Tunnel static excavation digital twin operation method and system and readable storage medium

Abstract

The invention provides a tunnel static excavation digital twin operation method and system and a computer readable storage medium, and the method comprises the steps: inputting the feature data of a rock mass in front into a pre-trained tunnel surrounding rock grade prediction module, enabling the tunnel surrounding rock grade prediction module to output a static excavation mode, and carrying out the construction according to the static excavation mode, the method comprises the steps that a surrounding rock grade prediction module is established, a real-scene model is established based on the three-dimensional laser scanning technology in the excavation process, then the real-scene model and a pre-established BIM reference model are aligned to form an over-excavation and under-excavation analysis model, then an over-excavation and under-excavation area is obtained through the over-excavation and under-excavation analysis model, and the surrounding rock grade prediction module and the over-excavation and under-excavation analysis module are linked in real time, so that the surrounding rock grade prediction accuracy is improved. According to the method, excavation deviation data and surrounding rock grades are automatically synchronized to a BIM reference model, support parameters are dynamically adjusted, model precision is continuously optimized by utilizing construction feedback data, a digital twin model covering a static excavation construction period of a tunnel is formed, tunneling efficiency, construction safety and resource utilization rate are further improved, and engineering cost is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of tunnel excavation and transportation construction, in particular to a static excavation method, more specifically, to a tunnel static excavation digital twin operation method and system, and a computer readable storage medium. BACKGROUND

[0002] Tunnels, as one of the important facilities of geotechnical engineering, have been widely used in the construction of subway, railway, highway and underground laboratory engineering. At present, drill and blast method is still the mainstream construction method for underground engineering due to its economy and efficiency, but the disturbance risk of blasting seismic waves to adjacent buildings and structures, and the safety hazards in the production and transportation of explosives, have always restricted the engineering application in densely populated urban areas and high sensitive environment. Since the 1980s, China has gradually carried out research and promotion of static excavation technology (such as hydraulic splitting, static expanding agent, etc.). This technology has advantages such as low disturbance, no flying rock, no toxic gas emission, and has achieved remarkable application results in municipal engineering, road excavation and high-risk mining fields. However, the existing static excavation construction faces some bottlenecks, (1) static excavation relies on manual experience to adjust the excavation parameters (such as splitting hole spacing, expanding agent dosage), but the uncertainty of the geological conditions in front of the working face leads to a lag in parameter adjustment, and geological mutations can easily cause surrounding rock collapse or support failure, especially in complex geology such as fault fracture zone and soft and hard interbedded layers, the collapse risk increases significantly; (2) the traditional construction concept of "never overbreak, but never underbreak" leads to frequent overbreak, and the overbreak directly increases the amount of blasting, slag and concrete for primary support, which not only increases the engineering cost, but also aggravates the safety risk due to the redistribution of surrounding rock stress; (3) the current construction relies on the experience of on-site personnel to judge, and lacks real-time correlation analysis of geological data, excavation parameters and support effect, making it difficult to form a closed-loop control mechanism of "detection-analysis-decision-execution".

[0003] Therefore, there is an urgent need for a tunnel static excavation digital twin operation scheme that can dynamically monitor the tunnel static excavation construction process, ensure the smooth breakthrough of tunnel excavation, solve the problems of lagging in surrounding rock identification and difficulty in controlling overbreak and underbreak in traditional tunnel engineering, significantly improve the excavation efficiency, construction safety and resource utilization rate, and reduce the engineering cost. SUMMARY

[0004] In view of the above problems, the purpose of the present application is to provide a tunnel static excavation digital twin operation method and system, and a computer readable storage medium, to solve the technical problems of lagging in surrounding rock identification and difficulty in controlling overbreak and underbreak in traditional tunnel engineering.

[0005] The tunnel static excavation digital twin operation method provided by the present application comprises the following steps:

[0006] By obtaining the characteristic data of the rock mass ahead to be excavated through advanced geological exploration, the characteristic data of the rock mass ahead is input into the pre-trained tunnel surrounding rock grade prediction module so that the tunnel surrounding rock grade prediction module outputs the static excavation mode.

[0007] Construction was carried out according to the described static excavation method, and a real-scene model was established based on three-dimensional laser scanning technology during the excavation process;

[0008] Align the real-world model with the pre-established BIM benchmark model to form an over-excavation and under-excavation analysis model;

[0009] The over- or under-excavation analysis model is used to obtain over- or under-excavation areas. Construction is carried out on the over- or under-excavation areas to form completed areas. After the completed areas are formed, the over- or under-excavation analysis model is updated to carry out the next round of construction. The parameter data of the completed areas are used as training samples and synchronously fed back to the tunnel surrounding rock grade prediction module to update the tunnel surrounding rock grade prediction module.

