Tunnel health evaluation method and system based on machine learning

Abstract

The invention provides a tunnel health evaluation method and system based on machine learning. The method comprises the following steps: constructing a tunnel health evaluation hierarchical structure model; obtaining criterion layer data based on the current tunnel index layer data, wherein the criterion layer data comprises lining crack, water leakage, lining material deterioration, lining back cavity, lining layer peeling and lining deformation data; calculating a lining deformation superposition amount according to the lining crack data and the water leakage data, and correcting the lining deformation data based on the superposition amount to obtain real lining deformation data; inputting real lining deformation data and other current tunnel criterion layer data into the machine learning model to obtain a current tunnel health evaluation result; wherein the machine learning model is constructed based on the criterion layer data and the target layer level, and the historical tunnel criterion layer data and the corresponding health level are utilized to train the machine learning model. According to the method, the superposition effect between the leakage water and the lining crack is considered, so that the accuracy of tunnel health evaluation is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel health assessment, and in particular to a tunnel health assessment method and system based on machine learning. Background Art

[0002] With the rapid development of transportation infrastructure, the number and operation time of tunnels have also increased. At the same time, many operating tunnels have different types of defects such as cracks, water leakage, cavities behind the lining and lining deformation. The development of these defects will not only destroy the integrity of the tunnel structure, but also threaten the safety of vehicles and personnel. Tunnel health assessment can ensure traffic safety and extend the service life of the tunnel. Traditional tunnel health assessment methods rely on manual regular inspections, which are not only time-consuming and labor-intensive, but also difficult to predict the development of tunnel diseases.

[0003] The machine learning model can comprehensively consider various factors that affect tunnel health. By training the model, it can capture the complex nonlinear relationship between disease indicators and health levels, thereby predicting the health status of the tunnel and improving the accuracy and efficiency of tunnel health evaluation.

[0004] The inventors found that the existing technologies mostly consider the relationship between each disease indicator and the health status of the tunnel independently. However, lining cracks and water leakage often do not exist in isolation, but are interrelated and influence each other: lining cracks and water leakage are two common diseases in tunnels. Lining cracks provide channels for water, making water leakage more likely to occur; and water leakage will further aggravate the corrosion and deterioration of the lining, causing the expansion and increase of cracks. This superposition situation will significantly aggravate the damage to the tunnel structure and reduce the service life and safety of the tunnel. Therefore, how to determine whether lining cracks and water leakage can produce a superposition effect, and consider the superposition effect when evaluating tunnel health to improve the accuracy of the evaluation, is an urgent problem to be solved. Summary of the invention

[0005] In order to solve the above problems, the present invention proposes a tunnel health assessment method and system based on machine learning. By giving priority to judging whether a superposition effect occurs and considering the superposition effect between leakage water and lining cracks, the actual health status of the tunnel can be more accurately reflected, thereby improving the accuracy of tunnel health assessment.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] In a first aspect, the present invention provides a tunnel health assessment method based on machine learning, comprising:

[0008] Construct a hierarchical model of tunnel health evaluation, defining the indicator layer, criterion layer and target layer;

[0009] Collect the current tunnel index layer data, and obtain the criterion layer data according to the index layer data; the criterion layer data includes lining cracks, water leakage, lining material degradation, lining back voids, lining peeling and lining deformation data;

[0010] Calculating the lining deformation superposition amount according to the lining crack data and the water leakage data, and correcting the lining deformation data based on the lining deformation superposition amount to obtain the real lining deformation data;

[0011] Input the real lining deformation data and other current tunnel criterion layer data into the trained machine learning model to obtain the current tunnel health assessment results;

[0012] Among them, the machine learning model is constructed based on criterion layer data and target layer levels, and the machine learning model is trained using historical tunnel criterion layer data and corresponding health levels; the target layer levels include four health levels.

[0013] Preferably, the indicator layer data corresponding to the criterion layer includes:

[0014] Lining cracks include crack length, crack width and crack depth data;

[0015] Leakage water includes leakage water status, pH value and frost damage status data;

[0016] Lining material degradation includes lining strength, lining thickness and reinforcement corrosion data;

[0017] The void behind the lining includes the void depth data;

[0018] Lining spalling includes the possibility of falling, spalling depth and spalling diameter data;

[0019] Lining deformation includes deformation amount and deformation velocity data.

