Method, system, device and medium for intelligent evaluation of health degree of rail transit structure in operation period

Abstract

The application discloses an operating period rail transit structure health degree intelligent evaluation method, system, equipment and medium, relates to the rail transit structure disease evaluation management technical field, and comprises the following steps: constructing a multi-dimensional health evaluation index system; collecting and processing structure detection data; carrying out health degree intelligent evaluation based on the analytic hierarchy process; outputting health degree grades and maintenance suggestions; through constructing a multi-dimensional dynamic weight index system fusing historical data and structure characteristics, and effectively integrating multi-source heterogeneous detection data, the accuracy of evaluation and data utilization efficiency are significantly improved. At the same time, the evaluation process can dynamically compensate the weight, so that the evaluation model has intelligent response and early warning capability; the finally output evaluation result not only contains quantitative health grades, but also can automatically generate visual maintenance suggestions and maintenance measures, converts the evaluation conclusion into specific and operable maintenance decision support, and improves the scientific nature and efficiency of rail transit structure operation and maintenance management.

Description

Technical Field

[0001] This invention relates to the field of structural defect assessment and management technology for rail transit, specifically to intelligent assessment methods, systems, equipment, and media for the health of rail transit structures during operation. Background Technology

[0002] During long-term operation, rail transit structures are susceptible to various defects such as water leakage, cracks, material deterioration, and lining deformation due to the combined effects of train loads, geological environment, and material aging. To ensure operational safety and extend the service life of the structure, a scientific assessment of its health status is necessary. Currently, the assessment methods commonly used in the industry mainly rely on regular manual inspections and expert judgment, using a pre-set fixed weight system (such as assigning fixed scores or weights to common defect types) to score and rate the inspection results.

[0003] Existing assessment weighting systems are typically static and rely on subjective experience, making it difficult to objectively reflect the differentiated impacts of different tunnel types, service life, environmental characteristics, and specific combinations of severe defects on the overall structural health. For example, when complex defects such as severe water leakage coupled with lining deformation occur, a fixed weighting system cannot dynamically amplify the impact of such high-risk combinations in the overall assessment, potentially leading to insufficient sensitivity of the assessment results to potential risks, thereby affecting the timeliness and targeted nature of maintenance decisions. Summary of the Invention

[0004] The purpose of this invention is to provide a method, system, equipment and medium for intelligent assessment of the structural health of rail transit during operation, in order to solve the problem that the assessment weight system in the prior art is usually relatively static and relies on subjective experience settings, making it difficult to objectively reflect the differentiated impact of different characteristics and specific combinations of serious defects on the overall structural health.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an intelligent assessment method for the structural health of rail transit during operation, comprising the following steps: S1. Construct a multi-dimensional health evaluation indicator system; S2. Collect and process structural inspection data; S3. Intelligent health assessment based on the analytic hierarchy process; S4. Output health level and maintenance suggestions; S1 includes the following sub-steps: S11. Determine the set of evaluation dimensions D = {d1, d2, d3, ..., d...} n}; S12: Assign a weight vector W=(w1,w2, ..., w) to each evaluation dimension. n ),satisfy =1; S13: Establish grading standards for each dimension, divide the disease status into five levels, and assign different deduction ranges accordingly.

[0006] Further, step S2 includes the following sub-steps: S21. Use at least two of the following devices for detection: laser profiler, ground-penetrating radar, rebound hammer, and crack width observation instrument; S22. Alignment and integration of multi-source heterogeneous detection data based on spatiotemporal multi-scale feature fusion mechanism; S23. Semantically associate and encode the integrated data with the evaluation dimensions determined in S11 to generate the current state dataset C.

[0007] Furthermore, the spatiotemporal multi-scale feature fusion mechanism in S22 includes, S221: Perform time series alignment and spatial grid mapping on the detection data of the same disease at different time and spatial scales; S222: An adaptive weighted fusion strategy is used to weight and integrate detection data at different scales; S223: Generate a disease feature vector with a unified representation.

