Subway tunnel water disaster risk evaluation method under extreme rainfall based on multi-source data

Abstract

The invention provides a subway tunnel water disaster risk evaluation method under extreme rainfall based on multi-source data, and belongs to the technical field of urban rail transit disaster prevention. The method comprises the following steps: establishing a grading evaluation system on the basis of field rainstorm monitoring data, hydrogeological parameters, field geological exploration, rock-soil physical and mechanical parameters, monitoring data of a tunnel lining structure and material strength parameters of a tunnel site area in combination with a plurality of developed tests, dividing corresponding grades and determining an influence index-; and S4, respectively carrying out weight calculation on the influence index-by adopting an analytic hierarchy process and an entropy weight method, then calculating a dynamic weight of the influence index-, correcting an abnormal dynamic weight, generating a comprehensive score, dividing risk grades, matching graded prevention and control measures, and optimizing and adjusting the prevention and control measures in real time through a parameter risk grade dynamic judgment and construction method matching mechanism. According to the method, minute-level accurate interpretation of a disaster evolution mechanism and dynamic optimization adaptation of a prevention and control scheme are realized.

Description

Technical Field

[0001] This invention relates to the field of urban rail transit disaster prevention technology, specifically to a method for assessing the risk of water damage in subway tunnels under extreme rainfall based on multi-source data. Background Technology

[0002] Under extreme rainfall, subway tunnel projects face the dual pressures of rapidly rising groundwater levels and backflow of surface runoff. The risks of water-related hazards, such as leakage in the lining structure and instability at segment joints, exhibit a highly non-linear growth trend. Current industry-reliant risk assessment systems, which depend on static geological parameter databases and periodic manual inspection data, suffer from severe timeliness bottlenecks: they struggle to quantify and analyze the dynamic coupling effects of geological and hydrological parameters (such as groundwater level and pore water pressure) under heavy rainfall conditions, and cannot accurately capture key disaster evolution factors such as seepage path expansion rates and structural stress redistribution characteristics (especially instantaneous pressure changes caused by hydraulic impact), severely hindering the understanding of disaster transient mechanisms. Traditional methods lack a real-time fusion mechanism for multi-source heterogeneous monitoring data. This makes it difficult to establish an effective transient correlation model between the real-time variation characteristics of formation permeability coefficients and the transient response characteristics of segment displacement / stress, resulting in the fragmentation of key information. At the same time, the extensive risk assessment system based on fixed threshold rules ignores the significant spatiotemporal heterogeneity of hydraulic shocks caused by extreme rainfall. This leads to delayed prevention and control measures in high-risk areas (such as geologically weak sections and low-lying sections), while emergency resources are misallocated in low-risk areas, resulting in a dual dilemma of insufficient timeliness and low cost-effectiveness of engineering projects.

[0003] Therefore, we propose a method for assessing the risk of water damage in subway tunnels under extreme rainfall based on multi-source data, in order to overcome the fusion barrier of multi-dimensional heterogeneous data such as dynamic disturbances of the geological environment, transient response of structural state, and spatial distribution of rainstorm intensity. Summary of the Invention

[0004] The purpose of this invention is to provide a method for assessing the risk of water damage in subway tunnels under extreme rainfall based on multi-source data, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: This invention provides a method for assessing the risk of water damage in subway tunnels under extreme rainfall based on multi-source data, comprising the following steps: S1: Based on on-site rainstorm monitoring data and hydrogeological parameters in the tunnel site area, and combined with multiple experimental tests, the rainstorm intensity abruptness level was classified and the impact index was determined. Classify surface runoff convergence intensity levels and determine impact indices. Classify the risk level of pipeline saturation overflow and determine the impact index. ; S2: Based on field geological exploration and geotechnical physical and mechanical parameters, combined with multiple experimental tests, the levels of formation permeability variation characteristics are classified and the influence index is determined. Classify the stress redistribution levels of the formation and determine the influence index. Classify interface slip risk levels and determine impact indices. ; S3: Based on monitoring data and material strength parameters of the tunnel lining structure, combined with multiple experimental tests, the concrete carbonation rate levels are classified and the influence index is determined. Classify the failure level of joint seals and determine the impact index. Classify the level of line smoothness degradation and determine the impact index. ; S4: The analytic hierarchy process and entropy weight method are used to analyze the influence index respectively. - Weights are calculated using both the analytic hierarchy process (AHP) and the entropy weight method to determine the influence index. - The dynamic weights are then adjusted to correct any abnormal dynamic weights, a comprehensive score is generated, and risk levels are assigned. These are then matched with tiered prevention and control measures. Through a mechanism that dynamically determines the risk level based on parameters and matches the construction method, prevention and control measures are optimized and adjusted in real time.

