Tunnel excavation geology-machine-structure multi-source perception and auxiliary decision method and system

Abstract

The application discloses a tunnel excavation geology-machine-structure multi-source sensing and auxiliary decision-making method and system, relates to the technical field of tunnel excavation auxiliary decision-making, and comprises the following steps: calculating a reflection coefficient according to a seismic wave, calculating a geological risk value according to a gas type and a concentration in the tunnel, extracting cumulative vibration energy according to a rock breaking vibration signal, calculating a cumulative wear amount according to a magnetic induction current of a cutter, and calculating a lining thickness according to an elastic reflection wave of a lining structure; constructing a spatial data matrix with a mileage as an axis according to the above results, constructing a multi-objective optimization function with a tunneling speed, a tunneling specific energy and a cutter wear amount as indexes within a feasible construction parameter range of tunneling parameters, adjusting weight factors of the indexes in the optimization function based on single-source risk assessment of the spatial data matrix, and determining an optimal tunneling parameter combination by taking the maximization of the optimization function as a target, so that multi-dimensional sensing integration of geology, equipment, environment and structure is realized, and intelligent collaborative management and control of the whole construction process is realized.

Description

Technical Field

[0001] This invention relates to the field of tunnel excavation auxiliary decision-making technology, and in particular to a method and system for multi-source perception and auxiliary decision-making of geological, mechanical and structural aspects in tunnel excavation. Background Technology

[0002] Tunnel boring machines (TBMs) have become core equipment for the construction of deep-buried and ultra-long tunnels, but existing technologies have significant drawbacks. First, the TBM's confined space and seismic wave crosstalk, coupled with the short detection range and low accuracy of traditional advanced geological prediction technologies, make it difficult to accurately identify adverse geological conditions such as karst caves, fault fracture zones, and water bodies, easily leading to safety accidents such as excavation face collapses. Second, the lack of integrated solutions for gas environment and dust monitoring results in significant cross-interference, threatening the safety of construction personnel. Third, the lagging identification of strata interfaces means that tunneling parameters cannot be adapted to changes in strata hardness in real time, leading to low construction efficiency. Fourth, cutter wear relies on manual inspection, making it impossible to monitor the status in real time, and excessive wear can easily cause equipment failure. Fifth, tunnel lining quality inspection is mostly done through post-construction sampling, which is inefficient, has a narrow coverage, and makes it difficult to guarantee construction quality.

[0003] Furthermore, current tunnel excavation construction utilizes independent and dispersed functional modules such as geological forecasting, equipment monitoring, environmental sensing, and structural inspection. This results in fragmented and severely isolated sensing data, failing to form a unified multi-source data fusion sensing system and lacking integrated sensing capabilities. Simultaneously, the absence of an intelligent decision-making system based on multi-source sensing data hinders dynamic construction optimization based on real-time geological conditions, equipment status, environmental parameters, and structural quality data. This leads to poor management and coordination, making it difficult to achieve intelligent, precise, and collaborative management throughout the entire tunnel excavation process. Summary of the Invention

[0004] To address the aforementioned issues, this invention proposes a multi-source perception and auxiliary decision-making method and system for tunnel excavation involving geology, equipment, environment, and structure. This system integrates multi-dimensional perception of geology, equipment, environment, and structure, enabling intelligent collaborative management and control throughout the entire construction process.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a multi-source sensing and auxiliary decision-making method for geological-mechanical-structural aspects of tunnel excavation, comprising: The reflection coefficient is calculated based on seismic wave data; the geological risk value is calculated based on the type and concentration of the target gas in the tunnel; the cumulative vibration energy is extracted based on the vibration signal during the rock breaking process; the cumulative wear is calculated based on the magnetic induction current of the tunnel boring machine cutter; and the lining thickness is calculated based on the elastic reflection wave of the lining structure. Based on the reflection coefficient, geological risk value, cumulative vibration energy, cumulative wear, and lining thickness, a spatial data matrix with mileage as the axis is constructed. A feasible construction parameter range is constructed based on the tunneling parameters. Within the feasible construction parameter range, a multi-objective optimization function is constructed with tunneling speed, tunneling specific energy, and cutter wear as indicators. Based on the single-source risk assessment of the spatial data matrix, the weight factors of each indicator in the multi-objective optimization function are adjusted, and the optimal combination of tunneling parameters is determined with the goal of maximizing the multi-objective optimization function.

[0006] As an alternative implementation method, a feasible range of construction parameters is constructed based on cutterhead rotation speed, thrust, torque, and tunneling speed. The multi-objective optimization function constructed within this feasible range is: ; In the formula, , , Respectively, the benefits of tunneling efficiency Energy consumption revenue Benefits of tool wear control The corresponding weighting factor; To be determined by the speed of the cutter head ,thrust Torque and tunneling speed A set of tunneling parameters; Among them, tunneling efficiency benefits for: ; Energy consumption benefits for: ; ; Benefits of tool wear control for: ; ; ; In the formula, s represents the mileage; Minimum propulsion speed; The maximum advancing speed is represented by A; A represents the cross-sectional area of ​​the tunnel excavation. This represents the maximum tunneling specific energy. Minimum tunneling specific energy; For tunneling specific energy; To predict tool wear; This refers to the wear threshold for scrapping the hob. This represents the current cumulative wear and tear. In order to combine tunneling parameters The increase in wear per unit distance of tunneling. The coefficient of formation erosion is denoted as .

[0007] As an alternative implementation method, the process of adjusting the weighting factors of each indicator in the multi-objective optimization function based on a single-source risk assessment of the spatial data matrix includes: Multi-source data in the spatial data matrix are uniformly converted into standardized data within the range of 0 to 1. Based on the degree to which the standardized data deviates from the normal range, the single-source risk characterization value of each data source is obtained. Based on the risk characterization values ​​of each individual source and the preset risk threshold, the abnormal state at the current mileage is determined, and the construction control mode is selected accordingly. Based on the construction control mode, the corresponding weight is called.

[0008] As an alternative implementation method, the single-source risk characterization value is: Let the first Data source in mileage The standardized eigenvalue at point is The normal range of values ​​is The minimum allowable threshold is The maximum allowed threshold is And satisfy ; when Within the normal range Within the time frame, single-source risk characterization value Set to 0; when hour, ; when hour, .

[0009] As an alternative implementation method, the process of selecting a construction control mode includes: Efficiency-first mode: When the single-source risk characterization value of all data sources... When both are 0, select the efficiency-first mode; When there is a single-source risk characterization value with one data source satisfy or single-source risk characterization values ​​from at least two data sources. satisfy However, when located at different mileage positions, the balanced tunneling mode should be selected; When there are single-source risk characterization values ​​from at least two data sources at the same mileage point satisfy Or a single-source risk characterization value from any one data source. satisfy When this happens, select the security priority mode; This is the risk threshold.

[0010] As an alternative implementation method, geological risk value The calculation formula is: In the formula, gas Geological weighting factor, gas The monitored concentration gas Normal formation background concentration, gas The standard deviation of the fluctuation, where n is the number of gas categories.