[0008] Further, wherein S3 includes the following sub-steps: S31: Based on the weights assigned in S12, construct the judgment matrix A; S32: Calculate the comprehensive score for each dimension using the analytic hierarchy process (AHP); S33: The structural health level is determined based on the comprehensive score and is divided into four levels.

[0009] Furthermore, the specific steps for constructing the judgment matrix A in step S31 are as follows: S311. Determine each element in the judgment matrix A. The judgment matrix A is represented as follows: ; Where a ij d represents the i-th evaluation dimension i Relative to the j-th evaluation dimension d j Importance scale, a ji =1 / a ij And when i=j, a ij =1.

[0010] Furthermore, the weight allocation in step S12 is specifically implemented through the following steps: S121. Establish a historical disease database H and statistically analyze each evaluation dimension d. i During the preset time period Thistory Frequency of occurrence F within i and average severity score S i S i The value ranges from 0 to 10, with higher values ​​indicating more severe conditions. S122. Calculate the initial weight coefficients W for each dimension. i （0） The W i （0） d represents dimension i The initial weights; ; Where α and β are empirical adjustment coefficients, representing the influence weights of disease frequency and severity, respectively. Here, the values ​​are α=0.4 and β=0.6. This represents the sum of the frequencies of all dimensions. This represents the sum of severity scores across all dimensions; S123. Based on the tunnel type T, service life Y, and environmental characteristics E, the weights are adjusted to obtain the final weight W. i ; Among them, the tunnel type correction factor f T For shield tunnels, f T =1.0; for cut-and-cover tunnels, f T =1.15; Service life correction factor f Y When the service life Y ≤ 15 years, f Y =1.0; when 15 years < Y ≤ 30 years, f Y =1.0 + (Y - 15) * 0.006; when Y > 30 years, f Y =1.09; Environmental characteristic correction factor f E Based on groundwater level and soil corrosivity, the environment is divided into three levels: Level 1, good, f E =1.0; Level 2, Medium, f E =1.05; Level 3, Severe, f E =1.12; The final weight is, ; Normalize the corrected weight vector so that ; When step S32 is executed, the weights of the judgment matrix A constructed in S31 are dynamically compensated based on the current state dataset C generated in S23.

[0011] Furthermore, the dynamic compensation includes the following specific steps: When severe water leakage (level 5 damage) is detected in the current state dataset, all row elements a in matrix A that are related to the water leakage dimension d1 will be evaluated. 1j Column element a j1 To strengthen the correction, specifically, for all j≠1, let , ,in This is the risk enhancement factor for water leakage. ; When the current state dataset detects that lining deformation (level 4 and above defects) and cracks (level 3 and above defects) are spatially coupled, the importance scale a of the lining deformation dimension d4 relative to the crack dimension d2 in the judgment matrix A will be used. 42 Improvement, specifically, , ,in The deformation-crack coupling risk coefficient. ; When tunnel floor structure collapse (level III or above defect) occurs and is accompanied by track bed detachment, the weight of the tunnel floor structure collapse dimension d6 in judgment matrix A relative to all other dimensions will be globally increased. Specifically, for all dimensions k except d6, let... , ,in The overall risk coefficient of the subgrade composite risk. ; Step S4 further includes the following sub-steps: S41. Based on the structural health level determined in S33, generate corresponding maintenance priority recommendations; S42. Output a visual assessment report, including a health distribution map, a disease statistics table, and recommended maintenance measures.

