A shield machine-mounted underground cavity real-time beam detection imaging method

Abstract

The present application belongs to the field of shield construction advanced geological exploration and real-time imaging of underground cavities, and discloses a shield machine-mounted real-time beam detection and imaging method for underground cavities, which comprises device deployment and pre-calibration, multi-frequency waveform directional emission, echo signal global perception, time-space synchronization and noise suppression, cavity parameter inversion and real-time imaging, dynamic early warning and data updating; the detection host machine mounted on the shield machine continuously emits low-frequency penetrating waves and high-frequency positioning waves to the front of the working face, the ground beam sensing array synchronously receives echo signals, the central imaging processing terminal combines the shield machine attitude dynamic data, the sensor spatial coordinates and the effective echo waveform data to invert the cavity spatial position, the cavity burial depth, the cavity boundary contour, the cavity volume and the distance from the working face, and generates a cavity two-dimensional profile, a cavity three-dimensional stereoscopic imaging diagram and a hierarchical early warning result, thereby improving the real-time performance, the anti-interference performance and the imaging accuracy of the shield construction cavity hidden danger detection.

Description

Technical Field

[0001] This invention belongs to the field of advanced geological exploration technology for tunnel boring machine (TBM) construction, and particularly relates to a real-time beamforming imaging method for underground cavities mounted on a TBM. Background Technology

[0002] During the construction of subway, highway, and railway tunnels, tunnel boring machines (TBMs) need to continuously advance forward towards the tunnel face. If there are karst caves, ground voids, loose soil cavities, or water-rich cavities in front of the tunnel face, it can easily lead to risks such as TBM head collapse, cutterhead jamming, surface subsidence, water and mud inrush, and tunnel structural deformation. Due to the enclosed space, strong mechanical vibration, and significant interference from mud, water, and slag during TBM construction, it is difficult for construction personnel to directly observe the ground conditions in front of the tunnel face. Therefore, a real-time beamforming imaging method for underground cavities mounted on TBMs is needed. This method can simultaneously complete underground directional transmission, ground-wide reception, echo signal processing, cavity parameter inversion, real-time imaging, and dynamic early warning during TBM excavation, providing continuous and advanced geological risk assessment for TBM construction.

[0003] The existing technology has at least the following problems that need to be improved: (1) Existing advanced geological exploration methods mostly adopt single-point ground deployment, fixed detection inside the shield chamber and intermittent detection. The detection signals are easily affected by the mechanical vibration of the shield machine, mud and slag, ground traffic noise, construction machinery noise and environmental electromagnetic noise. In addition, there is a lack of dynamic data of shield machine attitude, spatiotemporal synchronization calibration between sensor spatial coordinates and echo signals, resulting in insufficient accuracy of cavity location, cavity burial depth and cavity boundary contour inversion.

[0004] (2) Existing detection methods can usually only output abnormal prompts or two-dimensional detection results. It is difficult to generate two-dimensional cross-sectional diagrams and three-dimensional stereoscopic imaging diagrams of cavities in real time during the continuous tunneling process of the tunnel boring machine. Furthermore, there is a lack of a linkage mechanism for graded early warning based on the volume of the cavity, the burial depth of the cavity, the distance from the tunnel face, and the probability of the cavity's existence. This results in delayed detection results, insufficient support for construction decision-making, and untimely handling of dangerous cavities. Summary of the Invention

[0005] To address the above issues and overcome the shortcomings of existing technologies, this invention provides a real-time beamforming imaging method for underground cavities mounted on a tunnel boring machine (TBM). Utilizing principles of underground dual-band directional transmission, global reception via a ground-based beam array, attitude synchronization calibration, spatiotemporal matching, adaptive noise suppression, cavity identification feature fusion, and real-time imaging early warning, this method generates effective echo waveform data, cavity spatial location, cavity boundary contour, cavity volume, two-dimensional cross-sectional view of the cavity, and three-dimensional stereoscopic image of the cavity. This enables continuous advanced detection, accurate inversion, and dynamic early warning during the TBM tunneling process, solving the problems of insufficient detection coverage, strong noise interference, imaging lag, and insufficient support for construction decision-making in existing methods.