[0011] As an alternative implementation method, the formula for calculating the lining thickness is: T represents the lining thickness. For longitudinal wave velocity, Main frequency value; The formula for calculating the reflection coefficient wave is: In the formula, The rock density of the strata on one side of the impedance difference interface. V1 represents the rock density of the strata on the other side of the wave impedance difference interface; V2 represents the wave velocity of the strata on one side of the wave impedance difference interface; and V3 represents the wave velocity of the strata on the other side of the wave impedance difference interface.

[0012] As an optional implementation, the tunnel boring machine-structure multi-source sensing and auxiliary decision-making method further includes determining the geological state at the cutterhead location based on accumulated vibration energy, specifically including: Extract the root mean square value of the vibration signal; Calculate the ratio of the current cumulative vibration energy change rate to the average vibration energy change rate of the stable tunneling section. When the ratio is greater than the preset mutation coefficient, it is determined that an instantaneous mutation has occurred, and the cutterhead is currently at the interface of heterogeneous strata with uneven hardness. The strength determination result of the rock mass in the current area is obtained based on the range of the root mean square value and the cumulative vibration energy values ​​in the preset standard database.

[0013] As an optional implementation, the tunnel boring machine-machine-structure multi-source sensing and auxiliary decision-making method further includes determining the cumulative wear amount based on the magnetic induction current of the tunnel boring machine cutter head and the pre-calibrated standard correspondence curve of magnetic flux density-cutter wear amount, and identifying the uneven wear state based on the cumulative wear amount and the fluctuation frequency of the magnetic induction current within the cutter head rotation cycle; specifically, if the fluctuation frequency is consistent with the cutter head rotation frequency, it is determined that there is a fixed local wear point on the circumference of the cutter head, which is identified as uneven wear; when the cumulative wear amount is greater than or equal to the preset wear amount threshold, an alarm is automatically triggered.

[0014] Secondly, the present invention provides a multi-source sensing and auxiliary decision-making system for tunnel excavation geology-machine-structure, comprising: The multi-source sensing module is configured to calculate the reflection coefficient based on seismic wave data, calculate the geological risk value based on the type and concentration of target gas in the tunnel, extract the cumulative vibration energy based on the vibration signal during the rock breaking process, calculate the cumulative wear based on the magnetic induction current of the tunnel boring machine cutter, and calculate the lining thickness based on the elastic reflection wave of the lining structure. The auxiliary decision-making module is configured to construct a spatial data matrix with mileage as the axis based on reflection coefficient, geological risk value, cumulative vibration energy, cumulative wear, and lining thickness. It constructs a range of feasible construction parameters based on tunneling parameters. Within the range of feasible construction parameters, it constructs a multi-objective optimization function with tunneling speed, tunneling specific energy, and tool wear as indicators. Based on the single-source risk assessment of the spatial data matrix, it adjusts the weight factors of each indicator in the multi-objective optimization function and determines the optimal combination of tunneling parameters with the goal of maximizing the multi-objective optimization function.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention proposes a multi-source perception and auxiliary decision-making method and system for tunnel excavation, encompassing geology, equipment, environment, and structure. This achieves integrated perception across multiple dimensions, addressing the problems of fragmented modules and data isolation. Simultaneously, it constructs a multi-source data fusion intelligent decision-making system to realize intelligent collaborative management and control throughout the entire construction process. By integrating multi-dimensional data such as geological forecasts, geological conditions, equipment operating conditions, environmental parameters, and structural quality, a standardized spatial data matrix with mileage as the axis is constructed. A multi-objective optimization function is built, taking into account excavation speed, excavation energy density, and tool wear. Combined with dynamic adjustment of index weight factors based on single-source risk assessment, the system adaptively matches different geological and construction conditions to solve for the optimal combination of excavation parameters. This achieves multi-dimensional collaborative optimization of tunnel excavation safety, efficiency, quality, and equipment operation and maintenance, significantly improving the level of intelligent, precise, and integrated management and control of tunnel construction. It is particularly suitable for large-scale, high-quality construction of deep-buried, ultra-long, and complex geological tunnel projects.

[0016] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0018] Figure 1 This is a flowchart of the tunnel excavation geology-machine-structure multi-source perception and auxiliary decision-making method provided in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of a tunneling machine equipped with an advanced geological prediction system provided in Embodiment 1 of the present invention; Figure 3 This is a schematic diagram of the principle of the gas environment monitoring system mounted on the tunneling machine provided in Embodiment 1 of the present invention; Figure 4 This is a schematic diagram illustrating the sensor arrangement principle of the tunneling machine-mounted stratum interface identification system provided in Embodiment 1 of the present invention; Figure 5 This is a schematic diagram of the principle of the tunneling machine equipped with a cutter wear monitoring system provided in Embodiment 1 of the present invention; Figure 6 This is a schematic diagram of the principle of the tunneling machine-mounted lining quality detection system provided in Embodiment 1 of the present invention; Among them, 1. Detector; 2. Lining structure; 3. Controlled seismic source; 4. Tunneling machine; 5. Cutting cutter; 6. Excitation magnet at the transmitting end; 7. Magnetic field; 8. Receiving equipment; 9. Elastic wave excitation source; 10. Elastic wave receiving sensor; 11. Internal space of tunneling machine. Detailed Implementation

[0019] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0020] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, 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 invention pertains.

[0021] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, unless the context clearly indicates otherwise, the singular form is intended to include the plural form as well. Furthermore, it should be understood that the terms “comprising” and “including”, and any variations thereof, are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or apparatus that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0022] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0023] Example 1 This embodiment provides a multi-source perception and auxiliary decision-making method for geological, mechanical, and structural aspects of tunnel excavation, such as... Figure 1 As shown, it includes: The reflection coefficient is calculated based on seismic wave data; the geological risk value is calculated based on the type and concentration of the target gas in the tunnel; the cumulative vibration energy is extracted based on the vibration signal during the rock breaking process; the cumulative wear is calculated based on the magnetic induction current of the tunnel boring machine cutter; and the lining thickness is calculated based on the elastic reflection wave of the lining structure. Based on the reflection coefficient, geological risk value, cumulative vibration energy, cumulative wear, and lining thickness, a spatial data matrix with mileage as the axis is constructed. A feasible construction parameter range is constructed based on the tunneling parameters. Within the feasible construction parameter range, a multi-objective optimization function is constructed with tunneling speed, tunneling specific energy, and cutter wear as indicators. Based on the single-source risk assessment of the spatial data matrix, the weight factors of each indicator in the multi-objective optimization function are adjusted, and the optimal combination of tunneling parameters is determined with the goal of maximizing the multi-objective optimization function.

[0024] In this embodiment, the method described above adopts a modular integrated design, which is integrated into the tunneling machine body. Interconnection and interoperability of multi-source heterogeneous data are achieved through internal bus and wireless communication technology, forming a closed-loop system of perception-transmission-analysis-decision with auxiliary decision support as the central hub. The overall architecture includes: The geological sensing subsystem includes advanced geological forecasting, gas environment monitoring, and stratigraphic interface identification, enabling multi-scale collaborative sensing from the far field to the near field.

[0025] The tunneling machine sensing subsystem includes cutter wear monitoring, which senses the wear status of the rock-breaking terminal in real time and serves as an inversion reference for identifying formation hardness.

[0026] Tunnel structure sensing subsystem: includes lining quality detection, responsible for continuous scanning and quality assessment of the support structure after excavation.