[0006] Furthermore, step S1 specifically includes: S11: Based on on-site rainstorm monitoring data in the tunnel site area, combined with indoor rainstorm-runoff coupling simulation experiments, and referring to current standards, the rainstorm intensity abruptness level is divided into three levels: low abruptness, medium abruptness, and high abruptness, and then the impact index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.3 indicates high mutagenicity, 0.3 < ≤0.7 indicates moderate mutagenicity, 0.7 < ≤1.0 indicates low mutagenicity; S12: Based on the terrain slope, surface cover type, and catchment area parameters of the tunnel site area, combined with DEM data analysis and runoff simulation experiments, and referring to current specifications, the surface runoff convergence intensity level is divided into three levels: weak convergence, moderate convergence, and strong convergence, and then the influence index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.3 indicates strong convergence, 0.3 < ≤0.7 indicates convergence, 0.7 < ≤1.0 indicates weak convergence; S13: Based on the matching degree between rainfall intensity and drainage capacity of the pipe network, combined with the simulation results of the SWMM hydraulic model, and referring to current specifications, the risk level of saturated overflow of the pipe network is divided into three levels: low risk, medium risk, and high risk. Then, the impact index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.3 indicates high risk, 0.3 < ≤0.7 is considered medium risk, 0.7 < ≤1.0 indicates low risk.

[0007] Furthermore, step S2 specifically includes: S21: Based on the on-site geological borehole data and geotechnical physical and mechanical parameters of the tunnel site, combined with indoor soil column permeability and softening characteristic tests, and referring to current specifications, the formation permeability variation characteristics are divided into three levels: low variation, medium variation, and high variation. Then, the influence index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.3 indicates high variability, 0.3 < ≤0.7 indicates moderate variation, 0.7 < ≤1.0 indicates low variation; S22: Based on core wave velocity testing and in-situ stress measurement data, combined with triaxial stress-strain tests, and referring to current standards, the formation stress redistribution level is divided into three levels: weak adjustment, medium adjustment, and strong adjustment, and then the influence index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.3 indicates a strong adjustment, 0.3 < ≤0.7 is a medium adjustment, 0.7 < ≤1.0 indicates a weak adjustment; S23: Based on ground-penetrating radar data, combined with shear rheology tests and direct shear tests at the contact surface, and referring to current standards, the interface slip risk level is divided into three levels: low risk, medium risk, and high risk. Then, the impact index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.3 indicates high risk, 0.3 < ≤0.7 is considered medium risk, 0.7 < ≤1.0 indicates low risk.

[0008] Furthermore, step S3 specifically includes: S31: Based on the apparent crack development characteristics of tunnel lining structures, combined with the compressive strength test of lining concrete, environmental temperature and humidity monitoring records, and indoor accelerated carbonation test, and referring to the current specifications, the concrete carbonation rate is divided into five levels: negligible, slow, moderate, fast, and extremely fast. Then, the influence index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.3 is extremely fast, 0.3 < ≤0.5 is considered fast, 0.5 < ≤0.7 is considered moderate, 0.7 < ≤0.9 is considered slow, 0.9 < ≤1.0 is negligible; S32: Based on real-time monitoring data of segment joint displacement, combined with compression-rebound test records of sealing materials and joint watertightness pressure test records, and referring to current specifications, the failure levels of segment joint seals are classified into five levels: intact, slightly deteriorated, significantly degraded, critically failed, and functionally lost. Then, the influence index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.3 indicates loss of function, 0.3 < ≤0.5 indicates critical failure, 0.5 < ≤0.7 indicates significant decline, 0.7 < ≤0.9 indicates slight degradation, 0.9 < ≤1.0 indicates intact; S33: Based on dynamic monitoring data from a total station track inspection vehicle and referring to current specifications, the track smoothness deterioration level is divided into five levels: excellent, good, acceptable, poor, and very poor. Then, the influence index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.40 is the range, 0.40 < ≤0.60 is considered poor, 0.60 < ≤0.75 is considered acceptable; 0.75 < ≤0.90 is considered good, 0.90 < ≤1.00 is considered excellent.

[0009] Furthermore, step S4 specifically includes: S41: Using the Analytic Hierarchy Process (AHP) to analyze the influence index ~ Perform weight calculation; S42: Using the entropy weight method to evaluate the influence index ~ Perform weight calculation; S43: Impact index calculated based on step S41 ~ The weights and the influence index calculated in step S42 ~ Weight calculation affects the index ~ The dynamic weights are then adjusted to correct any abnormal dynamic weights, resulting in the influence index. ~ The final weight; S44: Based on the final weights calculated in step S43, calculate the comprehensive risk score for subway flooding, map the comprehensive score to a five-level risk system, and match graded prevention and control measures. Through the dynamic judgment of parameter risk level and the matching mechanism of construction methods, optimize and adjust the prevention and control measures in real time.