[0010] Preferably, the pre-training of the tunnel surrounding rock grade prediction module includes:

[0011] The raw data to be acquired will include at least the uniaxial compressive strength of rock, rock integrity index, deformation modulus, Poisson's ratio, different surrounding rock grade requirements given by existing excavation specifications and design data, and support methods.

[0012] The raw data is cleaned and labeled to form a dataset;

[0013] The pre-built neural network is repeatedly trained using 70% of the dataset as the training set to generate a rock mass level prediction network until the loss function of the rock mass level prediction network is lower than a preset loss threshold. Then, the last generated rock mass level prediction network is used as the tunnel rock mass level prediction module.

[0014] Preferably, the rock mass characteristic data ahead is input into a pre-trained tunnel surrounding rock grade prediction module, so that the tunnel surrounding rock grade prediction module outputs a static excavation method, including:

[0015] The characteristic data of the rock mass to be excavated ahead are input into the tunnel surrounding rock grade prediction module in real time; wherein, the characteristic data of the rock mass ahead includes the uniaxial compressive strength of the rock, the rock mass integrity index, the deformation modulus or Poisson's ratio;

[0016] The tunnel surrounding rock grade prediction module outputs the surrounding rock grade distribution based on the rock mass characteristic data ahead.

[0017] The distribution of surrounding rock grades exceeding a preset proportion is taken as the surrounding rock output grade;

[0018] The knowledge base is automatically retrieved based on the surrounding rock output level to determine the static excavation method and support scheme corresponding to the surrounding rock output level.

[0019] Preferably, construction is carried out according to the static excavation method, and a real-scene model is established based on three-dimensional laser scanning technology during the excavation process, including:

[0020] Point cloud data of the working face is obtained by three-dimensional laser scanning technology; wherein, each point in the point cloud data includes three-dimensional coordinates in the coordinate system of the point, color representing object attributes, classification value, laser reflection intensity, and normal vector;

[0021] The point cloud data is subjected to point cloud denoising and point cloud thinning to obtain a standard point cloud file;

[0022] Point cloud meshing technology is used to create triangular meshes from the standard point cloud file to form a three-dimensional reality model.

[0023] Preferably, a pre-established BIM benchmark model is included, comprising:

[0024] A three-dimensional integrated geological-structural dataset is established based on the pre-acquired tunnel design drawings, geological exploration reports, and construction specifications.

[0025] The Revit Dynamo plugin was used to perform parametric modeling based on the aforementioned 3D geological-structural integrated dataset to form a BIM benchmark model.

[0026] Preferably, the over- or under-excavation area is obtained through the over- or under-excavation analysis model, and the over- or under-excavation area is processed to form a completed area, including:

[0027] The over- and under-excavation values ​​of the cross section are determined by obtaining the distance from each point on the outline of the real scene model to each point on the outline of the BIM benchmark model through the over- and under-excavation analysis model.

[0028] Boolean operations are used to obtain irregular polygonal areas in the BIM benchmark model that are not covered by the real scene model based on the over-excavation and under-excavation values;

[0029] The irregular polygonal region is quantized to obtain the over-excavation and under-excavation area values;

[0030] Along the tunnel advancement direction, reference planes are set according to the preset construction advance intervals. Two adjacent reference planes are selected as the starting section and the ending section. The over-excavation and under-excavation area values ​​of the starting section and the ending section are obtained and the infinitesimal volume integral is used to obtain the over-excavation and under-excavation volume between the starting section and the ending section.

[0031] Irregular polygonal regions with the aforementioned over- or under-excavation volumes are designated as over- or under-excavation regions, and targeted over-excavation or under-excavation construction is carried out on these regions to form completed areas.

[0032] Preferably, updating the over-excavation and under-excavation analysis model after the completed area is formed includes:

[0033] After the completed area is formed, a new real-scene model is re-established based on three-dimensional laser scanning technology to update the real-scene model; and a completion assessment is performed on the completed area, and the BIM benchmark model is evaluated and adjusted based on the completion assessment results to update the BIM benchmark model.

[0034] Align the updated reality model with the updated BIM benchmark model to obtain the updated over-excavation and under-excavation analysis model.