[0006] The technical solution adopted in this invention is as follows: A real-time beam detection imaging method for underground cavities mounted on a tunnel boring machine (TBM), which is executed by a TBM-mounted detection host, a ground beam sensor array, a wireless data transmission unit, and a central imaging processing terminal, and includes the following steps: S1. Equipment Deployment and Pre-calibration: The detection host mounted on the tunnel boring machine is fixed inside the front shield diaphragm or on the back of the cutterhead spokes. The ground beam sensor array is deployed on the ground along the tunnel boring machine's tunneling axis. The wireless data transmission unit is connected to the detection host mounted on the tunnel boring machine, the ground beam sensor array, and the central imaging processing terminal. The central imaging processing terminal inputs the tunnel boring machine parameters, geological parameters, and sensor spatial coordinates to establish an initial detection calibration model. S2. Multi-frequency waveform directional transmission: The shield machine is equipped with a detection host that continuously transmits low-frequency penetrating waves and high-frequency positioning waves in front of the tunnel face during the shield machine tunneling process. The attitude synchronization calibration module collects the shield machine body attitude, cutterhead rotation angle and tunneling speed in real time, and dynamically corrects the beam transmission angle according to the shield machine body attitude, cutterhead rotation angle and tunneling speed. S3. Global Echo Signal Perception: The ground beam sensing array receives the echo signals formed after low-frequency penetrating waves and high-frequency positioning waves are reflected, refracted and attenuated by the strata, and collects the waveform distortion characteristics, amplitude attenuation characteristics, phase shift characteristics and arrival time characteristics in the echo signals. S4. Spatiotemporal synchronization and noise suppression: The central imaging processing terminal receives dynamic attitude data of the tunnel boring machine, echo signals collected by the ground beam sensor array, and sensor spatial coordinates. It performs spatiotemporal synchronization calibration according to a unified timestamp and adaptively filters out ground traffic noise, construction machinery noise, tunnel boring machine mechanical vibration noise, and environmental electromagnetic noise to generate effective echo waveform data. S5. Cavity parameter inversion and real-time imaging: The central imaging processing terminal extracts cavity identification features based on effective echo waveform data, shield machine attitude dynamic data and sensor spatial coordinates, inverts the spatial location, cavity depth, cavity boundary contour, cavity volume and distance from the tunnel face of the cavity, and generates a two-dimensional cross-sectional view and a three-dimensional stereoscopic image of the cavity. S6. Dynamic Early Warning and Data Update: The central imaging processing terminal compares the spatial location, burial depth, volume, distance from the tunnel face, and preset early warning threshold of the cavity to generate graded early warning results. As the tunnel boring machine continues to advance, it refreshes the detection data, two-dimensional cross-sectional view of the cavity, and three-dimensional stereoscopic image of the cavity.

[0007] As a preferred technical solution of this scheme, S1 specifically includes the installation of the detection host, the deployment of the ground array, the communication network, the spatial coordinate calibration, the input of construction parameters, and the establishment of the initial detection calibration model. In the installation of the detection host, the detection host is mounted on the tunnel boring machine and installed on the inside of the front shield bulkhead or the back of the cutterhead spokes of the tunnel boring machine using rigid fasteners. In the deployment of the ground array, multiple beam receiving sensors are deployed on the ground along the tunnel boring axis to form a linear sensor array or a grid sensor array. In the spatial coordinate calibration, the sensor spatial coordinates of each beam receiving sensor are collected by a spatial positioning calibration chip.

[0008] As a preferred technical solution of this scheme, S2 specifically includes transmission parameter reading, dual-band waveform generation, beam transmission angle calculation, attitude synchronization calibration, and directional beam transmission; in the dual-band waveform generation, the multi-frequency waveform transmission module generates low-frequency penetration waves and high-frequency positioning waves; in the attitude synchronization calibration, the attitude synchronization calibration module collects the shield machine body attitude, cutterhead rotation angle, and tunneling speed, and corrects the initial beam transmission angle according to the shield machine body attitude, cutterhead rotation angle, and tunneling speed to generate the corrected beam transmission angle.

[0009] As a preferred technical solution of this scheme, S3 and S4 specifically include echo signal reception, waveform distortion feature extraction, amplitude attenuation feature extraction, phase shift feature extraction, arrival time feature extraction, timestamp matching, attitude data binding, sensor coordinate binding, environmental noise identification, adaptive noise suppression, and effective echo waveform data generation; in the adaptive noise suppression, the central imaging processing terminal adaptively filters the echo signal according to the environmental noise identification result, and retains the waveform distortion features, amplitude attenuation features, phase shift features, and arrival time features related to the hole interface.

