Method and system for reinforcing tunnel based on fiber reinforced cement-based composite material

Abstract

The invention provides a method and system for reinforcing a tunnel based on a fiber reinforced cement-based composite material, and relates to the technical field of tunnel repair, and the method comprises the following steps: obtaining tunnel surface stress data information and three-dimensional point cloud data thereof; and inputting the stress data information into a preset damage judgment model to determine a tunnel damage area. And applying the damage area information to a damage expansion trend prediction model to obtain a future damage development trend. And preprocessing the three-dimensional point cloud data, and extracting damage features of the tunnel, including deformation and crack features. Damage features are analyzed and evaluated through the multi-scale fractal dimension, and reinforcement position analysis is carried out in combination with fractal dimension values of the damage features, damage area information and an extension trend prediction result. And generating a reinforcement scheme based on the to-be-reinforced position information and the historical reinforcement scheme, and finally determining a reinforcement scheme of the tunnel by using the fiber-reinforced cement-based composite material. The method effectively improves the safety and durability of the tunnel.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel repair, and more particularly, to a method and system for reinforcing tunnels based on fiber-reinforced cementitious composites. Background Art

[0002] With the acceleration of the urbanization process, tunnels, as important transportation infrastructure, are experiencing increasing usage frequencies and load-bearing pressures. Existing tunnel reinforcement technologies mainly rely on traditional concrete materials, which have obvious limitations in terms of stress-bearing capacity, durability, and construction flexibility. When tunnels are damaged, it is often difficult to achieve timely and effective repair. This has led to the emergence of structural defects in many tunnels during their use, seriously affecting safety and service life. Although traditional technologies can reinforce tunnels to a certain extent, they lack targeted solutions when faced with complex tunnel damage situations and are difficult to achieve the best reinforcement effect. Especially in the judgment of damage locations and the prediction of the expansion trend, the accuracy of existing technologies is insufficient, resulting in problems such as repetitive labor and resource waste in reinforcement plans.

[0003] Therefore, there is an urgent need for a tunnel reinforcement method and system based on fiber-reinforced cementitious composites to solve the above problems. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for reinforcing tunnels based on fiber-reinforced cementitious composites to improve the above problems. To achieve the above purpose, the technical solutions adopted by the present invention are as follows:

[0005] In a first aspect, the present application provides a method for reinforcing tunnels based on fiber-reinforced cementitious composites, including:

[0006] Obtaining tunnel surface stress data information and three-dimensional point cloud data of the tunnel;

[0007] Sending the tunnel surface stress data information to a preset tunnel damage judgment model for tunnel damage judgment to obtain tunnel damage area information;

[0008] Sending the tunnel damage area information to a preset tunnel damage expansion trend prediction model for damage expansion trend prediction to obtain a prediction result of the tunnel damage expansion trend;

[0009] Sending the three-dimensional point cloud data of the tunnel to a preset preprocessing model for preprocessing to obtain tunnel damage feature data information, where the tunnel damage feature data information includes the deformation feature and crack feature of the tunnel;

[0010] Send the damage characteristic data information of the tunnel to a preset analysis model for multi-scale fractal dimension analysis to obtain the fractal dimension value of the tunnel damage characteristics;

[0011] Send the fractal dimension value of the tunnel damage characteristics, the tunnel damage area information, and the prediction result of the tunnel damage expansion trend to the tunnel reinforcement analysis model for analysis of the positions to be reinforced of the tunnel to obtain the information on the positions to be reinforced of the tunnel;

[0012] Generate a reinforcement plan based on the information on the positions to be reinforced of the tunnel and a preset historical tunnel reinforcement plan, and optimize the generated reinforcement plan based on the simulated annealing algorithm and the parameters of the fiber-reinforced cementitious composite material to obtain an optimized plan for reinforcing the tunnel with the fiber-reinforced cementitious composite material;

[0013] Reinforce the tunnel with the fiber-reinforced cementitious composite material based on the optimized plan for reinforcing the tunnel with the fiber-reinforced cementitious composite material to obtain a reinforced tunnel.

[0014] In a second aspect, the present application also provides a system for reinforcing a tunnel with a fiber-reinforced cementitious composite material, including:

[0015] An acquisition unit for acquiring the tunnel surface stress data information and the three-dimensional point cloud data of the tunnel;

[0016] A judgment unit for sending the tunnel surface stress data information to a preset tunnel damage judgment model for tunnel damage judgment to obtain the tunnel damage area information;

[0017] A prediction unit for sending the tunnel damage area information to a preset tunnel damage expansion trend prediction model for damage expansion trend prediction to obtain the prediction result of the tunnel damage expansion trend;

[0018] A processing unit for sending the three-dimensional point cloud data of the tunnel to a preset preprocessing model for preprocessing to obtain the tunnel damage characteristic data information, where the tunnel damage characteristic data information includes the deformation characteristics and crack characteristics of the tunnel;

[0019] A first analysis unit for sending the tunnel damage characteristic data information to a preset analysis model for multi-scale fractal dimension analysis to obtain the fractal dimension value of the tunnel damage characteristics;

[0020] A second analysis unit for sending the fractal dimension value of the tunnel damage characteristics, the tunnel damage area information, and the prediction result of the tunnel damage expansion trend to the tunnel reinforcement analysis model for analysis of the positions to be reinforced of the tunnel to obtain the information on the positions to be reinforced of the tunnel;

[0021] An optimization unit, configured to generate a reinforcement plan based on the information of the position to be reinforced of the tunnel and a preset historical tunnel reinforcement plan, and optimize the generated reinforcement plan based on the simulated annealing algorithm and the parameters of the fiber-reinforced cementitious composite material to obtain an optimized plan for reinforcing the tunnel with the fiber-reinforced cementitious composite material;

[0022] A reinforcement unit, configured to reinforce the tunnel with the fiber-reinforced cementitious composite material based on the optimized plan for reinforcing the tunnel with the fiber-reinforced cementitious composite material to obtain a reinforced tunnel.