[0012] An intelligent assessment system for the structural health of rail transit during the operational phase, including: The data acquisition module is configured to use at least two of the following devices: laser profiler, ground-penetrating radar, rebound hammer, and crack width observation instrument, to collect detection data of the civil engineering structure of rail transit. The data processing and fusion module is communicatively connected to the data acquisition module and is configured to perform spatiotemporal multi-scale feature fusion and alignment processing on the acquired detection data. The indicator system and weight management module is configured to store and maintain a multi-dimensional health evaluation indicator system, which includes at least three evaluation dimensions among water leakage, lining cracks, material deterioration, lining deformation, deformation joints and tunnel bottom structure crushing, and assigns and manages weights for each dimension, wherein the weight of water leakage is not less than 0.25. The intelligent evaluation engine module is connected to the data processing and fusion module and the indicator system and weight management module, respectively. It is configured to be based on the analytic hierarchy process, and according to the indicator system and weights, it fuses the processed detection data to construct a judgment matrix and calculate a comprehensive score to quantitatively evaluate the health of the rail transit structure. The results output module is connected to the intelligent assessment engine module and is configured to automatically generate a health level of one to four based on the assessment results, and output the corresponding visual assessment report, maintenance priority suggestions and maintenance measures recommendations.

[0013] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the above-described method.

[0014] A computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when executed by a processor, implements the steps of the above-described method.

[0015] Compared with existing technologies, this invention significantly improves the accuracy and data utilization efficiency of rail transit structure health assessment by constructing a multi-dimensional dynamic weight index system that integrates historical data and structural features, and by combining spatiotemporal multi-scale feature fusion technology to effectively integrate multi-source heterogeneous detection data. Simultaneously, the assessment process can dynamically compensate weights based on real-time detected severe defects, enabling the assessment model to have responsiveness and early warning capabilities. The final assessment results not only include quantitative health levels but also automatically generate visualized maintenance suggestions and repair measures, thus directly transforming assessment conclusions into specific and actionable maintenance decision support, improving the scientific nature and efficiency of rail transit structure operation and maintenance management. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0017] Figure 1 This is a schematic diagram of the overall process of the evaluation method provided in the embodiments of the present invention. Detailed Implementation

[0018] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0019] As attached Figure 1 As shown: Example 1

[0020] Example 1: Health assessment of operational tunnel structures based on routine inspection data This embodiment provides a method for assessing the structural health of operational tunnels based on conventional testing data. The specific steps are as follows: S1. Construct a multi-dimensional health evaluation index system.

[0021] S11: The evaluation dimension set is determined as follows: water leakage (d1), lining cracks (d2), material deterioration (d3), lining deformation (d4), expansion joints (d5), and tunnel bottom structure crushing (d6).

[0022] S12: Through historical data analysis and expert consultation, assign initial weight vectors to each dimension: W0 = (0.25, 0.15, 0.15, 0.15, 0.15, 0.15).

