An adaptive spray-mix reinforcement method and system for TBM excavation face balancing

Abstract

The present invention relates to the technical field of tunnel construction, and in particular to an adaptive spray-mix reinforcement method and system for balancing the TBM excavation face. The spray-mix reinforcement method includes: real-time acquisition of excavation data, and calculation of a stability evaluation index based on the excavation data by fusing a weight coefficient and a dynamically updated threshold; when the stability evaluation index exceeds a preset threshold, pausing excavation and starting the spray-mix operation; acquiring images of the tunnel face before and after spraying and performing alignment processing to generate a differential mask, calculating the coverage, roughness, and void ratio based on the differential mask to comprehensively evaluate the spray-mix quality, and performing supplementary spraying or resuming excavation operations based on the evaluation results. The present invention automatically controls the timing, position, and amount of the spray-mix operation through real-time detection and intelligent evaluation of the surrounding rock state, and performs a spray-mix quality evaluation after the spray-mix is ​​completed, thereby achieving precise and efficient reinforcement and improving construction safety and efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel construction, and in particular to an adaptive spray-mix reinforcement method and system for balancing a TBM excavation face. Background Art

[0002] The TBM (Transport Brick Machine) is an automated tunnel construction machine that integrates excavation, support, and mucking functions. It is widely used in underground projects such as subways, railways, and water conservancy projects. During the excavation process, the stability of the surrounding rock mass directly affects construction safety and efficiency. In weak or broken strata, rock instability can lead to landslides, cutterhead jams, and other problems. To address these issues, spray-mix reinforcement is often used. This involves using shotcrete to rapidly support the surrounding rock mass and prevent instability.

[0003] Traditional shotcrete reinforcement methods usually rely on manual judgment of the surrounding rock conditions and operation of equipment. Its typical composition includes a spraying device installed on a TBM, and manual control of the timing and quality of shotcrete spraying based on the surrounding rock conditions.

[0004] Existing technologies suffer from the following problems: First, manual judgment of surrounding rock conditions is susceptible to subjective factors, resulting in slow response and difficulty in capturing dynamic changes in the surrounding rock. This can delay reinforcement and increase the risk of collapse. Second, manual operation has limited precision, making it difficult to achieve high-precision spraying positioning and concrete dosage control. This can lead to insufficient or excessive local reinforcement. The former fails to effectively support the surrounding rock, while the latter wastes materials and increases construction costs. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the purpose of the embodiments of the present invention is to provide an adaptive spray-mix reinforcement method for TBM excavation face balancing. Through real-time detection and intelligent evaluation of the surrounding rock status, the timing, position and amount of the spray-mix operation are automatically controlled, and the spray-mix quality is evaluated after the spray-mix is ​​completed, so as to achieve accurate and efficient reinforcement, improve construction safety and efficiency, and reduce the cost of manual intervention.

[0006] In order to achieve the above objectives, the embodiments of the present invention provide the following technical solutions:

[0007] An adaptive spray-mix reinforcement method for balancing a TBM excavation face comprises: collecting tunneling data in real time; calculating a stability evaluation index based on the tunneling data by fusing a weight coefficient and a dynamically updated threshold; pausing tunneling and initiating a spray-mix operation when the stability evaluation index exceeds a preset threshold; collecting and aligning tunnel face images before and after spraying to generate a differential mask; calculating coverage, roughness, and void ratio based on the differential mask to comprehensively evaluate the spray-mix quality; and performing supplemental spraying or resuming tunneling operations based on the evaluation results.

[0008] Optionally, the calculation of the stability evaluation index by fusing the weight coefficient and the dynamically updated threshold includes:

[0009] ;

[0010] in, Stability evaluation index, is the cutter head torque at the current moment, It is the reference torque for stable working conditions; is the cutter head pressure at the current moment, is the design pressure; is the shield deformation at the current moment, is the historical normal deformation; are the weight coefficients of cutter head torque, pressure, and shield deformation in comprehensive evaluation; is the threshold value, which is updated using statistical methods;

[0011] The specific update formula is as follows:

[0012] ;

[0013] in, , For nearly N samples The mean and standard deviation of the data, k is the confidence coefficient, the smaller the k value, the more conservative it is.

[0014] Optionally, a graded judgment is made based on the stability evaluation index, and the specific steps include: when the stability evaluation index is within the normal range, maintain normal excavation; when the stability evaluation index exceeds the normal range but does not reach the warning threshold, reduce the advancement speed for a preset time, and do not issue a warning if the stability evaluation index falls back to the safe range; if the cumulative time that the stability evaluation index exceeds the normal range within the preset time exceeds the set value, issue a warning and start spray-mix reinforcement; when the stability evaluation index of multiple consecutive sampling points exceeds the normal range, stop excavation directly and start the spray-mix reinforcement task.