[0023] The beneficial effects of the present invention are as follows:

[0024] The present invention first obtains the stress data information and three-dimensional point cloud data on the surface of the tunnel, identifies the damaged area by establishing a tunnel damage judgment model, and analyzes the future damage path by using a damage propagation trend prediction model. Then, the damage characteristics of the tunnel are extracted by processing the three-dimensional point cloud data, and multi-scale fractal dimension analysis is performed to accurately obtain the damage characteristic parameters. These data will be comprehensively applied to the tunnel reinforcement analysis model to identify the position to be reinforced and generate a plan. Finally, the method of the present application optimizes the generated reinforcement plan through the simulated annealing algorithm to ensure the maximum utilization efficiency of the reinforcement material, significantly improving the efficiency and effect of tunnel reinforcement and overcoming the deficiencies in the application of traditional technologies.

[0025] Other features and advantages of the present invention will be described in the subsequent specification, and some of them will become apparent from the specification or can be understood by implementing the embodiments of the present invention. The objectives and other advantages of the present invention can be realized and obtained by the structures specifically pointed out in the written specification, claims, and drawings. Description of the Drawings

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the embodiments. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0027] Figure 1 It is a schematic flow chart of the method for reinforcing a tunnel based on a fiber-reinforced cementitious composite material described in the embodiments of the present invention;

[0028] Figure 2 It is a schematic structural diagram of the system for reinforcing a tunnel based on a fiber-reinforced cementitious composite material described in the embodiments of the present invention.

[0029] In the figure: 701, acquisition unit; 702, judgment unit; 703, prediction unit; 704, processing unit; 705, first analysis unit; 706, second analysis unit; 707, optimization unit; 708, reinforcement unit. Specific implementation mode

[0030] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Components of the embodiments of the present invention usually described and illustrated in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0031] It should be noted that: similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of the present invention, terms such as "first", "second", etc. are only used for differential description and cannot be construed as indicating or implying relative importance.

[0032] Embodiment 1:

[0033] This embodiment provides a method for strengthening a tunnel based on a fiber-reinforced cementitious composite material.

[0034] See Figure 1 , the figure shows that this method includes steps S1, S2, S3, S4, S5, S6, S7 and S8.

[0035] Step S1, acquire the tunnel surface stress data information and the three-dimensional point cloud data of the tunnel;

[0036] It can be understood that in this step, the tunnel surface stress data information is usually obtained by installing stress sensors. These sensors can monitor the stress changes inside and outside the tunnel in real time and provide important dynamic data. These data can reveal potential damage or deformation conditions and ensure timely identification of the structural problems of the tunnel.

[0037] Meanwhile, the three-dimensional point cloud data is obtained through laser scanning technology or photogrammetry technology. Laser scanning can provide high-precision three-dimensional spatial information, accurately capturing the geometric shape and surface features of the tunnel. The acquisition of point cloud data is crucial for subsequent analysis, providing a rich information basis for damage identification and model establishment. By performing subsequent processing on the point cloud data, such as denoising and feature extraction, detailed geometric information and damage characteristics of the tunnel can be obtained.

[0038] Step S2: Send the tunnel surface stress data information to a preset tunnel damage judgment model for tunnel damage judgment to obtain tunnel damage area information;

[0039] It can be understood that in this process, the model performs multi-scale analysis on the stress data to extract stress characteristics at different scales. This method can help identify local stress concentration areas, indicating potential damage. The model requires a large amount of historical data and labeled samples during the training phase to improve its accurate recognition ability for new data. By establishing a mapping relationship between stress and damage, the model can automatically judge the damage state of the current tunnel. In this step, step S2 includes step S21, step S22, and step S23.

[0040] Step S21: Perform two-dimensional wavelet transform processing on the tunnel surface stress data information to obtain a multi-scale decomposition result of tunnel stress, where the multi-scale decomposition result of tunnel stress includes stress characteristic data at at least two scales;

[0041] Furthermore, in this process, a suitable wavelet basis function (such as Daubechies wavelet or Haar wavelet) is first selected to transform the original stress data. Wavelet transform decomposes the signal into a low-frequency part and a high-frequency part, enabling the model to observe stress characteristics at different scales. By selecting decompositions at at least two scales, analysts can obtain multi-level information, such as the relationship between macroscopic stress distribution and local stress concentration phenomena.

[0042] The generated multi-scale decomposition result includes stress characteristic data, which will be used for subsequent damage judgment. Specifically, the high-frequency part can reveal the location of stress anomaly points, while the low-frequency part reflects the overall stress state. In this way, the model can not only detect possible damage areas but also provide a detailed view of stress distribution, helping with subsequent prediction of damage propagation trends and formulation of reinforcement plans. Thus, two-dimensional wavelet transform not only improves the accuracy of data analysis but also enhances the understanding of the health status of the tunnel structure.