[0020] Preferably, the collecting of the current tunnel indicator layer data and obtaining the criterion layer data according to the indicator layer data specifically includes:

[0021] According to the important influence relationship between the characteristic parameters in each level, the corresponding index scale is selected to establish the judgment matrix between the indicator layer and the criterion layer. If the consistency of the judgment matrix is ​​less than the set value, the weight value between the indicator layer and the criterion layer in the judgment matrix is ​​calculated;

[0022] Calculate the criterion layer data based on the indicator layer data and weight values.

[0023] Preferably, before calculating the lining deformation superposition amount according to the lining crack data and the water leakage data, it also includes judging whether the superposition effect of lining cracks and water leakage occurs:

[0024] Acquire images of lining cracks and water leakage, identify contours of lining cracks and water leakage, and establish a detection frame based on the contours;

[0025] The midpoint of the detection frame is obtained. If the distance between the midpoints of the detection frames of adjacent lining cracks and water leakage is less than a preset length, it is determined that the lining cracks and water leakage in the detection frame have a superposition effect.

[0026] Preferably, the calculation of the lining deformation superposition amount according to the lining crack data and the water leakage data specifically includes:

[0027] Input the current lining crack data and water leakage data into the linear regression model to obtain the lining deformation superposition;

[0028] Among them, the training process of the linear regression model is to use the lining crack data and water leakage data at the first moment of the historical period, and the lining crack data and water leakage data at the second moment as input values, and the lining deformation in the corresponding period as output values, and train the linear regression model until the model loss function is minimized.

[0029] Preferably, the lining deformation data is corrected based on the lining deformation superposition to obtain the real lining deformation data, which specifically includes:

[0030] Setting weights for the lining deformation data and the lining deformation superposition amount respectively;

[0031] The lining deformation data and the lining deformation superposition are multiplied by the corresponding weights and then added to obtain the real lining deformation data.

[0032] Preferably, the machine learning model includes: random forest, XGBOOST, decision tree, support vector machine and neural network.

[0033] In a second aspect, the present invention provides a tunnel health assessment system based on machine learning, comprising:

[0034] Hierarchical model building module, used to build a hierarchical model of tunnel health evaluation, defining the indicator layer, criterion layer and target layer;

[0035] The criterion layer data acquisition module is used to collect the current tunnel indicator layer data and obtain the criterion layer data according to the indicator layer data; the criterion layer data includes lining cracks, water leakage, lining material degradation, lining back voids, lining peeling and lining deformation data;

[0036] A deformation data correction module is used to calculate the lining deformation superposition amount according to the lining crack data and the water leakage data, and to correct the lining deformation data based on the lining deformation superposition amount to obtain the real lining deformation data;

[0037] The health assessment module is used to input the real lining deformation data and other current tunnel criterion layer data into the trained machine learning model to obtain the current tunnel health assessment result; wherein, the machine learning model is constructed based on the criterion layer data and the target layer level, and the machine learning model is trained using the historical tunnel criterion layer data and the corresponding health level; the target layer level includes four health levels.

[0038] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in a tunnel health assessment method based on machine learning described in the first aspect.

[0039] In a fourth aspect, the present invention provides a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in the tunnel health assessment method based on machine learning described in the first aspect are implemented.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] This application takes into account that lining cracks and water leakage in tunnels are often not isolated diseases, and there may be interactions between them. Therefore, considering this superposition effect when evaluating the health of the tunnel based on the collected disease characteristic data, and judging whether the superposition effect has occurred before considering the superposition effect, can more accurately reflect the actual health status of the tunnel and avoid the one-sidedness that may be caused by a single factor evaluation.

[0042] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The accompanying drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their description are used to explain the present invention but do not constitute a limitation of the present invention.