[0010] As a preferred technical solution of this scheme, S5 specifically includes cavity identification feature fusion, cavity existence probability calculation, cavity spatial location inversion, cavity boundary contour generation, cavity volume calculation, distance from the tunnel face calculation, and visualization imaging output. In the cavity identification feature fusion, the central imaging processing terminal fuses the waveform distortion features, amplitude attenuation features, phase shift features, arrival time features, shield machine attitude dynamic data, and sensor spatial coordinates in the effective echo waveform data to generate cavity identification features. In the visualization imaging output, the central imaging processing terminal generates a two-dimensional cross-sectional view and a three-dimensional stereoscopic image of the cavity based on the cavity spatial location, cavity burial depth, cavity boundary contour, cavity volume, and distance from the tunnel face.

[0011] As a preferred technical solution of this scheme, S6 specifically includes reading the early warning threshold, judging the hazard level, triggering the graded early warning, refreshing the imaging screen, generating tunneling parameter suggestions, and storing the detection data; in judging the hazard level, the central imaging processing terminal generates the hazard level judgment result based on the cavity volume, cavity burial depth, distance from the working face, and probability of cavity existence; in generating tunneling parameter suggestions, the central imaging processing terminal generates tunneling speed adjustment suggestions, propulsion force adjustment suggestions, cutterhead speed adjustment suggestions, and grouting reinforcement suggestions based on the graded early warning results.

[0012] The beneficial effects of the present invention after adopting the above technical solution are as follows: The present invention enables continuous advanced detection of cavities in front of the tunnel face by continuously emitting low-frequency penetrating waves and high-frequency positioning waves as the tunnel boring machine (TBM) advances, and receiving echo signals across the entire area through a ground beam sensing array; by synchronizing the dynamic attitude data of the TBM, the spatial coordinates of the sensors, and the echo signals in time and space, the impact of the mechanical movement of the TBM and the noise of the construction environment on the detection results can be reduced; by fusing cavity identification features, inverting cavity parameters, real-time imaging, and hierarchical early warning, two-dimensional cross-sectional images of cavities, three-dimensional stereoscopic images of cavities, and construction early warning information can be generated, improving the real-time performance, accuracy, and construction decision support capabilities for identifying cavity hazards in TBM construction.

[0013] The beneficial effects of this invention are as follows: (1) By utilizing the principle of directional transmission of the detection host mounted on the shield machine and the global reception of the ground beam sensor array, the detection host mounted on the shield machine continuously transmits low-frequency penetrating waves and high-frequency positioning waves in front of the tunnel face during the tunneling process. The ground beam sensor array receives the echo signals formed after reflection, refraction and energy attenuation of the strata. The technical effect is to obtain multi-source echo data covering the detection area in front of the tunnel face, which solves the problems of insufficient detection coverage, easy to miss small cavities and difficulty in identifying the boundaries of cavities in the existing single-point reception method.

[0014] (2) By using the principles of attitude synchronization calibration, unified timestamp matching and adaptive noise suppression, the attitude of the shield machine body, the rotation angle of the cutterhead, the tunneling speed, the spatial coordinates of the sensor and the echo signal are spatiotemporally synchronized and calibrated. Ground traffic noise, construction machinery noise, shield machine mechanical vibration noise and environmental electromagnetic noise are adaptively filtered out. The technical effect is to improve the stability of effective echo waveform data and the reliability of void identification features, and solve the problem of increased detection deviation caused by the inconsistency of shield machine mechanical movement, construction environment noise and receiver spatial reference.

[0015] (3) By utilizing the principles of cavity identification feature fusion, cavity parameter inversion, real-time visualization imaging and hierarchical early warning linkage, cavity identification features are generated based on waveform distortion features, amplitude attenuation features, phase shift features, arrival time features, shield machine attitude dynamic data and sensor spatial coordinates. The spatial location, cavity depth, cavity boundary contour, cavity volume and distance from the tunnel face are inverted, and a two-dimensional cross-sectional view of the cavity, a three-dimensional stereoscopic image of the cavity and hierarchical early warning results are generated. The technical effect is to transform echo anomalies into visualized cavity risk information, which solves the problems of existing detection results being lagging in imaging, incomplete parameter expression and difficulty in supporting shield dynamic construction decision-making. Attached Figure Description

[0016] Figure 1 This is a diagram showing the overall architecture of the device of the present invention; Figure 2 This is a flowchart of the real-time detection and imaging process proposed in this invention. Detailed Implementation

[0017] Example 1: The present invention provides a real-time beamforming imaging method for underground cavities mounted on a tunnel boring machine (TBM). The method is executed by a TBM-mounted detection host, a ground beamforming sensor array, a wireless data transmission unit, and a central imaging processing terminal. The method includes equipment deployment and pre-calibration steps, multi-frequency waveform directional transmission steps, echo signal global perception steps, spatiotemporal synchronization and noise suppression steps, cavity parameter inversion and real-time imaging steps, and dynamic early warning and data update steps.