[0043] Step S22: Perform local outlier detection processing on the multi-scale decomposition result of tunnel stress, where stress anomaly areas are obtained by identifying stress anomaly points on the tunnel surface;

[0044] It is understandable that in this step, the Local Outlier Factor (LOF) algorithm identifies potential outliers by calculating the local density of each point and comparing its density difference with that of surrounding points. Specifically, points with significantly lower local density than their neighborhoods are considered potential outliers. First, a suitable neighborhood range is set, and then the density and relative density of each stress point are calculated. If the relative density of a stress point is significantly lower than that of surrounding points, it is marked as a stress outlier. These outliers are often associated with structural defects, cracks, or other forms of damage.

[0045] The formula for calculating the local density is as follows:

[0046]

[0047] where ρ(x) represents the local density of point x, N k (x) is the set of k neighbors of point x, K(d(x, y)) is the Gaussian kernel function, and |N k (x)| is the number of points in the neighborhood of point x.

[0048] The formula for calculating the relative density is as follows:

[0049]

[0050] where LOF(x) represents the local outlier factor of point x, N k (x) is the set of k neighbors of point x, |N k (x)| is the number of points in the neighborhood of point x, |N k (x)| is the number of points in the neighborhood of point x, and ρ(y) represents the local density of point y.

[0051] Step S23: Mark the stress anomaly region as an anomaly, and use the marked region as the tunnel damage region.

[0052] It is understandable that this step ensures the clear definition of the damage region, providing a reliable data basis for predicting the subsequent damage expansion trend. Through clear marking, the efficiency and accuracy of the model can be improved, errors that may occur in subsequent analysis can be reduced, and ultimately, it helps to optimize the tunnel reinforcement strategy.

[0053] Step S3: Send the tunnel damage region information to a preset tunnel damage expansion trend prediction model for predicting the damage expansion trend, and obtain the prediction result of the tunnel damage expansion trend;

[0054] It can be understood that this step provides an important basis for subsequent reinforcement solutions. By understanding the possible expansion trends of damage, targeted measures can be taken in advance to optimize the reinforcement design and avoid potential safety hazards. In this step, step S3 includes step S31, step S32, and step S33.

[0055] Step S31: Process the tunnel damage area information based on a preset marking algorithm to obtain damage concentration areas.

[0056] It can be understood that in this step, morphological operations (such as dilation and erosion) are applied to preprocess the damage area to eliminate noise and isolated small areas. In this way, the damage features can be strengthened and made more obvious. Next, using the DBSCAN algorithm, clustering is performed according to the density distribution of the damage area to determine each aggregated damage area. The advantage of this algorithm is that it does not require setting the number of clusters in advance and can flexibly adapt to different damage distribution situations. By marking the damage concentration areas, not only more accurate data input is provided for subsequent trend prediction, but also potential key damage points can be effectively identified, facilitating targeted analysis and the design of reinforcement measures.

[0057] Step S32: Perform random sampling and damage path evolution on the damage concentration areas based on a preset Markov chain Monte Carlo algorithm to obtain a set of damage evolution paths.

[0058] It can be understood that in this step, first, a state space is established for the damage concentration areas, and different damage states are defined. For example, the damage degree of the tunnel can be divided into multiple levels, such as "no damage", "slight damage", "moderate damage", and "severe damage". These states will be used as the input of the Markov chain Monte Carlo algorithm. Next, based on historical data or expert experience, a state transition probability matrix is constructed. This matrix describes the possibility of transitioning from one damage state to another. For example, if the current state is "slight damage", the probability of transitioning to "moderate damage" can be obtained through the analysis of historical damage data. Then, the Markov chain Monte Carlo algorithm is used to start from the initial state and perform multiple samplings through the state transition probability matrix. In each iteration, the next state is determined according to the transition probability of the current state. For example, if the current state is "slight damage", then randomly select to transition to the state of "no damage" or "moderate damage" and make the selection based on the corresponding probability. Repeat this process to gradually generate a damage evolution path.

[0059] Step S33: Predict the damage expansion trend in a preset future time period for the set of damage evolution paths based on a preset dynamic Bayesian network to obtain the prediction result of the tunnel damage expansion trend.

[0060] It can be understood that in this step, first, a dynamic Bayesian network is constructed. This network consists of nodes and edges. The nodes represent different variables (such as damage states, environmental factors, applied stresses, etc.), and the edges represent the dependencies between these variables. The dynamic nature of the network allows for the processing of time series data. By introducing the time dimension, it can capture the evolution of damage over time. In the dynamic Bayesian network, key state variables are defined, which usually include the current damage state, historical damage states, and other factors affecting damage propagation (such as temperature, humidity, etc.). Then, through the historical damage data and the damage evolution path obtained in step S32, parameter learning is carried out to estimate the probabilities of each state transition and conditional probabilities. This process involves using techniques such as the expectation-maximization (EM) algorithm to adjust the probability parameters in the network so that the model can truly reflect the evolution process of damage. Finally, the dynamic Bayesian network is constructed and the parameters are learned. Next, the prediction of the future damage propagation trend can be carried out. Given the current state, the likelihood of the damage state within a preset future time period is calculated through forward inference (i.e., inferring the future state from the known state). The conditional probability distribution of the future state is calculated for each node in the network.