[0035] This invention also provides a digital twin operation system for static tunnel excavation, which implements the digital twin operation method for static tunnel excavation as described above, including:

[0036] The tunnel surrounding rock grade prediction module is used to receive the characteristic data of the rock mass to be excavated ahead of the tunnel obtained through advanced geological exploration, and output the static excavation method based on the characteristic data of the rock mass ahead.

[0037] The model building module is used to carry out construction according to the static excavation method, and to build a real scene model based on three-dimensional laser scanning technology during the excavation process, and to align the real scene model with the pre-established BIM benchmark model to form an over-excavation and under-excavation analysis model.

[0038] The construction feedback linkage module is used to obtain the over- or under-excavation area through the over- or under-excavation analysis model, carry out construction processing on the over- or under-excavation area to form a completed area, update the over- or under-excavation analysis model after the completed area is formed to carry out the next round of construction, and synchronously feed back the parameter data of the completed area as a training sample to the tunnel surrounding rock grade prediction module to update the tunnel surrounding rock grade prediction module.

[0039] Preferably, the updated over-excavation and under-excavation analysis model includes a relatively homogeneous updated reality model and an updated BIM benchmark model; wherein,

[0040] After the completed area is formed, a new real-world model is re-established based on 3D laser scanning technology to obtain the updated real-world model; a completion assessment is performed on the completed area, and the BIM benchmark model is evaluated and adjusted based on the completion assessment results to obtain the updated BIM benchmark model.

[0041] In addition, the present invention provides a computer-readable storage medium storing a computer program, characterized in that the computer program, when executed by a processor, implements the aforementioned digital twin operation method for static tunnel excavation.

[0042] As can be seen from the above technical solution, the digital twin operation method, system, and computer-readable storage medium for static tunnel excavation provided by this invention first inputs the rock mass characteristic data ahead into a pre-trained tunnel surrounding rock grade prediction module, so that the tunnel surrounding rock grade prediction module outputs a static excavation mode. Then, construction is carried out according to the static excavation mode, and a real-scene model is established based on three-dimensional laser scanning technology during the excavation process. The real-scene model is then aligned with a pre-established BIM benchmark model to form an over-excavation / under-excavation analysis model. Subsequently, over-excavation / under-excavation areas are obtained through the over-excavation / under-excavation analysis model, and construction processing is performed on these areas to form completed areas. After the completed areas are formed, the over-excavation / under-excavation analysis model is updated for the next round. Construction is carried out, and the parameter data of the completed area is used as training samples to be synchronously fed back to the tunnel surrounding rock grade prediction module to update the tunnel surrounding rock grade prediction module. In this way, the rock strata data of the completed area is fed back to the tunnel surrounding rock grade prediction module to continuously improve the model accuracy and update the relevant parameters of the tunnel surrounding rock grade prediction module and the BIM benchmark model. By linking the surrounding rock grade prediction module with the over-excavation and under-excavation analysis module in real time, the excavation deviation data and surrounding rock grade are automatically synchronized to the BIM benchmark model, the support parameters are dynamically adjusted, and the model accuracy is continuously optimized using construction feedback data, forming a digital twin model covering the static excavation construction cycle of the tunnel, thereby improving tunneling efficiency, construction safety and resource utilization, and reducing project costs. Attached Figure Description

[0043] Other objects and results of the invention will become more apparent and readily understood by referring to the following description taken in conjunction with the accompanying drawings, and with a more complete understanding of the invention. In the drawings:

[0044] Figure 1 A flowchart of a digital twin operation method for static tunnel excavation according to an embodiment of the present invention;

[0045] Figure 2 This is a schematic diagram illustrating the execution flow of a specific embodiment of the digital twin operation method for static tunnel excavation according to an embodiment of the present invention;

[0046] Figure 3 This is a system block diagram of a digital twin operation system for static tunnel excavation according to an embodiment of the present invention. Detailed Implementation

[0047] The existing standard axle load conversion method is no longer suitable for heavy-duty traffic.

[0048] To address the aforementioned problems, this invention provides a digital twin operation method, system, and computer-readable storage medium for static tunnel excavation. Specific embodiments of this invention will be described in detail below with reference to the accompanying drawings.

[0049] To illustrate the digital twin operation method, system, and computer-readable storage medium for static tunnel excavation provided by this invention... Figures 1-3 The embodiments of the present invention are illustrated by way of example.