[0023] S13: Establish grading standards for each dimension, classifying the condition of the lining into five levels, with corresponding deduction ranges. Here, a level one (minor) lining crack will incur a deduction of 5-10 points, while a level five (extremely severe) crack will incur a deduction of >50 points; as shown in the table below: Lining cracks Generally cracked or without development. Crack width δ < 0.3 mm in reinforced concrete lining; crack width δ < 3 mm and length L < 5 m in ordinary concrete lining. The width of cracks in reinforced concrete lining is 0.5 mm ≥ δ ≥ 0.3 mm; the width of cracks in ordinary concrete lining is 5 mm ≥ δ ≥ 3 mm, the length is L = 5 m and the cracks are developing, but at a slow rate. Through cracks appeared in the lining; the crack width δ>0.5 mm in reinforced concrete lining; the crack width δ>5 mm in ordinary concrete lining, the length ≥10mL≥5m and the cracks were dense. Cracks in reinforced concrete linings are δ>0.5mm wide; cracks in ordinary concrete linings are δ>5mm wide and L>10m long, with deformation continuing to develop; the arch cracks into blocks and may detach. Lining deformation It may deform, but it doesn't develop further, and it has no impact on its use. Deformation occurs, and the velocity V < 3 mm / year Deformation or movement speed ≥ 10 mm / year ≥ V ≥ 3 mm / year, and new deformation occurs, expressed as deformation amount. The clearance must not decrease to ensure no intrusion into the clearance. Deformation or movement speed V > 10 mm / year The lining is deforming, moving, and settling rapidly, threatening traffic safety. Expansion joint The expansion joint strip is secure, intact, and undamaged. Mostly intact, with no damage, but some areas are hollow, cracked, or peeling. The expansion joint strip is slightly warped and damaged, with the missing parts showing hollowness, cracks, and peeling. Expansion joint molding strips are warped, detached, damaged, or severely incomplete. Most of the expansion joint molding strips have warped, fallen off, or are severely damaged. Leakage There is water leakage, but it poses no threat to driving safety and does not affect the tunnel's functionality; there is slight corrosion on the concrete surface. Water leakage causes rail corrosion, shortens the maintenance cycle, and if it continues to develop, it will be upgraded to level three; the concrete surface is prone to becoming brittle and rough. Water dripping, seeping, and poor drainage in the tunnel caused localized deterioration of the track bed condition; the concrete surface became uneven within a short period of time. Water is seeping from the tunnel floor, dripping in a line from the arch, and flowing down the sidewalls, jeopardizing normal operations; the cement is dissolving, and the concrete may crack. Water inrush in the tunnel endangers driving safety. The table continues as follows: Material deterioration The concrete has some roughness or honeycomb-like texture, but it is not serious. The concrete is peeling and deteriorating, but the process is slow. Concrete spalling, material deterioration, reduced lining thickness, and a certain degree of reduction in concrete strength. (1) Material deterioration; it will collapse or peel off with slight external force or vibration, which will have a significant impact on traffic. (2) Corrosion depth is 10mm and the area is 0.3㎡. (3) The effective thickness of the lining is about 2 / 3 of the design thickness. (4) Pitting corrosion or complete corrosion of the steel reinforcement surface. The materials are severely degraded, frequently spalling off and endangering driving safety; the lining thickness is only 3 / 5 of the original design thickness, resulting in a significant decrease in concrete strength; due to corrosion, the cross-sectional area of ​​the reinforcing bars is significantly reduced, impairing the structure's function. S2. Collect and process structural inspection data.

[0024] S21: Ground-penetrating radar and crack width monitoring instruments are used to detect the target tunnel section.

[0025] S22: Spatial-temporal alignment of ground-penetrating radar images and crack data acquired at different time points and from different survey lines; fusion of point crack data and area radar anomaly area data through gridded mapping to generate a unified tunnel lining defect distribution feature map.

[0026] S23. Semantically associate and encode the integrated data with the evaluation dimensions determined in S11 to generate the current state dataset C. That is, according to the preset evaluation dimension classification rules, automatically identify the dimension category (water leakage, cracks, material deterioration, etc.) to which each defect feature belongs, assign the corresponding defect level label and deduction value according to the grading standard, and generate a structured, machine-readable current state dataset C as the standardized input for subsequent analytic hierarchy process (AHP) evaluation.

[0027] Specifically, the tunnel lining defect distribution feature map generated by S22 is semantically associated and encoded. For each detection section, the system automatically identifies the evaluation dimension category (water leakage, cracks, material deterioration, etc.) to which each pixel or region in the map belongs, and assigns corresponding defect level labels and deduction values ​​to each dimension according to the preset grading standards, generating a structured, machine-readable current state dataset C.

[0028] S3. Intelligent health assessment based on the analytic hierarchy process.

[0029] S31: Based on the weight of S12, and considering that the tunnel is a shield tunnel with a service life of 20 years (f T =1.0, f Y =1.03, f E =Level 1 takes 1.0), calculate the corrected weight W; S311. Construct the judgment matrix A; based on the weight vector determined in S12, construct the n-dimensional judgment matrix A=[a] using the analytic hierarchy process. ij ], where a ij d represents the i-th evaluation dimension i Relative to the j-th evaluation dimension The importance scale. Matrix A satisfies: when i=j, a ij =1; when i≠j, a ji =1 / a ij . Each element a ij The value is determined according to the Saaty nine-level scale, and is assigned after pairwise comparisons based on the relative importance of each dimension. S32: Construct the initial judgment matrix A based on the corrected weights. init ; S33: Semantically encode the defect feature map generated in S22 to identify multiple secondary water seepage points and local tertiary cracks.