[0027] Ultimately, the decision support system integrates disparate geological, tunneling machine, and tunnel structure data from multiple sources and transforms them into executable construction instructions.

[0028] The specific technical solution is as follows.

[0029] 1. Geological Sensing Subsystem.

[0030] 1.1 Far-field scale: Based on active seismic sources, this advanced geological prediction system aims to solve the challenge of wavefield identification within the confined space of a tunnel boring machine. Through high-frequency controlled excitation and high-sensitivity detection, it enables the detection of adverse geological conditions such as karst caves, fault fracture zones, and water bodies within a certain range ahead of the tunnel. It mainly includes a controlled excitation unit, a receiving sensor array, and a data interpretation terminal.

[0031] like Figure 2As shown, the controlled excitation unit employs a controlled seismic source 3 with broadband modulation capability, installed at the lateral position of the tunneling machine 4. The controlled seismic source 3 generates high-penetration seismic wave signals through high-energy impacts, ensuring that the seismic wave signals can effectively propagate to the rock mass in front of the excavation face.

[0032] Receiving sensor array: Inside the completed tunnel lining structure 2, several (e.g., 8) geophones 1 are arranged in a circumferential minimum offset mode. Minimal offset means that the distance between the geophone 1 and the controlled seismic source 3 in the spatial projection is much smaller than the detection depth, usually less than 3m. This arrangement can make the most of the near-vertical incident characteristics of the reflected waves and reduce signal distortion caused by wave path refraction.

[0033] The receiving sensor array and the data interpretation terminal integrated in the main control cabinet are connected wirelessly to achieve high-speed synchronous transmission of raw seismic wave data.

[0034] Following the principles of elastic dynamics, seismic waves propagate forward in rock media as longitudinal (P-waves) and transverse (S-waves). When encountering interfaces with significant differences in wave impedance, such as the water-filled interface of a karst cave or a fault fracture zone, strong reflection signals are generated. The wave impedance difference is described as follows: In the formula, R is the reflection coefficient, and the subscripts 1 and 2 represent the strata on both sides of the interface of wave impedance difference, respectively. The rock density of the strata on one side of the impedance difference interface. V1 represents the rock density of the strata on the other side of the wave impedance difference interface; V2 represents the wave velocity of the strata on one side of the wave impedance difference interface; and V3 represents the wave velocity of the strata on the other side of the wave impedance difference interface.

[0035] Therefore, by calculating the abrupt changes in the R value in real time, the abrupt change characteristics of the stratigraphic structure can be inverted.

[0036] To extract useful information in a cluttered environment, the following processing algorithm is executed: (1) The received three-component seismic wave data is divided into time windows with set intervals, and the motion trajectory of the particles is analyzed separately in each time window. In each time window, the covariance matrix is ​​constructed by using the correlation of the seismic wave data in the three directions recorded by the three-component seismic wave data, and the eigenvalues ​​are solved to identify the main vibration direction with the strongest particle vibration energy. If the angle between the main vibration direction and the propagation direction is close to 0°, the signal is determined to be a P-wave; if the angle is close to 90°, it is determined to be an S-wave.

[0037] (2) Common bias superposition: The wave field of multiple seismic wave data collected by the receiving sensor array is superimposed, and the random background signal is canceled out by statistical averaging effect, thereby improving the signal-to-noise ratio of weak reflection signal.

[0038] (3) Since seismic wave detection provides time, the wave travel time is converted into the actual distance between the geological body and the excavation face by constructing a rock mass wave velocity field model. First, the direct wave signal between the controlled seismic source 3 and the geophone 1 is used to calculate the rock mass velocity in the current area by the ratio of the distance between them to the wave arrival time. Then, the velocity of the unknown strata ahead is estimated by combining the amplitude and polarity of the wave reflected back from the interface ahead (specifically, the amplitude represents the velocity change, and the positive or negative polarity represents the velocity increase or decrease), thus forming the rock mass wave velocity field model.

[0039] Based on the rock mass wave velocity field model and seismic wave recording time, the time-domain information is converted into spatial coordinate information required for geological exploration using an imaging formula. The imaging formula is as follows: ; Where L is the distance between the target geological body and the working face. is a wave velocity function that varies along the path.

[0040] As the first link in the sensing system, the advanced geological prediction module pushes its detection results to the auxiliary decision support system in real time. It also triggers an automatic detection every 5m of tunneling to form a new rock mass wave velocity field model and replace the original model, thus achieving continuous updates.

[0041] 1.2 Near-field scale: Geological structure-gas feature mapping perception, by capturing and interpreting the dynamics of gas composition in the tunneling chamber in real time, establishes a nonlinear mapping model between gas characteristics and the geological structure ahead of the tunnel. This serves not only as a construction safety barrier but also as a key logical link in the geological perception system.

[0042] like Figure 3 As shown, it specifically includes: (1) Two-stage component separation and multi-gas detection mechanism: In the gas environment to be tested (i.e., inside the tunnel), the gas to be tested, such as oxygen (O2), carbon monoxide (CO), carbon dioxide (CO2), methane (CH4), ammonia (NH3), sulfur dioxide (SO2), nitrogen dioxide (NO2), hydrogen sulfide (H2S), etc., is detected, and the concentration of the gas to be tested is also detected. In order to solve the cross-interference of multiple harmful gases under the complex working conditions of the tunnel, a coupling separation strategy of modified separation membrane and specific catalyst is adopted.

[0043] Specifically, it includes: (1-1) At the physical level, physical screening is carried out based on differences in molecular diameter / diffusion coefficient, etc.: For the eight target gases to be tested, namely oxygen, carbon monoxide, carbon dioxide, methane, ammonia, sulfur dioxide, nitrogen dioxide and hydrogen sulfide, the following design is achieved by adjusting the modification process: First, the average effective pore size of the modified separation membrane is controlled between 0.33 nm and 0.45 nm. By utilizing the significant difference in molecular dynamic diameter between carbonaceous and small-diameter gases such as carbon dioxide (0.33 nm) and oxygen (0.346 nm) and large-diameter sulfurous gases such as sulfur dioxide (0.41 nm), preliminary physical-level sieving and interception are achieved.

[0044] Secondly, by modifying polar functional groups on the surface of the modified separation membrane, the strong polarity of nitrogen gases such as ammonia and nitrogen dioxide interacts with the membrane surface, causing a significant lag in their travel velocity (diffusion coefficient) within the membrane channels.

[0045] Through the above design, different types of gases exhibit a significant time difference when passing through the membrane layer, thereby completing the major classification at the physical layer: carbon gases pass through quickly, sulfur gases pass through slowly, and nitrogen gases pass through even more slowly, eliminating cross-interference caused by different components rushing to the sensor at the same time.

[0046] (1-2) At the chemical level, cross-interference of explosive gases is eliminated through specific catalytic oxidation.

[0047] For explosive gases with very similar properties, a special catalyst is used to induce selective catalytic oxidation of specific components, further eliminating interference terms and achieving precise extraction of harmful gas concentrations.