[0061] Step S4: Send the three-dimensional point cloud data of the tunnel into a preset preprocessing model for preprocessing to obtain the damage feature data information of the tunnel. The damage feature data information of the tunnel includes the deformation features and crack features of the tunnel.

[0062] It can be understood that through systematic data preprocessing in this step, the quality and accuracy of the data are improved, making subsequent analysis and damage assessment more reliable. In particular, by combining algorithms such as kernel density estimation and Mahalanobis distance, noise can be effectively removed to ensure the accuracy of feature extraction, providing an important basis for formulating the tunnel reinforcement plan. In this step, step S4 includes step S41, step S42, and step S43.

[0063] Step S41: Smooth the three-dimensional point cloud data based on a preset kernel density estimation algorithm. Specifically, by calculating the position distribution of other points within the neighborhood of each point and combining a preset kernel function, the local density change of the data is adjusted to obtain the filtered three-dimensional point cloud data.

[0064] It can be understood that in this step, first, for each point in the three-dimensional point cloud data, a neighborhood area is defined, usually selected in a spherical or cubic manner. The size of the neighborhood (i.e., the radius or side length) can be preset according to the data distribution to ensure that there are enough points within this range to participate in the calculation. After determining the neighborhood, the local structural information around the point is obtained by calculating the position distribution of all points within the neighborhood. This process is completed by statistically counting the number of points within the neighborhood and their distribution characteristics. Next, a preset Gaussian kernel function is applied to smooth each point. The kernel function is used to assign weights to each point within the neighborhood, and the size of the weight is related to the distance from the point to the center point. Using the kernel function, the contributions of each point within the neighborhood can be aggregated into a new value, thereby adjusting the local density of the point. Finally, the original value of each point is combined with its local density to calculate the filtered three-dimensional point cloud data.

[0065] Step S42: Calculate the anomaly degree of the filtered three-dimensional point cloud data based on the preset Mahalanobis distance algorithm, and delete the abnormal points based on the anomaly degree of each point relative to its neighborhood to obtain the three-dimensional point cloud data after deleting the abnormal points.

[0066] It can be understood that in this step, a reference point (usually a point in the point cloud data) is selected, and the Mahalanobis distance between this point and all other points is calculated. The Mahalanobis distance measures the similarity between points by considering the covariance of the data. Among them, after calculating the Mahalanobis distance of each point, the anomaly degree of each point is defined according to the distance value. Usually, points with a distance exceeding a certain threshold are considered abnormal points. A dynamic threshold can be set. For example, based on the mean and standard deviation, points exceeding twice the standard deviation can be marked as abnormal. The formula for the Mahalanobis distance is shown as follows:

[0067]

[0068] where a and b represent the points to be compared, and R represents the covariance matrix of the point cloud data. D M (a, b) represents the similarity between a and b.

[0069] Step S43: Perform spatial partitioning on the three-dimensional point cloud data after deleting the abnormal points based on the octree algorithm, and extract feature points from the partitioned data based on the preset image feature detection algorithm to obtain the damage feature data information of the tunnel.

[0070] It can be understood that in this step, first, the processed three-dimensional point cloud data is input into the octree algorithm. An octree is a tree-shaped data structure specifically used to represent point data in three-dimensional space. Its basic principle is to recursively divide the three-dimensional space into eight subspaces to form a hierarchical structure. In specific implementation, first, a cube (root node) containing all the point cloud data is defined. Then, according to the set conditions (such as point cloud density or the maximum number of points within a node), the cube is further divided into eight equal sub-cubes (child nodes), and each child node stores the point data it contains. Next, within each octree node, a preset image feature detection algorithm is applied for feature point extraction. Among them, the preset image feature detection algorithm is the SIFT (Scale-Invariant Feature Transform) algorithm, which can be used to identify regions with significant changes in the point cloud data. Taking the Harris corner detection as an example, the algorithm will identify points with high curvature by calculating the autocorrelation matrix within the local neighborhood, and these points are usually related to damage features such as cracks and deformations in the tunnel.

[0071] Step S5: Send the damage feature data information of the tunnel to a preset analysis model for multi-scale fractal dimension analysis to obtain the fractal dimension value of the tunnel damage feature;

[0072] It can be understood that the fractal dimension value not only provides a quantitative basis for subsequent damage assessment and reinforcement plans but also helps engineers understand the development trend and potential risks of the damage. In this step, step S5 includes step S51, step S52, and step S53.

[0073] Step S51: Perform binarization processing on the damage feature data information of the tunnel. Specifically, convert the deformation feature and crack feature of the tunnel into images, and convert the deformation feature image and crack feature image into black, and the background into white to obtain the binarized tunnel damage image;

[0074] It can be understood that in this step, the binarization processing not only effectively highlights the damage features of the tunnel but also simplifies the subsequent analysis steps, making the process of feature extraction and fractal dimension calculation more efficient and accurate. The finally generated binary image provides clear and definite visual information for further damage assessment and analysis, ensuring the accuracy of the subsequent steps.

[0075] Step S52: Based on a preset size, establish boxes and grids on the binarized tunnel damage image, and calculate the number of boxes containing tunnel damage within each grid;

[0076] It can be understood that in this step, first, on the generated binary image, a preset grid size needs to be set. The selection of this grid size should take into account the size and distribution of tunnel damage characteristics to ensure that the details of the damage characteristics can be fully captured. The grid is usually designed in the form of a square or rectangle, dividing the entire image into several small regions.