[0030] The evaluation engine calls the preset benchmark matrix A based on the real-time data from S33. base Used to highlight the importance of water leakage, and related to A init Weighted fusion (β1=0.3, β2=0.7) generates the final judgment matrix A. final ; Based on A final Calculate the scores for each dimension and combine them with the deduction criteria of S13 to obtain a comprehensive score for the health of this interval. During the evaluation process, the judgment matrix is ​​dynamically adjusted based on real-time detected current status data. Specifically, this includes the following scenarios: when severe water leakage is detected, the importance scale of the water leakage dimension relative to all other dimensions is increased; when spatial coupling of lining deformation and cracks is detected, the importance scale of the lining deformation dimension relative to the crack dimension is increased; when tunnel floor structure collapse is detected accompanied by track bed voiding, the comprehensive influence weight of the tunnel floor structure collapse dimension relative to all other dimensions is increased. All adjustments are within a preset reasonable range and are achieved through adjustment coefficients. The dynamically compensated and corrected judgment matrix is ​​used for subsequent comprehensive score calculations.

[0031] S4. Output health level and maintenance suggestions.

[0032] S41: Based on the preset threshold, 85 and above is Level 1, 70-84 is Level 2, 55-69 is Level 3, and below 55 is Level 4. The health level is determined to be Level 2.

[0033] S42: The system automatically generates an assessment report, highlighting areas with concentrated water leakage and cracks, and recommending "preventive maintenance, mainly focusing on leak sealing and crack sealing" as a priority measure. Example 2

[0034] This embodiment is basically the same as the previous embodiment, except that it provides a method for depth assessment of high-risk tunnel sections with complex coupled defects. The specific steps are as follows: S1. Construct an enhanced multi-dimensional health evaluation index system.

[0035] S11: Based on the six dimensions of Example 1, add two dimensions: “slab defects” (d7) and “joint misalignment” (d8).

[0036] S12: Based on the intensive monitoring data of this high-risk area over the past three years, the frequency of each dimension was statistically analyzed. Compared with average severity The initial weight W is calculated. i （0） If water leakage is frequent and severe, its W i （0） =0.30; S121: Establish a historical disease database; Statistically analyze the frequency of occurrence and average severity score of each evaluation dimension in this section within a preset time period. The frequency unit is the number of times the disease occurs per unit length of tunnel per unit time period, and the severity score ranges from 0 to 10 points, with higher values ​​indicating more severe disease. S122: Calculate the initial weights; using the formula: ; Calculate the initial weight coefficients for each dimension; S123: Weight Correction. Based on the tunnel type (cut-and-cover or shield tunnel), service life, and environmental characteristics of this section, a tunnel type correction factor, a service life correction factor, and an environmental characteristic correction factor are introduced respectively to calculate the corrected weights. Finally, the corrected weight vector is normalized so that the sum of the weights of each dimension is 1. S13: This section is a cut-and-cover tunnel (f T =1.15), served for 28 years (f Y =1.078), in a corrosive environment (f E =1.12); After correction and normalization, the final weights W include leakage weight W1=0.32 and track bed defect weight W7=0.10; S2. Collect and process multi-source heterogeneous detection data.

[0037] S21: Comprehensive testing is conducted using a 3D laser scanner, ground-penetrating radar, and rebound hammer.

[0038] S22: The millimeter-precision deformation point cloud obtained by laser scanning, the internal void images revealed by ground-penetrating radar, and the concrete strength data measured by the rebound method are fused under the same spatial coordinate system and unified time reference. Through feature extraction algorithms, the spatial coupling areas of "track bed voids," "lining deformation," and "insufficient strength" are identified. S23: Semantically associate and encode the coupled regions of "slab bed void", "lining deformation" and "insufficient strength" identified in S22; map each defect feature to the corresponding evaluation dimension (slab bed defect d7, lining deformation d4, material deterioration d3), assign defect level labels and deduction values ​​to each dimension according to the grading standard, and record the spatial positional relationship of the coupled defects, such as the distance between the geometric centers of two defects, to generate a structured state dataset C.