[0048] Specifically, manganese copper oxide catalysts are used for the elimination of carbon monoxide and hydrogen sulfide. Under ambient temperature conditions, these catalysts preferentially induce the oxidation of CO and H₂S, thus eliminating low-ignition-point interference terms with similar chemical properties before detecting high-ignition-point gases such as methane. For the specific conversion of ammonia and nitrogen dioxide, vanadium pentoxide (V₂O₅)-based catalysts are used. Utilizing their specific catalytic activity towards nitrogen-containing compounds, they are converted into chemically stable inert components, avoiding cross-interference with the electrochemical detection of sulfur dioxide (SO₂).

[0049] (2) Establish a nonlinear mapping model between gas characteristics and geological structures ahead of the tunnel. The core is to establish a geological structure inversion logic based on the gas environment. The occurrence form of gas in the strata is highly correlated with the geological evolution process. By monitoring the abnormal fluctuation gradient of gas concentration, the geological properties ahead can be inverted.

[0050] Specifically: Establish a geological risk assessment function to calculate the geological risk value. : ; In the formula, gas Geological weighting factor, gas The monitored concentration gas Normal formation background concentration, gas The standard deviation of the fluctuation, where n is the number of gas categories.

[0051] Based on the calculated geological risk values, the geological risks are divided into three levels, and corresponding handling logic is implemented: (1) Safe and stable zone: The formation was determined to be in a normal background state, with stable geological structures ahead, and the regular monitoring frequency was maintained.

[0052] (2) Risk warning zone: The gas concentration was determined to be abnormally trending, suggesting that we are entering a geologically variable zone or that there are minute fracture structures ahead.

[0053] (3) High-risk warning area: The gas concentration was determined to have experienced a magnitude jump, indicating a potential major geological risk ahead.

[0054] This embodiment uses a dynamic setting method based on statistical background values ​​to calculate the threshold. and This involves automatically collecting gas concentration data within a known stable surrounding rock section during tunnel excavation, calculating the average concentration of each target gas as the normal formation background concentration, and calculating its standard deviation of fluctuation. Typically, this is set... This is set to three standard deviations of the background fluctuation level to exclude random noise interference. It is typically set to... The background fluctuation level is set at 5-10 standard deviations, depending on the geological complexity of the specific project. Furthermore, the geological weighting factor is adjusted... To achieve dynamic optimization of the early warning sensitivity for specific gases in specific sensitive formations such as high-gas formations.

[0055] 1.3 Near-field scale: Based on the vibration response of the cutterhead breaking rock, the formation interface is identified. By monitoring the vibration response characteristics of the tunneling machine during the cutting and breaking of rock in real time, the soft and hard interfaces of the formation and the strength of the rock mass ahead of the tunneling are accurately identified.

[0056] like Figure 4 As shown, a highly sensitive vibration sensing matrix is ​​constructed using the internal space of the tunneling machine. Vibration signal acquisition sensor arrays (A1-A4, B1-B4) are spatially arranged according to circumferential and radial dimensions (e.g., distances from the muck outlet of 2.6m, 3.0m, etc.). Sensors are installed at key stress and vibration transmission locations such as behind the tunneling machine's soil chamber and on the front shield bulkhead, ensuring maximum capture of the raw dynamic feedback of the cutterhead breaking rock. The acquired vibration signals are transmitted in real-time and stably via a wireless transmission unit, avoiding the limitations imposed by complex mechanical structures on wired paths.

[0057] Due to differences in the strength and hardness of the rock mass, the resistance experienced by the cutterhead during rock breaking varies significantly, and this variation is directly reflected in the vibration characteristics. In hard rock formations, the rock-breaking resistance is high, and the excited vibration signals typically have high characteristic frequencies and large peak amplitudes; in soft rock formations, the vibration signals are low-frequency and have relatively stable amplitudes.

[0058] Therefore, the specific processing procedure is as follows: (1) First, the collected vibration signal is adaptively filtered and noise reduced to eliminate periodic interference caused by the rotation of the spindle motor.

[0059] (2) Then, the root mean square value of the vibration signal is extracted as an indicator to reflect the overall intensity of the vibration signal.

[0060] (3) Establish an energy accumulation model based on vibration signals to calculate the cumulative vibration energy within a time window. : ; in, for Vibration signal at any given moment.

[0061] (4) Accumulated vibration energy is obtained in real time through continuous calculation. And its rate of change. When the cutterhead encounters a heterogeneous stratum interface with uneven hardness, due to the instantaneous step change in rock-breaking resistance, the root mean square value of the vibration signal and the cumulative vibration energy... It will produce obvious transient changes.

[0062] Therefore, the root mean square value and cumulative vibration energy are extracted and calculated. After determining the formation state and its rate of change, the following logic is used to complete the determination: The above-mentioned indicators, which are calculated in real time, are combined with the tunneling parameters collected synchronously and input into the built-in multidimensional database of tunneling parameters, formation information and vibration response characteristics. The database pre-stores various response model data under known standard formation conditions.

[0063] The judgment logic is as follows: By matching and comparing the measured data with the standard model in the database, the following criteria are established: The ratio of the current cumulative vibration energy change rate to the average vibration energy change rate of the previous stable tunneling section is calculated in real time. When this ratio is greater than a preset abrupt change coefficient (the abrupt change coefficient ranges from 1.5 to 3.0), an instantaneous abrupt change is determined, indicating that the cutterhead is currently at the interface of a heterogeneous stratum with uneven hardness. Further analysis is performed based on the root mean square value and cumulative vibration energy... The values ​​are compared with the range in the database, and the strength determination results of the rock mass in the current area are output.

[0064] As a near-field verification link in the geological-mechanical-structural sensing system, the geological sensing system provides real-time basis for the dynamic adjustment of tunneling parameters, ensuring that the tunneling parameters can adapt to changes in the softness and hardness of the strata, and solving the pain point of lagging strata interface identification in traditional construction.

[0065] 2. Tunneling machine sensing subsystem.

[0066] The hobbing cutter is mounted on the cutterhead. During rock breaking, the metal profile of the hobbing cutter ring undergoes cumulative wear. Because the hobbing cutter ring is made of ferromagnetic material, reductions in its size and shape directly distort the surrounding magnetic field. Therefore, as... Figure 5 As shown, the transmitting end excites the magnet 6 to emit a signal, and an electromagnetic induction sensor is installed inside the cutter box of the cutter 5 to capture the disturbance characteristics of the magnetic field 7 caused by the change in the size of the cutter ring of the cutter 5. The receiving end device 8 receives the magnetic induction current to realize the quantitative perception of the cumulative wear and uneven wear state of the cutter ring. Taking a tunnel boring machine with a diameter of 8m as an example, one sensor is arranged inside each cutter box, with a total of 36 monitoring points deployed throughout the machine.

[0067] Electromagnetic induction sensors capture changes in magnetic field strength in real time through electromagnetic mutual inductance. Their electromagnetic conversion relationship satisfies the following physical formula: ;in, Magnetic flux density is used to describe the strength and direction of a magnetic field; Vacuum permeability is the proportionality constant between magnetic flux and magnetic field strength in a vacuum environment, and its value is [value missing]. ; The current intensity in the conductor; The perpendicular distance from the point to the conductor; The angle between the ray drawn from the field point to one end of the conductor and the perpendicular line from the field point to the conductor; Let the angle between the ray drawn from the field point to the other end of the conductor and the perpendicular line from the field point to the conductor be the angle between them.

