold filtering method is used to filter out noise signals higher than the preset frequency threshold in the high-frequency components, and only significant geological feature signals are retained. Further, by analyzing the mutation points and intensity of the denoised high-frequency components, the rock, pore and water features in the target area are accurately identified by using a pattern recognition algorithm (such as support vector machine, neural network, etc.), and by analyzing the fluctuations of the low-frequency components, the stratum heterogeneity features in the target area are accurately identified by using a pattern recognition algorithm.

[0076] It should be noted that the preset frequency threshold can be adjusted according to experiments and actual stratum conditions, and the parameters of each multi-source sensor and the parameters of each algorithm model can also be corrected regularly to adapt to different geological condition changes, and the embodiments of the present application do not limit this.

[0077] The wavelet transform in the above process (steps S111-S113): through multi-resolution analysis, the low-frequency and high-frequency features of the stratum structure signal are extracted, and large-scale stratum heterogeneity (low-frequency) and local geological details (high-frequency) are accurately identified. Filter high-frequency noise and detect mutation points to enhance the recognition ability of geological features such as rocks, pores and water, and provide high-precision geological basis for direction adjustment.

[0078] Finally, the preprocessed inclination angle, the preprocessed acceleration, the preprocessed stratum structure signal, the preprocessed tunneling face image and the preprocessed mud parameters are fused by using an extended Kalman filter algorithm (EKF), the stratum structure, image, parameter and current state of the shield machine are fused, and unified real-time monitoring data is generated to enhance the comprehensive perception ability of the system to complex geological environment.

[0079] Step S110 above utilizes multi-source sensors to comprehensively collect construction data from the tunnel boring machine and the target area, ensuring the comprehensiveness and real-time nature of the data. A genetic algorithm is used to optimize sensor placement, improving data acquisition efficiency. Simultaneously, algorithms such as Kalman filtering, wavelet transform, convolutional neural networks, and support vector machines are used to preprocess the data, improving its accuracy and reliability. An extended Kalman filter algorithm is then used to fuse the preprocessed data, generating unified real-time monitoring data to provide a foundation for subsequent analysis.

[0080] Step S120: Obtain historical tunneling error data of the target area, analyze the historical tunneling error data and the real-time monitoring data using a deep learning model, and generate tunneling direction adjustment instructions.

[0081] In one alternative implementation, such as Figure 3 As shown, step S120 includes steps S121 to S124:

[0082] Step S121: Use a support vector regression model to perform regression analysis on the historical tunneling error data to determine the tunneling direction deviation trend and tunneling direction deviation factors.

[0083] Understandably, the historical tunneling error data is first analyzed using a Support Vector Regression (SVR) model to generate an error prediction curve. The changing trend of the error prediction curve is observed to identify the linear trend, periodic fluctuations, and abrupt changes in the tunneling displacement error, thus obtaining the trend of tunneling direction deviation. Specifically, if the tunneling displacement error shows a linear increase over time, it may indicate a persistent source of deviation in the system, such as equipment wear or uneven thrust distribution; if the tunneling displacement error shows periodic changes, it may be related to periodic geological features (such as changes in formation density) or equipment operating cycles; if there are abrupt changes in the tunneling displacement error on the error prediction curve, it may be due to sudden changes in geological conditions (such as entering hard rock formations) or equipment failure.

[0084] Next, additional data (such as thrust, vibration, and formation density) collected by additional sensors are acquired. Correlation analysis is performed on the additional data and historical tunneling error data to identify basic factors such as geology or equipment that are related to changes in tunneling displacement error. Then, through feature weight analysis of the support vector regression model, target factors among the basic factors whose correlation with tunneling displacement error is greater than a preset correlation threshold are determined. These are the main factors that have a significant impact on tunneling displacement error. For example, changes in formation properties may be the main factor leading to an increase in tunneling displacement error.

[0085] Optionally, the embodiment can use charts (such as error trend chart, deviation heat map) to intuitively display the tunneling displacement error change and possible factors, and evaluate the prediction accuracy of the support vector regression model and the actual impact of directional deviation by comparing the difference between the actual tunneling path and the planned path.