[0044] Figure 1 A main flow chart of a tunnel health assessment method based on machine learning provided by an embodiment of the present invention;

[0045] Figure 2 A health evaluation index system provided by an embodiment of the present invention;

[0046] Figure 3 Iteration diagram of 10-fold cross validation of the machine learning model provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0047] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0048] Embodiment 1

[0049] like Figure 1 As shown, this embodiment discloses a tunnel health assessment method based on machine learning, comprising the following steps:

[0050] S1: Construct a hierarchical model for tunnel health evaluation and define the indicator layer, criterion layer, and target layer;

[0051] S2: Collect the index layer data of the current tunnel, and obtain the criterion layer data according to the index layer data; the criterion layer data includes lining cracks, water leakage, lining material degradation, lining back cavities, lining peeling and lining deformation data;

[0052] S3: calculating the lining deformation superposition amount according to the lining crack data and the water leakage data, and correcting the lining deformation data based on the lining deformation superposition amount to obtain the real lining deformation data;

[0053] S4: Input the real lining deformation data and other current tunnel criterion layer data into the trained machine learning model to obtain the current tunnel health evaluation result; wherein, the machine learning model is constructed based on the criterion layer data and the target layer level, and the machine learning model is trained using the historical tunnel criterion layer data and the corresponding health level; the target layer level includes four health levels.

[0054] Specifically, a hierarchical model of tunnel health evaluation is first constructed to define the indicator layer, criterion layer, and target layer.

[0055] When establishing a highway tunnel health status indicator system, we follow the principles of scientificity, completeness, simplicity, independence, hierarchy, and operability to determine the key indicators that affect the health status of the tunnel.

[0056] Establish as Figure 2 The health evaluation index system of highway tunnels in operation period shown in the figure adopts the hierarchical analysis method to divide the health evaluation index system of highway tunnels into three levels.

[0057] The first layer is the target layer, which contains a target object, namely the health status of the highway tunnel, which is characterized by the health level. When establishing the health level of the highway tunnel, according to relevant domestic and international standards, a four-level classification method is adopted to divide the health status of the tunnel structure into four levels: level 1 (no disease or light disease), level 2 (general disease), level 3 (more serious disease), and level 4 (severe disease), with clear boundaries between each level.

[0058] The second layer is the criterion layer, which is composed of factors that affect the health status of highway tunnels, including lining cracks, water leakage, lining material deterioration, lining back voids, lining delamination, peeling, lining deformation, movement, and settlement.

[0059] The third layer is the indicator layer, which is composed of various indicators that affect the factors in the criterion layer. Lining cracks include three indicators: crack length, width and depth; water leakage includes three indicators: leakage water status, pH value and frost damage status; lining material degradation includes three indicators: lining strength ratio, lining thickness ratio and steel corrosion; the cavity behind the lining includes one indicator: cavity depth; lining delamination and peeling include three indicators: possibility of falling, depth and diameter; lining deformation, movement and settlement include two indicators: internal limit ratio and deformation speed.

[0060] For the original data set of tunnel diseases obtained from on-site inspection, the box plot method is used to detect and remove data outliers, the interpolation method is used to fill missing values, and the Z-score method is used to standardize the data to form a standardized evaluation data set to provide a data basis for model training. At the same time, based on the standardized data set, the Karl Pearson correlation coefficient calculation formula is used to perform correlation analysis between different feature indicators. The calculation formula of the Pearson correlation coefficient is:

[0061]

[0062] Where: x and y are the observed values ​​of two variables, is the mean of variables x and y.

[0063] Generally speaking, the absolute value of the correlation coefficient is in the range of 0.0-0.2, 0.2-0.4, 0.4-0.6, 0.6-0.8, and 0.8-1.0. The correlations between different characteristic indicators are "very weak correlation or no correlation", "weak correlation", "moderate correlation", "strong correlation", and "very strong correlation", respectively. The closer the absolute value of the correlation coefficient is to 1, the stronger the correlation is. After calculating the correlation coefficients between different indicators, a thermal matrix of correlation coefficients was constructed to verify the rationality of the selection of indicators for health evaluation of highway tunnels.

[0064] The processed data is stored in the database, and Python is used for data call and further processing to ensure the integrity and availability of the data set.