[0010] Furthermore, step S41 specifically includes: S411: Using a nine-level scaling method to measure the influence index ~ Assign pairwise importance scales, construct a judgment matrix, and normalize the column vectors of the judgment matrix. S412: Calculate the influence index using the eigenvector method. ~ The weights; S413: Check the consistency of the matrix through a consistency test. If the test passes, proceed to the influence index calculated in step S412. ~ If the weights are reasonable, and the test fails, repeat steps S411 to S413.

[0011] Furthermore, step S42 specifically includes: S421: Impact Elimination Index ~ The difference in the dimensions of the source data is used to generate a dimensionless dataset; S422: Statistical Impact Index ~ The proportion of gravity; S423: Calculate the impact index ~ The weight.

[0012] Furthermore, step S43 specifically includes: S431: Calculating the influence index based on weights calculated using the analytic hierarchy process (AHP) and the entropy weight method. ~ Dynamic weights; S432: Will affect the index ~ The dynamic weights are compared with the standard weights in the historical engineering database, and the influence index is calculated accordingly. ~ The offset is used to correct the dynamic weights of the influence index for offsets that do not meet the requirements, thus obtaining the influence index. ~ The final dynamic weight; S433: Impact index in step S432 ~ The final dynamic weights are normalized to obtain the influence index. ~ The final weight.

[0013] Furthermore, step S44 specifically includes: S441: Based on the calculated final weights, calculate the comprehensive risk score for subway flooding, map the comprehensive score to a five-level risk system, and match it with graded prevention and control measures; S442: Set independent risk level thresholds for individual parameters. When the combination of multiple parameters reaches the preset risk level, trigger the superimposed prevention and control measures and adjust the weight allocation to update the comprehensive score and adjust the prevention and control measures. S443: Based on historical engineering cases and real-time monitoring data, establish and update a database of emergency construction methods with strong parameter correlations, screen out high-frequency associated water hazard parameter combinations, establish a direct correspondence between parameter anomalies and emergency construction methods through manual experience and data statistics, form a basic construction method association list, and then form a construction method matching degree list based on historical engineering cases, and divide the execution priority according to the matching degree.

[0014] Compared with the prior art, the present invention has the following technical effects: In this invention, the method constructs a multi-dimensional dynamic correlation model of geological permeability characteristics, segment structure status, and hydrological environmental factors. This model interprets the spatiotemporal evolution of seepage paths in real time and generates a dynamic risk score. Based on the score results, and combined with a pre-set multi-level trigger threshold rule base and material performance engineering constraints, it intelligently matches the optimal combination of prevention and control measures, including the priority execution sequence of grouting, sealing reinforcement, and structural strengthening strategies. This achieves minute-level accurate interpretation of disaster evolution mechanisms and dynamic optimization and adaptation of prevention and control schemes, overcoming the shortcomings of existing static assessment models and experience-driven approaches. It improves the accuracy and timeliness of engineering response measures under extreme rainfall conditions. Furthermore, by driving iterative optimization of model parameters and the scheme library through multi-dimensional prevention and control effectiveness verification indicators, it significantly improves the accuracy and timeliness of water hazard prevention and control. Attached Figure Description

[0015] Figure 1 This is a flowchart of the evaluation method according to an embodiment of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of the present invention.

[0017] In this article, terms such as "left," "right," "up," "down," "front," and "back" are established based on the positional relationships shown in the attached drawings. Depending on the attached drawings, the corresponding positional relationships may also change. Therefore, they should not be interpreted as an absolute limitation on the scope of protection.

[0018] Please see Figure 1 This embodiment provides a method for assessing the risk of water damage in subway tunnels under extreme rainfall based on multi-source data, including the following steps: S1: Based on on-site rainstorm monitoring data and hydrogeological parameters in the tunnel site area, and combined with multiple experimental tests, the rainstorm intensity abruptness level was classified and the impact index was determined. Classify surface runoff convergence intensity levels and determine impact indices. Classify the risk level of pipeline saturation overflow and determine the impact index. .

[0019] Specifically, step S1 includes: S11: Based on on-site rainstorm monitoring data in the tunnel site area, combined with indoor rainstorm-runoff coupling simulation experiments, and referring to the classification principles of rainfall abrupt change characteristics in current specifications, classify the rainstorm intensity abruptness levels and determine the impact index. The intensity of rainstorms is classified into three levels: low, medium, and high variability, and then the impact index is determined. Impact Index Standardized values ​​of 0 to 1 were used for scoring (0 represents the highest mutation risk, and 1 represents the lowest mutation risk), where 0 ≤ ≤0.3 indicates high mutagenicity, 0.3 < ≤0.7 indicates moderate mutagenicity, 0.7 < ≤1.0 indicates low mutation rate, as shown in Table 1.