[0086] The above step S121 can effectively identify the linear growth, periodic fluctuation and mutation point of the tunneling direction deviation, and reveal potential problems such as equipment wear and tear and stratum change. Through feature weight analysis, the main error source (such as stratum property change) is located, and targeted basis is provided for the steering strategy.

[0087] In step S122, the geological parameters of the stratum structure signal are comprehensively analyzed by using the Bayesian network to determine the influence weight of each geological layer section in the target area on the tunneling direction.

[0088] Understandably, in the embodiment, the geological parameters of the stratum structure signal include density, porosity, water content, compressive strength and stratum thickness. First, a Bayesian network is constructed, each geological parameter represents a node, and the conditional dependence relationship between nodes is defined according to historical tunneling error data and geological knowledge, for example, water content affects porosity, and porosity affects the compressive strength of the stratum. According to the historical tunneling error data and expert knowledge, the conditional probability table is defined to describe the probability distribution of a node under the given condition of its parent node.

[0089] Then, input each geological parameter monitored in real time into the Bayesian network, and use Bayesian inference algorithms (such as Gibbs sampling, variational inference, etc.) to calculate the probability of tunneling direction deviation under the given condition of each parameter. Further, the influence weight of each geological layer section on the tunneling direction is calculated by using weighted average or other statistical methods in combination with the probability of tunneling direction deviation under the given condition of each geological parameter and the influence degree of each tunneling direction deviation on the tunneling direction. The formula is as follows:

[0090]

[0091] In the formula, wi is the influence weight of the i-th geological layer section on the tunneling direction, pi is the probability of tunneling direction deviation under the given condition of the geological parameter, wi is the influence weight of the i-th geological layer section on the tunneling direction, pi is the probability of tunneling direction deviation under the given condition of the geological parameter, wi is the influence weight of the i-th geological layer section on the tunneling direction, pi is the probability of tunneling direction deviation under the given condition of the geological parameter,

[0092] Optionally, in the embodiment, the structure and conditional probability table of the Bayesian network can be updated regularly to adapt to new geological information and tunneling experience. Through the feedback mechanism, the prediction accuracy and applicability of the model are gradually improved.

[0093] ​The step S122 quantifies the influence weight of each geological section (density, porosity, etc.) on the tunneling direction based on the conditional probability table and the real-time geological parameters. The expert knowledge is combined to dynamically update the network, improve the adaptability of the model to complex geological conditions, and enhance the scientificity of the steering decision.

[0094] In step S123, a fuzzy adaptive control algorithm is used to determine the thrust distribution of each jacking cylinder in the shield machine according to the historical tunneling error data and the stratum structure signal.

[0095] Understandably, the geological parameters, historical tunneling error data, and real-time tunneling error data are converted into fuzzy sets. For example, the stratum density in the geological parameters can be divided into three fuzzy sets: "low", "medium", and "high". Then, a preset membership function (such as a triangular or trapezoidal function) is used to define the range of each fuzzy set. Based on expert experience and historical tunneling error data, fuzzy rules are developed, such as: if the stratum density is high and the tunneling error is large, increase the thrust; if the stratum density is low and the tunneling error is small, reduce the thrust.

[0096] Next, a fuzzy adaptive control algorithm (such as the Mamdani reasoning algorithm) is used to apply the fuzzy rules to each fuzzy set, calculate the fuzzy thrust distribution of each jacking cylinder, and use a defuzzification algorithm to process the fuzzy thrust distribution of each jacking cylinder. The fuzzy thrust distribution is converted into a specific numerical output, and the thrust distribution of each jacking cylinder is obtained.

[0097] Optionally, in this embodiment, the thrust feedback during tunneling can be monitored in real time to evaluate the effect of the current thrust distribution. The preset membership function and fuzzy rules are adjusted through the feedback mechanism to improve the adaptability of the control system. According to the feedback data, the parameters of the fuzzy adaptive control algorithm are dynamically adjusted to adapt to different geological conditions and tunneling requirements, such as appropriately increasing the threshold value of the preset membership function in high-density strata to provide more thrust.

[0098] The step S123 above converts the geological parameters and historical errors into fuzzy rules to dynamically adjust the thrust of each jacking cylinder. Through the feedback mechanism, the membership function and the rules are optimized to adapt to different geological conditions, ensuring the balance and response speed of the thrust distribution.