[0068] As the wear of the hob increases, the parameters r and the included angle... The continuous change causes a quantifiable shift in the magnetic flux density B, which in turn induces a differentiated magnetic flux current signal in the sensing coil.

[0069] Based on the physical correlation logic of wear change, electromagnetic change, and current change, the complex wear behavior of hobbing tools can be accurately identified through in-depth processing of magnetic induction current signals.

[0070] Specifically: The acquired magnetic induction current signal is amplified by a high factor and low-pass filtered to eliminate the high-power electromagnetic noise generated by the cutter head drive system.

[0071] The cumulative wear is calculated based on the pre-calibrated standard curve corresponding to the magnetic flux density and hob wear.

[0072] By analyzing the fluctuation frequency of the magnetic induction current during the rotation cycle of the cutter head, it is possible to identify whether the hob is experiencing uneven wear (i.e., excessive local wear caused by uneven force on the cutter ring), which is a pain point that is difficult to detect with traditional manual inspection.

[0073] Specifically, if the hob's circumference wears evenly, the distance between the hob's surface and the sensor remains constant during rotation. In this case, the magnetic induction current signal exhibits a stable DC component without significant periodic fluctuations. However, if the hob experiences uneven wear, the hob's circumference will no longer be a perfect circle. As the hob rotates, the distance *r* between its surface and the sensor will periodically fluctuate, causing the magnetic induction current to generate an AC component that fluctuates synchronously with the rotation.

[0074] Therefore, the frequency of magnetic induction current fluctuation is analyzed using the fast Fourier transform method. If the fluctuation frequency is consistent with the rotation frequency of the hob, it is determined that there are fixed local wear points on the circumference of the hob, which is identified as uneven wear.

[0075] In addition, when the cumulative wear of the cutter reaches or exceeds a preset threshold (such as 10mm), an alarm is automatically triggered to remind the operator to replace the cutter in time to avoid damage to the cutter box or a sudden drop in tunneling efficiency due to excessive wear.

[0076] 3. Tunnel structure sensing subsystem.

[0077] Synthetic aperture impact echo lining detection is integrated behind the tunnel boring machine trolley. By conducting real-time and continuous non-destructive testing on the installed lining structure, quantitative evaluation of the tunnel lining structure thickness, backfill grouting density, and internal defects can be achieved.

[0078] like Figure 6 As shown, the mobile mounting platform is installed on a reserved bracket above the tunneling machine trolley, driving the elastic wave excitation source 9 and the elastic wave receiving sensor 10 to achieve continuous circumferential and longitudinal scanning of the lining structure 2. The probe maintains a constant distance of 5-8 cm from the surface of the lining structure 2, adapting to the narrow internal space 11 of the tunneling machine. Using an electromagnetic vibratory hammer as the excitation source replaces traditional manual excitation, ensuring that the emitted high-frequency elastic wave signal source is stable, adjustable, and has a controllable waveform. This overcomes the limitations of low efficiency and narrow coverage of traditional manual sampling inspection, achieving automated detection synchronized with tunneling.

[0079] Based on the principle of the impact echo method, this method utilizes the reflection characteristics of elastic waves at the interface of a non-homogeneous medium for sensing. A vibrating hammer emits high-frequency elastic waves towards the surface of the lining structure, which then propagate within the structure. When encountering interfaces with impedance differences, such as voids or loose areas, significant reflection signals are generated. A high-frequency single-component sensor receives the reflected echoes in real time with high sensitivity and high resolution for spectral analysis. Specifically, by analyzing the propagation time and amplitude of the reflected wave, combined with the propagation velocity of the elastic wave within the lining structure, the lining thickness is calculated. In the formula, T is the lining thickness. For longitudinal wave velocity, The dominant frequency value.

[0080] Furthermore, spectral analysis was performed on the reflected waves from all measuring points within the detection area to extract the dominant frequency value of each measuring point. and main frequency amplitude Calculate the average of the dominant frequency value and the dominant frequency amplitude of all measuring points in this area. , ) and standard deviation ( , The lining quality condition is determined using the following quantitative criteria; specifically: Complete density determination: If the dominant frequency value of the measuring point is within And the main frequency amplitude is at Within the specified range, the lining of the area at the measuring point is determined to be intact and dense. Determination of non-compactness: If the dominant frequency of the measuring point is too low (meeting the requirements) And accompanied by an abnormal decrease in amplitude (satisfying) The measurement point area was determined to be poorly liningd. Void defect determination: If the main frequency of the measuring point is too high (meets the requirements) And accompanied by an abnormally high amplitude (satisfying) This indicates that there are voids inside the lining of the measuring point area.

[0081] 4. Decision support system.

[0082] By unifying the perception results from multiple dimensions such as geology, environment, equipment, and structure into the same mileage axis and the same decision-making framework, a closed loop of auxiliary decision-making is formed, which includes multi-source perception, matrix fusion, risk coupling, parameter optimization, and feedback correction.

[0083] Employing an industrial-grade layered architecture, it ensures real-time processing of massive amounts of data and reliable operation of complex algorithms. It integrates embedded data coupling units, high-speed cloud platform servers, edge computing terminals, and front-end interactive terminals. Utilizing a LoRa+Ethernet dual-mode communication mode, the LoRa module handles high-frequency, low-power, long-distance transmission between sensor nodes; the Ethernet interface serves as a high-speed backup link, ensuring efficient interaction between high-definition imaging data and cloud models. The front-end terminal is installed in the TBM operator's cab for real-time control by the driver; the data coupling unit connects to the main control cabinet, achieving deep integration with the tunneling machine's power and control systems.

[0084] By incorporating an industrial-grade data processing chip, the system enables the access, mileage mapping, dimensionless transformation, multi-source anomaly pattern recognition, and real-time construction parameter evaluation of heterogeneous data from multiple dimensions such as geology, environment, equipment, and structure. The intelligence is mainly reflected in the automatic determination of the construction control mode (including efficiency-first mode, balanced tunneling mode, and safety-first mode) based on the results of multi-source data analysis. The system then calls the corresponding objective function weights according to the construction control mode, and finally generates a recommended combination of construction parameters within the parameter range jointly defined by equipment capabilities, construction boundaries, and construction control mode.

[0085] The specific implementation process is as follows: (1) Real-time reception of seismic wave data, target gas concentration and type, cumulative vibration energy of cutterhead vibration signal, magnetic induction current of cutterhead and elastic reflection wave of lining structure.

[0086] (2) Establish a unified tunnel spatial coordinate system. Based on the real-time mileage of the tunnel boring machine, coordinate conversion is performed according to the installation position, detection distance and acquisition timestamp of each sensor. Among them, seismic wave data, target gas concentration and type are mapped to the mileage in front of the tunnel face, the cumulative vibration energy of the cutterhead vibration signal and the magnetic induction current of the cutter head are mapped to the current tunnel face mileage, and the elastic reflection wave of the lining structure is mapped to the mileage already constructed behind the trolley. In this way, the historical, current and future perception information are unified to the same mileage axis.