[0018] In the equipment deployment and pre-calibration steps, the shield machine-mounted detection host is fixed inside the shield front shield diaphragm or on the back of the cutterhead spokes. The ground beam sensor array is deployed on the ground along the shield tunneling axis. The wireless data transmission unit is connected to the shield machine-mounted detection host, the ground beam sensor array, and the central imaging processing terminal. The central imaging processing terminal inputs the shield machine parameters, geological parameters, and sensor spatial coordinates to establish an initial detection calibration model.

[0019] In the multi-frequency waveform directional transmission step, the tunnel boring machine (TBM) is equipped with a detection host that continuously transmits low-frequency penetrating waves and high-frequency positioning waves in front of the tunnel face during the TBM tunneling process. The attitude synchronization calibration module collects the TBM body attitude, cutterhead rotation angle and tunneling speed in real time, and dynamically corrects the beam transmission angle according to the TBM body attitude, cutterhead rotation angle and tunneling speed.

[0020] In the echo signal global sensing step, the ground beam sensing array receives the echo signals formed after low-frequency penetrating waves and high-frequency positioning waves are reflected, refracted and attenuated by the strata, and collects the waveform distortion characteristics, amplitude attenuation characteristics, phase shift characteristics and arrival time characteristics of the echo signals.

[0021] In the spatiotemporal synchronization and noise suppression steps, the central imaging processing terminal receives the dynamic attitude data of the tunnel boring machine, the echo signal collected by the ground beam sensor array, and the spatial coordinates of the sensor. It performs spatiotemporal synchronization calibration according to a unified timestamp and adaptively filters out ground traffic noise, construction machinery noise, tunnel boring machine mechanical vibration noise, and environmental electromagnetic noise to generate effective echo waveform data.

[0022] In the cavity parameter inversion and real-time imaging steps, the central imaging processing terminal extracts cavity identification features based on effective echo waveform data, shield machine attitude dynamic data and sensor spatial coordinates, inverts the spatial location of the cavity in front of the tunnel face, cavity depth, cavity boundary contour, cavity volume and distance from the tunnel face, and generates a two-dimensional cross-sectional view and a three-dimensional stereoscopic image of the cavity.

[0023] In the dynamic early warning and data update process, the central imaging processing terminal compares the spatial location, burial depth, volume, distance from the tunnel face, and preset early warning threshold of the cavity to generate graded early warning results. As the tunnel boring machine continues to advance, it refreshes the detection data, two-dimensional cross-sectional view of the cavity, and three-dimensional stereoscopic image of the cavity.

[0024] Through the above steps, this method can form a continuous detection chain of underground directional transmission, ground-wide reception, real-time terminal imaging, and dynamic early warning updates, enabling the tunnel boring machine to obtain information on potential cavities in front of the tunnel face in real time during continuous tunneling, thus solving the problems of data lag, weak construction compatibility, and unclear cavity boundary identification in traditional intermittent detection methods.

[0025] Example 2: This example is based on all the above examples. The equipment deployment and pre-calibration steps specifically include the installation of the detection host, the deployment of the ground array, the communication network setup, the spatial coordinate calibration, the input of construction parameters, and the establishment of the initial detection calibration model.

[0026] During the installation of the detection host, the detection host is mounted on the shield machine and installed on the inside of the shield front shield bulkhead or the back of the cutterhead spokes using rigid fasteners, so that the detection host mounted on the shield machine does not interfere with the cutterhead cutting operation and the excavated soil transportation process.

[0027] In the ground array deployment, multiple beam receiving sensors are deployed on the ground along the tunnel boring machine axis. Adjacent beam receiving sensors maintain a preset spacing and form a linear sensor array or a grid sensor array.

[0028] In the communication network, a data communication link is established between the tunnel boring machine-mounted detection host, the ground beam sensor array, and the central imaging processing terminal through a wireless data transmission unit.

[0029] In spatial coordinate calibration, the spatial coordinates of each beam receiving sensor are collected by the spatial positioning calibration chip and uploaded to the central imaging processing terminal.

[0030] During the construction parameter input process, the central imaging processing terminal inputs the shield machine diameter, front shield position, cutterhead position, tunneling axis, current face position, geological parameters, and preset early warning thresholds.