[0015] Optionally, key point features of the images before and after spraying are extracted, and the key point features are extracted and matched by a sparse feature matching method, and a homography matrix is ​​calculated to achieve pixel-level alignment of the post-spraying image and the reference image; the aligned reference image and the post-spraying image are subjected to grayscale difference, color difference and edge gradient analysis respectively to generate a comprehensive differential feature score map, and the optimal threshold is automatically determined by an adaptive threshold segmentation technique to extract a binary mask of the sprayed concrete coverage area to generate a differential mask for spray-mix quality evaluation.

[0016] Optionally, graded processing is performed according to the comprehensive evaluation index, and the specific steps include: when the comprehensive evaluation index reaches the qualified standard, automatically instructing to retract the nozzle and resume excavation construction; when the comprehensive evaluation index is in the critical range, automatically identifying the defective area and guiding the supplementary spraying operation, and then re-evaluating until the qualified standard is reached; when the comprehensive evaluation index is lower than the qualified standard, automatically executing the second supplementary spraying operation. If the standard is still not met after the second supplementary spraying, manual on-site confirmation is carried out.

[0017] An embodiment of the present invention also provides an adaptive spray-mix reinforcement system for TBM excavation face balancing, including an excavation face stability detection module, a spray-mix reinforcement module and a spray-mix effect evaluation module; the excavation face stability detection module is used to collect excavation data in real time, and calculate the stability evaluation index based on the excavation data by fusing weight coefficients and dynamically updating thresholds; the spray-mix reinforcement module is used to suspend excavation and start spray-mixing operations when the stability evaluation index exceeds a preset threshold; the spray-mix effect evaluation module is used to collect images of the tunnel face before and after spraying and align them to generate a differential mask, calculate the coverage, roughness and void ratio based on the differential mask, and then comprehensively evaluate the spray-mix quality, and perform supplementary spraying or resume excavation operations according to the evaluation results.

[0018] Optionally, the excavation face stability detection module includes a torque sensor installed on the cutterhead drive spindle, a pressure sensor in the cutterhead cavity, a displacement sensor at a key position of the shield, and a data processing terminal connected to the torque sensor, pressure sensor and displacement sensor.

[0019] Optionally, the spray-mix reinforcement module includes a receiving plate and a nozzle; the receiving plate is rotatably mounted on the main shaft of the cutter disc and is located on the rear side of the main shaft, and an electric telescopic rod and a receiving chamber are provided on the side of the receiving plate facing the cutter disc, the nozzle is installed in the receiving chamber, and a through hole is provided on the cutter disc; during normal excavation, the electric telescopic rod is inserted into the through hole, and during spray-mix reinforcement, the receiving plate rotates relative to the cutter disc, and the nozzle is moved out of the receiving chamber and passes through the through hole of the cutter disc for spray-mix reinforcement.

[0020] Optionally, the spray-mix reinforcement module also includes a hopper, a mixing system, a cement injection pump and an air compressor; the hopper serves as a storage space for dry materials, the mixing system mixes the dry materials with water and additives, the cement injection pump uses pumping pressure to transport the mixed concrete through a pipeline to the nozzle, and the air compressor provides the required air power for injection.

[0021] Optionally, a guide rail is provided in the storage chamber, a base is slidably mounted on the guide rail, a spray arm is mounted on the base, and the nozzle is mounted on the spray arm.

[0022] One or more technical solutions provided in the embodiments of the present invention have at least the following technical effects or advantages:

[0023] This spray-mix reinforcement method collects excavation data in real time and then calculates a stability evaluation index by fusing weight coefficients and dynamically updating thresholds. This method can promptly capture changes in excavation face stability and ensure construction safety. When the stability evaluation index exceeds the preset threshold, excavation operations are suspended and spray-mixing operations are immediately initiated to reinforce the excavation face, promptly addressing unstable conditions and preventing accidents. By collecting images of the tunnel face before and after spraying, aligning them and generating a differential mask, coverage, roughness, and void ratio are calculated based on this data, thereby comprehensively evaluating the spray-mix quality. Based on the evaluation results, it is determined whether additional spraying is required or whether excavation operations can be resumed. This method automatically determines the stability of the surrounding rock by fusing excavation data in real time and combines it with image intelligent evaluation technology to achieve closed-loop control of spray-mix quality. This method solves the problems of delayed response, insufficient accuracy, and low construction efficiency in existing spray-mix reinforcement technologies, ensuring spray-mix quality, reducing resource waste, and improving construction efficiency. Ultimately, it enhances the continuous excavation capability of the TBM while ensuring excavation face stability.

[0024] Additional advantages of the present invention will be given in the description which follows, and in part will be obvious from the description which follows, or may be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] To more clearly illustrate the embodiments of this application or the technical solutions in the prior art, the following briefly introduces the drawings required for the embodiments or the description of the prior art. Obviously, the drawings described below are only some of the embodiments described in this application. For those skilled in the art, other drawings can be derived from these drawings without inventive effort. In addition, the spacing or size of the components are exaggerated to show the position of the components, and the schematic diagrams are for illustrative purposes only.