[0065] As a further implementation method, considering that lining cracks and water seepage are two major diseases that affect tunnel health, their superposition will significantly aggravate lining corrosion, thereby causing a more serious impact on the overall health of the tunnel. Therefore, this embodiment takes into account the superposition effect between lining cracks and water seepage.

[0066] If the lining crack is far away from the leakage area, there is no possibility of interaction. Therefore, before calculating the superposition effect, first confirm whether the superposition effect has occurred. The specific steps are:

[0067] S301: Acquire images of lining cracks and water leakage, identify contours of lining cracks and water leakage, and establish a detection frame based on the contours;

[0068] S302: Obtain the midpoint of the detection frame. If the distance between the midpoints of the detection frames of adjacent lining cracks and water leakage is less than a preset length, it is determined that the lining cracks and water leakage in the detection frame have a superposition effect.

[0069] After determining the occurrence of superposition effect, calculate the lining deformation caused by the superposition effect; the specific steps are:

[0070] S311: inputting the current lining crack data and water leakage data into a linear regression model to obtain the lining deformation superposition amount;

[0071] Among them, the training process of the linear regression model is to use the lining crack data and water leakage data at the first moment of the historical period, and the lining crack data and water leakage data at the second moment as input values, and the lining deformation in the corresponding period as output values, and train the linear regression model until the model loss function is minimized.

[0072] After obtaining the lining deformation superposition, the initial lining deformation data is corrected. The specific steps are as follows:

[0073] S321: setting weights for the lining deformation data and the lining deformation superposition amount respectively;

[0074] S322: The lining deformation data and the lining deformation superposition are multiplied by the corresponding weights and then added together to obtain the real lining deformation data.

[0075] In S4, the current tunnel health assessment result is obtained based on machine learning.

[0076] In Python environment, the random forest and XGBOOST algorithms are implemented using the scikit-learn library.

[0077] First, stratified sampling was performed according to a ratio of 7:3 to divide the data set into a training set and a test set. The training set was used to fit the model and adjust the model parameters, and the test set was used to evaluate the final model.

[0078] The selection of hyperparameters is done by combining manual tuning and grid search. Manual tuning first determines the parameters that have a greater impact on model performance, and then refines these parameters through grid search. A 10-fold cross validation is used to select the optimal hyperparameter combination, such as Figure 3 As shown. The traversal of grid search and the evaluation of cross-validation help to find the model configuration that best represents the data characteristics, thereby improving the prediction accuracy and generalization ability of the model.

[0079] The model is evaluated by 10-fold cross validation to accurately estimate the performance of the model in practical applications. First, the training set is divided into 10 mutually exclusive subsets of similar size and distribution, and then the cycle is repeated 10 times, merging 9 of the subsets into new training sets each time, and the remaining 1 subset is used as the validation set. In each cycle, the model is trained with the training set, and the performance of the model is evaluated with the validation set. Finally, the average of the 10 training and validation results is taken as the actual performance of the model.

[0080] Hyperparameter optimization of machine learning models is performed based on a combination of manual parameter adjustment and grid search. Before performing grid search parameter adjustment, hyperparameters that have a greater impact on model performance should be determined manually as much as possible to reduce the amount of calculation in the grid search process. In the grid search, the performance of each hyperparameter combination is evaluated through 10-fold cross validation, which helps to reduce the risk of overfitting. Through grid search traversal and cross-validation evaluation, the optimal hyperparameter combination is finally selected to ensure maximum model performance.

[0081] After determining the optimal hyperparameter combination, verify the performance of the machine learning model on the test set, and use the four evaluation indicators of accuracy, precision, recall, and F1-Score to make a final evaluation of the machine learning model. The calculation method of the evaluation indicators is as follows:

[0082] 1. Accuracy

[0083] Accuracy refers to the ratio of the number of samples predicted correctly to the total number of samples.

[0084]

[0085] 2. Precision

[0086] Precision refers to the proportion of samples predicted to be positive samples that are actually positive samples.

[0087]

[0088] 3. Recall

[0089] The recall rate refers to the proportion of samples that are correctly predicted to be positive samples among the samples that are actually positive samples.