[0050] The following description of exemplary embodiments is merely illustrative and is in no way intended to limit the invention or its application or use. Techniques and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques and equipment should be considered part of the specification.

[0051] like Figure 1 As shown, this invention provides a digital twin operation method for static tunnel excavation, which mainly includes the following steps:

[0052] S1: Obtain the characteristic data of the rock mass to be excavated by advanced geological exploration, and input the characteristic data of the rock mass to be excavated into the pre-trained tunnel surrounding rock grade prediction module so that the tunnel surrounding rock grade prediction module outputs the static excavation mode.

[0053] S2: Carry out the construction according to the static excavation method, and establish a real-scene model based on three-dimensional laser scanning technology during the excavation process;

[0054] S3: Align the real-world model with the pre-established BIM benchmark model to form an over-excavation and under-excavation analysis model;

[0055] S4: Obtain the over- or under-excavation area through the over- or under-excavation analysis model, carry out construction processing on the over- or under-excavation area to form a completed area, and update the over- or under-excavation analysis model after the completed area is formed to carry out the next round of construction, and synchronously feed back the parameter data of the completed area as a training sample to the tunnel surrounding rock grade prediction module to update the tunnel surrounding rock grade prediction module.

[0056] Step S1 involves obtaining characteristic data of the rock mass ahead of the excavation site through advanced geological exploration, and inputting this characteristic data into a pre-trained tunnel surrounding rock grade prediction module, so that the tunnel surrounding rock grade prediction module outputs a static excavation method; wherein,

[0057] The pre-training of the tunnel surrounding rock grade prediction module includes:

[0058] S01: The pre-acquired data shall include at least the uniaxial compressive strength of rock, rock integrity index, deformation modulus, Poisson's ratio, different surrounding rock grade requirements given by existing excavation specifications and design data, and support methods as raw data;

[0059] S02: Perform data cleaning and label generation on the raw data to form a dataset;

[0060] S03: Use 70% of the dataset as the training set to repeatedly train the pre-built neural network to generate a rock mass level prediction network until the loss function of the rock mass level prediction network is lower than a preset loss threshold. Then, use the last generated rock mass level prediction network as the tunnel rock mass level prediction module. Use 30% of the dataset as the validation set. The specific training and validation methods are not described here. The commonly used training and validation methods can be used.

[0061] Specifically, the rock mass characteristic data ahead is input into a pre-trained tunnel surrounding rock grade prediction module, so that the tunnel surrounding rock grade prediction module outputs a static excavation method, including:

[0062] S11: Input the characteristic data of the rock mass to be excavated ahead into the tunnel surrounding rock grade prediction module in real time; wherein, the characteristic data of the rock mass ahead includes the uniaxial compressive strength of the rock, the rock mass integrity index, the deformation modulus or Poisson's ratio;

[0063] S12: The tunnel surrounding rock grade prediction module outputs the surrounding rock grade distribution based on the rock mass characteristic data ahead;

[0064] S13: The distribution of surrounding rock grades exceeding a preset ratio is taken as the surrounding rock output grade;

[0065] S14: Automatically retrieve the knowledge base based on the surrounding rock output level to determine the static excavation method and support scheme corresponding to the surrounding rock output level.

[0066] Step S2 involves constructing the excavation according to the static excavation method and establishing a real-world model based on 3D laser scanning technology during the excavation process, including:

[0067] S21: Obtain point cloud data of the working face using three-dimensional laser scanning technology; wherein, each point in the point cloud data includes three-dimensional coordinates in the coordinate system of the point, color representing object attributes, classification value, laser reflection intensity, and normal vector;

[0068] S22: Perform point cloud denoising and point cloud thinning on the point cloud data to obtain a standard point cloud file;

[0069] S23: Use point cloud meshing technology to create a triangular mesh from the standard point cloud file to form a three-dimensional real-world model.

[0070] Step S3 is the process of aligning the real-world model with the pre-established BIM benchmark model to form an over-excavation / under-excavation analysis model, wherein the pre-established BIM benchmark model includes:

[0071] S31: Establish a three-dimensional integrated geological-structural dataset based on the pre-acquired tunnel design drawings, geological exploration reports, and construction specifications.

[0072] S32: Parametric modeling is performed using the Revit Dynamo plugin based on the aforementioned 3D geological-structural integrated dataset to form a BIM benchmark model.