[0099] In step S124, the real-time monitoring data, the tunneling direction deviation trend, the tunneling direction deviation factors, the influence weight of each geological section in the target area on the tunneling direction, and the thrust distribution of each jacking cylinder in the shield machine are processed using the deep learning model to generate the tunneling direction adjustment instruction.

[0100] Understandably, the deep learning model includes a convolutional neural network module (CNN), a long short-term memory network module (LSTM), and a comprehensive output module. First, through the convolutional layer, the pooling layer, and the fully connected layer inside the convolutional neural network module, the spatial feature data of the real-time monitoring data (such as the formation texture, density change, etc.) is extracted. Then, the long short-term memory network module analyzes the time correlation of the historical excavation error data and the real-time monitoring data. The long short-term memory network module can capture the long-term dependence and dynamic trend of the data, is suitable for predicting the formation change trend and the excavation deviation risk, and can effectively capture the time series data. Then, through the comprehensive output module, the spatial feature data and the time series data are combined, and through the fully connected layer and the activation function, the prediction result of the future formation change and the deviation risk is output.

[0101] Further, the target excavation direction adjustment amount is calculated according to the prediction result, for example, the prediction result shows that a certain formation has a high deviation risk, and the advancing direction of the shield machine needs to be adjusted. The target excavation direction adjustment amount is converted into a specific excavation direction adjustment instruction by using a PID control algorithm, and the advancing angle of the shield machine is controlled. Based on the excavation direction deviation trend and the excavation direction deviation factor, a targeted excavation direction adjustment instruction is formulated, such as modifying the thrust distribution strategy or adjusting the excavation speed. According to the influence weight of each geological formation in the target area on the excavation direction, the excavation direction adjustment instruction is adjusted in real time, such as changing the advancing force of the shield machine or correcting the direction. According to the thrust distribution of each jacking cylinder in the shield machine, the excavation direction adjustment instruction is adjusted in real time to ensure the balance and effectiveness of the overall thrust.

[0102] Optionally, in the present embodiment, the actual effect of the excavation process can be monitored in real time, and the feedback data can be used for continuous optimization and instruction adjustment of various models and algorithms. By continuously updating the model algorithm and adjusting the instruction, the accuracy and response speed of the direction adjustment instruction are improved.

[0103] The above step S124 predicts the formation change and the deviation risk by fusing the spatial features (formation texture) and the time series (historical error trend), and converts the prediction result into an accurate excavation direction adjustment instruction (such as the advancing angle and the thrust distribution), thereby realizing closed-loop control.

[0104] In step S130, the advancing speed and the steering angle of the shield machine are adjusted according to the excavation direction adjustment instruction, and the thrust distribution of the multiple jacking cylinders in the shield machine is adjusted according to the excavation direction adjustment instruction.

[0105] Understandably, the advancing speed and the steering angle of the shield machine are accurately adjusted by the PID control algorithm according to the excavation direction adjustment instruction, intelligent steering is realized, the advancing direction of the shield machine is ensured to remain within the predetermined trajectory, the thrust distribution of each jacking cylinder is dynamically adjusted, the overall thrust configuration is optimized, and stable advancing of the shield machine is realized.

[0106] The step S130 precisely adjusts the propulsion parameters through the PID control algorithm to ensure that the shield machine excavates along the predetermined trajectory and reduces the error caused by manual intervention.

[0107] In step S140, a mud cake prevention instruction is generated according to the real-time monitoring data, and the removal parameters of the mud cake removal device are adjusted according to the mud cake prevention instruction. The running state and the excavation direction of the adjusted shield machine are displayed in real time through a man-machine interaction interface.

[0108] Understandably, during the shield tunneling process, mud cake refers to the mud blockage formed at the cutter head, conveying pipeline and the like due to soil conditions, improper mud-water ratio and the like, which affects the normal operation of the equipment. The mud cake position, mud cake thickness, mud cake distribution range and mud cake physical properties (such as viscosity and hardness) in the cutter head and conveying pipeline of the shield machine can be identified through the mud parameters collected by the mud parameter sensor and the excavation face images collected by the camera, and then a mud cake prevention instruction is generated. The removal intensity and frequency of the mud cake removal device are adjusted according to the mud cake prevention instruction, the mud cake removal device (such as a mixer and a mud pump) is controlled to operate, and the mud cake is removed in time to ensure the smooth flow of the mud and prevent the mud cake from blocking the operation of the shield machine.