[0020] Table 1 Evaluation of Sudden Changes in Rainstorm Intensity

[0021] S12: Based on the terrain slope, surface cover type, and catchment area parameters of the tunnel site area, combined with DEM data analysis and runoff simulation experiments, and referring to the surface runoff calculation principles in the current specifications, the surface runoff convergence intensity level is classified and the influence index is determined. The intensity of surface runoff convergence is classified into three levels: weak convergence, moderate convergence, and strong convergence, and then the impact index is determined. Impact Index Standardized values ​​of 0 to 1 are used for scoring (0 represents the highest confluence risk, and 1 represents the lowest confluence risk), where 0 ≤ ≤0.3 indicates strong convergence, 0.3 < ≤0.7 indicates convergence, 0.7 < ≤1.0 indicates weak aggregation, as shown in Table 2.

[0022] Table 2 Evaluation of Surface Runoff Convergence Intensity Levels

[0023] S13: Based on the matching degree between rainfall intensity and drainage capacity of the pipe network, combined with the simulation results of the SWMM hydraulic model, and referring to the pipe network carrying capacity assessment method in the current specifications, classify the saturation overflow risk level of the pipe network and determine the impact index. The risk level of pipeline saturation overflow is divided into three levels: low risk, medium risk, and high risk, and then the impact index is determined. Impact Index A standardized score of 0 to 1 is used (0 represents the highest overflow risk, and 1 represents the lowest overflow risk), where 0 ≤ ≤0.3 indicates high risk, 0.3 < ≤0.7 is considered medium risk, 0.7 < ≤1.0 indicates low risk, as shown in Table 3.

[0024] Table 3 Risk Level Assessment of Pipeline Saturation Overflow

[0025] S2: Based on field geological exploration and geotechnical physical and mechanical parameters, combined with multiple experimental tests, the levels of formation permeability variation characteristics are classified and the influence index is determined. Classify the stress redistribution levels of the formation and determine the influence index. Classify interface slip risk levels and determine impact indices. .

[0026] Specifically, step S2 includes: S21: Based on the on-site geological borehole data and geotechnical physical and mechanical parameters of the tunnel site, combined with indoor soil column permeability and softening characteristic tests, and referring to the permeability characteristic classification principles in the current specifications, classify the strata permeability variation characteristics and determine the influence index. The formation permeability variation characteristics are classified into three levels: low variation, medium variation, and high variation. Then, the influence index is determined. Impact Index Standardized values ​​of 0 to 1 are used for scoring (0 represents the highest risk of penetration variation, and 1 represents the lowest risk of penetration variation), where 0 ≤ ≤0.3 indicates high variability, 0.3 < ≤0.7 indicates moderate variation, 0.7 < ≤1.0 indicates low variability, as shown in Table 4.

[0027] Table 4 Evaluation of Formation Permeability Variation Characteristics

[0028] S22: Based on core wave velocity testing and in-situ stress measurement data, combined with triaxial stress-strain tests, and referring to the stress adjustment coefficient calculation method in current specifications, classify the formation stress redistribution levels and determine the influence index. The formation stress redistribution levels are divided into three categories: weak adjustment, medium adjustment, and strong adjustment, and then the influence index is determined. Impact Index A standardized value of 0 to 1 is used for scoring (0 represents the strongest risk of stress redistribution, and 1 represents the weakest risk of stress redistribution), where 0 ≤ ≤0.3 indicates a strong adjustment, 0.3 < ≤0.7 is a medium adjustment, 0.7 < ≤1.0 indicates a weak adjustment, as shown in Table 5.

[0029] Table 5 Evaluation of Formation Stress Redistribution Level

[0030] S23: Based on ground-penetrating radar data, combined with shear rheological tests and direct shear tests at the contact surface, and referring to the principle of reducing the shear strength of structural surfaces in current specifications, classify the interface slip risk level and determine the influence index. The interface slippage risk level is divided into three levels: low risk, medium risk, and high risk, and then the impact index is determined. Impact Index A standardized score of 0 to 1 is used (0 represents the highest risk of interface slippage, and 1 represents the lowest risk of interface slippage), where 0 ≤ ≤0.3 indicates high risk, 0.3 < ≤0.7 is considered medium risk, 0.7 < ≤1.0 indicates low risk, as shown in Table 6.

[0031] Table 6. Assessment of Interface Slippage Risk Level

[0032] S3: Based on monitoring data and material strength parameters of the tunnel lining structure, combined with multiple experimental tests, the concrete carbonation rate levels are classified and the influence index is determined. Classify the failure level of joint seals and determine the impact index. Classify the level of line smoothness degradation and determine the impact index. .