[0090]

[0091] 4. F1-Score

[0092] The F1 value is the harmonic mean of precision and recall.

[0093]

[0094] In the formula, TP represents the positive sample predicted by the model as positive; FP represents the negative sample predicted by the model as positive; TN represents the negative sample predicted by the model as negative; FN represents the positive sample predicted by the model as negative.

[0095] Based on the above model evaluation results, the machine learning model with the best performance in terms of generalization ability and prediction accuracy is selected. The newly collected tunnel data is input into the model, and the model will use the learned rules and patterns based on the input feature data to predict the health level of each tunnel section. Compared with traditional tunnel health assessment methods, the present invention can achieve higher assessment accuracy and efficiency, timely capture minor changes in tunnel health status, effectively predict and manage the development of tunnel diseases, thereby greatly improving the operational safety and maintenance efficiency of the tunnel.

[0096] The purpose of this specific embodiment is to improve the accuracy and efficiency of tunnel health status assessment and realize early prediction of tunnel disease development. First, a highway tunnel health evaluation index system covering the target layer, criterion layer and indicator layer is established, and multiple aspects such as lining cracks, water seepage, lining material degradation, lining back voids, lining peeling, lining deformation, movement and settlement are considered in detail. The health status of the tunnel structure is classified according to the four-level classification method, and the boundaries between each level are clarified. On this basis, the tunnel disease data obtained through on-site detection is preprocessed to form a standardized evaluation data set. The data set is trained using machine learning algorithms such as random forest and XGBOOST, the optimal model parameters are selected, and it is continuously iterated and optimized to construct a machine learning model for predicting the tunnel health level. Finally, the model performance is verified on the test set and the model is comprehensively evaluated to ensure the high accuracy and reliability of the prediction. The present invention not only improves the scientificity and accuracy of highway tunnel health assessment and disease prediction, but also helps to optimize the allocation of tunnel maintenance resources and improve the level of tunnel operation management.

[0097] Embodiment 2

[0098] This embodiment provides a tunnel health assessment system based on machine learning, including:

[0099] Hierarchical model building module, used to build a hierarchical model of tunnel health evaluation, defining the indicator layer, criterion layer and target layer;

[0100] The criterion layer data acquisition module is used to collect the current tunnel indicator layer data and obtain the criterion layer data according to the indicator layer data; the criterion layer data includes lining cracks, water leakage, lining material degradation, lining back voids, lining peeling and lining deformation data;

[0101] A deformation data correction module is used to calculate the lining deformation superposition amount according to the lining crack data and the water leakage data, and to correct the lining deformation data based on the lining deformation superposition amount to obtain the real lining deformation data;

[0102] The health assessment module is used to input the real lining deformation data and other current tunnel criterion layer data into the trained machine learning model to obtain the current tunnel health assessment result; wherein, the machine learning model is constructed based on the criterion layer data and the target layer level, and the machine learning model is trained using the historical tunnel criterion layer data and the corresponding health level; the target layer level includes four health levels.

[0103] Embodiment 3

[0104] This embodiment provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps in the tunnel health assessment method based on machine learning as described in the above-mentioned embodiment 1 are implemented.

[0105] Embodiment 4

[0106] This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps in the tunnel health assessment method based on machine learning as described in the first embodiment above are implemented.

[0107] The steps or modules involved in the above embodiments 2 to 4 correspond to those in embodiment 1. For the specific implementation, please refer to the relevant description of embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood to include any medium that can store, encode or carry an instruction set for execution by a processor and enable the processor to execute any method in the present invention.

[0108] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A tunnel health assessment method based on machine learning, characterized in that: include: Construct a hierarchical model of tunnel health evaluation, defining the indicator layer, criterion layer and target layer; Collect the current tunnel index layer data, and obtain the criterion layer data based on the index layer data; The standard layer data include lining cracks, water leakage, lining material degradation, lining back voids, lining peeling and lining deformation data; Calculating the lining deformation superposition amount according to the lining crack data and the water leakage data, and correcting the lining deformation data based on the lining deformation superposition amount to obtain the real lining deformation data; Input the real lining deformation data and other current tunnel criterion layer data into the trained machine learning model to obtain the current tunnel health assessment results; Among them, the machine learning model is constructed based on criterion layer data and target layer levels, and the machine learning model is trained using historical tunnel criterion layer data and corresponding health levels; the target layer levels include four health levels.