[0077] Next, the image is divided into blocks according to the set grid size. Each grid represents a region in the image, and the pixel information within this region is concentrated together. Then, for each grid, calculate the number of black pixels it contains, and these black pixels correspond to the damage characteristics. This not only provides the necessary basic data for the calculation of the fractal dimension but also reveals the distribution characteristics of tunnel damage characteristics in different regions.

[0078] Step S53: Process the number of boxes containing tunnel damage in each grid based on a preset fractal dimension calculation formula to obtain the fractal dimension value of the tunnel damage characteristics.

[0079] It can be understood that in this step, the fractal dimension is a mathematical tool used to quantify the self-similarity and irregularity of complex structures. To calculate the fractal dimension of tunnel damage characteristics, the box-counting method needs to be used. Among them, according to the preset fractal dimension calculation formula, determine multiple boxes of different sizes. These box sizes should gradually decrease from larger to capture all levels of the damage characteristics. For each size of the box, count the number of boxes contained in each grid. Usually, as the box size decreases, the number of required boxes will increase. Record these data to form the relationship between the number of boxes and the box size. Then, perform a linear regression analysis on the logarithm of the number of boxes and the logarithm of the box size to form a linear equation. Through the linear regression analysis, obtain the fitted slope. This slope is the fractal dimension value of the tunnel damage characteristics, reflecting the complexity and self-similar characteristics of the damage. The linear equation is as follows:

[0080] log(N) = -D·log(∈) + C

[0081] Where, N represents the number of boxes, ∈ represents the box size, log(N) represents taking the logarithm of the number of boxes, (log(ε)) represents taking the logarithm of the box size, D is the fractal dimension, and C is a constant.

[0082] Step S6: Send the fractal dimension value of the tunnel damage characteristics, the tunnel damage area information, and the prediction result of the tunnel damage expansion trend to the tunnel reinforcement analysis model for analyzing the positions to be reinforced of the tunnel, and obtain the information on the positions to be reinforced of the tunnel;

[0083] It can be understood that this step not only provides an accurate positioning basis for the tunnel reinforcement work, but also optimizes the allocation of reinforcement resources through data analysis, ensuring the effectiveness and pertinence of the reinforcement measures, thereby improving the safety and service life of the tunnel structure. In this step, step S6 includes step S61 and step S62.

[0084] Step S61: Standardize the fractal dimension value of the tunnel damage characteristics, the tunnel damage area information, and the prediction result of the tunnel damage expansion trend, and assign initial weights to the fractal dimension value of the tunnel damage characteristics, the tunnel damage area information, and the prediction result of the tunnel damage expansion trend.

[0085] It can be understood that to ensure the comparison of different types of data under the same dimension, each data item is first standardized. After data standardization, initial weights are then assigned to each input feature. These weights reflect the importance of each feature in the tunnel reinforcement analysis. The weight assignment can be made according to the experience of domain experts.

[0086] Step S62: Generate at least two groups of damage point combinations based on the fractal dimension value of the tunnel damage characteristics, the tunnel damage area information, and the prediction result of the tunnel damage expansion trend, and use all the damage point combinations as particles to input into a preset particle swarm model for area selection. Among them, the particle swarm model performs area selection through a preset objective function to obtain the area position information to be reinforced.

[0087] It can be understood that in this step, first, based on the fractal dimension value of the tunnel damage characteristics, the damage area information, and the prediction result of the damage expansion trend, key damage points are extracted. These points may represent the most severe damage positions or potential vulnerable areas in the tunnel. Subsequently, at least two different groups of damage point combinations are generated using a combination algorithm, and all the generated damage point combinations are used as particles to input into a preset particle swarm optimization (PSO) model. Among them, particle swarm optimization is an optimization algorithm based on swarm intelligence, which searches for the optimal solution by simulating the foraging behavior of bird flocks. Each particle represents a potential reinforcement plan in the search space, and by continuously adjusting its position (damage point combination), the optimal area selection is found. Finally, through the iterative calculation of the particle swarm model, the particles will continuously update their positions to find the best reinforcement position for the damage area. Among them, the objective function in the particle swarm model is as follows:

[0088] F = w1·R d + w2·R f + w3·R t

[0089] Where, R d represents the severity of the damage area, R fThe risk level indicating the damage propagation trend, R t Indicates the estimated cost of each reinforcement point, where w1, w2, and w3 represent the weight coefficients of the corresponding indicators, and F is the effectiveness value for evaluating the reinforced area.

[0090] Step S7: Generate a reinforcement plan based on the tunnel's position information to be reinforced and the preset historical tunnel reinforcement plans, and optimize the generated reinforcement plan based on the simulated annealing algorithm and the parameters of fiber-reinforced cementitious composites to obtain an optimized plan for reinforcing the tunnel with fiber-reinforced cementitious composites;

[0091] It can be understood that this step not only considers the structural characteristics and damage conditions of the tunnel, but also optimizes according to the actual parameters to ensure the maximization of the reinforcement effect and the minimization of the cost, thereby achieving the goal of effectively protecting the tunnel and extending its service life. In this step, step S7 includes step S71, step S72, and step S73.