[0033] S31: Based on the apparent crack development characteristics of tunnel lining structures, combined with the compressive strength test of lining concrete, environmental temperature and humidity monitoring records, and indoor accelerated carbonation test, and referring to the classification technical rules for the effects of carbonation environment in current specifications and industry standards, the concrete carbonation rate is classified and the influence index is determined. The carbonation rate of concrete was classified into five levels: negligible, slow, moderate, fast, and extremely fast. Then, the influence index was determined. Impact Index A standardized score of 0 to 1 is used (0 indicates extremely rapid carbonization and degradation, 1 indicates no risk of carbonization), where 0 ≤ ≤0.3 is extremely fast, 0.3 < ≤0.5 is considered fast, 0.5 < ≤0.7 is considered moderate, 0.7 < ≤0.9 is considered slow, 0.9 < Values ​​≤1.0 are negligible, as shown in Table 7.

[0034] Table 7 Evaluation of Concrete Carbonation Rate Grades

[0035] S32: Based on real-time monitoring data of segment joint displacement, combined with compression-rebound test records of sealing materials and joint watertightness pressure test records, and referring to the grading principles of sealing system durability in current specifications, industry standards, and national standards, classify the joint seal failure level and determine the impact index. The failure levels of segment joint seals are classified into five levels: intact, slightly deteriorated, significantly degraded, critically failed, and functionally lost. Then, the impact index is determined. Impact Index A standardized score of 0 to 1 is used (0 represents complete loss of sealing function, and 1 represents good sealing performance), where 0 ≤ ≤0.3 indicates loss of function, 0.3 < ≤0.5 indicates critical failure, 0.5 < ≤0.7 indicates significant decline, 0.7 < ≤0.9 indicates slight degradation, 0.9 < ≤1.0 indicates good condition, as shown in Table 8.

[0036] Table 8 Evaluation of Joint Seal Failure Levels

[0037] S33: Based on dynamic monitoring data from a total station track inspection vehicle and referring to the smoothness control standards in current specifications, classify the track smoothness deterioration levels and determine the impact index. The smoothness degradation level of the line is divided into five levels: excellent, good, acceptable, poor, and very poor. Then, the influence index is determined. Impact Index Standardized values ​​of 0 to 1 are used for scoring (0 represents the worst level of line smoothness degradation, and 1 represents the best level of line smoothness degradation), where 0 ≤ ≤0.40 is the range, 0.40 < ≤0.60 is considered poor, 0.60 < ≤0.75 is considered acceptable; 0.75 < ≤0.90 is considered good, 0.90 < ≤1.00 is considered excellent, as shown in Table 9.

[0038] Table 9 Evaluation of Line Smoothness Deterioration Level

[0039] S4: The analytic hierarchy process and entropy weight method are used to analyze the influence index respectively. - Weights are calculated using both the analytic hierarchy process (AHP) and the entropy weight method to determine the influence index. - The dynamic weights are then adjusted to correct any abnormal dynamic weights, a comprehensive score is generated, and risk levels are assigned. These are then matched with tiered prevention and control measures. Through a mechanism that dynamically determines the risk level based on parameters and matches the construction method, prevention and control measures are optimized and adjusted in real time.

[0040] In this embodiment, 10 groups of generated influence index samples are used as examples, as shown in Table 10.

[0041] Table 10 Sample Examples

[0042] Specifically, step S4 includes: S41: Using the Analytic Hierarchy Process (AHP) to analyze the influence index ~ Weight calculation specifically includes: S411: Using a nine-level scaling method to measure the influence index ~ Assign pairwise importance scales, construct a judgment matrix, and then normalize the column vectors of the judgment matrix.

[0043] Taking sample 6 as an example, the influence index of sample 6 ~ The judgment matrix is ​​shown in Table 11.

[0044] Table 11 Example of the judgment matrix for Sample 6

[0045] The column vector normalization of the judgment matrix is ​​calculated using the following formula: , , in, To determine the elements of a normalized matrix; i Representative line; j Representative column; To determine the elements of a matrix; These are the elements of the weight vector; n It is of matrix order.

[0046] S412: Calculate the influence index using the eigenvector method. ~ The weights are used to generate an AHP weight vector that meets the consistency requirements. =[ ;~; ].

[0047] Calculate the impact index of sample 6 ~ The weights are: =[ ; ; ; ; ; ; ; ; ]T=[0.102;0.148;0.055;0.082;0.032;0.123;0.041;0.245;0.172]T S413: Check the consistency of the judgment matrix through a consistency test (CR < 0.1) to eliminate logically conflicting abnormal judgment data and ensure that the judgment matrix that passes the test is used for weight parsing. If the test passes, the influence index calculated in step S412 is performed. ~ If the weights are reasonable, and the test fails, repeat steps S411 to S413.