[0031] During the initial detection and calibration model establishment, the central imaging processing terminal establishes the initial detection and calibration model based on the shield machine parameters, geological parameters, sensor spatial coordinates, and the current position of the tunnel face.

[0032] Through the above processing, the equipment deployment and pre-calibration steps can unify the underground transmitter, the ground receiver, and the terminal imaging end into the same spatial coordinate system, solving the problem of inconsistent spatial reference between the shield machine-mounted equipment and the ground sensor array.

[0033] Example 3: This example is based on all the above examples. The multi-frequency waveform directional transmission steps specifically include transmission parameter reading, dual-band waveform generation, beam transmission angle calculation, attitude synchronization calibration, and directional beam transmission.

[0034] During the launch parameter reading, the tunnel boring machine (TBM) carries a detection host to read the geological parameters of the strata, the current position of the tunnel face, the preset detection depth, and the preset launch power from the initial detection calibration model.

[0035] In the dual-band waveform generation, the multi-frequency waveform transmission module generates low-frequency penetration waves and high-frequency positioning waves. The low-frequency penetration waves are suitable for deep penetration detection, and the high-frequency positioning waves are suitable for cavity boundary positioning.

[0036] In the beam emission angle calculation, the tunnel boring machine (TBM) is equipped with a detection host that generates an initial beam emission angle based on the current position of the tunnel face and the preset detection direction.

[0037] During attitude synchronization calibration, the attitude synchronization calibration module collects the shield machine body attitude, cutterhead rotation angle and tunneling speed, and corrects the initial beam emission angle based on the shield machine body attitude, cutterhead rotation angle and tunneling speed to generate the corrected beam emission angle.

[0038] During directional beam transmission, the multi-frequency waveform transmission module continuously transmits dual-band directional beams in front of the tunnel face according to the corrected beam transmission angle.

[0039] Through the above processing, the multi-frequency waveform directional transmission step can maintain the consistency between the beam transmission direction and the tunneling direction during the tunnel boring machine's excavation and cutterhead rotation, thus solving the problem of the detection beam direction shift caused by the mechanical movement of the tunnel boring machine.

[0040] Example 4: This example is based on all the above examples. The echo signal global sensing step specifically includes echo signal reception, waveform distortion feature extraction, amplitude attenuation feature extraction, phase shift feature extraction, arrival time feature extraction, and valid echo uploading.

[0041] In echo signal reception, the ground beam sensing array receives the echo signal formed after the dual-band directional beam propagates through the strata.

[0042] In waveform distortion feature extraction, the beam receiving sensor extracts the waveform distortion features of the echo signal relative to the transmitted waveform.

[0043] In amplitude attenuation feature extraction, the beam receiving sensor extracts the amplitude attenuation features of the echo signal relative to the transmitted waveform.

[0044] In phase offset feature extraction, the beam receiving sensor extracts the phase offset features of the echo signal relative to the transmitted waveform.

[0045] In the time of arrival feature extraction, the beam receiving sensor extracts the time of arrival features of the same echo signal received by different beam receiving sensors.

[0046] During effective echo uploading, the ground beam sensing array uploads waveform distortion characteristics, amplitude attenuation characteristics, phase shift characteristics, arrival time characteristics, and sensor spatial coordinates to the central imaging processing terminal via a wireless data transmission unit.

[0047] Through the above processing, the echo signal global sensing step can synchronously acquire waveform changes caused by the cavity interface from multiple receiving locations on the ground, solving the problems of missing small cavities and being unable to determine cavity boundaries when receiving at a single point.

[0048] Example 5: This example is based on all the above examples. The spatiotemporal synchronization and noise suppression steps specifically include timestamp matching, attitude data binding, sensor coordinate binding, environmental noise identification, adaptive noise suppression, and effective echo waveform data generation.

[0049] During timestamp matching, the central imaging processing terminal matches the tunnel boring machine's dynamic attitude data, echo signals, and sensor spatial coordinates based on a unified timestamp.

[0050] In attitude data binding, the central imaging processing terminal binds the shield machine's body attitude, cutterhead rotation angle, and tunneling speed with the echo signal at the same acquisition time.

[0051] In sensor coordinate binding, the central imaging processing terminal binds the sensor spatial coordinates of each beam receiving sensor with the corresponding echo signal.

[0052] In environmental noise identification, the central imaging processing terminal identifies ground traffic noise, construction machinery noise, tunnel boring machine vibration noise, and environmental electromagnetic noise.

[0053] In adaptive noise suppression, the central imaging processing terminal adaptively filters out echo signals based on the environmental noise identification results, while retaining waveform distortion features, amplitude attenuation features, phase shift features, and arrival time features related to the hole interface.