[0026] Figure 1 1 is a flow chart of a spray-mix reinforcement method according to an embodiment of the present invention;

[0027] Figure 2 is a schematic diagram of a spray-mix reinforcement system provided by an embodiment of the present invention;

[0028] Figure 3 This is a schematic diagram of a normal excavation state provided by an embodiment of the present invention;

[0029] Figure 4 1 is a schematic diagram of the end face of the cutter head provided by an embodiment of the present invention;

[0030] Figure 5 is a schematic diagram of the end face of a storage tray provided in an embodiment of the present invention;

[0031] Figure 6 This is a schematic diagram of the installation of a spray arm provided by an embodiment of the present invention;

[0032] Figure: 1. Cutter head; 2. Electric telescopic rod; 3. Storage tray; 4. Connecting tray; 5. Air compressor; 6. Hopper; 7. Mixing system; 8. Jet pump; 9. Spindle; 10. Through hole; 11. Storage chamber; 12. Nozzle; 13. Camera; 14. Hose; 15. Spray arm; 16. Base. DETAILED DESCRIPTION

[0033] It should be noted that the following detailed description is illustrative and is intended to further illustrate the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as commonly understood by those of ordinary skill in the art to which the present invention belongs. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0034] Example 1

[0035] like Figure 1 As shown, this embodiment proposes an adaptive spray-mix reinforcement method for TBM excavation face balancing, including: real-time acquisition of excavation data, and calculation of a stability evaluation index based on the excavation data by fusing a weight coefficient and a dynamically updated threshold; when the stability evaluation index exceeds a preset threshold, pausing excavation and starting the spray-mix operation; acquiring and aligning the tunnel face images before and after spraying to generate a differential mask, calculating the coverage, roughness, and void ratio based on the differential mask to comprehensively evaluate the spray-mix quality, and performing supplementary spraying or resuming excavation operations according to the evaluation results.

[0036] This method addresses the lag in response of traditional manual judgment by collecting excavation data in real time and dynamically updating thresholds. By integrating weighting coefficients (such as cutterhead torque, pressure, and shield deformation), combined with normalization and data fusion algorithms, it more accurately reflects the stability of the surrounding rock mass. When the stability evaluation index exceeds the threshold, excavation is automatically suspended and the shotcrete mixing operation is initiated, avoiding the risk of collapse caused by manual delays. The shotcrete mixing performance evaluation utilizes image alignment and differential mask generation techniques to quantitatively analyze the coverage, roughness, and void ratio of the shotcrete layer, forming a closed-loop control system to ensure that support quality meets standards. The coverage ratio quantifies the effective area of ​​concrete coverage on the excavation face, ensuring that the surrounding rock mass is fully encapsulated and avoiding the risk of instability caused by localized exposure. The roughness ratio evaluates the uniformity and smoothness of the concrete spraying by analyzing the edge gradient difference between the shotcrete layer surface and the baseline image, preventing stress concentration or subsequent construction obstacles caused by surface irregularities. The void ratio identifies areas of missed shotcrete or void defects, ensuring the continuity of the support structure and preventing local defects from becoming a breakthrough point for surrounding rock failure. By combining "coverage, roughness and void ratio", the spray-mix quality can be comprehensively evaluated from the three dimensions of coverage, surface quality and structural integrity. This can comprehensively and accurately quantify the construction effect of sprayed concrete, provide a reliable basis for the closed-loop control of spray-mix quality, ensure that the support quality meets the standards, reduce resource waste and improve construction efficiency.

[0037] The excavation data includes cutterhead torque, cutterhead pressure, and shield deformation. Cutterhead torque reflects the ease with which the cutterhead 1 breaks rock, cutterhead pressure reflects the stress state at the tunnel face, and shield deformation indicates whether the shield is evenly stressed. By monitoring this data, the stability of the excavation face can be comprehensively and accurately assessed, providing a reliable basis for subsequent spray-concrete reinforcement decisions.

[0038] A graded judgment is made based on the stability evaluation index. The specific steps include: when the stability evaluation index is within the normal range, maintain normal excavation; when the stability evaluation index exceeds the normal range but does not reach the warning threshold, reduce the advancement speed for a preset time. If the stability evaluation index falls back to the safe range, no warning is issued; if the cumulative time that the stability evaluation index exceeds the normal range within the preset time exceeds the set value, a warning is issued and spray-mix reinforcement is started; when the stability evaluation index of multiple consecutive sampling points exceeds the normal range, excavation is stopped directly and the spray-mix reinforcement task is started.

[0039] A graded assessment mechanism improves construction efficiency by responding to surrounding rock instability risks in stages. When indicators exceed normal ranges but fall below warning thresholds, the propulsion speed is reduced rather than shut down, minimizing the impact of frequent starts and stops on equipment life. Spray-mix reinforcement is initiated after the cumulative time exceeds the limit, avoiding misjudgments caused by short-term fluctuations. If multiple sampling points exceed the limit consecutively, the system is shut down immediately, ensuring safety in high-risk conditions. This logic, combined with dynamic threshold adjustment, significantly reduces missed and false alarm rates.

