Tunnel primary support flatness automatic trimming device and method

Abstract

The present application relates to a kind of tunnel primary support flatness automatic finishing equipment and method, equipment is configured on wet spraying mechanical arm, include: concrete nozzle, for completing primary support once spraying forming;Three-dimensional laser scanning system is used to carry out three-dimensional scanning to excavation section before and after spraying respectively, central control computer system is used to construct primary support three-dimensional model based on point cloud data after spraying, to determine the datum plane of current spraying forming surface in statistical regression mode, identify the unqualified point set deviating from the datum plane, milling and digging execution device is installed on the end of wet spraying mechanical arm small arm, for implementing layer-by-layer milling and planing finishing to the region corresponding to the unqualified point set according to the control instruction, so that spraying, detection and finishing are closed loop completed in the same construction cycle.The present application realizes the automation closed-loop control of spraying-detection-finishing-reinspection by intelligent and accurate transformation to existing wet spraying trolley.

Description

Technical Field

[0001] This invention relates to the field of tunnel construction technology, specifically to an automatic equipment and method for adjusting the flatness of the initial support of a tunnel. Background Technology

[0002] In the initial support construction of tunnels, the smoothness treatment after shotcrete formation typically employs the traditional "three-spray, two-scrape" process. This involves an operator manually scraping and smoothing uneven areas after the initial support spraying using a scraper mounted on a wet-spraying robotic arm. This process is widely used in current engineering practice. The smoothing path, area, and scraping depth rely heavily on the operator's on-site experience. It demands high levels of skill, visual judgment, and rhythm control from the operator during construction, making it a common and representative smoothness treatment method in current tunnel initial support shotcrete construction.

[0003] However, the aforementioned traditional processes have significant limitations. First, the procedures are complex and involve frequent changes; the "three sprayings and two scrapings" often account for 15% to 20% of a single cycle, significantly impacting cycle efficiency. Second, the forming quality is highly dependent on manual experience, making it difficult to control the variance of flatness, often resulting in over-excavation or under-excavation, leading to difficulties in subsequent waterproofing membrane laying and excessive consumption of sprayed concrete, among other quality risks. Third, operators must work at close range on the tunnel face for extended periods, resulting in high labor intensity and safety risks. Fourth, repeated scraping and multiple rework processes easily lead to excessive removal of already sprayed concrete, resulting in generally high material waste and poor economic efficiency. These problems make it difficult for traditional flatness treatment techniques to simultaneously ensure construction quality while maintaining efficiency, safety, and material utilization, leaving room for improvement on-site. Summary of the Invention

[0004] To address the aforementioned deficiencies in existing technologies, this invention provides an automatic tunnel initial support smoothness adjustment device and method, which intelligently and precisely transforms existing wet spraying trolleys to achieve automated closed-loop control of spraying, detection, adjustment, and re-inspection.

[0005] To achieve the above objectives, the technical solution of the present invention is as follows:

[0006] An automatic tunnel initial support smoothness adjustment device, configured on a wet spraying robotic arm, includes:

[0007] Shotcrete nozzles are used to complete the initial spraying and shaping of the concrete structure in one pass.

[0008] The three-dimensional laser scanning system is used to perform three-dimensional scanning of the excavation section before and after spraying. Before spraying, it is used to calculate the trolley posture and establish the reference coordinate system of the excavation section by looking back at the known control points. After spraying, it is used to collect massive point cloud data of the sprayed and formed initial support surface.

[0009] A central control computer system is configured to construct a three-dimensional model of the primary support based on the point cloud data after spraying, determine the reference surface of the current spray forming surface in a statistical regression manner, identify the unqualified point set deviating from the reference surface, set the path priority in combination with the downward-to-upward solidification strength gradient of the sprayed concrete, and generate a bottom-up trimming control instruction.

[0010] A milling and excavating execution device is installed on the end arm of the wet spraying mechanical arm and is configured to implement layer-by-layer milling and planing trimming on the region corresponding to the unqualified point set according to the control instruction, so that spraying, detection and trimming are closed-loop completed in the same construction cycle.

[0011] Preferably, the three-dimensional laser scanning system is a phase or pulse laser scanner, the scanning rate is not less than 1 million points per second, the accuracy is millimeter level, and 360° full coverage scanning is achieved through a trolley mast or an independent gimbal.