[0054] During the generation of effective echo waveform data, the central imaging processing terminal generates effective echo waveform data based on the echo signal after noise filtering.

[0055] Through the above processing, the spatiotemporal synchronization and noise suppression steps can eliminate the time deviation between underground transmission, ground reception and shield attitude acquisition, reduce the impact of construction environment noise on cavity identification, and solve the problems of echo signal distortion and unstable cavity feature extraction.

[0056] Example 6: This example is based on all the above examples. The cavity parameter inversion and real-time imaging steps specifically include cavity identification feature fusion, cavity existence probability calculation, cavity spatial location inversion, cavity boundary contour generation, cavity volume calculation, distance from the working face calculation, and visualization imaging output.

[0057] In the cavity identification feature fusion process, the central imaging processing terminal fuses the waveform distortion features, amplitude attenuation features, phase shift features, arrival time features, shield machine attitude dynamic data, and sensor spatial coordinates in the effective echo waveform data to generate cavity identification features.

[0058] In calculating the probability of a cavity's existence, the central imaging processing terminal calculates the probability of a cavity's existence based on its cavity identification features. The probability of a cavity's existence is calculated according to Formula 1:

[0059] In Formula 1, The probability of a void existing. For normalized mapping functions, This is a waveform distortion characteristic. It is an amplitude attenuation characteristic. This is a phase shift characteristic. For arrival time characteristics, This refers to the spatial attitude fusion feature formed by the sensor spatial coordinates and the dynamic attitude data of the tunnel boring machine. , , , and These are the weight coefficients for the corresponding features. This is the bias parameter.

[0060] In the cavity spatial location inversion, the central imaging processing terminal inverts the cavity spatial location based on the cavity existence probability, arrival time characteristics, sensor spatial coordinates, and the current position of the tunnel face.

[0061] During the generation of the cavity boundary contour, the central imaging processing terminal generates the cavity boundary contour based on the cavity existence probability distribution and waveform distortion feature distribution corresponding to different beam receiving sensors.

[0062] In the calculation of cavity volume, the central imaging processing terminal calculates the cavity volume based on the cavity boundary contour and cavity burial depth.

[0063] In the calculation of the distance from the working face, the central imaging processing terminal calculates the distance from the working face based on the location of the cavity space and the current position of the working face.

[0064] In the visualization imaging output, the central imaging processing terminal generates a two-dimensional cross-sectional view and a three-dimensional stereoscopic image of the cavity based on the cavity's spatial location, burial depth, boundary contour, volume, and distance from the working face.

[0065] Regarding parameter adjustments: The first step is to increase the weighting coefficient corresponding to the amplitude attenuation feature when the echo signal corresponding to the low-frequency transmission wave is too weak.

[0066] The second step is to increase the weighting coefficients of waveform distortion features and phase shift features when the boundary reflection corresponding to the high-frequency positioning wave is obvious.

[0067] The third step is to increase the weighting coefficient of the spatial attitude fusion feature formed by the spatial coordinates of the sensors and the dynamic attitude data of the tunnel boring machine when the spacing of the ground beam sensor array increases.

[0068] Fourth, when increased noise from construction machinery causes fluctuations in arrival time characteristics, the weighting coefficient corresponding to the arrival time characteristics is reduced, and the adaptive noise suppression strength is increased.

[0069] Through the above processing, the cavity parameter inversion and real-time imaging steps can transform the changes in echo signals into the cavity spatial location, cavity depth, cavity boundary contour, and cavity volume, solving the problem that traditional detection methods can only provide anomaly alerts but cannot output real-time three-dimensional cavity imaging results.

[0070] Example 7: This example is based on all the above examples. The dynamic early warning and data update steps specifically include reading the early warning threshold, judging the danger level, triggering the graded early warning, refreshing the imaging screen, generating tunneling parameter suggestions, and storing the detection data.

[0071] During the reading of the warning threshold, the central imaging processing terminal reads the preset warning threshold, which includes the cavity volume threshold, the cavity burial depth threshold, the distance from the working face threshold, and the cavity existence probability threshold.

[0072] In the hazard level assessment, the central imaging processing terminal generates the hazard level assessment result based on the cavity volume, cavity burial depth, distance from the working face, and probability of cavity existence.

[0073] During the tiered early warning triggering process, when the hazard level assessment result reaches the preset early warning threshold, the central imaging processing terminal triggers the tiered early warning and pushes the tiered early warning result to the construction control platform.