[0109] Optionally, in the present embodiment, the adjusted shield machine state and the mud cake removal effect can be monitored in real time, a deep reinforcement learning (Deep Q-Network, DQN) model is used for feedback learning to continuously optimize the steering and removal strategies, and the adaptive ability and decision accuracy of the system are improved.

[0110] The above process identifies the mud cake position, thickness and physical properties in combination with the mud parameters and image data, dynamically adjusts the intensity and frequency of the removal device, and effectively prevents blockage.

[0111] The remote control end can display the running state, position information, ground humidity and mud cake condition of the shield machine in real time through a man-machine interaction interface, improve the monitoring experience of the operator, and use a heat map algorithm to display the excavation direction adjustment instruction, execution result and starting condition of the mud cake prevention measure in real time to help the operator quickly understand the working state of the shield machine. The monitoring data is analyzed in real time by using an isolation forest algorithm to detect system abnormal conditions (such as serious mud cake accumulation and ground mutation), and an alarm prompt is sent in time through a multi-level alarm mechanism (sound, vision, SMS / email notification) to ensure the safety of the construction process, take countermeasures in time, and manually intervene through an artificial intelligence assisted decision system when necessary to adjust the excavation parameters or start the backup mud cake prevention measure.

[0112] The above process can display the tunneling state and abnormalities (such as stratum mutation) in real time through a human-computer interaction interface, assisting the operator to make a quick decision. High-risk events (such as serious mud cake accumulation) are warned in time through sound, vision and remote notification, ensuring construction safety. The state of the device, stratum humidity and the execution result of the steering instruction are displayed intuitively, improving monitoring efficiency and operation experience.

[0113] The shield tunneling intelligent steering method for the volcanic ash stratum provided by the embodiments of the present application comprehensively utilizes multi-source sensor data, historical tunneling error analysis, uneven geological distribution in the volcanic ash stratum, thrust distribution of different jacking oil cylinders and advanced intelligent algorithms, solves the shield tunneling direction control problem, realizes real-time regulation and control of the advancing speed and the steering angle, and ensures accurate tunneling and construction safety of the shield machine under complex geological conditions.

[0114] In several embodiments provided in the present application, it should be understood that the disclosed apparatus and method can also be implemented by other manners. The apparatus embodiments described above are only schematic, for example, the flow charts and structural diagrams in the drawings show the possible implementation architectures, functions and operations of the apparatus, method and computer program product according to the embodiments of the present application. In this regard, each block in the flow chart or block diagram can represent a module, a program segment or a part of code, which contains one or more executable instructions for implementing the specified logical function. It should also be noted that, in alternative implementation manners, the functions noted in the blocks can also occur in different orders from those noted in the drawings. For example, two consecutive blocks can actually be executed substantially in parallel, and sometimes they can be executed in reverse order, depending on the functions involved. It should also be noted that each block in the structural diagram and / or flow chart, and the combination of blocks in the structural diagram and / or flow chart, can be implemented by a dedicated hardware-based system for executing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0115] In addition, each functional module or unit in each embodiment of the present application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0116] If the functions are implemented in the form of software function modules and sold or used as independent products, the functions can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the parts of the technical solutions that essentially contribute to the prior art can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a smart phone, a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium can be a non-volatile storage medium or a volatile storage medium, for example, the storage medium can be a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0117] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. A method for intelligent steering of a tunneling shield in a volcanic ash formation, characterized in that, The method comprises: acquiring construction data of the shield tunneling machine and the target area through a multi-source sensor, preprocessing the construction data, and fusing the preprocessed construction data to obtain real-time monitoring data; acquiring historical tunneling error data of the target area, analyzing the historical tunneling error data and the real-time monitoring data by using a deep learning model, and generating a tunneling direction adjustment instruction; adjusting the advancing speed and steering angle of the shield tunneling machine according to the tunneling direction adjustment instruction, and adjusting the thrust distribution of multiple jacking cylinders in the shield tunneling machine according to the tunneling direction adjustment instruction; generating a mud cake prevention instruction according to the real-time monitoring data, adjusting the cleaning parameters of a mud cake cleaning device according to the mud cake prevention instruction, and displaying the running state and the tunneling direction of the shield tunneling machine in real time through a human-computer interaction interface.