[0012] Preferably, the central control computer system generates a milling path planning instruction that is bottom-up and advances layer by layer with a 1 cm step according to the unqualified point set, and directly drives the milling and excavating execution device to complete local quantitative milling and trimming. After the trimming is completed, the three-dimensional laser scanning system is triggered again to perform re-scanning and re-inspection, so that spraying, detection, trimming and re-inspection form a closed control loop in the same construction cycle.

[0013] Preferably, the milling and excavating execution device is driven by hydraulic pressure or a combination of hydraulic pressure and electricity, and the milling head is provided with a plurality of tungsten carbide hard alloy cutters for milling C20-C30 sprayed concrete. The milling head is in a spherical structure to realize smooth transition between the milled region and the unmilled region.

[0014] Preferably, the milling and excavating execution device is connected to the wet spraying mechanical arm or an independent mechanical arm through a high-rigidity support and has at least three degrees of freedom for pose adjustment.

[0015] Preferably, the central control computer system performs curvature filtering and normal vector clustering processing on the point cloud based on a specification threshold of D / L≤1 / 20, calculates the depth-length ratio of the protrusions for flatness determination, and introduces an AI abnormality recognition module to improve the reliability of the detection result, with an error rate of less than or equal to 2%.

[0016] Preferably, the central control computer system triggers forced ventilation and smoke exhaust for 5-10 minutes after spraying to improve the visibility quality of the point cloud.

[0017] Preferably, the central control computer system automatically archives the re-scanning point cloud data and the flatness evaluation result as a traceable electronic construction record after the trimming is completed.

[0018] In another aspect, the application also discloses an automatic trimming method for flatness of a tunnel primary support, which is applied to the automatic trimming device for flatness of a tunnel primary support and includes the following steps:

[0019] S1, device in place and initialization: the wet spraying mechanical arm travels with the trolley to the tunnel working face, connects the wind, water and electricity, the central control computer system starts, and the self-checking of each sensor and device is completed;

[0020] S2, positioning and modeling before spraying: the three-dimensional laser scanning system calculates the trolley attitude through the rear view known control point before spraying and establishes the reference coordinate system of the excavation section;

[0021] S3, one-time concrete spraying: the spraying concrete nozzle completes the primary spraying forming of the primary support according to the reference coordinate system and the design section;

[0022] S4, environmental preparation: after spraying is completed, the central control computer system triggers forced ventilation and smoke exhaust for 5-10 minutes, so that the scanning visibility quality meets the point cloud collection requirements;

[0023] S5, post-spraying point cloud collection: the three-dimensional laser scanning system performs full-section high-density scanning on the sprayed primary support surface to obtain massive point cloud data;

[0024] S6, intelligent analysis and automatic trimming: the central control computer system constructs a three-dimensional model of the primary support based on the point cloud data, identifies unqualified point sets deviating from the reference plane, and sets path priorities in combination with the downward strength gradient of the sprayed concrete to generate bottom-up trimming control instructions to drive the milling and digging execution device to implement layer-by-layer milling and trimming on the regions corresponding to the unqualified point sets;

[0025] S7, re-inspection and archiving: after trimming is completed, the three-dimensional laser scanning system performs scanning re-inspection again, and the central control computer system automatically archives the re-scanning point cloud data and the flatness evaluation results as traceable electronic construction records;

[0026] S8, end processing: after trimming and re-inspection are completed, stop the operation of each execution device, disconnect the wind, water and electricity connections, make the device exit the working face, and perform cleaning and routine maintenance operations.

[0027] Preferably, in the step S6, the point spacing of the massive point cloud data is ≤1cm, the full-section data collection of a 60m section can be completed within 15 minutes, the objective evaluation results are automatically generated in combination with the AI algorithm, and the construction optimization is guided through the visualization platform.