[0040] The key point features of the images before and after injection are extracted, pixel-level alignment is performed using the homography matrix, and the difference mask is generated by adaptive threshold segmentation.

[0041] Key point feature extraction and homography matrix alignment accurately match tunnel face images before and after spraying, resolving image mismatches caused by camera 13 pose offsets and ensuring the accuracy of the differential mask. Adaptive threshold segmentation automatically determines the optimal threshold by analyzing grayscale, color, and edge gradient differences, effectively distinguishing between areas covered and uncovered by shotcrete. For example, in low-contrast environments, adaptive thresholding can avoid the subjectivity of manually set thresholds and improve the robustness of defect detection. The application of these image processing techniques improves the accuracy and reliability of spray-mix quality assessment, enabling the system to accurately identify subtle differences in spray quality and defective areas, providing strong support for subsequent re-spraying decisions.

[0042] Grading processing is carried out according to the comprehensive evaluation index, and the specific steps include: when the comprehensive evaluation index reaches the qualified standard, automatically instructing to retract the nozzle 12 and resume excavation construction; when the comprehensive evaluation index is in the critical range, automatically identifying the defective area and guiding the supplementary spraying operation, and then re-evaluating until the qualified standard is reached; when the comprehensive evaluation index is lower than the qualified standard, automatically executing the second supplementary spraying operation. If the standard is still not met after the second supplementary spraying, manual on-site confirmation is carried out.

[0043] When the index reaches the qualified standard, the system automatically instructs to retract the nozzle 12 and resume excavation construction, indicating that the spray-mix quality meets the requirements and avoids excessive intervention. If the index is in the critical range, the system automatically identifies the defective area and guides the supplementary spraying operation, and then re-evaluates until it is qualified to ensure that the spray-mix quality meets the standard. If the index is too low, the system automatically performs a second supplementary spraying. If the standard is still not met after the second supplementary spraying, the manual confirmation procedure is initiated. This hierarchical processing mechanism realizes the automated management and control of spray-mix quality, improves construction efficiency, and ensures the quality of spray-mix construction. Moreover, the hierarchical logic cooperates with the image evaluation module to form a "detection-repair-reinspection" closed loop, which significantly improves the reliability of spray-mix quality.

[0044] The specific steps of the spray-mix reinforcement method include:

[0045] 1. Stability evaluation:

[0046] Sensor data collection: monitor the instantaneous value of torque T(t), the instantaneous value of pressure P(t), the absolute value of shield deformation 𝛿(t), and normalize the above data.

[0047] The data fusion judgment introduces the normalized data into the stability evaluation formula: ;

[0048] in, Stability evaluation index, is the cutter head torque at the current moment, To stabilize the working condition, the average value of stable excavation for 30 minutes is taken before starting the work. is the cutter head pressure at the current moment, is the design pressure, is the shield deformation at the current moment, is the historical normal deformation; These are the weight coefficients of cutterhead torque, pressure, and shield deformation in the comprehensive assessment, determined by engineering experience or statistical models. Cutterhead torque reflects the difficulty of rock breaking, cutterhead pressure reflects the stress condition of the tunnel face, and shield deformation indicates whether the shield is evenly stressed. By monitoring these data, the stability of the excavation face can be comprehensively and accurately assessed, providing a reliable basis for subsequent spray-mix reinforcement decisions. is the threshold value and is updated using statistical methods.

[0049] The specific update formula is as follows:

[0050] ;

[0051] in, , For nearly N samples The mean and standard deviation of the data are obtained by taking different k values ​​to obtain different confidence levels. When k is 2.33, the confidence level is about 99%. The smaller the k value, the more conservative it is.

[0052] According to different The value is graded and judged. When the normal excavation, When the speed is too slow, it is necessary to slow down and observe, reduce the propulsion speed to a certain extent, and continue for 20 seconds. If the value falls back to the safe range, no warning will be given; if the cumulative time is greater than 30s within 60s, a warning will be given and spray-mix reinforcement will be carried out. , immediately stop excavation and carry out spray-mix reinforcement. The threshold can be fine-tuned based on the on-site missed alarm and false alarm rates.

[0053] When the stability assessment reaches the instability threshold, the TBM will slow down or suspend excavation and start the spray-mixing operation.

[0054] 2. Spray mixing operation:

[0055] The electric telescopic rod 2 of the storage tray 3 retracts, revealing the through-holes 10 reserved in the cutterhead 1. The storage tray 3 is rotated to align each nozzle 12 with the through-holes 10. The spraying arm 15 drives the nozzle 12 out of the cutterhead 1 to start spraying and mixing. The shotcrete operation then begins. The mixed wet shotcrete slurry is pumped from the rear of the TBM to the nozzle 12 at the front via a hose 14. Compressed air is used to spray the concrete at high speed onto the rock mass at the excavation face, forming a support layer. During the spraying process, the distance and angle of the nozzle 12 relative to the rock face are adjusted as needed to ensure that the spraying distance and angle of incidence remain within the optimal range. The nozzle 12 is essentially perpendicular to the rock face and maintained at an appropriate distance to reduce rebound.