2. The intelligent steering method for shield tunneling in a volcanic ash stratum according to claim 1, characterized in that, The multi-source sensor comprises an attitude sensor, a geological radar, a camera, and a mud parameter sensor. The construction data comprises a tilt angle, an acceleration, a stratum structure signal, a tunneling face image, pressure data, flow data, viscosity data, and stratum humidity data. The construction data of the shield tunneling machine and the target area is acquired through the multi-source sensor, which comprises: optimizing the installation position of the multi-source sensor on the shield tunneling machine by using a genetic algorithm; acquiring the tilt angle and the acceleration of the shield tunneling machine through the attitude sensor, and acquiring the stratum structure signal of the target area through the geological radar; acquiring the tunneling face image of the target area through the camera, and acquiring the mud parameters of the target area through the mud parameter sensor, including at least one of pressure data, flow data, viscosity data, and stratum humidity data.

3. The intelligent steering method for shield tunneling in a volcanic ash stratum according to claim 2, characterized in that, The construction data is preprocessed, and the preprocessed construction data is fused to obtain real-time monitoring data, which comprises: smoothing the tilt angle and the acceleration of the shield tunneling machine by using a Kalman filtering algorithm to obtain preprocessed tilt angle and preprocessed acceleration; performing multi-resolution analysis on the stratum structure signal of the target area by using a wavelet transform algorithm to obtain preprocessed stratum structure signal; performing image analysis on the tunneling face image of the target area by using a convolutional neural network to obtain preprocessed tunneling face image; classifying and predicting the mud parameters of the target area by using a support vector machine to obtain preprocessed mud parameters; performing fusion processing on the preprocessed tilt angle, the preprocessed acceleration, the preprocessed stratum structure signal, the preprocessed tunneling face image, and the preprocessed mud parameters by using an extended Kalman filtering algorithm to generate the real-time monitoring data.

4. The intelligent steering method for shield tunneling in a volcanic ash stratum according to claim 3, characterized in that, The preprocessed stratum structure signal comprises rock, pore, moisture, and stratum heterogeneity. The wavelet transform algorithm is used to perform multi-resolution analysis on the stratum structure signal of the target area to obtain the preprocessed stratum structure signal, which comprises: determining a target wavelet function according to the frequency characteristics and the target resolution of the stratum structure signal; performing wavelet decomposition on the stratum structure signal by using the target wavelet function to obtain wavelet components, including low-frequency components and high-frequency components; Filter out noise signals higher than a preset frequency threshold in the high-frequency components, identify rocks, pores and water in the target area by analyzing the mutation points and intensity of the denoised high-frequency components, and identify the unevenness of the stratum in the target area by analyzing the fluctuations of the low-frequency components.

5. The intelligent steering method for shield tunneling in a volcanic ash stratum according to claim 2, wherein, The deep learning model is used to analyze the historical tunneling error data and the real-time monitoring data to generate a tunneling direction adjustment instruction. A support vector regression model is used to analyze the historical tunneling error data to determine the tunneling direction deviation trend and the tunneling direction deviation factors. A Bayesian network is used to comprehensively analyze the geological parameters of the stratum structure signal to determine the influence weight of each geological layer in the target area on the tunneling direction. A fuzzy adaptive control algorithm is used to determine the thrust distribution of each jacking oil cylinder in the shield machine based on the historical tunneling error data and the stratum structure signal. The deep learning model is used to process the real-time monitoring data, the tunneling direction deviation trend, the tunneling direction deviation factors, the influence weight of each geological layer in the target area on the tunneling direction, and the thrust distribution of each jacking oil cylinder in the shield machine to generate the tunneling direction adjustment instruction.