[0087] (3) Map the feature recognition results extracted by each module at each mileage location to a spatial coordinate system to construct a spatial data matrix with mileage s as the axis. : ; in: R The reflection coefficient is obtained from advanced geological prediction (characterizing the abrupt change in forward wave impedance). This refers to the geological risk value calculated during gas environment monitoring. This refers to the cumulative vibration energy calculated during stratigraphic interface identification. This is the current cumulative wear amount calculated based on the magnetic induction current of the hob. The lining thickness is calculated based on the elastic reflected wave.

[0088] The fusion decision-making process based on the spatial data matrix is ​​as follows: (3-1) Based on the allowable range of the tunneling machine equipment, the current construction control boundary, and the construction control mode, construct a system including the cutterhead rotation speed. ,thrust Torque and propulsion speed Real-time feasible construction parameter range The range of real-time feasible construction parameters is dynamically updated based on the geological, environmental, equipment, and structural anomalies at mileage s.

[0089] ; in, and These are the minimum and maximum cutter head speeds, respectively. and These are the minimum thrust and the maximum thrust, respectively. and These are the minimum torque and the maximum torque, respectively. and These are the minimum propulsion speed and the maximum propulsion speed, respectively.

[0090] Within the range of real-time feasible construction parameters Inside, for each set of parameters Calculate the benefits of tunneling efficiency, energy consumption, and tool wear control separately, and construct a multi-objective optimization function. : ; In the formula, , , These are the weighting factors corresponding to tunneling efficiency gains, energy consumption gains, and tool wear control gains, respectively. The efficiency gain in tunneling is represented by a combination of parameters. propulsion speed Normalization yields the result; This represents the energy consumption benefit, derived from the tunneling specific energy. Normalization yields the tunneling specific energy. The lower, The larger the value, the higher the tunneling energy. According to thrust Torque Cutter head speed and propulsion speed The conversion yielded the result. This represents the benefit of tool wear control, derived from the predicted tool wear amount. The normalized result shows that the lower the predicted tool wear, the better. The larger the value, the more likely the tool wear will be. Based on the current cumulative wear and cutter head speed ,thrust and propulsion speed The result is obtained through conversion.

[0091] Specifically: (1) Tunneling efficiency benefits for: ; in, Minimum propulsion speed; This represents the maximum propulsion speed.

[0092] (2) Energy consumption revenue for: ; ; Where A is the cross-sectional area of ​​the tunnel excavation; This represents the maximum tunneling specific energy. This represents the minimum tunneling specific energy.

[0093] (3) Benefits of tool wear control for: ; ; ; in, The wear threshold for scrapping the hob (e.g., 10mm); This represents the current measured cumulative wear amount; To predict in parameter combinations The increase in wear per unit distance of tunneling. The coefficient of formation erosion is denoted as .

[0094] (3-2) The access data with different units and different value ranges in the spatial data matrix are uniformly converted into standardized data within the range of 0 to 1; then, based on the degree to which the standardized data deviates from the normal range, the single-source risk characterization value of each data source is obtained. This is used for subsequent multi-source anomaly pattern recognition.

[0095] Specifically: For the Data source type, assuming it is at the tunneling mileage The standardized eigenvalue at point is The normal range of values ​​is The minimum allowable threshold is The maximum allowed threshold is And satisfy .

[0096] when Within the normal range Within the time frame, single-source risk characterization value Set to 0; when At that time, according to calculate; when At that time, according to calculate.

[0097] Therefore, the single-source risk characterization value Limited to the range of 0 to 1, used to represent the first The degree of deviation of the data source from normal operating conditions.

[0098] (3-3) Single-source risk characterization values ​​calculated based on each data source (Value range 0~1), determine the abnormal status at the current mileage and select the construction control mode according to the following rules: Efficiency-first mode: When the risk value of all data sources is 0 (i.e., all are within the normal range) (Inside) when entering; Balanced tunneling mode: When there is one data source in a slightly abnormal state (i.e. , The risk threshold can be set to 0.6 (an empirical value set manually, which can also be adjusted according to actual working conditions), or it can be entered when multiple data sources are slightly abnormal but located at different mileage positions; Safety Priority Mode: When two or more data sources at the same mileage are showing minor anomalies ( ), or any one of the data sources reaches a high-risk state ( Enter when ).

[0099] (3-4) Weight Invocation: Based on the identified construction control mode, invoke the corresponding objective function weight.

[0100] In efficiency-first mode, take =0.5、 =0.3、 =0.2; In balanced tunneling mode, take =0.4、 =0.3、 =0.3; In security priority mode, take =0.3、 =0.3、 =0.4.

[0101] The above weights can be preset based on engineering experience and meet the following requirements. + + =1.

[0102] (3-5) Within the range of real-time feasible construction parameters Internally, the rotational speed of the tool turret ,thrust Torque and propulsion speed Evaluate different combinations, that is , , Substitution Select The largest parameter combination is used as the recommended construction parameters; due to the range of real-time feasible construction parameters. The space is defined by the allowable range of equipment, construction control boundaries, and construction control mode. Any combination of parameters within the space meets the current safety constraints.

[0103] (3-6) Based on the new round of seismic wave data, target gas concentration and type, cumulative vibration energy of cutterhead vibration signal, magnetic induction current of cutterhead and elastic reflection wave of lining structure after the recommended construction parameters are executed, update the spatial data matrix, update the background value, construction control mode judgment result and real-time feasible construction parameter range, and realize the closed-loop iteration of multi-source perception-pattern recognition-assisted decision-feedback correction.

[0104] The decision output layer consists of a front-end interactive terminal and a remote monitoring platform, directly serving the construction site. It displays real-time geological cloud maps, multi-source anomaly states, construction control modes, equipment status curves, harmful gas trends, lining quality heat maps, and parameter evaluation results. When any single-source indicator triggers a threshold, two or more data sources simultaneously malfunction, or the construction control mode switches to safety-first mode, an early warning is issued in a prominent position on the visualization terminal, and a decision adjustment plan including propulsion speed, cutterhead rotation speed, thrust, torque, cutter replacement, and reinforcement grouting suggestions is simultaneously pushed to the driver's operating interface.

[0105] Example 2 This embodiment provides a multi-source sensing and auxiliary decision-making system for tunnel excavation geology-machine-structure, including: The multi-source sensing module is configured to calculate the reflection coefficient based on seismic wave data, calculate the geological risk value based on the type and concentration of target gas in the tunnel, extract the cumulative vibration energy based on the vibration signal during the rock breaking process, calculate the cumulative wear based on the magnetic induction current of the tunnel boring machine cutter, and calculate the lining thickness based on the elastic reflection wave of the lining structure. The auxiliary decision-making module is configured to construct a spatial data matrix with mileage as the axis based on reflection coefficient, geological risk value, cumulative vibration energy, cumulative wear, and lining thickness. It constructs a range of feasible construction parameters based on tunneling parameters. Within the range of feasible construction parameters, it constructs a multi-objective optimization function with tunneling speed, tunneling specific energy, and tool wear as indicators. Based on the single-source risk assessment of the spatial data matrix, it adjusts the weight factors of each indicator in the multi-objective optimization function and determines the optimal combination of tunneling parameters with the goal of maximizing the multi-objective optimization function.

[0106] It should be noted that the above modules correspond to the steps described in Embodiment 1, and the examples and application scenarios implemented by the above modules and the corresponding steps are the same, but are not limited to the content disclosed in Embodiment 1. It should also be noted that the above modules, as part of the system, can be executed in a computer system such as a set of computer-executable instructions.