[0092] Step S71: Match based on the preset historical tunnel reinforcement plans and the tunnel's position information to be reinforced. Among them, by performing an association analysis on the preset historical tunnel reinforcement plans and the tunnel's position information to be reinforced, the historical tunnel reinforcement plan with the highest degree of association is used as the generated reinforcement plan;

[0093] It can be understood that in this step, first, organize the relevant data of the position information to be reinforced and the historical tunnel reinforcement plans. The position information to be reinforced usually includes specific position information, damage characteristics, surrounding environment, etc.; while the historical reinforcement plans include different reinforcement materials, methods, structural characteristics, etc. The key to this step is to extract the features that can effectively represent these two types of information, and through numerical processing, convert the text information into vector representation for subsequent association analysis. Then, quantitatively evaluate the relationship between the position information to be reinforced and the historical reinforcement plans. Finally, after calculating the degree of association between all historical reinforcement plans and the position information to be reinforced, select the historical reinforcement plan with the highest degree of association as the generated reinforcement plan. Among them, the following formula is used to calculate the degree of association:

[0094]

[0095] Where K represents the degree of association, X represents the feature vector of the position information to be reinforced, Y represents the feature vector of the historical reinforcement plan, Represents the mean value of the feature vector of the position information to be reinforced, Represents the mean value of the feature vector of the historical reinforcement plan.

[0096] Step S72: Encode the reinforcement parameters in the generated reinforcement plan to obtain initial encoded information;

[0097] It can be understood that in this step, for discrete parameters, "One-Hot Encoding" is adopted to represent each category with a binary vector; for continuous parameters, the numerical values can be directly used or standardized to scale to a specific range.

[0098] Step S73: Construct an energy function based on the material usage and material effectiveness in the reinforcement plan, and iterate the initial coding information and the energy function according to a preset simulated annealing algorithm. Among them, by adjusting the laying thickness and direction of the fiber-reinforced material, new candidate plans are generated until the preset number of iterations is reached, and the final reinforcement plan is obtained.

[0099] It can be understood that in this step, an energy function is first constructed, and its formula is as follows:

[0100] G = c·Q + d·V

[0101] Among them, G represents the total energy, Q represents the material usage, c is the weight coefficient of the material usage, V represents the material effectiveness (strength and durability), and d represents the weight coefficient of the material effectiveness.

[0102] After constructing the energy function, the initial coding information obtained in the previous step is input into the energy function. By calculating the original energy value, the energy performance of the initial plan is obtained. The preset simulated annealing algorithm is used for iterative optimization. Then, by adjusting the laying thickness and direction of the fiber-reinforced material, new candidate plans are generated. The generated new plan is input into the energy function again to calculate its new energy value. According to the acceptance criterion of simulated annealing, it is decided whether to accept the new plan. Usually, if the new energy value is less than the original energy value, the new plan is accepted; otherwise, the new plan is rejected. Finally, the above steps are repeated until the preset number of iterations or the convergence condition of the energy function is reached.

[0103] Step S8: Reinforce the tunnel with fiber-reinforced cementitious composite materials based on the optimized plan for reinforcing the tunnel with fiber-reinforced cementitious composite materials, and obtain the reinforced tunnel.

[0104] It can be understood that in this step, first, according to the optimized reinforcement plan, the required fiber-reinforced cementitious composite materials are prepared to ensure that the material ratio and characteristics meet the requirements of the plan. Then, the surface of the tunnel is cleaned and pre-treated to ensure that the reinforcement materials can adhere to the tunnel structure sufficiently. This step usually includes removing loose materials, oil stains and moisture on the surface, so as to improve the reinforcement effect.

[0105] Then, according to the regulations in the optimization plan, the fiber-reinforced cementitious composite material is applied layer by layer. This process may involve multiple procedures, such as laying, compaction, and curing. By controlling the laying thickness and direction, ensure the uniformity and effectiveness of the material to achieve the best reinforcement effect. At the same time, it is necessary to monitor the environmental conditions (such as temperature and humidity) during the construction process to avoid adverse effects on the material curing process.

[0106] Finally, the reinforced tunnel needs to undergo acceptance tests, including the detection of material properties and the assessment of structural stability. The goal of this stage is to ensure that the reinforcement measures meet the expected strength and durability requirements, thus providing guarantee for the long-term use of the tunnel.

[0107] Example 2:

[0108] As Figure 2 shown, this embodiment provides a system for reinforcing a tunnel based on a fiber-reinforced cementitious composite material. Refer to Figure 2 The system includes an acquisition unit 701, a judgment unit 702, a prediction unit 703, a processing unit 704, a first analysis unit 705, a second analysis unit 706, an optimization unit 707, and a reinforcement unit 708.

[0109] The acquisition unit 701 is used to acquire the tunnel surface stress data information and the three-dimensional point cloud data of the tunnel;

[0110] The judgment unit 702 is used to send the tunnel surface stress data information to a preset tunnel damage judgment model for tunnel damage judgment to obtain the tunnel damage area information;

[0111] The prediction unit 703 is used to send the tunnel damage area information to a preset tunnel damage propagation trend prediction model for damage propagation trend prediction to obtain the prediction result of the tunnel damage propagation trend;

[0112] The processing unit 704 is used to send the three-dimensional point cloud data of the tunnel into a preset preprocessing model for preprocessing to obtain the tunnel damage feature data information, and the tunnel damage feature data information includes the deformation feature and the crack feature of the tunnel;