[0028] The application realizes the mode transition of the initial support flatness from "experience dependence, manual scraping" to "data driven, automatic trimming" by adding a three-dimensional laser scanning system, a central control computer system and a milling and digging execution device on the basis of the existing wet spraying trolley. Compared with the prior art, the application has the following advantages: (1) low modification investment and fast investment recovery, only adding the "scanning-milling-digging-control" three modules to the existing wet spraying trolley can form the equipment capacity, which is significantly lower than purchasing special tooling; (2) significantly simplified process and doubled efficiency, the operation link is restructured from the traditional "three spraying and two scraping" to "one spraying and one milling", which can shorten the single cycle operation time by 20%-30%; (3) digital precise trimming and quality leap, based on the millimeter level point cloud data, the quantitative milling is performed, so that the flatness of the initial support forming tends to be "measurable and closed loop", and the one-time compliance rate of the flatness can be improved to more than 98%, which provides a reliable geometric accuracy basis for subsequent waterproof board laying and secondary lining construction; (4) significantly reduced material loss and realized green construction, only the protruding part is removed to achieve the designed section, avoiding the waste of concrete caused by "blind scraping", and it is expected that the material loss can be reduced by 15%-25%; (5) significantly reducing skill dependence and safety exposure risk, mechanization replaces close-range manual operation, which can reduce the time of personnel close to the working face, reduce labor load and improve the substantive safety level. In summary, the application can directly bring system-level cost reduction and efficiency increase value to the initial support operation organization, quality control mechanism and construction economy. BRIEF DESCRIPTION OF DRAWINGS

[0029] Figure 1 FIG. 1 is a structural schematic diagram of a tunnel initial support flatness automatic trimming device according to an embodiment of the application.

[0030] Figure 2 FIG. 2 is a flowchart of a tunnel initial support flatness automatic trimming method according to an embodiment of the application.

[0031] FIG. 1 is a structural schematic diagram of a tunnel initial support flatness automatic trimming device according to an embodiment of the application. DETAILED DESCRIPTION

[0032] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, not all the embodiments of the application. Based on the embodiments of the application, all other embodiments obtained by those skilled in the art belong to the application.

[0033] Moreover, the described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a thorough understanding of embodiments of the application. One skilled in the relevant art will recognize, however, that the application can be practiced without one or more of the specific details, or with other methods, components, materials, and so forth. In other instances, well-known structures, devices, implementations, or operations are not shown or described in detail to avoid obscuring aspects of the application.

[0034] As shown in the drawings: Figure 1 The embodiment discloses a tunnel primary support flatness automatic trimming device, which is arranged on a wet spraying mechanical arm 1 and comprises:

[0035] A spraying concrete nozzle 2 is used to complete primary spraying forming; the spraying concrete nozzle 2 actually directly undertakes the original spraying function of the wet spraying trolley, and the track coverage of the spraying concrete is completed through the linkage of the wet spraying mechanical arm 1, so that it is unnecessary to rebuild the spraying main body, and only the spraying section is retained on the basis of the existing operation process without changing the control logic of the original spraying link;

[0036] A three-dimensional laser scanning system 3 is used to perform three-dimensional scanning on the excavation section before and after spraying, is used to solve the trolley posture through the rear view of the known control point and establish the reference coordinate system of the excavation section before spraying, and is used to collect massive point cloud data of the sprayed primary support surface after spraying; the three-dimensional laser scanning system 3 is arranged in a visible domain stable region of the wet spraying mechanical arm 1 and ensures that there is a visual condition with the control point in the pre-spraying stage to solve the trolley posture, the post-spraying scanning stage uses the same hardware to realize high-density point cloud collection, and the two use modes of “establishing the reference surface” before spraying and “quality detection surface” after spraying are realized in seamless switching in a single device;

[0037] A central control computer system 4 is used to construct a primary support three-dimensional model based on the point cloud data after spraying, determine the reference surface of the current sprayed forming surface in a statistical regression manner, identify an unqualified point set deviating from the reference surface, set a path priority in combination with the solidification strength gradient of the sprayed concrete from bottom to top, and generate a top-down trimming control instruction; here, the central control computer system 4 is equivalent to the decision center of the entire construction closed loop, which performs reference surface regression from the post-spraying point cloud, highlights point identification and path planning, and directly determines the execution sequence of milling and planing; “from bottom to top” is to adapt to the solidification gradient of the gradually increasing concrete strength with the height, otherwise, if the upper area is milled and planed first, the low part will be re-contaminated because of the re-flowing of the high-slump slurry, and repeated milling and planing is required;

[0038] The milling and digging execution device 5 is installed on the end arm of the wet spraying mechanical arm 1, and is used to implement layer-by-layer milling and planing finishing on the region corresponding to the set of unqualified points according to the control instruction, so that spraying, detection and finishing are closed loop completed in the same construction cycle. The milling and digging execution device 5 is connected to the existing wet spraying operation chain by being mounted on the end arm, thereby achieving the lowest modification cost. Without additional independent robots or additional mechanical arms, the "single-arm spraying + scanning + milling closed loop" can be formed. This structure realizes true one-time upper arm deployment, one-time positioning and cycle operation closed loop.