[0056] 3. Spray mix quality evaluation:

[0057] Through adaptive alignment of pre-spraying reference and post-spraying images, color-texture segmentation, and statistical index calculation, rapid assessment of spray layer coverage, roughness, and missed defects is achieved. This approach significantly reduces system complexity and cost while ensuring high efficiency and real-time feedback on the construction site.

[0058] The specific steps include:

[0059] (1) Reference acquisition: Before spraying concrete, an industrial camera 13, mounted on the sidewall of the nozzle 12, is placed approximately 1 m from the tunnel face and captures a reference image of the tunnel face at a predetermined shooting posture and angle. A lightweight semantic segmentation network is used to quickly identify bare rock areas on the tunnel face, which serve as detection areas for subsequent evaluation. To ensure image quality, the acquisition process must be performed under uniform lighting conditions.

[0060] (2) Spraying: After the spraying is completed, at the time of initial setting of the concrete (approximately 30–60 seconds), camera 13 maintains the same position and posture as during the baseline acquisition and recaptures the post-spraying tunnel face image. At this point, the visual characteristics of the concrete are essentially stable, which helps to more accurately evaluate the spraying effect.

[0061] (3) Image alignment: Using sparse feature matching methods such as ORB or SIFT, key point information in the image, such as edges or corners, is used to extract and match feature points, providing a basis for precise image alignment. The homography matrix is ​​then calculated using the RANSAC algorithm, and the post-injection tunnel face image is precisely aligned at the pixel level with the reference image, resulting in a fully matched image sequence. This alignment method effectively eliminates image mismatches caused by minor movements or posture errors of the camera 13.

[0062] (4) Differential mask generation: Grayscale difference, color difference, and edge gradient analysis are performed on the baseline image and the sprayed image, respectively, to calculate a comprehensive differential feature score map. Adaptive threshold segmentation technology is then used to automatically calculate the optimal threshold and automatically determine the optimal threshold to extract the binary mask of the sprayed concrete coverage area. This method can accurately identify subtle differences in the spraying effect and defective areas.

[0063] (5) Index calculation: Based on the differential mask, a detailed statistical analysis is performed to obtain the following key indicators:

[0064] Coverage C: The ratio of the area covered by shotcrete to the total area of ​​bare rock within the test area, which accurately quantifies the effective coverage of shotcrete;

[0065] Roughness S: By calculating the difference in edge gradient between the sprayed image and the reference image, the smoothness of the sprayed layer is quantitatively evaluated to detect the construction quality of the concrete surface;

[0066] Void ratio V: The binary mask is processed using image connected domain analysis technology to calculate the ratio of the area of ​​unsprayed patches or missed spraying areas to the total area of ​​the assessment area, so as to effectively identify and locate construction defects.

[0067] Calculation method of coverage C:

[0068] The differential mask is binarized, where the area with a pixel value of 1 represents the shotcrete-covered area, and the area with a pixel value of 0 represents the bare rock area. The number of pixels in the covered area Nc and the total number of pixels in the detection area N are calculated using a pixel statistics algorithm. The coverage rate C is calculated according to the formula C=Nc / N×100%, where Nc is the total number of pixels with a pixel value of 1 in the differential mask, and N is the total number of pixels in the detection area.

[0069] Furthermore, to improve calculation accuracy, the actual pixel-physical size mapping relationship can be combined to convert the number of pixels into the actual coverage area for calculation.

[0070] Calculation method of roughness S:

[0071] The roughness S is calculated using an edge gradient analysis method. This method involves first performing Sobel or Canny edge detection on both the baseline and post-spray images to extract their respective edge gradient maps. Then, within the sprayed area defined by the differential mask, the difference between the two edge gradient maps is calculated pixel by pixel to construct a gradient difference matrix. Finally, the standard deviation or root mean square (RMS) of this matrix is ​​calculated as the roughness index S. Larger S values ​​indicate a rougher sprayed surface, while smaller S values ​​indicate a smoother sprayed surface. Furthermore, to improve evaluation accuracy, the gradient difference values ​​can be normalized to make the roughness indices more comparable.

[0072] Calculation method of void ratio V:

[0073] Morphological image processing technology is used to calculate the void ratio V. The specific steps are as follows: first, a morphological opening operation is performed on the differential mask to remove small noise points; then, a connected domain labeling algorithm is used to identify all connected regions with a value of 0 in the mask, that is, unsprayed or missed spraying regions; for each connected domain, the number of pixels it contains is counted, and an area threshold T is set to filter out valid void regions; finally, the ratio of the total number of pixels Nv in all valid void regions to the total number of pixels Nt in the detection area is calculated, that is, V = Nv / Nt × 100%.