[0074] As the imaging screen refreshes, the central imaging processing terminal updates the current position of the tunnel face as the tunnel boring machine continues to advance, and refreshes the two-dimensional profile and three-dimensional stereoscopic imaging of the cavity based on the new echo signals.

[0075] The tunneling parameter suggestions are being generated. The central imaging processing terminal generates suggestions for adjusting tunneling speed, thrust, cutterhead speed, and grouting reinforcement based on the graded early warning results.

[0076] During the data storage process, the central imaging processing terminal records the shield machine's dynamic attitude data, echo signals, sensor spatial coordinates, effective echo waveform data, two-dimensional profile of the cavity, three-dimensional stereoscopic imaging of the cavity, hazard level judgment results, graded early warning results, and tunneling parameter suggestions.

[0077] Through the above processing, the dynamic early warning and data update steps can continuously refresh the cavity detection results as the tunnel boring machine advances, and transform the cavity risk into actionable early warning information and tunneling parameter suggestions, solving the problems of lagging traditional detection results and difficulty in supporting dynamic construction decisions.

Claims

1. A method for real-time beamforming imaging of underground cavities mounted on a tunnel boring machine (TBM), comprising a TBM-mounted detection host, a ground beam sensor array, a wireless data transmission unit, and a central imaging processing terminal, characterized in that: Includes the following steps: S1. Equipment Deployment and Pre-calibration: The detection host mounted on the tunnel boring machine is fixed inside the front shield diaphragm or on the back of the cutterhead spokes. The ground beam sensor array is deployed on the ground along the tunnel boring machine's tunneling axis. The wireless data transmission unit is connected to the detection host mounted on the tunnel boring machine, the ground beam sensor array, and the central imaging processing terminal. The central imaging processing terminal inputs the tunnel boring machine parameters, geological parameters, and sensor spatial coordinates to establish an initial detection calibration model. S2. Multi-frequency waveform directional transmission: The shield machine is equipped with a detection host that continuously transmits low-frequency penetrating waves and high-frequency positioning waves in front of the tunnel face during the shield machine tunneling process. The attitude synchronization calibration module collects the shield machine body attitude, cutterhead rotation angle and tunneling speed in real time, and dynamically corrects the beam transmission angle according to the shield machine body attitude, cutterhead rotation angle and tunneling speed. S3. Global Echo Signal Perception: The ground beam sensing array receives the echo signals formed after low-frequency penetrating waves and high-frequency positioning waves are reflected, refracted and attenuated by the strata, and collects the waveform distortion characteristics, amplitude attenuation characteristics, phase shift characteristics and arrival time characteristics in the echo signals. S4. Spatiotemporal synchronization and noise suppression: The central imaging processing terminal receives dynamic attitude data of the tunnel boring machine, echo signals collected by the ground beam sensor array, and sensor spatial coordinates. It performs spatiotemporal synchronization calibration according to a unified timestamp and adaptively filters out ground traffic noise, construction machinery noise, tunnel boring machine mechanical vibration noise, and environmental electromagnetic noise to generate effective echo waveform data. S5. Cavity parameter inversion and real-time imaging: The central imaging processing terminal extracts cavity identification features based on effective echo waveform data, shield machine attitude dynamic data and sensor spatial coordinates, inverts the spatial location, cavity depth, cavity boundary contour, cavity volume and distance from the tunnel face of the cavity, and generates a two-dimensional cross-sectional view and a three-dimensional stereoscopic image of the cavity. S6. Dynamic Early Warning and Data Update: The central imaging processing terminal compares the spatial location, burial depth, volume, distance from the tunnel face, and preset early warning threshold of the cavity to generate graded early warning results. As the tunnel boring machine continues to advance, it refreshes the detection data, two-dimensional cross-sectional view of the cavity, and three-dimensional stereoscopic image of the cavity.

2. The method for real-time beamforming imaging of underground cavities mounted on a tunnel boring machine according to claim 1, characterized in that: S1 specifically includes the installation of the detection host, the deployment of the ground array, the communication network, the spatial coordinate calibration, the input of construction parameters, and the establishment of the initial detection calibration model. In the installation of the detection host, the detection host is mounted on the shield machine and installed on the inside of the shield front shield bulkhead or the back of the cutterhead spokes of the shield machine using rigid fasteners. In the deployment of the ground array, multiple beam receiving sensors are deployed on the ground along the shield tunneling axis, with a preset spacing between adjacent beam receiving sensors, forming a linear sensor array or a grid sensor array. In spatial coordinate calibration, the spatial coordinates of each beam receiving sensor are collected by the spatial positioning calibration chip and uploaded to the central imaging processing terminal.