[0039] Further, the three-dimensional laser scanning system 3 is a phase or pulse laser scanner, the scanning rate is not less than 1 million points per second, the precision is millimeter level, and 360° full coverage scanning is realized through the trolley mast or independent gimbal. The three-dimensional laser scanning system 3 is configured in a phase or pulse mode to maintain stable measurement quality in a high-dust and high-light-interference environment such as a tunnel. The high sampling rate of more than 1 million points per second ensures that the cross-section point cloud density is sufficient to support subsequent regression modeling. The millimeter-level precision ensures that the concave-convex deviation evaluation has sufficient engineering effectiveness. By installing the three-dimensional laser scanning system 3 on the trolley mast or independent gimbal, a 360° full coverage scanning angle around the tunnel face can be realized to avoid side wall blind areas, ensuring the integrity and reliability of the flatness detection results.

[0040] Further, the central control computer system 4 generates milling path planning instructions from bottom to top and layer by layer with a step of 1 cm according to the set of unqualified points. The central control computer system 4 advances with a step of 1 cm, which is a priority strategy based on the setting of the solidification strength distribution characteristics of sprayed concrete from bottom to top. Because the bottom coagulates fast and has high mechanical stability, milling the lower part first can avoid the risk of local overall peeling caused by disturbance when the upper section is still in the plastic stage. The central control computer system 4 directly drives the milling and digging execution device 5 to complete local quantitative milling and finishing. The milling and digging execution device 5 does not need manual secondary transition and does not have random errors such as experienced over-scratching and under-scratching. After finishing, the three-dimensional laser scanning system 3 is triggered again for re-scanning and re-inspection, so that spraying, detection, finishing and re-inspection form a closed control loop in the same construction cycle. The significance of the closed loop is to avoid the traditional "shift spraying, shift scraping and rework" time difference mismatch, so that the flatness control no longer drifts across shifts and time, thereby forming an instant digital closed control chain within one-time spraying.

[0041] In this embodiment, the milling and excavating device 5 is driven by hydraulic pressure or a combination of hydraulic pressure and electricity, the milling head is provided with a plurality of tungsten carbide hard alloy cutters for milling C20-C30 sprayed concrete, the milling head is spherical in structure to realize smooth transition between the milled area and the unmilled area, the milling and excavating device 5 is driven by hydraulic pressure or a combination of hydraulic pressure and electricity to ensure sufficient output stiffness and response speed under the high resistance characteristics of the concrete in the initial setting period within a short time after spraying, and the tungsten carbide alloy cutters and the spherical milling head are used to ensure that the cutters do not break and there is no hard step difference between the "flat-convex point" under the conditions of high friction milling and large local stress. The milling and excavating device 5 is connected to the wet spraying mechanical arm 1 or an independent mechanical arm through a high-rigidity support, has at least three degrees of freedom for posture adjustment, the high-rigidity support ensures that there is no "flexible displacement-data distortion" coupling during milling, and the three degrees of freedom for posture adjustment ensure that the milling and excavating device 5 can make local normal adjustment for local convex points, so that the control quantity is no longer dependent on manual posture matching, thereby maintaining the geometric accurate landing point of the algorithm generated path.

[0042] Moreover, the central control computer system 4 performs curvature filtering and normal vector clustering processing on the point cloud based on the specification threshold of D / L≤1 / 20, and calculates the depth-length ratio of the protrusions for flatness determination and introduces an AI anomaly recognition module to improve the detection result reliability error rate≤2%. The central control computer system 4 references the threshold of D / L≤1 / 20, which belongs to the industry standardization experience interval, is established based on the forming upper limit range of the allowable perturbation curvature and local uplift shape of the sprayed concrete primary support surface under the action of vehicle load and surrounding rock, and is a determination model that converts "whether the protrusion needs manual rework" into "whether the depth / length ratio of the protrusion exceeds the allowed plastic rebound window"; curvature filtering and normal vector clustering reconstruct the local disturbance surface in the massive point cloud into geometric segments for qualitative and quantitative analysis, so that each convex point is not a "scatter point cloud segment" but a "structured object"; the AI anomaly recognition module is further introduced because random interference such as suspended ventilation dust, light spot mis-scanning, and fog droplet noise inevitably exists in the original point cloud, which easily causes "false convex point" false positives, so that the light spot density and normal consistency of the local suspicious segments are learned through the model for secondary confidence screening, so that the overall error rate is stable≤2%, and unnecessary milling and excavating actions are prevented due to noise points, thereby ensuring the reliability and convergence of the generated repair path.