[0113] The first analysis unit 705 is used to send the tunnel damage feature data information into a preset analysis model for multi-scale fractal dimension analysis to obtain the fractal dimension value of the tunnel damage feature;

[0114] The second analysis unit 706 is used to send the fractal dimension value of the tunnel damage feature, the tunnel damage area information, and the prediction result of the tunnel damage propagation trend to a tunnel reinforcement analysis model for tunnel to-be-reinforced position analysis to obtain the to-be-reinforced position information of the tunnel;

[0115] An optimization unit 707, configured to generate a reinforcement plan based on the tunnel location information to be reinforced and a preset historical tunnel reinforcement plan, and optimize the generated reinforcement plan based on the simulated annealing algorithm and fiber-reinforced cementitious composite material parameters, so as to obtain an optimized plan for reinforcing the tunnel with fiber-reinforced cementitious composite material;

[0116] A reinforcement unit 708, configured to reinforce the tunnel with fiber-reinforced cementitious composite material based on the optimized plan for reinforcing the tunnel with fiber-reinforced cementitious composite material, so as to obtain a reinforced tunnel.

[0117] It should be noted that regarding the system in the above embodiments, the specific manners in which each module performs operations have been described in detail in the embodiments related to the method, and will not be elaborated herein.

[0118] The above are only the preferred embodiments of the present invention, and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

[0119] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present invention, and all should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A method for strengthening a tunnel based on fiber-reinforced cementitious composite materials, characterized in that, Including: Obtaining tunnel surface stress data information and three-dimensional point cloud data of the tunnel; Sending the tunnel surface stress data information to a preset tunnel damage judgment model for tunnel damage judgment to obtain tunnel damage area information; Sending the tunnel damage area information to a preset tunnel damage expansion trend prediction model for damage expansion trend prediction to obtain a prediction result of the tunnel damage expansion trend; Sending the three-dimensional point cloud data of the tunnel to a preset preprocessing model for preprocessing to obtain tunnel damage feature data information, where the tunnel damage feature data information includes tunnel deformation features and crack features; Sending the tunnel damage feature data information to a preset analysis model for multi-scale fractal dimension analysis to obtain a fractal dimension value of the tunnel damage features; Sending the fractal dimension value of the tunnel damage features, the tunnel damage area information, and the prediction result of the tunnel damage expansion trend to a tunnel reinforcement analysis model for analyzing the positions to be reinforced of the tunnel to obtain the positions to be reinforced information of the tunnel; Generating a reinforcement plan based on the positions to be reinforced information of the tunnel and a preset historical tunnel reinforcement plan, and optimizing the generated reinforcement plan based on the simulated annealing algorithm and fiber-reinforced cement-based composite material parameters to obtain an optimized plan for reinforcing the tunnel with fiber-reinforced cement-based composite materials; Reinforcing the tunnel with fiber-reinforced cement-based composite materials based on the optimized plan for reinforcing the tunnel with fiber-reinforced cement-based composite materials to obtain a reinforced tunnel.

2. The method for strengthening a tunnel based on a fiber-reinforced cementitious composite material according to claim 1, wherein ,Sending the tunnel surface stress data information to a preset tunnel damage judgment model for tunnel damage judgment, including: Performing two-dimensional wavelet transform processing on the tunnel surface stress data information to obtain a multi-scale decomposition result of the tunnel stress, where the multi-scale decomposition result of the tunnel stress includes stress feature data at least two scales; Performing local outlier detection processing on the multi-scale decomposition result of the tunnel stress, where stress abnormal areas are obtained by identifying stress abnormal points on the tunnel surface; Performing abnormal marking on the stress abnormal areas, and taking the abnormally marked areas as tunnel damage areas.

3. The method for strengthening a tunnel based on a fiber-reinforced cementitious composite material according to claim 1, wherein ,Sending the tunnel damage area information to a preset tunnel damage expansion trend prediction model for damage expansion trend prediction, including: Processing the tunnel damage area information based on a preset marking algorithm to obtain damage concentration areas; Performing random sampling and damage path evolution on the damage concentration areas based on a preset Markov chain Monte Carlo algorithm to obtain a set of damage evolution paths; Performing damage expansion trend prediction for a preset future time period on the set of damage evolution paths based on a preset dynamic Bayesian network to obtain a prediction result of the tunnel damage expansion trend.

4. The method for strengthening a tunnel based on a fiber-reinforced cementitious composite material according to claim 1, characterized in that ,Sending the three-dimensional point cloud data of the tunnel to a preset preprocessing model for preprocessing, including: Performing smoothing processing on the three-dimensional point cloud data based on a preset kernel density estimation algorithm, where the local density change of the data is adjusted by calculating the position distribution of other points within the neighborhood of each point and combining a preset kernel function to obtain filtered three-dimensional point cloud data; Calculate the anomaly degree of the filtered three-dimensional point cloud data based on the preset Mahalanobis distance algorithm, and delete the abnormal points based on the anomaly degree of each point relative to its neighborhood to obtain the three-dimensional point cloud data after deleting the abnormal points; Perform spatial block processing on the three-dimensional point cloud data after deleting the abnormal points based on the octree algorithm, and extract feature points from the blocked data based on the preset image feature detection algorithm to obtain the damage feature data information of the tunnel.