[0073] Step S4 involves obtaining over- or under-excavation areas through the over- or under-excavation analysis model, performing construction processing on these areas to form completed areas, updating the over- or under-excavation analysis model after forming the completed areas for the next round of construction, and synchronously feeding the parameter data of the completed areas back to the tunnel surrounding rock grade prediction module as training samples to update the module. The process of obtaining over- or under-excavation areas through the over- or under-excavation analysis model and performing construction processing on these areas to form completed areas includes:

[0074] S41: Obtain the distance from each point on the outline of the real scene model to each point on the outline of the BIM benchmark model through the over-excavation and under-excavation analysis model to determine the over-excavation and under-excavation value of the section;

[0075] S42: Using Boolean operations, the irregular polygonal areas in the BIM benchmark model that are not covered by the real scene model are obtained based on the over-excavation and under-excavation values;

[0076] S43: Quantize the irregular polygonal region to obtain the over- or under-excavation area value;

[0077] S44: Along the tunnel advancement direction, set reference planes according to the preset construction advance intervals, select two adjacent reference planes as the starting section and the ending section, obtain the over-excavation and under-excavation area values ​​of the starting section and the ending section, and perform infinitesimal volume integration to obtain the over-excavation and under-excavation volume between the starting section and the ending section.

[0078] S45: The irregular polygonal region with the over- or under-excavation volume is designated as the over- or under-excavation region, and targeted over-excavation or under-excavation construction is carried out on the over- or under-excavation region to form the completed area.

[0079] After the completed area is formed, the over-excavation and under-excavation analysis model is updated, including:

[0080] S46: After the completed area is formed, a new real-scene model is re-established based on three-dimensional laser scanning technology to update the real-scene model; and a completion assessment is performed on the completed area, and the BIM benchmark model is evaluated and adjusted based on the completion assessment results to update the BIM benchmark model.

[0081] S47: Align the updated reality model with the updated BIM baseline model to obtain the updated over-excavation and under-excavation analysis model.

[0082] In this embodiment, there is no restriction on the timing of establishing the BIM baseline model and the real-world model. Either one can be established first, as long as it can be aligned and combined. In this embodiment, the BIM baseline model is established in advance, and the real-world model is established in real time during construction, so it can be directly aligned and used with the established BIM baseline model.

[0083] Specifically, such as Figure 2 As shown, in one specific embodiment, prior to construction, it is necessary to collect survey data and relevant information from existing projects, such as data on uniaxial compressive strength of rock, rock mass integrity index, deformation modulus, and Poisson's ratio, as well as requirements for different surrounding rock grades and support methods provided by specifications and design documents, as data to be used for subsequent neural network training. Since the prediction model established using deep learning algorithms may be confused by too many input features, leading to overfitting of the model's prediction results and low model accuracy and generalization. Therefore, in this embodiment, when training the deep learning model, the parameters input to the model are screened, and sensitivity analysis methods, grey relational analysis methods, etc., are used to select appropriate model input features. Then, a knowledge base for static excavation methods is established based on the mechanical parameters of the surrounding rock. For example, for soft rock (E<5GPa), the combination of hydraulic rock splitter and static expansion agent is preferred; for medium-hard rock (5GPa≤E<15GPa), carbon dioxide blasting and cantilever excavator are selected for coordinated operation; for hard rock (E≥15GPa), thermal rock splitting is preferred. This allows the tunnel surrounding rock grade prediction module to automatically search the knowledge base according to the surrounding rock output grade to determine the static excavation method and support scheme corresponding to the surrounding rock output grade.

[0084] After the preparatory work was completed, deep learning algorithms such as deep convolutional networks were used to train and establish a tunnel surrounding rock grade prediction module, using surrounding rock classification, static excavation methods, and support schemes as output features. Advanced geological surveys were used to obtain rock mass feature data ahead, which was then input into the tunnel surrounding rock grade prediction module. Based on this data, the module (tunnel surrounding rock grade prediction module) could intelligently analyze the current geological conditions, output the corresponding surrounding rock grade, and provide the corresponding static excavation method and support parameters. The module was compared with the design documents; if it matched the design documents, construction was carried out according to the design documents; otherwise, construction was carried out according to the construction scheme provided by the prediction module, which essentially meant construction was carried out according to the static excavation method output by the tunnel surrounding rock grade prediction module.

[0085] It is also necessary to pre-integrate design drawings (including tunnel centerline coordinates and slope parameters), geological survey reports (including rock mass mechanical properties, integrity coefficient Kv, and other parameters), and construction specification documents to establish a three-dimensional geological-structural integrated dataset. Parametric modeling is achieved using the Revit Dynamo plugin to establish a tunnel BIM benchmark model and automatically associate geological layers with support component family libraries (such as anchor bolt length / spacing, steel arch spacing, and other parameters).