2. A tunnel health assessment method based on machine learning as claimed in claim 1, characterized in that: The indicator layer data corresponding to the criterion layer includes: Lining cracks include crack length, crack width and crack depth data; Leakage water includes leakage water status, pH value and frost damage status data; Lining material degradation includes lining strength, lining thickness and reinforcement corrosion data; The void behind the lining includes the void depth data; Lining spalling includes the possibility of falling, spalling depth and spalling diameter data; Lining deformation includes deformation amount and deformation velocity data.

3. A tunnel health assessment method based on machine learning as claimed in claim 1, characterized in that: The collecting of the current tunnel index layer data and obtaining the criterion layer data according to the index layer data specifically includes: According to the important influence relationship between the characteristic parameters in each level, the corresponding index scale is selected to establish the judgment matrix between the indicator layer and the criterion layer. If the consistency of the judgment matrix is ​​less than the set value, the weight value between the indicator layer and the criterion layer in the judgment matrix is ​​calculated; Calculate the criterion layer data based on the indicator layer data and weight values.

4. The tunnel health assessment method based on machine learning according to claim 1, characterized in that: Before calculating the lining deformation superposition amount according to the lining crack data and the water leakage data, it also includes judging whether the superposition effect of lining cracks and water leakage occurs: Acquire images of lining cracks and water leakage, identify contours of lining cracks and water leakage, and establish a detection frame based on the contours; The midpoint of the detection frame is obtained. If the distance between the midpoints of the detection frames of adjacent lining cracks and water leakage is less than a preset length, it is determined that the lining cracks and water leakage in the detection frame have a superposition effect.

5. The tunnel health assessment method based on machine learning according to claim 1, characterized in that: The calculation of the lining deformation superposition amount according to the lining crack data and the water leakage data specifically includes: Input the current lining crack data and water leakage data into the linear regression model to obtain the lining deformation superposition; Among them, the training process of the linear regression model is to use the lining crack data and water leakage data at the first moment of the historical period, and the lining crack data and water leakage data at the second moment as input values, and the lining deformation in the corresponding period as output values, and train the linear regression model until the model loss function is minimized.

6. The tunnel health assessment method based on machine learning according to claim 1, characterized in that: The method of correcting the lining deformation data based on the lining deformation superposition to obtain the real lining deformation data specifically includes: Setting weights for the lining deformation data and the lining deformation superposition amount respectively; The lining deformation data and the lining deformation superposition are multiplied by the corresponding weights and then added to obtain the real lining deformation data.

7. The tunnel health assessment method based on machine learning according to claim 1, characterized in that: The machine learning models include: random forest, XGBOOST, decision tree, support vector machine and neural network.

8. A tunnel health assessment system based on machine learning, characterized in that: include: Hierarchical model building module, used to build a hierarchical model of tunnel health evaluation, defining the indicator layer, criterion layer and target layer; The criterion layer data acquisition module is used to collect the current tunnel indicator layer data and obtain the criterion layer data according to the indicator layer data; the criterion layer data includes lining cracks, water leakage, lining material degradation, lining back voids, lining peeling and lining deformation data; A deformation data correction module is used to calculate the lining deformation superposition amount according to the lining crack data and the water leakage data, and to correct the lining deformation data based on the lining deformation superposition amount to obtain the real lining deformation data; The health assessment module is used to input the real lining deformation data and other current tunnel criterion layer data into the trained machine learning model to obtain the current tunnel health assessment result; wherein, the machine learning model is constructed based on the criterion layer data and the target layer level, and the machine learning model is trained using the historical tunnel criterion layer data and the corresponding health level; the target layer level includes four health levels.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by the processor, the steps in the tunnel health assessment method based on machine learning as described in any one of claims 1-7 are implemented.

10. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps in the tunnel health assessment method based on machine learning as described in any one of claims 1-7 are implemented.

Images