5. The method for strengthening a tunnel based on a fiber-reinforced cementitious composite material according to claim 1, wherein ,Send the damage feature data information of the tunnel to a preset analysis model for multi-scale fractal dimension analysis to obtain the fractal dimension value of the tunnel damage feature, including: Perform binarization processing on the damage feature data information of the tunnel. Specifically, convert the deformation feature and crack feature of the tunnel into images, convert the deformation feature image and crack feature image into black, and convert the background into white to obtain the binarized tunnel damage image; Establish boxes and grids on the binarized tunnel damage image based on the preset size, and calculate the number of boxes containing tunnel damage in each grid; Process the number of boxes containing tunnel damage in each grid based on the preset fractal dimension calculation formula to obtain the fractal dimension value of the tunnel damage feature.

6. A system for strengthening tunnels based on fiber-reinforced cementitious composites, characterized in that, Including: An acquisition unit for acquiring the tunnel surface stress data information and the three-dimensional point cloud data of the tunnel; A judgment unit for sending the tunnel surface stress data information to a preset tunnel damage judgment model for tunnel damage judgment to obtain the tunnel damage area information; A prediction unit for sending the tunnel damage area information to a preset tunnel damage expansion trend prediction model for damage expansion trend prediction to obtain the prediction result of the tunnel damage expansion trend; A processing unit for sending the three-dimensional point cloud data of the tunnel to a preset preprocessing model for preprocessing to obtain the damage feature data information of the tunnel, where the damage feature data information of the tunnel includes the deformation feature and crack feature of the tunnel; A first analysis unit for sending the damage feature data information of the tunnel to a preset analysis model for multi-scale fractal dimension analysis to obtain the fractal dimension value of the tunnel damage feature; A second analysis unit for sending the fractal dimension value of the tunnel damage feature, the tunnel damage area information, and the prediction result of the tunnel damage expansion trend to a tunnel reinforcement analysis model for analyzing the position to be reinforced of the tunnel to obtain the information of the position to be reinforced of the tunnel; An optimization unit for generating a reinforcement plan based on the information of the position to be reinforced of the tunnel and the preset historical tunnel reinforcement plan, and optimizing the generated reinforcement plan based on the simulated annealing algorithm and the parameters of the fiber-reinforced cementitious composite material to obtain the optimized plan for reinforcing the tunnel with the fiber-reinforced cementitious composite material; A reinforcement unit for reinforcing the tunnel with the fiber-reinforced cementitious composite material based on the optimized plan for reinforcing the tunnel with the fiber-reinforced cementitious composite material to obtain the reinforced tunnel.

7. The system for strengthening a tunnel based on a fiber-reinforced cementitious composite material according to claim 6, characterized in that, The judgment unit includes: The first judgment subunit is used to perform two-dimensional wavelet transform processing on the tunnel surface stress data information to obtain a multi-scale decomposition result of the tunnel stress, where the multi-scale decomposition result of the tunnel stress includes stress characteristic data at least two scales; The second judgment subunit is used to perform local outlier detection processing on the multi-scale decomposition result of the tunnel stress, where by identifying stress abnormal points on the tunnel surface, a stress abnormal area is obtained; The third judgment subunit is used to perform abnormal marking on the stress abnormal area, and the marked area is used as the tunnel damage area.

8. The system for strengthening a tunnel based on a fiber-reinforced cementitious composite material according to claim 6, characterized in that, The prediction unit includes: The first prediction subunit is used to process the tunnel damage area information based on a preset marking algorithm to obtain a damage concentration area; The second prediction subunit is used to perform random sampling and damage path evolution on the damage concentration area based on a preset Markov chain Monte Carlo algorithm to obtain a set of damage evolution paths; The third prediction subunit is used to predict the damage expansion trend in a preset future time period for the set of damage evolution paths based on a preset dynamic Bayesian network to obtain a prediction result of the tunnel damage expansion trend.

9. The system for reinforcing a tunnel based on a fiber-reinforced cementitious composite material according to claim 6, wherein The processing unit includes: The first processing subunit is used to perform smoothing processing on the three-dimensional point cloud data based on a preset kernel density estimation algorithm, where by calculating the position distribution of other points within the neighborhood of each point and combining a preset kernel function, the local density change of the data is adjusted to obtain filtered three-dimensional point cloud data; The second processing subunit is used to calculate the abnormality degree of the filtered three-dimensional point cloud data based on a preset Mahalanobis distance algorithm, and delete abnormal points based on the abnormality degree of each point relative to its neighborhood to obtain three-dimensional point cloud data after deleting abnormal points; The third processing subunit is used to perform spatial partitioning processing on the three-dimensional point cloud data after deleting abnormal points based on an octree algorithm, and extract feature points from the partitioned data based on a preset image feature detection algorithm to obtain tunnel damage feature data information.

10. The system for strengthening a tunnel based on a fiber-reinforced cementitious composite material according to claim 6, wherein The first analysis unit includes: The first analysis subunit is used to perform binarization processing on the tunnel damage feature data information, where the deformation feature and crack feature of the tunnel are converted into images, and the deformation feature image and crack feature image are converted into black, and the background is converted into white to obtain a binarized tunnel damage image; The second analysis subunit is used to establish boxes and grids on the binarized tunnel damage image based on a preset size, and calculate the number of boxes containing tunnel damage in each grid; The third analysis subunit is used to process the number of boxes containing tunnel damage in each grid based on a preset fractal dimension calculation formula to obtain a fractal dimension value of the tunnel damage feature.

Images