[0086] After construction using the static excavation method, point cloud data of the working face is acquired through 3D laser scanning technology. Each point in the point cloud data contains a wealth of information, including its 3D coordinates (XYZ), color (RGB) representing object attributes, classification value, laser reflection intensity, and normal vector. When using laser scanning, the specific locations and number of scanning stations should be determined based on the site survey. During data acquisition, the laser scanner should be positioned at the optimal distance and angle according to the on-site construction environment.

[0087] The point cloud data obtained through scanning should be processed using professional point cloud processing software, including but not limited to point cloud denoising and point cloud thinning, to generate a processed XYZ coordinate file. After point cloud compression and thinning, triangular meshing is used to build a 3D reality model. Under the same coordinate system, the reality model is aligned with the BIM benchmark model to form an over-excavation and under-excavation analysis model.

[0088] By calculating the over- and under-excavation analysis model, the distance from each point on the outline of the real model to the outline of the design model is compared to calculate the over- and under-excavation value of the section. At the same time, the model is trimmed so that the over- and under-excavation areas of the two models can be obtained by Boolean subtraction. Based on this, a reference plane for the advance is selected, and the over- and under-excavation volume is obtained by solving the model volume to the reference plane.

[0089] After processing the over- and under-excavation areas, point cloud data of the tunnel face was re-acquired using a laser scanner to construct a 3D reality model and update the BIM benchmark model. Simultaneously, rock strata data for the completed areas were fed back to the tunnel surrounding rock grade prediction module to continuously improve model accuracy and update relevant parameters in both the module and the BIM benchmark model. By linking the surrounding rock grade prediction module and the over- and under-excavation analysis module in real time, excavation deviation data and surrounding rock grades were automatically synchronized to the BIM benchmark model, support parameters were dynamically adjusted, and construction feedback data was used to continuously optimize model accuracy, forming a digital twin model covering the entire static excavation construction cycle of the tunnel.

[0090] Therefore, the digital twin operation method for static tunnel excavation in this embodiment integrates design drawings, geological data, and construction specifications to construct a three-dimensional geological-structural integrated model. Combined with parametric modeling and deep learning algorithms, it achieves intelligent prediction of surrounding rock grade and dynamic matching of optimal excavation technology. By utilizing three-dimensional laser scanning to acquire point cloud data of the tunnel face in real time, and accurately calculating deviations through an over-excavation / under-excavation analysis module, it dynamically adjusts support parameters in conjunction with the BIM model, forming a digital twin system covering the entire construction cycle. This method effectively solves problems such as delayed surrounding rock identification and difficulty in controlling over-excavation / under-excavation in traditional tunnel engineering, significantly improving tunneling efficiency, construction safety, and resource utilization, reducing project costs, and contributing significantly to the stable development of the social economy.

[0091] like Figure 3 As shown, the present invention also provides a digital twin operation system 100 for static tunnel excavation, which implements the aforementioned digital twin operation method for static tunnel excavation, including:

[0092] The tunnel surrounding rock grade prediction module 110 is used to receive the rock mass characteristic data to be excavated ahead of the tunnel obtained through advanced geological exploration, and output the static excavation mode according to the rock mass characteristic data ahead.

[0093] The model building module 120 is used to carry out construction according to the static excavation method, and to build a real scene model based on three-dimensional laser scanning technology during the excavation process, and to align the real scene model with the pre-established BIM benchmark model to form an over-excavation and under-excavation analysis model.

[0094] The construction feedback linkage module 130 is used to obtain the over-excavation and under-excavation area through the over-excavation and under-excavation analysis model, carry out construction processing on the over-excavation and under-excavation area to form a completed area, update the over-excavation and under-excavation analysis model after the completed area is formed to carry out the next round of construction, and synchronously feed back the parameter data of the completed area as a training sample to the tunnel surrounding rock grade prediction module to update the tunnel surrounding rock grade prediction module.

[0095] The updated over- and under-excavation analysis model includes a relatively homogeneous updated reality model and an updated BIM benchmark model; among them,

[0096] After the completed area is formed, a new real-world model is re-established based on 3D laser scanning technology to obtain the updated real-world model; a completion assessment is performed on the completed area, and the BIM benchmark model is evaluated and adjusted based on the completion assessment results to obtain the updated BIM benchmark model.