[0074] Furthermore, the location distribution map and size statistics of the voids can be output to provide accurate guidance for subsequent re-spraying operations.

[0075] In the above method, the differential mask is used as a calculation basis to achieve quantitative evaluation of the construction quality of shotcrete through different image processing algorithms and statistical analysis methods.

[0076] (6) Comprehensive evaluation and grading: The above indicators are combined using a linear weighted method to calculate the comprehensive evaluation index Q:

[0077] ;

[0078] Among them, the weight coefficients 0.5, 0.3, and 0.2 can be obtained by comparing and analyzing the on-site construction test data with the manual acceptance rating results, and using linear regression or multi-objective optimization methods such as random forest and genetic algorithms to obtain preliminary recommended values; the specific values ​​can be further adjusted according to construction site conditions, historical data or expert experience to ensure the accuracy and reliability of the automatic evaluation results. The Q value is the empirical bias threshold, and the recommended value is 0.25, but it can be dynamically adjusted according to the actual construction scenario and historical data. Based on the specific range of the Q value, the following grading and feedback measures are carried out:

[0079] When Q≥0, the evaluation is qualified, and the system automatically instructs to retract the nozzle 12 and resume the excavation construction;

[0080] When -0.1≤Q<0, the evaluation is critical, and the system automatically identifies the defective area and guides the robot arm to perform precise re-spraying operations, and then re-evaluates until it meets the qualified standards;

[0081] When Q<-0.1, the evaluation is unqualified, and the system automatically performs a second spraying operation; if the standard is still not met after the second spraying, the manual on-site confirmation procedure is started to ensure the quality of the spray-mixing construction.

[0082] Through this closed-loop detection, the quality of the spray-mix support is guaranteed to meet the standards. After the effect evaluation is completed and the support is qualified, the remaining material is cleared by air spraying, and the spray arm 15 drives the nozzle 12 to return to the storage tray 3. Then the storage tray 3 is rotated until the automatic telescopic rod is aligned with the hole entrance, and the automatic telescopic rod is reinserted into the cutter disc 1 to ensure that there is no gap in the cutter disc 1 and complete the work.

[0083] Example 2

[0084] This embodiment provides an adaptive spray-mix reinforcement system for TBM excavation face balancing, including an excavation face stability detection module, a spray-mix reinforcement module, and a spray-mix effect evaluation module;

[0085] The excavation face stability detection module is used to collect excavation data in real time and, based on this data, calculate a stability evaluation index by integrating weight coefficients and dynamically updating thresholds. The spray-mix reinforcement module is used to suspend excavation and initiate spray-mixing operations when the stability evaluation index exceeds a preset threshold. The spray-mixing effect evaluation module is used to collect and align images of the tunnel face before and after spraying, generate a differential mask, calculate coverage, roughness, and void ratio based on this differential mask, and comprehensively evaluate spray-mixing quality. Based on the evaluation results, additional spraying or resumption of excavation operations is performed. The excavation face stability detection module is capable of predicting excavation face stability and controlling the spray-mixing reinforcement module based on the prediction results.

[0086] The system integrates stability testing, spray-mix reinforcement, and performance evaluation through a modular design. The stability testing module's multi-sensor data fusion addresses the issue of single sensor susceptibility to interference; the spray-mix reinforcement module's automatic start / stop and parameter adjustment reduces manual intervention; and the performance evaluation module's image difference technology replaces traditional visual inspection, improving inspection efficiency. These modules work closely together to form a complete adaptive spray-mix reinforcement system, enabling automated control of the entire process, from stability testing to spray-mix reinforcement and quality evaluation, effectively improving the safety and efficiency of TBM construction.

[0087] The excavation face stability detection module includes a torque sensor installed on the drive spindle 9 of the cutterhead 1, a pressure sensor in the cavity of the cutterhead 1, a displacement sensor at a key position of the shield, and a data processing terminal connected to the torque sensor, pressure sensor and displacement sensor.

[0088] A torque sensor, mounted on the drive spindle 9 of cutterhead 1, measures cutterhead torque in real time. Multiple pressure sensors, positioned at various locations within the cutterhead cavity, provide real-time information on cutterhead pressure. Displacement sensors, installed at key locations such as the front, middle, and rear shields, capture displacement data resulting from uneven stress on the shields and determine local deformation. This layout enables accurate, real-time acquisition of key mechanical parameters during TBM excavation. The data processing terminal processes and analyzes the data collected by these sensors, calculating stability evaluation indicators to promptly identify any abnormal changes in excavation face stability and provide accurate data support for subsequent spray-mix reinforcement decisions.