[0107] Example 3 This embodiment provides an intelligent linkage early warning and decision push system for tunneling parameters of a tunnel boring machine. The system includes, at the hardware level, a multi-source heterogeneous sensor array deployed on the cutterhead panel, around the shield body, and inside the soil chamber.

[0108] Specifically, a torque sensor is installed in the cutter head drive motor circuit to collect the current signal of the drive motor in real time and convert it into an analog torque quantity; a thrust sensor is embedded in the pressure line of the propulsion cylinder to monitor the hydraulic pressure distribution of each group of cylinders; a high-resolution speed encoder is integrated at the main bearing to capture the instantaneous angular velocity and speed fluctuation characteristics of the cutter head; and tool wear electromagnetic induction detection units are distributed on the cutter head spokes to detect the remaining height and wear pattern of the tool cutting edge in a non-contact manner through the principle of electromagnetic induction.

[0109] Each of the aforementioned sensors continuously acquires raw analog signals with a millisecond-level sampling period. The signals are then transmitted via shielded cables to an edge computing node near the data acquisition end. The node's built-in A / D conversion circuit performs analog-to-digital conversion and preliminary filtering, and the signals are subsequently sent to the central data processing unit in the form of timestamped data packets.

[0110] The data processing unit is configured to perform asynchronous timestamp alignment and Kalman filtering denoising on the received multi-source data streams to eliminate phase differences and signal delays caused by differences in the physical installation locations of the sensors and the length of the transmission paths.

[0111] Based on this, the data processing unit executes three exception detection logic steps: Firstly, there is a single-source threshold exceeding the limit judgment, which triggers an early warning when the real-time parameter of any physical sensor exceeds the adaptive threshold zone dynamically calibrated based on the current geological profile. This threshold zone is not a fixed value, but is adjusted in real time according to the composite strength characteristics of the strata ahead of the tunnel face fed back by the advanced geological prediction system. This allows the threshold tolerance to be automatically tightened in the area where soft strata and hard rock strata meet, avoiding missed reports of abnormal tool wear caused by sudden changes in strata. Secondly, there is a multi-source synchronization anomaly detection. When two or more physical sensors synchronously output signals that deviate from their respective normal operating ranges within a preset time window, and the cross-validation algorithm trained based on historical tunneling data confirms that the deviation is not caused by a single sensor failure or electromagnetic interference, a linkage warning is triggered, thereby solving the problem of false alarms caused by asynchronous multi-source data. Thirdly, for the control mode switching determination, when the main control system of the tunnel boring machine switches from normal tunneling mode to safety priority mode due to the detection of an unknown obstacle ahead or the shield body attitude exceeding the limit, the data processing unit automatically increases the sampling frequency from the default period to the encrypted period and shortens the monitoring time window from the normal duration to the emergency duration, so as to achieve intensive monitoring of the equipment status.

[0112] When any of the above-mentioned abnormal conditions are met, the system performs multimodal physical output on the human-machine interface (HMI) of the driver's cab visualization terminal.

[0113] Specifically, the graphics processor of the touch screen receives the warning level signal from the data processing unit and performs pixel-level real-time rendering with a bright red border and flashing icons in a prominent position on the main interface of the tunneling parameters. This rendering operation directly affects the display driving circuit, changing the pixel charge distribution of the LCD panel to form a visual warning. At the same time, the system drives the buzzer to emit a graded audio alarm associated with the type of abnormality. The first frequency is for single-source over-limit, the second frequency is for multi-source linkage abnormalities, and the third frequency is for mode switching abnormalities. In addition, the bright warning light strip integrated into the control panel indicates the location of the abnormality source with different flashing modes. For example, the high-frequency flashing on the left side of the light strip corresponds to an abnormality in the cutterhead area, and the low-frequency flashing on the right side of the light strip corresponds to an abnormality in the propulsion system.

[0114] At the same time, the control command issuing module establishes a communication connection with the tunnel boring machine PLC control cabinet through hard wire or industrial Ethernet, converts the decision adjustment scheme into directly executable electrical signals and writes them into the controllers of the corresponding actuators.

[0115] Specifically, the propulsion speed adjustment value is calculated based on the ratio of the current cutterhead torque to the thrust and the formation cohesion parameters provided by the advanced geological prediction system. This adjustment value is written to the speed setpoint port of the propulsion system inverter in the form of an analog voltage signal or a digital communication message, directly changing the extension and retraction rate of the propulsion cylinder. The cutterhead rotation speed adjustment value is calculated based on the cutting edge wear amount fed back by the electromagnetic inductance detection unit for cutter wear and the vibration spectrum characteristics at the current rotation speed, through a built-in cutter life prediction model. This model comprehensively considers the hardness of the cutter material, the cumulative number of tunneling rings, and the formation abrasiveness index. The calculation results are written to the opening control terminal of the hydraulic proportional valve driven by the cutterhead in the form of an electrical signal to adjust the displacement of the hydraulic motor. The upper limit of thrust adjustment value is determined by coupling calculation of real-time detection data of the shield tail gap and monitoring data of the slag discharge volume of the screw conveyor, which is used to prevent shield tail seal failure due to excessive thrust. The torque limit value is combined with the slag temperature sensor in the soil chamber and the temperature monitoring data of the cutterhead, and output through a thermo-mechanical coupling analysis model. This model establishes a physical mapping relationship between the frictional heat of the cutterhead and the torque load. When the temperature monitoring value is close to the annealing temperature of the tool material, the torque limit is automatically reduced to avoid thermal damage to the tool.

[0116] In terms of reinforcement grouting, the system calculates the pressure, rate and grout mix ratio parameters for secondary grouting based on the location and density data of the void area behind the wall fed back by the synchronous grouting detection system.

[0117] Specifically, the data processing unit receives the three-dimensional coordinates and void volume data of the voided area obtained by the back wall radar or acoustic detection device. Combined with the current formation permeability coefficient and groundwater pressure monitoring value, it calculates the required upper limit of grouting pressure and grout gelation time through the grouting diffusion model. The generated control parameters are written into the PID controller of the synchronous grouting pump, which directly adjusts the speed of the grouting pump and the opening of the pressure regulating valve of the grouting pipeline, so that the grout is injected into the designated back wall area at the calculated flow rate and pressure, realizing closed-loop control from geological defect detection to physical grouting action execution.

[0118] Through the complete technical chain described above, from geological sensing, physical acquisition of multi-source signals, synchronous fusion of heterogeneous data, visual and audio output of abnormal states, to the direct physical writing of control commands to frequency converters, hydraulic proportional valves and grouting pump PID controllers, this system solves the technical problems of shield tunneling in complex strata, such as missed reports of abnormal cutter wear due to fixed thresholds of a single sensor, false reports due to asynchronous multi-source data and electromagnetic interference, and shield instability caused by lag in control parameter adjustment when switching safety modes. It achieves deep technical integration between tunneling parameter monitoring and equipment physical control.