[0097] For a more specific implementation of the digital twin operation system for static tunnel excavation, please refer to the above-described embodiment of the digital twin operation method for static tunnel excavation, which will not be elaborated here.

[0098] This invention also provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements the aforementioned digital twin operation method for static tunnel excavation. The readable storage medium can be non-volatile or volatile. The readable storage medium stores a device setting self-tuning program, which, when executed by a processor, implements:

[0099] By obtaining the characteristic data of the rock mass ahead to be excavated through advanced geological exploration, the characteristic data of the rock mass ahead is input into the pre-trained tunnel surrounding rock grade prediction module so that the tunnel surrounding rock grade prediction module outputs the static excavation mode.

[0100] Construction was carried out according to the described static excavation method, and a real-scene model was established based on three-dimensional laser scanning technology during the excavation process.

[0101] Align the real-world model with the pre-established BIM benchmark model to form an over-excavation and under-excavation analysis model;

[0102] The over- or under-excavation analysis model is used to obtain over- or under-excavation areas. Construction is carried out on the over- or under-excavation areas to form completed areas. After the completed areas are formed, the over- or under-excavation analysis model is updated to carry out the next round of construction. The parameter data of the completed areas are used as training samples and synchronously fed back to the tunnel surrounding rock grade prediction module to update the tunnel surrounding rock grade prediction module.

[0103] Specifically, the specific implementation method when the computer program is executed by the processor can be referred to the description of the relevant steps in the embodiment of the digital twin operation method for static tunnel excavation, which will not be repeated here.

[0104] In the several embodiments provided by this invention, it should be understood that the disclosed methods, systems, and readable storage media can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0105] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0106] The method, system, and computer-readable storage medium for static tunnel excavation proposed in this invention have been described above by way of example with reference to the accompanying drawings. However, those skilled in the art should understand that various modifications can be made to the method, system, and computer-readable storage medium for static tunnel excavation proposed in this invention without departing from the scope of the invention. Therefore, the scope of protection of this invention should be determined by the appended claims.

Claims

1. A digital twin operation method for static tunnel excavation, characterized in that, include: By obtaining the characteristic data of the rock mass ahead to be excavated through advanced geological exploration, the characteristic data of the rock mass ahead is input into the pre-trained tunnel surrounding rock grade prediction module so that the tunnel surrounding rock grade prediction module outputs the static excavation mode. Construction was carried out according to the described static excavation method, and a real-scene model was established based on three-dimensional laser scanning technology during the excavation process. Align the real-world model with the pre-established BIM benchmark model to form an over-excavation and under-excavation analysis model; The over- or under-excavation analysis model is used to obtain over- or under-excavation areas. Construction is carried out on the over- or under-excavation areas to form completed areas. After the completed areas are formed, the over- or under-excavation analysis model is updated to carry out the next round of construction. The parameter data of the completed areas are used as training samples and synchronously fed back to the tunnel surrounding rock grade prediction module to update the tunnel surrounding rock grade prediction module.

2. The digital twin operation method for static tunnel excavation as described in claim 1, characterized in that, The pre-training of the tunnel surrounding rock grade prediction module includes: The raw data to be acquired will include at least the uniaxial compressive strength of rock, rock integrity index, deformation modulus, Poisson's ratio, different surrounding rock grade requirements given by existing excavation specifications and design data, and support methods. The raw data is cleaned and labeled to form a dataset; The pre-built neural network is repeatedly trained using 70% of the dataset as the training set to generate a rock mass level prediction network until the loss function of the rock mass level prediction network is lower than a preset loss threshold. Then, the last generated rock mass level prediction network is used as the tunnel rock mass level prediction module.

3. The digital twin operation method for static tunnel excavation as described in claim 2, wherein the rock mass characteristic data ahead is input into a pre-trained tunnel surrounding rock grade prediction module, so that the tunnel surrounding rock grade prediction module outputs a static excavation mode, including: The characteristic data of the rock mass to be excavated ahead are input into the tunnel surrounding rock grade prediction module in real time; wherein, the characteristic data of the rock mass ahead includes the uniaxial compressive strength of the rock, the rock mass integrity index, the deformation modulus or Poisson's ratio; The tunnel surrounding rock grade prediction module outputs the surrounding rock grade distribution based on the rock mass characteristic data ahead. The distribution of surrounding rock grades exceeding a preset proportion is taken as the surrounding rock output grade; The knowledge base is automatically retrieved based on the surrounding rock output level to determine the static excavation method and support scheme corresponding to the surrounding rock output level.