6. The intelligent steering method for shield tunneling in a volcanic ash stratum according to claim 5, characterized in that, The support vector regression model is used to analyze the historical tunneling error data to determine the tunneling direction deviation trend and the tunneling direction deviation factors, including: The support vector regression model is used to analyze the historical tunneling error data to generate an error prediction curve. The linear trend, periodic fluctuations and mutation points of the tunneling displacement error are identified based on the error prediction curve to obtain the tunneling direction deviation trend. Additional data collected by additional sensors is obtained, and the additional data and the historical tunneling error data are analyzed for correlation to determine basic factors that have a correlation with the tunneling displacement error, and the support vector regression model is used to determine target factors that have a correlation greater than a preset correlation threshold with the tunneling displacement error to obtain the tunneling direction deviation factors.

7. The intelligent steering method for shield tunneling in a volcanic ash stratum according to claim 5, wherein, The Bayesian network is used to comprehensively analyze the geological parameters of the stratum structure signal to determine the influence weight of each geological layer in the target area on the tunneling direction, including: The nodes of the Bayesian network are defined based on the geological parameters, and the conditional dependency relationships and conditional probability tables between the nodes are defined based on the historical tunneling error data, where the conditional probability table is used to describe the probability distribution of a node given the condition of its parent node. Each geological parameter is input into the Bayesian network to calculate the probability of tunneling direction deviation given the condition of each geological parameter. The influence weight of each geological layer in the target area on the tunneling direction is calculated based on the probability of tunneling direction deviation given the condition of each geological parameter and the influence degree of each tunneling direction deviation on the tunneling direction.

8. The intelligent steering method for shield tunneling in a volcanic ash stratum according to claim 5, wherein, The fuzzy adaptive control algorithm is used to determine the thrust distribution of each jacking oil cylinder in the shield machine based on the historical tunneling error data and the stratum structure signal, including: The geological parameters of the stratum structure signal, the historical tunneling error data and the real-time tunneling error data are converted into fuzzy sets, and a preset membership function is used to define the range of each fuzzy set; According to the historical tunneling error data, fuzzy rules are determined, and the fuzzy adaptive control algorithm is used to apply the fuzzy rules to each fuzzy set to calculate the fuzzy thrust distribution of each jacking oil cylinder; The fuzzy thrust distribution of each jacking oil cylinder is de-fuzzied using a de-fuzzification algorithm to obtain the thrust distribution of each jacking oil cylinder.

9. The intelligent steering method for shield tunneling in a volcanic ash stratum according to claim 5, wherein, The deep learning model includes a convolutional neural network module, a long short-term memory network module, and a comprehensive output module. The deep learning model is used to process the real-time monitoring data, the tunneling direction deviation trend, the tunneling direction deviation factors, the influence weight of each geological layer section in the target area on the tunneling direction, and the thrust distribution of each jacking oil cylinder in the shield machine to generate the tunneling direction adjustment instruction, including: The spatial feature data of the real-time monitoring data is extracted through the convolutional neural network module; The time correlation between the historical tunneling error data and the real-time monitoring data is analyzed through the long short-term memory network module to obtain time series data; The prediction result is generated by combining the spatial feature data and the time series data through the comprehensive output module. The prediction result is used to represent the prediction of the stratum change trend and the deviation risk; According to the prediction result, the target tunneling direction adjustment amount is calculated, and the PID control algorithm is used to combine the target tunneling direction adjustment amount, the tunneling direction deviation trend, the tunneling direction deviation factors, the influence weight of each geological layer section in the target area on the tunneling direction, and the thrust distribution of each jacking oil cylinder in the shield machine to generate the tunneling direction adjustment instruction.

10. The intelligent steering method for shield tunneling in a volcanic ash stratum according to claim 2, wherein, The mud cake prevention instruction is generated according to the real-time monitoring data, and the removal parameters of the mud cake removal device are adjusted according to the mud cake prevention instruction, including: According to the tunneling face image, the mud cake position, mud cake thickness, mud cake distribution range, and mud cake physical properties are identified; According to the mud cake position, mud cake thickness, mud cake distribution range, mud cake physical properties, and mud parameters, the mud cake prevention instruction is generated; According to the mud cake prevention instruction, the removal intensity and frequency of the mud cake removal device are adjusted.

Images