[0039] S3. Intelligent health assessment based on the analytic hierarchy process with dynamic compensation.

[0040] S31: Construct the basic judgment matrix A based on the final weight W of S1.2. base ; S32: The data processing module identifies the coupled defect characteristics in S22, triggering a dynamic compensation rule. Since "lining deformation (level 4)" and "crack (level 3)" are simultaneously detected coupled at the same location, the system automatically adjusts the importance scale 'a' of deformation relative to crack in the judgment matrix. 42 Increase by 20% (Y2=1.2); When severe water leakage (level 5 damage) is detected in the current state dataset C, the water leakage risk enhancement compensation rule is triggered, and all dimensions in the judgment matrix A related to water leakage are considered. Related row elements and column elements A strengthening correction is performed, specifically by setting the condition for all j≠1. , ,in This is the risk enhancement factor for water leakage. ; When lining deformation (level 4 and above defects) and cracks (level 3 and above defects) are detected to be spatially coupled in the current state dataset C, the deformation-crack coupling reinforcement compensation rule is triggered, and the importance scale a of the lining deformation dimension d4 relative to the crack dimension d2 in the judgment matrix A is applied. 42 Improvement, specifically, , ,in The deformation-crack coupling risk coefficient. ; When tunnel floor structure crushing (level 3 or above) is detected in the current state dataset C, accompanied by track bed detachment, the global compensation rule for composite risk of the track bed is triggered. The comprehensive impact weight of the tunnel floor structure crushing dimension d6 in the judgment matrix A relative to all other dimensions is globally increased. Specifically, for all dimensions k except d6, let... , ,in The overall risk coefficient of the subgrade composite risk. ; When both "tunnel bottom crushing (level 3)" and "track bed delamination" are identified simultaneously, another rule is triggered, which increases the global impact weight of the crushing dimension d6 relative to other dimensions by 25% (Y3=1.25). A comprehensive score is calculated using a judgment matrix that has been dynamically compensated and corrected. S4. Output the health level and maintenance recommendations in emergency situations.

[0041] S41: When the overall score is lower than the level 4 threshold (55 points), the system determines the health level as level 4 (very poor) and triggers an alarm.

[0042] S42: In addition to the health level, the output report clearly points out that the "deformation-crack coupling zone" and the "subsoil composite disease zone" are the key areas for emergency reinforcement, and automatically generates a complete set of emergency treatment and repair suggestions including "emergency support", "grouting reinforcement" and "long-term monitoring plan". Example 3

[0043] This embodiment is basically the same as the previous embodiment, except that it provides an evaluation method suitable for establishing baselines for newly constructed tunnels and tracking long-term performance. The specific steps are as follows: S1. Construct a benchmark-based multi-dimensional health evaluation index system.

[0044] S11: Determine the same six core evaluation dimensions as in Example 1.

[0045] S12: As this is a newly built tunnel with no historical defect data, the initial weights are determined entirely based on design specifications, expert consensus, and similar engineering cases. Forming the benchmark weight W baseline = (0.28, 0.18, 0.12, 0.17, 0.10, 0.15); the weight of leakage water remains prominent; S13: Establish strict grading and deduction standards as the performance baseline for the entire life cycle of the tunnel; S2: Collect and process high-precision baseline detection data; S21: Before the tunnel is put into operation, full-line three-dimensional laser scanning and full-coverage ground-penetrating radar scanning are used to obtain the "texture" data of the initial geometric state and internal compaction state of the tunnel structure. S22: Establish the initial scan data into a high-precision reference 3D model and a reference dielectric constant spectrum, and store them in the database; S23: Semantically associate and encode the baseline 3D model and dielectric constant spectrum established in S22. Semantically segment the tunnel structure according to the evaluation dimensions to form a baseline feature vector library for each dimension. During subsequent periodic inspections, compare and encode the current inspection data with the baseline data one by one to generate the current state dataset C; S3. Trend-based health assessment based on baseline comparison.