[0089] like Figure 2 、 Figure 3 、 Figure 4 、 Figure 5 As shown, the spray-mix reinforcement module includes a storage tray 3 and a nozzle 12; the storage tray 3 is rotatably mounted on the main shaft 9 of the cutterhead 1 and is located behind the main shaft 9. The side of the storage tray 3 facing the cutterhead 1 is provided with an electric telescopic rod 2 and a storage chamber 11, and the nozzle 12 is installed in the storage chamber 11. The cutterhead 1 is provided with a through hole 10; during normal excavation, the electric telescopic rod 2 is inserted into the through hole 10, and the nozzle 12 is folded and placed in the storage tray 3 behind the cutterhead 1 to protect the nozzle 12 and maintain the integrity of the cutterhead 1. When a sign of instability is detected and sprayed concrete support is required, the storage tray 3 rotates relative to the cutterhead 1, and the nozzle 12 moves out of the storage chamber 11 and passes through the through hole 10 of the cutterhead 1 to start the spray-mix operation. This design can quickly and efficiently perform spray-mix reinforcement without affecting normal excavation operations, and the telescopic nozzle 12 ensures the flexibility and accuracy of the spray-mix operation.

[0090] Furthermore, a rotary hydraulic cylinder is mounted on the spindle 9. The rotating seat of the rotary hydraulic cylinder is fixedly connected to the receiving tray 3, thereby driving the receiving tray 3 to rotate relative to the spindle 9. During normal operation, the receiving tray 3 rotates with the spindle 9. During spray mixing, the spindle 9 stops rotating, and the rotary hydraulic cylinder drives the receiving tray 3 to rotate a predetermined angle relative to the spindle 9 (the cutterhead 1) to allow the nozzle 12 to pass through the cutterhead 1 for spray mixing reinforcement.

[0091] The spray-mix reinforcement module also includes a hopper 6, a mixing system 7, a cement injection pump 8, and an air compressor 5. The hopper 6 serves as a storage space for dry materials, the mixing system 7 mixes the dry materials with water and additives, the cement injection pump 8 uses pumping pressure to transport the mixed concrete through a pipe to the nozzle 12, and the air compressor 5 provides the air power required for injection. These components work together to automate the entire process of concrete preparation, transportation, and injection, providing a stable and efficient material supply and power support for the spray-mix reinforcement operation, ensuring the continuity and quality of the spray-mix operation.

[0092] like Figure 5 、 Figure 6 As shown, a guide rail is provided in the storage chamber 11 , a base 16 is slidably mounted on the guide rail, a spray arm 15 is mounted on the base 16 , and the nozzle 12 is mounted on the spray arm 15 .

[0093] Specifically, the spray arm 15 adopts a modular design and is connected to the connecting rod in stages through movable joints. The movable joints have built-in electric servo drives to achieve high-precision posture adjustment and control. The drive motor of the spray arm 15 is connected to the electronic control system in the storage tray 3 via a cable, realizing real-time control. The base 16 of the spray arm 15 is mounted on a guide rail in the storage chamber 11, so that the spray arm 15 can send the nozzle 12 out of the storage tray 3, thereby realizing the deployment and recovery of the nozzle 12. When the storage tray 3 rotates, the spray arm 15 moves synchronously. When concrete spraying is required, the spray arm 15 extends the nozzle 12 through the multi-degree-of-freedom movable joint and positions it to the designated work area for concrete spraying. After the operation is completed, the nozzle 12 is recovered to the storage chamber 11 in the storage tray 3 through the reverse action of the movable joint, realizing the safe storage of the spray arm 15. The spray arm 15 ensures that the nozzle 12 has full and flexible adjustment capabilities during operation, thereby ensuring the accuracy of the concrete spraying operation.

[0094] Because the receiving tray 3 can rotate with the cutterhead 1, a connecting tray 4 is added between the jet pump 8 and the nozzle 12 to ensure the normal operation of the concrete delivery pipeline. The connecting tray 4 can deliver concrete to each nozzle 12. The connecting tray 4 adopts an existing structure, mainly including a rotating part and a fixed part. The rotating part is mounted on the main shaft 9 and rotates with the main shaft 9 and the receiving tray 3, so that the pipeline between the connecting tray 4 and the nozzle 12 can rotate with the receiving tray 3. The fixed part is connected to the jet pump 8, and the pipeline between the jet pump 8 and the connecting tray 4 is fixed.

[0095] The spray mixing effect evaluation module includes a spray mixing evaluation camera 13 installed next to the nozzle 12. The camera 13 should be provided with an outer cover and can be rotated at a certain angle to cover the entire working surface.

[0096] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it 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 on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.

Claims

1. An adaptive spray-mix reinforcement method for TBM excavation face balancing, characterized in that: include: collecting excavation data in real time, and calculating a stability evaluation index based on the excavation data by fusing a weight coefficient and a dynamically updated threshold; When the stability evaluation index exceeds a preset threshold, the excavation is suspended and the spray-mixing operation is started; Collect and align tunnel face images before and after blasting to generate a differential mask. Calculate coverage, roughness, and void ratio based on the differential mask to comprehensively evaluate the blasting quality. Re-blasting or resumption of tunneling operations is performed based on the evaluation results. The calculation of the stability evaluation index by fusing the weight coefficient and the dynamically updated threshold value includes: ； in, Stability evaluation index, is the cutter head torque at the current moment, It is the reference torque for stable working conditions; is the cutter head pressure at the current moment, is the design pressure; is the shield deformation at the current moment, is the historical normal deformation; are the weight coefficients of cutter head torque, pressure, and shield deformation in comprehensive evaluation; is the threshold value, which is updated using statistical methods; The specific update formula is as follows: ； in, , For nearly N samples The mean and standard deviation of the data, k is the confidence coefficient, the smaller the k value, the more conservative it is.