[0043] In some embodiments, the central computer system 4 triggers forced ventilation smoke exhaust 5-10 minutes after spraying to improve the point cloud visibility quality. The reason why the central computer system 4 triggers forced ventilation smoke exhaust is that there is a large amount of “wet state atomized water film + dust plume” mixture on the primary support surface immediately after spraying, which can cause local high scattering points, virtual false convex surfaces and point cloud holes in the three-dimensional laser scanning system 3. Therefore, it is necessary to perform forced ventilation smoke exhaust for 5-10 minutes immediately after spraying to form a visible light + laser visibility window, so that the point cloud reflection enters the high consistency interval.

[0044] Further, the central computer system 4 automatically archives the re-scanned point cloud data and the flatness evaluation results as traceable electronic construction records after finishing the trimming. The automatic archiving of the central computer system 4 not only saves the “result data” of a single construction, but also includes the scanning time, the trolley pose, the specific path sequence of the milling and digging execution device 5 and the planing depth matrix, so that the entire operation process forms a complete digital link for engineering operation and maintenance and later detection. This archiving method essentially converts the traditional manual oral inspection into data-based inspection, which can meet the data-based inspection requirements for the primary support surface flatness in rail transit / highway / municipal engineering.

[0045] As shown in Figure 2 Another embodiment of the present application also discloses an automatic trimming method based on tunnel primary support flatness, which is applied to the automatic trimming device for tunnel primary support flatness.

[0046] S1, equipment positioning and initialization: the wet spraying mechanical arm 1 drives with the trolley to the tunnel working face, connects the wind, water and electricity, the central computer system 4 is started, and the self-checking of each sensor and device is completed. The wet spraying mechanical arm 1 in this step not only serves as a carrier of the spraying execution main body, but also is used for task period coordinate unification, so that the subsequent scanning, spraying and trimming actions are all based on the same mechanical reference frame;

[0047] S2, pre-spraying positioning modeling: the three-dimensional laser scanning system 3 calculates the trolley pose by the rear-view known control point before spraying and establishes the reference coordinate system of the excavation section.

[0048] S3, one-time concrete spraying: the spraying concrete nozzle 2 completes the one-time spraying forming of the primary support according to the reference coordinate system and the design section.

[0049] S4, environment preparation: the central computer system 4 triggers forced ventilation smoke exhaust for 5-10 minutes after spraying, so that the scanning visibility quality meets the point cloud acquisition requirements. The forced ventilation action is used to remove the atomized slurry mist and floating dust interference, so that the subsequent point cloud spatial reflection echo is stable in the millimeter error band;

[0050] S5, post-spraying point cloud collection: the three-dimensional laser scanning system 3 performs full-section high-density scanning on the surface of the spray forming primary support to obtain massive point cloud data. This step is different from the traditional local sampling inspection and belongs to full-section digital sampling;

[0051] S6, intelligent analysis and automatic trimming: the central control computer system 4 constructs a three-dimensional model of the primary support based on the point cloud data, identifies the unqualified point set deviating from the reference surface, sets the path priority in combination with the bottom-up solidification strength gradient of the sprayed concrete, generates a bottom-up trimming control instruction to drive the milling and digging execution device 5 to implement layer-by-layer milling and digging trimming on the region corresponding to the unqualified point set. In this step, the logic of processing from bottom to top is because the lower part of the concrete solidifies first, so processing the lower section first can avoid the collapse of the upper section which has not been solidified.

[0052] S7, re-inspection and archiving: after trimming is completed, the three-dimensional laser scanning system 3 performs scanning and re-inspection again, and the central control computer system 4 automatically archives the re-scanning point cloud data and the flatness evaluation results as traceable electronic construction records. This archiving is not simply storing photos, but storing point cloud matrices + cutter execution matrices, which is a data certificate level archiving for future engineering quality inspection;

[0053] S8, end processing: after trimming and re-inspection are completed, stop the operation of each execution device, disconnect the wind, water and electricity connections, make the equipment exit the working face, and perform cleaning and routine maintenance operations, so as to complete a complete construction cycle and restore the initial state of the equipment for the next cycle.