[0048] The calculations for the maximum eigenvalue, consistency index, and consistency test value are as follows: , , , in, A norm The normalized matrix is ​​the AHP judgment matrix; λ max The largest eigenvalue of the matrix; CI As a consistency indicator; RI For the random consistency index, look up the table value. RI Standard table n When =9, RI =1.45.

[0049] The judgment matrix of sample 6 is tested, and the largest eigenvalue is calculated. λ max =9.62, consistency index CI=0.077, according to the RI standard table when n=9, RI =1.45, calculate CR=0.053, CR<0.1, the judgment matrix passes the consistency test, and the expert judgment logic is self-consistent.

[0050] S42: Using the entropy weight method to evaluate the influence index ~ Weight calculations are performed to ensure the objectivity of weight allocation. Specifically, this includes: Step S421, Eliminate the impact index ~ The source data's dimensional differences are used to generate a dimensionless dataset.

[0051] (Positive indicator) (Negative indicator) in, For standardized samples m The i Dimensionless data affecting the index; For the sample m The i The original data of the impact index; max( x i ), min( x i ) are samples m The i The item affects the maximum and minimum values ​​of the index range.

[0052] Since all influence index data fall within the [0,1] interval, the original data is used directly as the standardized values. Taking sample 6 as an example:

[0053] S422: Statistical Impact Index ~ The proportion of.

[0054] The formulas for calculating the influence of the index weight and entropy value are shown below: , , in, E i For the tenth sample i The entropy value of the index that affects the item; pmi For the sample m The Middle i The weight of each item affecting the index, when pmi When =0, pmi ln pmi =0.

[0055] S423: Calculate the impact index ~ The weight.

[0056] The formulas for calculating the coefficient of variation and the normalized weights affecting the index are shown below: , , in, g i For the tenth sample i The coefficient of difference in the impact index.

[0057] Table 12 Examples of Entropy Values ​​and Weights

[0058] The final sample 6 influence index was obtained. ~ for: =[ ; ; ; ; ; ; ; ; ]T=[0.107;0.095;0.125;0.137;0.111;0.182;0.099;0.078;0.066]T.

[0059] S43: Impact index calculated based on step S41 ~ The weights and the influence index calculated in step S42 ~ Weight calculation affects the index ~ The dynamic weights are then adjusted to correct any abnormal dynamic weights, resulting in the influence index. ~ The final weights. Specifically, they include: S431: Calculating the influence index based on weights calculated using the analytic hierarchy process (AHP) and the entropy weight method. ~ The dynamic weights are calculated using the following formula: , in, For the first i The dynamic weights of the factors affecting the index; These are non-quantitative indicators representing a percentage or coefficient. For the first i The AHP weights of the impact index; For the first i The entropy weight method is used to determine the weight of the index.

[0060] Based on the proportion of quantitative indicators (7 / 9 = 77.8% ≥ 70%), =0.3 =[ ; ; ; ; ; ; ; ; ]T=[0.106;0.111;0.104;0.121;0.087;0.164;0.082;0.128;0.098]T S432: Will affect the index ~ The dynamic weights are compared with the standard weights in the historical engineering database, and the influence index is calculated accordingly. ~ The offset is used to correct the dynamic weights of the influence index for offsets that do not meet the requirements, thus obtaining the influence index. ~ The final dynamic weights can be omitted if there is no historical engineering database or if it is considered unnecessary to adjust them.

[0061] The following description uses sample 6 as an example to illustrate this process.

[0062] The baseline weights for each influence index in the historical engineering database are as follows: =[ ; ; ; ; ; ; ; ; ]T=[0.115;0.121;0.103;0.130;0.092;0.175;0.105;0.092;0.067]T The formula for offset analysis is as follows: , in, For the first i The offset of the factor affecting the index; For the first i The dynamic weights of the factors affecting the index; For the first in the historical engineering database i The weight of each item affects the index.

[0063] Table 13 Example of Weighted Warning

[0064] According to Table 15, when When ≤0.2, the warning level is Level 1, which is within the normal range and no correction is needed. When 0.2 < When the value is ≤0.4, the warning level is Level II, requiring expert correction. The specific process is as follows: When the dynamic weight value of an indicator is lower than the benchmark weight (i.e., dynamic weight < historical database benchmark value), the system automatically identifies it as a key monitoring indicator, triggering a weight adjustment mechanism, and the dynamic weight is strengthened using a correction coefficient of 1.1; when the dynamic weight value of an indicator is higher than the benchmark weight (i.e., dynamic weight > historical database benchmark value), the system automatically identifies it as a high-risk indicator, triggering a weight adjustment mechanism, and the dynamic weight is balanced using a correction coefficient of 0.9. When the value is greater than 0.4, the warning level is level three, and the weight needs to be reconstructed. The specific process is as follows: immediately freeze the current weight value, temporarily assign a value of 0.6 times the current weight, generate an anomaly report and send it to the engineering data center, start the on-site data re-collection process, and reconstruct the weight within 72 hours.