[0046] S31: After the annual routine inspections, the current inspection data will be automatically compared with the baseline model in S22; S32: The system quantitatively analyzes the evolution of indicators in various dimensions, including calculating the cumulative deformation, the total length of newly added cracks, the increase in the number of leakage points, etc., and uses these "changes" as evaluation inputs; S33: The evaluation engine uses a fixed benchmark judgment matrix A baseline A baseline The score is generated from the baseline weights of S12 and focuses on scoring the "trend of change". For example, if the annual deformation rate exceeds the limit in a certain interval, the corresponding score will be deducted in the "lining deformation" dimension.

[0047] Calculate the trend health score.

[0048] S41: Not only does it output the current level, but it also outputs the health score curve over time, clearly showing the performance degradation trend.

[0049] S42: Based on the trend prediction model, when the system detects that the growth rate of a certain indicator (such as crack density) is accelerating, it will issue an early warning in the maintenance recommendations and recommend targeted non-destructive testing and preventive repair in the next maintenance cycle, thus realizing the transformation from post-maintenance to predictive maintenance.

[0050] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. An intelligent assessment method for the structural health of rail transit during the operational phase, characterized in that, Includes the following steps: S1. Construct a multi-dimensional health evaluation indicator system; S2. Collect and process structural inspection data; S3. Intelligent health assessment based on the analytic hierarchy process; S4. Output health level and maintenance suggestions; S1 includes the following sub-steps: S11. Determine the set of evaluation dimensions D = {d1, d2, d3, ..., d...} n }; S12: Assign a weight vector W=(w1,w2, ..., w) to each evaluation dimension. n ),satisfy =1; S13: Establish grading standards for each dimension, divide the disease status into five levels, and assign different deduction ranges accordingly.

2. The intelligent assessment method for the structural health of rail transit during the operational period according to claim 1, characterized in that, in, Step S2 includes the following sub-steps: S21. Use at least two of the following devices for detection: laser profiler, ground-penetrating radar, rebound hammer, and crack width observation instrument; S22. Alignment and integration of multi-source heterogeneous detection data based on spatiotemporal multi-scale feature fusion mechanism; S23. Semantically associate and encode the integrated data with the evaluation dimensions determined in S11 to generate the current state dataset C.

3. The intelligent assessment method for the structural health of rail transit during the operational period according to claim 2, characterized in that, in, The spatiotemporal multi-scale feature fusion mechanism in S22 includes, S221: Perform time series alignment and spatial grid mapping on the detection data of the same disease at different time and spatial scales; S222: An adaptive weighted fusion strategy is used to weight and integrate detection data at different scales; S223: Generate a disease feature vector with a unified representation.

4. The intelligent assessment method for the structural health of rail transit during the operational period according to claim 3, characterized in that, in, S3 includes the following sub-steps: S31: Based on the weights assigned in S12, construct the judgment matrix A; S32: Calculate the comprehensive score for each dimension using the analytic hierarchy process (AHP); S33: The structural health level is determined based on the comprehensive score and is divided into four levels.

5. The intelligent assessment method for the structural health of rail transit during the operational period according to claim 4, characterized in that, in, The specific steps for constructing the judgment matrix A in step S31 are as follows: S311. Determine each element a in the judgment matrix A. ij The judgment matrix A is represented as follows: ； Where a ij d represents the i-th evaluation dimension i Relative to the j-th evaluation dimension d j Importance scale, a ji =1 / a ij And when i=j, a ij =1.