2. The adaptive spray-mix reinforcement method for TBM excavation face balancing according to claim 1, characterized in that: Grading is performed based on stability evaluation indicators. The specific steps include: When the stability evaluation index is within the normal range, maintain normal excavation; When the stability evaluation index exceeds the normal range but does not reach the warning threshold, the propulsion speed is reduced for a preset time. If the stability evaluation index falls back to the safe range, no warning is issued. If the cumulative time that the stability evaluation index exceeds the normal range exceeds the set value within the preset time, a warning will be issued and spray-mix reinforcement will be started; When the stability evaluation indicators of multiple consecutive sampling points exceed the normal range, excavation is stopped directly and the spray-mix reinforcement task is started.

3. The adaptive spray-mix reinforcement method for TBM excavation face balancing according to claim 1, characterized in that: Extracting key point features of the images before and after injection, extracting and matching the key point features using a sparse feature matching method, and calculating a homography matrix to achieve pixel-level alignment of the image after injection and the reference image; The aligned reference image and the sprayed image are subjected to grayscale difference, color difference, and edge gradient analysis, respectively, to generate a comprehensive differential feature score map. The optimal threshold is automatically determined through adaptive threshold segmentation technology, and the binary mask of the sprayed concrete coverage area is extracted to generate a differential mask for spray-mix quality evaluation.

4. The adaptive spray-mix reinforcement method for TBM excavation face balancing according to claim 1, characterized in that: Grading is carried out according to the comprehensive evaluation index. The specific steps include: Calculate the comprehensive evaluation index Q: ； Among them, C is coverage, S is roughness, V is void ratio, is the empirical bias threshold; When Q≥0, the evaluation is qualified, and the nozzle is automatically retracted and the excavation construction is resumed; When -0.1≤Q<0, the evaluation is critical, the defective area is automatically identified and the re-spraying operation is guided, and then re-evaluation is carried out until the qualified standard is met; When Q<-0.1, the evaluation is unqualified and a second spraying operation is automatically performed. If the standard is still not met after the second spraying, manual on-site confirmation is carried out.

5. A reinforcement system using the adaptive spray-mix reinforcement method for TBM excavation face balancing according to any one of claims 1 to 4, characterized in that: It includes excavation face stability detection module, spray-mix reinforcement module and spray-mix effect evaluation module; The excavation face stability detection module is used to collect excavation data in real time and calculate the stability evaluation index based on the excavation data by integrating the weight coefficient and the dynamically updated threshold; The spray-mix reinforcement module is configured to suspend excavation and initiate spray-mixing operations when the stability evaluation index exceeds a preset threshold; The spray-mixing effect evaluation module is used to collect and align tunnel face images before and after spraying, generate a differential mask, calculate coverage, roughness, and void ratio based on the differential mask, and then comprehensively evaluate the spray-mixing quality. Re-spraying or resumption of tunneling operations is performed based on the evaluation results.

6. The reinforcement system according to claim 5, wherein: The excavation face stability detection module includes a torque sensor installed on the cutterhead drive spindle, a pressure sensor in the cutterhead cavity, a displacement sensor at a key position of the shield, and a data processing terminal connected to the torque sensor, pressure sensor and displacement sensor.

7. The reinforcement system according to claim 5, wherein: The spray-mix reinforcement module includes a receiving tray and a nozzle; The receiving disc is rotatably mounted on the main shaft of the cutter disc and is located at the rear side of the main shaft. An electric telescopic rod and a receiving chamber are provided on the side of the receiving disc facing the cutter disc. The nozzle is installed in the receiving chamber. A through hole is provided on the cutter disc. During normal excavation, the electric telescopic rod is inserted into the through hole. During spray-mixing reinforcement, the receiving disc rotates relative to the cutter disc, and the nozzle moves out of the receiving chamber and passes through the through hole of the cutter disc for spray-mixing reinforcement.

8. The reinforcement system according to claim 7, wherein: The spray-mix reinforcement module also includes a hopper, a mixing system, a cement injection pump and an air compressor; The hopper serves as a storage space for dry materials, the mixing system mixes the dry materials with water and additives, the cement injection pump uses pumping pressure to transport the mixed concrete to the nozzle through a pipeline, and the air compressor provides the air power required for injection.

9. The reinforcement system according to claim 7, wherein: A guide rail is provided in the storage chamber, a base is slidably mounted on the guide rail, a spray arm is mounted on the base, and the nozzle is mounted on the spray arm.

Images