[0065] S433: Impact index in step S432 ~ The final dynamic weights are normalized to obtain the influence index. ~ The final weight.

[0066] The formula for weight normalization is: , in, For the first i The final weight of each item affects the index; For the revised firsti The dynamic weights of the factors affecting the index.

[0067] The final weights are: =[ ; ; ; ; ; ; ; ; ]T=[0.113;0.118;0.111;0.129;0.093;0.175;0.096;0.123;0.042]T S44: Based on the final weights calculated in step S43, calculate the comprehensive risk score for subway flooding, map the comprehensive score to a five-level risk system, and match tiered prevention and control measures. Through a dynamic judgment mechanism of parameter risk level and construction method matching, optimize and adjust prevention and control measures in real time. Specifically, this includes: S441: Based on the calculated final weights, calculate the comprehensive score of subway flood risk, map the comprehensive score to a five-level risk system, and match it with graded prevention and control measures.

[0068] The formula for calculating the overall score is as follows: , Where S represents the comprehensive water damage score; For the first i The value of the impact index. ∈[0,1]; For the first i The final weight of each item affects the index.

[0069] Table 14 Examples of Tiered Prevention and Control Measures

[0070] S442: Set independent risk level thresholds for individual parameters (as shown in the example of the risk level table for key parameters in Table 15). When the combination of multiple parameters reaches the preset risk level, trigger the superimposed prevention and control measures and adjust the weight allocation (as shown in the example in Table 16) to update the comprehensive score of step S441 and adjust the prevention and control measures.

[0071] Table 15 Example of a Risk Level Table for Key Parameters

[0072] Table 16 Examples of Overlay Trigger Rules

[0073] The weight allocation is adjusted using the following formula: , in, This refers to the adjustment coefficient corresponding to the superposition conditions; This is a conditional trigger flag; it takes a value of 1 when the trigger condition is met, and 0 otherwise. The adjusted weights; The weights before adjustment.

[0074] The overall score will be adjusted using the following formula: , in, This is the adjusted overall score; To influence the index.

[0075] S443: Based on historical engineering cases and real-time monitoring data, establish and update a database of emergency construction methods with strong parameter correlations. Screen out frequently correlated water hazard parameter combinations. Through manual experience and data statistics, establish a direct correspondence between parameter anomalies and emergency construction methods, forming a basic construction method correlation list, as shown in Table 17. Then, based on historical engineering cases, form a construction method matching degree list, and divide the execution priority according to the matching degree (high: matching degree > 0.8; medium: 0.4 < matching degree ≤ 0.8; low: matching degree ≤ 0.4), as shown in Table 18.

[0076] Table 17 Example of Construction Method Association List

[0077] The following formula is used to quantify the method matching degree (DM) and generate the execution priority: , in, For the degree of matching of construction methods; , All are weighting coefficients; Normalized value for the historical frequency of construction method combinations; This represents the probability of successful application of the construction method.

[0078] Table 18 Example of Construction Method Matching List

[0079] Specifically, in this invention, the method constructs a multi-dimensional dynamic correlation model of geological permeability characteristics, segment structure status, and hydrological environmental factors. This model interprets the spatiotemporal evolution of seepage paths in real time and generates a dynamic risk score. Based on the score results, combined with a pre-set multi-level trigger threshold rule base and material performance engineering constraints, it intelligently matches the optimal combination of prevention and control measures, including the priority execution sequence of grouting, sealing reinforcement, and structural strengthening strategies. This achieves minute-level accurate interpretation of disaster evolution mechanisms and dynamic optimization and adaptation of prevention and control schemes, overcoming the shortcomings of existing static assessment models and experience-driven approaches, and improving the accuracy and timeliness of engineering response measures under extreme rainfall conditions. Furthermore, by driving iterative optimization of model parameters and the scheme library through multi-dimensional prevention and control effectiveness verification indicators, the accuracy and timeliness of water hazard prevention and control are significantly improved.

[0080] The above embodiments merely illustrate the basic principles and characteristics of the present invention, but are not limited to the above implementation schemes. It should be understood that those skilled in the art can make various changes and modifications to the present invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined freely by the appended claims and their equivalents.