[0054] Further, in step S6, the point spacing of massive point cloud data is ≤1 cm, which can complete full-section data collection of a 60m section within 15 minutes, automatically generate objective evaluation results by combining AI algorithms, and guide construction optimization through a visualization platform. The point spacing ≤1 cm here means that the three-dimensional laser scanning system 3 can form a very high-density data grid in one scanning coverage, so that the macro and micro undulations of the primary support surface are digitally captured and entered into the calculation layer. The completion of 60m section within 15 minutes means that this method is not a laboratory-level calculation, but an efficiency executable based on the continuous advancement period of the project. Under the current supporting rhythm of tunnel primary support operation, the point cloud acquisition time will not become a bottleneck. Automatically generating objective evaluation results by combining AI algorithms means that the central control computer system 4 does not rely on manual threshold judgment after completing three-dimensional modeling, but performs abnormal identification and quantitative grading through a pre-trained model to avoid the influence of subjective judgment by the operator. And guiding construction optimization through a visualization platform means that the calculation results are converted into graphical presentations such as color coding and local deviation amount labeling, so that the construction personnel can directly locate the three-dimensional coupling relationship of “deviation + depth + range”, thereby directly guiding the wet spraying mechanical arm 1 and the milling and digging execution device 5 to perform trimming actions.

[0055] In summary, the application discloses a tunnel primary support flatness automatic trimming device and method, the device is integrated with a sprayed concrete nozzle 2, a three-dimensional laser scanning system 3, a central control computer system 4 and a milling and digging execution device 5 on a wet spraying mechanical arm 1, realizes automatic closed-loop control of the whole process from pre-spraying positioning modeling, one-time spraying forming, post-spraying point cloud collection, intelligent analysis and trimming to re-scanning and archiving, establishes a primary support "reference surface" based on millimeter-level massive point clouds and performs layer-by-layer milling and trimming from bottom to top, completes the process revolution from traditional "three spraying and two scraping" to "one spraying and one milling", and achieves the comprehensive technical effects of reducing manual intervention, improving trimming precision, speeding up the cycle pace and reducing material loss. Under the action of the technical scheme of the application, spraying-detection-trimming-retesting is no longer operated separately, but is uniformly scheduled by the central control computer system 4 into a continuous link in the same construction cycle, so that the flatness control is upgraded from "subjective experience method" to "data-driven instruction execution", and the flatness quality, construction efficiency and material utilization rate are simultaneously improved. The technical scheme has important significance in the field of tunnel primary support quality guarantee, can directly support the improvement of waterproof board laying quality, reduce the overconsumption of secondary lining, reduce the construction safety exposure risk, and provide a landing type "engineering level technical fulcrum" for the production mode evolution of tunnel engineering from mechanization to digitization and intelligentization, and has significant popularization value and engineering application prospect.

[0056] The above examples are only used to illustrate the technical solutions of the application, and not to limit them; under the idea of the application, the technical features in the above examples or different examples can also be combined, the steps can be implemented in any order, and there are many other changes of different aspects of the application as described above, which are not provided in details for simplicity; although the application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that the technical solutions recorded in the foregoing examples can be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the application.