[0119] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A multi-source perception and auxiliary decision-making method for geological-mechanical-structural aspects in tunnel excavation, characterized in that: include: The reflection coefficient is calculated based on seismic wave data; the geological risk value is calculated based on the type and concentration of the target gas in the tunnel; the cumulative vibration energy is extracted based on the vibration signal during the rock breaking process; the cumulative wear is calculated based on the magnetic induction current of the tunnel boring machine cutter; and the lining thickness is calculated based on the elastic reflection wave of the lining structure. Based on the reflection coefficient, geological risk value, cumulative vibration energy, cumulative wear, and lining thickness, a spatial data matrix with mileage as the axis is constructed. A feasible construction parameter range is constructed based on the tunneling parameters. Within the feasible construction parameter range, a multi-objective optimization function is constructed with tunneling speed, tunneling specific energy, and cutter wear as indicators. Based on the single-source risk assessment of the spatial data matrix, the weight factors of each indicator in the multi-objective optimization function are adjusted, and the optimal combination of tunneling parameters is determined with the goal of maximizing the multi-objective optimization function.

2. The tunnel excavation geological-mechanical-structural multi-source perception and auxiliary decision-making method as described in claim 1, characterized in that, Based on the cutterhead rotation speed, thrust, torque, and tunneling speed, a feasible range of construction parameters is established. Within this range, a multi-objective optimization function is constructed as follows: ； In the formula, , , Respectively, the benefits of tunneling efficiency Energy consumption revenue Benefits of tool wear control The corresponding weighting factor; To be determined by the speed of the cutter head ,thrust Torque and tunneling speed A set of tunneling parameters; Among them, tunneling efficiency benefits for: ; Energy consumption benefits for: ； ； Benefits of tool wear control for: ； ； ； In the formula, s represents the mileage; Minimum propulsion speed; The maximum advancing speed is represented by A; A represents the cross-sectional area of ​​the tunnel excavation. This represents the maximum tunneling specific energy. Minimum tunneling specific energy; For tunneling specific energy; To predict tool wear; This refers to the wear threshold for scrapping the hob. This represents the current cumulative wear and tear. In order to combine tunneling parameters The increase in wear per unit distance of tunneling. The coefficient of formation erosion is denoted as .

3. The tunnel excavation geological-mechanical-structural multi-source perception and auxiliary decision-making method as described in claim 2, characterized in that, Based on a single-source risk assessment of the spatial data matrix, the process of adjusting the weighting factors of each indicator in the multi-objective optimization function includes: Multi-source data in the spatial data matrix are uniformly converted into standardized data within the range of 0 to 1. Based on the degree to which the standardized data deviates from the normal range, the single-source risk characterization value of each data source is obtained. Based on the risk characterization values ​​of each individual source and the preset risk threshold, the abnormal state at the current mileage is determined, and the construction control mode is selected accordingly. Based on the construction control mode, the corresponding weight is called.

4. The tunnel excavation geological-mechanical-structural multi-source perception and auxiliary decision-making method as described in claim 3, characterized in that, The single-source risk characterization value is: Let the first Data source in mileage The standardized eigenvalue at point is The normal range of values ​​is The minimum allowable threshold is The maximum allowed threshold is And satisfy ; when Within the normal range Within the time frame, single-source risk characterization value Set to 0; when hour, ; when hour, .

5. The multi-source perception and auxiliary decision-making method for tunnel excavation geology-machine-structure as described in claim 3, characterized in that, The process of selecting a construction control mode includes: Efficiency-first mode: When the single-source risk characterization value of all data sources... When both are 0, select the efficiency-first mode; When there is a single-source risk characterization value with one data source satisfy or single-source risk characterization values ​​from at least two data sources. satisfy However, when located at different mileage positions, the balanced tunneling mode should be selected; When there are single-source risk characterization values ​​from at least two data sources at the same mileage point satisfy Or a single-source risk characterization value from any one data source. satisfy When this happens, select the security priority mode; This is the risk threshold.

6. The tunnel excavation geological-mechanical-structural multi-source perception and auxiliary decision-making method as described in claim 1, characterized in that, Geological risk value The calculation formula is: In the formula, gas Geological weighting factor, gas The monitored concentration gas Normal formation background concentration, gas The standard deviation of the fluctuation, where n is the number of gas categories.

7. The multi-source sensing and auxiliary decision-making method for tunnel excavation geology-machine-structure as described in claim 1, characterized in that, The formula for calculating the lining thickness is: T represents the lining thickness. For longitudinal wave velocity, Main frequency value; The formula for calculating the reflection coefficient wave is: In the formula, The rock density of the strata on one side of the impedance difference interface. V1 represents the rock density of the strata on the other side of the wave impedance difference interface; V2 represents the wave velocity of the strata on one side of the wave impedance difference interface; and V3 represents the wave velocity of the strata on the other side of the wave impedance difference interface.

8. The multi-source perception and auxiliary decision-making method for tunnel excavation geology-machine-structure as described in claim 1, characterized in that, The tunnel boring machine-machine-structure multi-source sensing and auxiliary decision-making method also includes determining the geological state at the cutterhead location based on accumulated vibration energy, specifically including: Extract the root mean square value of the vibration signal; Calculate the ratio of the current cumulative vibration energy change rate to the average vibration energy change rate of the stable tunneling section. When the ratio is greater than the preset mutation coefficient, it is determined that an instantaneous mutation has occurred, and the cutterhead is currently at the interface of heterogeneous strata with uneven hardness. The strength determination result of the rock mass in the current area is obtained based on the range of the root mean square value and the cumulative vibration energy values ​​in the preset standard database.

9. The multi-source sensing and auxiliary decision-making method for tunnel excavation geology-machine-structure as described in claim 1, characterized in that, The tunnel boring machine-machine-structure multi-source perception and auxiliary decision-making method further includes determining the cumulative wear amount based on the magnetic induction current of the tunnel boring machine cutter head and the pre-calibrated standard curve corresponding to the magnetic flux density and cutter head wear amount, and identifying the uneven wear state based on the cumulative wear amount and the fluctuation frequency of the magnetic induction current within the cutter head rotation cycle; specifically, if the fluctuation frequency is consistent with the cutter head rotation frequency, it is determined that there is a fixed local wear point on the circumference of the cutter head, which is identified as uneven wear; when the cumulative wear amount is greater than or equal to the preset wear amount threshold, an alarm is automatically triggered.

10. A multi-source sensing and auxiliary decision-making system for geology, machinery, and structure in tunnel excavation, characterized in that: include: The multi-source sensing module is configured to calculate the reflection coefficient based on seismic wave data, calculate the geological risk value based on the type and concentration of target gas in the tunnel, extract the cumulative vibration energy based on the vibration signal during the rock breaking process, calculate the cumulative wear based on the magnetic induction current of the tunnel boring machine cutter, and calculate the lining thickness based on the elastic reflection wave of the lining structure. The auxiliary decision-making module is configured to construct a spatial data matrix with mileage as the axis based on reflection coefficient, geological risk value, cumulative vibration energy, cumulative wear, and lining thickness. It constructs a range of feasible construction parameters based on tunneling parameters. Within the range of feasible construction parameters, it constructs a multi-objective optimization function with tunneling speed, tunneling specific energy, and tool wear as indicators. Based on the single-source risk assessment of the spatial data matrix, it adjusts the weight factors of each indicator in the multi-objective optimization function and determines the optimal combination of tunneling parameters with the goal of maximizing the multi-objective optimization function.

Images