4. The digital twin operation method for static tunnel excavation as described in claim 1, characterized in that, Construction will be carried out according to the described static excavation method, and a real-world model will be established based on three-dimensional laser scanning technology during the excavation process, including: Point cloud data of the working face is obtained by three-dimensional laser scanning technology; wherein, each point in the point cloud data includes three-dimensional coordinates in the coordinate system of the point, color representing object attributes, classification value, laser reflection intensity, and normal vector; The point cloud data is subjected to point cloud denoising and point cloud thinning to obtain a standard point cloud file; Point cloud meshing technology is used to create triangular meshes from the standard point cloud file to form a three-dimensional reality model.

5. The digital twin operation method for static tunnel excavation as described in claim 4, characterized in that, Pre-established BIM baseline model, including: A three-dimensional integrated geological-structural dataset is established based on the pre-acquired tunnel design drawings, geological exploration reports, and construction specifications. The Revit Dynamo plugin was used to perform parametric modeling based on the aforementioned 3D geological-structural integrated dataset to form a BIM benchmark model.

6. The digital twin operation method for static tunnel excavation as described in claim 5, characterized in that, The over- or under-excavation analysis model is used to obtain the over- or under-excavation areas, and construction processing is carried out on the over- or under-excavation areas to form the completed areas, including: The over- and under-excavation values ​​of the cross section are determined by obtaining the distance from each point on the outline of the real scene model to each point on the outline of the BIM benchmark model through the over- and under-excavation analysis model. Boolean operations are used to obtain irregular polygonal areas in the BIM benchmark model that are not covered by the real scene model based on the over-excavation and under-excavation values; The irregular polygonal region is quantized to obtain the over-excavation and under-excavation area values; Along the tunnel advancement direction, reference planes are set according to the preset construction advance intervals. Two adjacent reference planes are selected as the starting section and the ending section. The over-excavation and under-excavation area values ​​of the starting section and the ending section are obtained and the infinitesimal volume integral is used to obtain the over-excavation and under-excavation volume between the starting section and the ending section. Irregular polygonal regions with the aforementioned over- or under-excavation volumes are designated as over- or under-excavation regions, and targeted over-excavation or under-excavation construction is carried out on these regions to form completed areas.

7. The digital twin operation method for static tunnel excavation as described in claim 6, characterized in that, After the completed area is formed, the over-excavation and under-excavation analysis model is updated, including: After the completed area is formed, a new real-scene model is re-established based on three-dimensional laser scanning technology to update the real-scene model; and a completion assessment is performed on the completed area, and the BIM benchmark model is evaluated and adjusted based on the completion assessment results to update the BIM benchmark model. Align the updated reality model with the updated BIM benchmark model to obtain the updated over-excavation and under-excavation analysis model.

8. A digital twin operation system for static tunnel excavation, characterized in that, Implementing the digital twin operation method for static tunnel excavation as described in any one of claims 1-7 includes: The tunnel surrounding rock grade prediction module is used to receive the characteristic data of the rock mass to be excavated ahead of the tunnel obtained through advanced geological exploration, and output the static excavation method based on the characteristic data of the rock mass ahead. The model building module is used to carry out construction according to the static excavation method, and to build a real scene model based on three-dimensional laser scanning technology during the excavation process, and to align the real scene model with the pre-established BIM benchmark model to form an over-excavation and under-excavation analysis model. The construction feedback linkage module is used to obtain the over- or under-excavation area through the over- or under-excavation analysis model, carry out construction processing on the over- or under-excavation area to form a completed area, update the over- or under-excavation analysis model after the completed area is formed to carry out the next round of construction, and synchronously feed back the parameter data of the completed area as a training sample to the tunnel surrounding rock grade prediction module to update the tunnel surrounding rock grade prediction module.

9. The digital twin operation system for static tunnel excavation as described in claim 8, characterized in that, The updated over-excavation and under-excavation analysis model includes a relatively homogeneous updated reality model and an updated BIM benchmark model; among them, After the completed area is formed, a new real-world model is re-established based on 3D laser scanning technology to obtain the updated real-world model; a completion assessment is performed on the completed area, and the BIM benchmark model is evaluated and adjusted based on the completion assessment results to obtain the updated BIM benchmark model.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the digital twin operation method for static tunnel excavation as described in any one of claims 1 to 7.

Images