Claims

1. A tunnel primary support flatness automatic trimming device, configured on a wet spraying mechanical arm (1), characterized in that, Comprise: Spraying concrete nozzle (2) for completing the primary support once spray forming; Spraying concrete nozzle (2) directly accepts the original spraying function of wet spraying trolley, and the track coverage of concrete spraying is completed through the linkage wet spraying mechanical arm (1); Three-dimensional laser scanning system (3) for three-dimensional scanning of the excavation section before and after spraying, respectively, for solving the trolley attitude through the rear view known control point before spraying and establishing the reference coordinate system of the excavation section, and for collecting massive point cloud data of the sprayed primary support surface after spraying; Central control computer system (4) for constructing the three-dimensional model of the primary support based on the point cloud data after spraying, determining the reference surface of the current spraying forming surface in a statistical regression manner, identifying the unqualified point set deviating from the reference surface, and setting the path priority in combination with the downward strength gradient of the sprayed concrete to generate the top-down trimming control instruction; Milling and digging execution device (5) installed on the wet spraying mechanical arm end arm, with at least three degrees of freedom of adjusting attitude, for implementing layer-by-layer milling and digging trimming of the corresponding region of the unqualified point set according to the control instruction, so that spraying, detection and trimming are closed loop completed in the same construction cycle; The central control computer system (4) generates milling path planning instruction which is from bottom to top and advances layer by layer with 1cm as a ladder according to the unqualified point set, and directly drives the milling and digging execution device (5) to complete local quantitative milling and trimming, and after the trimming is completed, the three-dimensional laser scanning system is triggered again to perform rescan and retest, so that spraying, detection, trimming and retest form a closed control loop in the same construction cycle; The central control computer system (4) performs curvature filtering and normal vector clustering processing on the point cloud based on the specification threshold of D / L≤1 / 20, and calculates the depth-length ratio of the protrusions for flatness determination, and introduces an AI abnormality recognition module to improve the reliability of the detection result, with an error rate of ≤2%; The central control computer system (4) triggers forced ventilation and smoke exhaust for 5-10 minutes after spraying to improve the point cloud visibility quality.

2. The tunnel primary support flatness automatic trimming device according to claim 1, characterized in that, The three-dimensional laser scanning system (3) is a phase or pulse laser scanner with a scanning rate of not less than 1 million points / second and a precision of millimeter level, and realizes 360° full coverage scanning through the trolley mast or independent gimbal.

3. The tunnel primary support flatness automatic trimming device according to claim 1, characterized in that, The milling and digging execution device (5) is driven by hydraulic pressure or electro-hydraulic hybrid, and the milling head is provided with a plurality of tungsten carbide hard alloy cutters for milling C20-C30 sprayed concrete, and the milling head is in spherical structure to realize smooth transition between the milled area and the unmilled area.

4. The device according to claim 3, wherein The milling and digging execution device (5) is connected with the wet spraying mechanical arm or independent mechanical arm through a high-rigidity support.

5. The tunnel primary support flatness automatic trimming device according to claim 1, characterized in that, The central control computer system (4) automatically archives the rescan point cloud data and flatness evaluation results as traceable electronic construction records after the trimming is completed.

6. A method for automatically modifying the flatness of a tunnel primary support based on the flatness of the tunnel primary support, applied to the device for automatically modifying the flatness of a tunnel primary support according to any one of claims 1-5, characterized in that, Comprise the following steps: S1, equipment in place and initialization: the wet spraying mechanical arm drives the trolley to the tunnel working face, connects the wind, water and electricity, the central control computer system (4) starts, and completes the self-checking of each sensor and device; S2, pre-spraying positioning modeling: the three-dimensional laser scanning system (3) calculates the trolley attitude through the rear-view known control points before spraying and establishes a reference coordinate system of the excavation section; S3, one-time concrete spraying: the sprayed concrete nozzle (2) completes the primary support one-time spraying forming according to the reference coordinate system and the design section; S4, environmental preparation: after spraying, the central control computer system (4) triggers forced ventilation and smoke exhaust for 5-10 minutes to make the scanning visibility quality meet the point cloud collection requirements; S5, post-spraying point cloud collection: the three-dimensional laser scanning system (3) performs full-section high-density scanning on the sprayed primary support surface to obtain massive point cloud data; S6, intelligent analysis and automatic trimming: the central control computer system (4) constructs a three-dimensional model of the primary support based on the point cloud data, identifies unqualified point sets deviating from the reference plane, and sets path priorities in combination with the downward solidification strength gradient of sprayed concrete to generate bottom-up trimming control instructions to drive the milling and digging execution device (5) to implement layer-by-layer milling and trimming on the regions corresponding to the unqualified point sets; S7, re-inspection and archiving: after trimming, the three-dimensional laser scanning system (3) performs scanning re-inspection again, and the central control computer system (4) automatically archives the re-scanned point cloud data and the flatness evaluation results as traceable electronic construction records; S8, end processing: after trimming and re-inspection are completed, stop the operation of each execution device, disconnect the wind, water and electricity connections, make the equipment exit the working face, and perform cleaning and routine maintenance operations.

7. The method of claim 6, wherein, In the step S6, the point spacing of the massive point cloud data is ≤1cm, the full-section data collection of a 60m section can be completed within 15 minutes, the objective evaluation results are automatically generated in combination with the AI algorithm, and the construction optimization is guided through the visualization platform.

Images