6. The intelligent assessment method for the structural health of rail transit during the operational period according to claim 5, characterized in that, in, The weight allocation in step S12 is specifically implemented through the following steps: S121. Establish a historical disease database H and statistically analyze each evaluation dimension d. i During the preset time period T history Frequency of occurrence F i and average severity score S i S i The value ranges from 0 to 10, with higher values ​​indicating more severe conditions. S122. Calculate the initial weight coefficients W for each dimension. i （0） The W i （0） d represents dimension i The initial weights; ; Where α and β are empirical adjustment coefficients, representing the influence weights of disease frequency and severity, respectively. Here, the values ​​are α=0.4 and β=0.

6. This represents the sum of the frequencies of all dimensions. This represents the sum of severity scores across all dimensions; S123. Based on the tunnel type T, service life Y, and environmental characteristics E, the weights are adjusted to obtain the final weight W. i ; Among them, the tunnel type correction factor f T For shield tunnels, f T =1.0; for cut-and-cover tunnels, f T =1.15; Service life correction factor f Y When the service life Y ≤ 15 years, f Y =1.0; when 15 years < Y ≤ 30 years, f Y =1.0 + (Y - 15) * 0.006; when Y > 30 years, f Y =1.09; Environmental characteristic correction factor f E Based on groundwater level and soil corrosivity, the environment is divided into three levels: Level 1, good, f E =1.0; Level 2, Medium, f E =1.05; Level 3, Severe, f E =1.12; The final weight is, ； Normalize the corrected weight vector so that ； When step S32 is executed, the weights of the judgment matrix A constructed in S31 are dynamically compensated based on the current state dataset C generated in S23.

7. The intelligent assessment method for the structural health of rail transit during the operational period according to claim 6, characterized in that, The dynamic compensation includes the following specific steps: When a serious water leakage is detected in the current state dataset, all row elements a in matrix A that are related to the water leakage dimension d1 will be evaluated. 1j Column element a j1 To strengthen the correction, specifically, for all j≠1, let , ,in This is the risk enhancement factor for water leakage. ; When spatial coupling between lining deformation and cracks is detected in the current state dataset, the importance scale a of the lining deformation dimension d4 relative to the crack dimension d2 in matrix A will be determined. 42 Improvement, specifically, , ,in The deformation-crack coupling risk coefficient. ; When tunnel floor structure collapse occurs and is accompanied by track bed detachment, the weight of the tunnel floor structure collapse dimension d6 in judgment matrix A relative to all other dimensions is globally increased. Specifically, for all dimensions k except d6, let... , ,in The overall risk coefficient of the subgrade composite risk. ; Step S4 further includes the following sub-steps: S41. Based on the structural health level determined in S33, generate corresponding maintenance priority recommendations; S42. Output a visual assessment report, including a health distribution map, a disease statistics table, and recommended maintenance measures.

8. An intelligent assessment system for the structural health of rail transit during operation, characterized in that: include, The data acquisition module is configured to use at least two of the following devices: laser profiler, ground-penetrating radar, rebound hammer, and crack width observation instrument, to collect detection data of the civil engineering structure of rail transit. The data processing and fusion module is communicatively connected to the data acquisition module and is configured to perform spatiotemporal multi-scale feature fusion and alignment processing on the acquired detection data. The indicator system and weight management module is configured to store and maintain a multi-dimensional health evaluation indicator system, which includes at least three evaluation dimensions among water leakage, lining cracks, material deterioration, lining deformation, deformation joints and tunnel bottom structure crushing, and assigns and manages weights for each dimension, wherein the weight of water leakage is not less than 0.

25. The intelligent evaluation engine module is connected to the data processing and fusion module and the indicator system and weight management module, respectively. It is configured to be based on the analytic hierarchy process, and according to the indicator system and weights, it fuses the processed detection data to construct a judgment matrix and calculate a comprehensive score to quantitatively evaluate the health of the rail transit structure. The results output module is connected to the intelligent assessment engine module and is configured to automatically generate a health level of one to four based on the assessment results, and output the corresponding visual assessment report, maintenance priority suggestions and maintenance measures recommendations.

9. A computer device, characterized in that, It includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images