3. The method for real-time beamforming imaging of underground cavities mounted on a tunnel boring machine according to claim 2, characterized in that: S2 specifically includes transmission parameter reading, dual-band waveform generation, beam transmission angle calculation, attitude synchronization calibration, and directional beam transmission. In the transmission parameter reading, the tunnel boring machine (TBM) carries a detection host to read the geological parameters of the strata, the current position of the tunnel face, the preset detection depth, and the preset transmission power from the initial detection calibration model. In the dual-band waveform generation, the multi-frequency waveform transmission module generates low-frequency penetration waves and high-frequency positioning waves. In the attitude synchronization calibration, the attitude synchronization calibration module collects the attitude of the TBM body, the cutterhead rotation angle, and the tunneling speed, and corrects the initial beam transmission angle to generate the corrected beam transmission angle.

4. The method for real-time beamforming imaging of underground cavities mounted on a tunnel boring machine according to claim 3, characterized in that: S3 specifically includes echo signal reception, waveform distortion feature extraction, amplitude attenuation feature extraction, phase shift feature extraction, arrival time feature extraction, and effective echo uploading. In echo signal reception, the ground beam sensing array receives the echo signal formed after the dual-band directional beam propagates through the strata. In effective echo uploading, the ground beam sensing array uploads the waveform distortion features, amplitude attenuation features, phase shift features, arrival time features, and sensor spatial coordinates to the central imaging processing terminal through the wireless data transmission unit.

5. The method for real-time beamforming imaging of underground cavities mounted on a tunnel boring machine according to claim 4, characterized in that: S4 specifically includes timestamp matching, attitude data binding, sensor coordinate binding, environmental noise identification, adaptive noise suppression, and effective echo waveform data generation; in timestamp matching, the central imaging processing terminal matches the shield machine attitude dynamic data, echo signal, and sensor spatial coordinates according to a unified timestamp; In adaptive noise suppression, the central imaging processing terminal adaptively filters out echo signals based on the environmental noise identification results, while retaining waveform distortion features, amplitude attenuation features, phase shift features, and arrival time features related to the hole interface.

6. The method for real-time beamforming imaging of underground cavities mounted on a tunnel boring machine according to claim 5, characterized in that: S5 specifically includes cavity identification feature fusion, cavity existence probability calculation, cavity spatial location inversion, cavity boundary contour generation, cavity volume calculation, distance from the tunnel face calculation, and visualization imaging output. In the cavity identification feature fusion, the central imaging processing terminal fuses the waveform distortion features, amplitude attenuation features, phase shift features, arrival time features, shield machine attitude dynamic data, and sensor spatial coordinates in the effective echo waveform data to generate cavity identification features. In the visualization imaging output, the central imaging processing terminal generates a two-dimensional cross-sectional view and a three-dimensional stereoscopic image of the cavity based on the cavity spatial location, cavity depth, cavity boundary contour, cavity volume, and distance from the tunnel face.

7. The method for real-time beamforming imaging of underground cavities mounted on a tunnel boring machine according to claim 6, characterized in that: In the cavity existence probability calculation, the central imaging processing terminal calculates the cavity existence probability based on cavity identification features; in the cavity spatial location inversion, the central imaging processing terminal inverts the cavity spatial location based on the cavity existence probability, arrival time characteristics, sensor spatial coordinates, and the current position of the tunnel face; in the cavity boundary contour generation, the central imaging processing terminal generates the cavity boundary contour based on the cavity existence probability distribution and waveform distortion feature distribution corresponding to different beam receiving sensors.

8. The method for real-time beamforming imaging of underground cavities mounted on a tunnel boring machine according to claim 7, characterized in that: S6 specifically includes reading the warning threshold, judging the hazard level, triggering the graded warning, refreshing the imaging screen, generating tunneling parameter suggestions, and storing detection data; in judging the hazard level, the central imaging processing terminal generates the hazard level judgment result based on the cavity volume, cavity burial depth, distance from the working face, and probability of cavity existence; in triggering the graded warning, when the hazard level judgment result reaches the preset warning threshold, the central imaging processing terminal triggers the graded warning and pushes the graded warning result to the construction control platform; The tunneling parameter suggestions are being generated. The central imaging processing terminal generates suggestions for adjusting tunneling speed, thrust, cutterhead speed, and grouting reinforcement based on the graded early warning results.

Images