Claims

1. A method for assessing the risk of water hazard in subway tunnels under extreme rainfall based on multi-source data, characterized in that... Includes the following steps: S1: Based on on-site rainstorm monitoring data and hydrogeological parameters in the tunnel site area, and combined with multiple experimental tests, the rainstorm intensity abruptness level was classified and the impact index was determined. Classify surface runoff convergence intensity levels and determine impact indices. Classify the risk level of pipeline saturation overflow and determine the impact index. ; S2: Based on field geological exploration and geotechnical physical and mechanical parameters, combined with multiple experimental tests, the levels of formation permeability variation characteristics are classified and the influence index is determined. Classify the stress redistribution levels of the formation and determine the influence index. Classify interface slip risk levels and determine impact indices. ; S3: Based on monitoring data and material strength parameters of the tunnel lining structure, combined with multiple experimental tests, the concrete carbonation rate levels are classified and the influence index is determined. Classify the failure level of joint seals and determine the impact index. Classify the level of line smoothness degradation and determine the impact index. ; S4: The analytic hierarchy process and entropy weight method are used to analyze the influence index respectively. ~ Weights are calculated using both the analytic hierarchy process (AHP) and the entropy weight method to determine the influence index. ~ The dynamic weights are then adjusted to correct any abnormal dynamic weights, a comprehensive score is generated, and risk levels are assigned. These are then matched with tiered prevention and control measures. Through a mechanism that dynamically determines the risk level based on parameters and matches the construction method, prevention and control measures are optimized and adjusted in real time.

2. The method for assessing the risk of water hazard in subway tunnels under extreme rainfall based on multi-source data according to claim 1, characterized in that, Step S1 specifically includes: S11: Based on on-site rainstorm monitoring data in the tunnel site area, combined with indoor rainstorm-runoff coupling simulation experiments, and referring to current standards, the rainstorm intensity abruptness level is divided into three levels: low abruptness, medium abruptness, and high abruptness, and then the impact index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.3 indicates high mutagenicity, 0.3 < ≤0.7 indicates moderate mutagenicity, 0.7 < ≤1.0 indicates low mutagenicity; S12: Based on the terrain slope, surface cover type, and catchment area parameters of the tunnel site area, combined with DEM data analysis and runoff simulation experiments, and referring to current specifications, the surface runoff convergence intensity level is divided into three levels: weak convergence, moderate convergence, and strong convergence, and then the influence index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.3 indicates strong convergence, 0.3 < ≤0.7 indicates convergence, 0.7 < ≤1.0 indicates weak convergence; S13: Based on the matching degree between rainfall intensity and drainage capacity of the pipe network, combined with the simulation results of the SWMM hydraulic model, and referring to current specifications, the risk level of saturated overflow of the pipe network is divided into three levels: low risk, medium risk, and high risk. Then, the impact index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.3 indicates high risk, 0.3 < ≤0.7 is considered medium risk, 0.7 < ≤1.0 indicates low risk.

3. The method for assessing the risk of water hazard in subway tunnels under extreme rainfall based on multi-source data according to claim 1, characterized in that, Step S2 specifically includes: S21: Based on the on-site geological borehole data and geotechnical physical and mechanical parameters of the tunnel site, combined with indoor soil column permeability and softening characteristic tests, and referring to current specifications, the formation permeability variation characteristics are divided into three levels: low variation, medium variation, and high variation. Then, the influence index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.3 indicates high variability, 0.3 < ≤0.7 indicates moderate variation, 0.7 < ≤1.0 indicates low variation; S22: Based on core wave velocity testing and in-situ stress measurement data, combined with triaxial stress-strain tests, and referring to current standards, the formation stress redistribution level is divided into three levels: weak adjustment, medium adjustment, and strong adjustment, and then the influence index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.3 indicates a strong adjustment, 0.3 < ≤0.7 is a medium adjustment, 0.7 < ≤1.0 indicates a weak adjustment; S23: Based on ground-penetrating radar data, combined with shear rheology tests and direct shear tests at the contact surface, and referring to current standards, the interface slip risk level is divided into three levels: low risk, medium risk, and high risk. Then, the impact index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.3 indicates high risk, 0.3 < ≤0.7 is considered medium risk, 0.7 < ≤1.0 indicates low risk.

4. The method for assessing the risk of water hazard in subway tunnels under extreme rainfall based on multi-source data according to claim 1, characterized in that, Step S3 specifically includes: S31: Based on the apparent crack development characteristics of tunnel lining structures, combined with the compressive strength test of lining concrete, environmental temperature and humidity monitoring records, and indoor accelerated carbonation test, and referring to the current specifications, the concrete carbonation rate is divided into five levels: negligible, slow, moderate, fast, and extremely fast. Then, the influence index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.3 is extremely fast, 0.3 < ≤0.5 is considered fast, 0.5 < ≤0.7 is considered moderate, 0.7 < ≤0.9 is considered slow, 0.9 < ≤1.0 is negligible; S32: Based on real-time monitoring data of segment joint displacement, combined with compression-rebound test records of sealing materials and joint watertightness pressure test records, and referring to current specifications, the failure levels of segment joint seals are classified into five levels: intact, slightly deteriorated, significantly degraded, critically failed, and functionally lost. Then, the influence index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.3 indicates loss of function, 0.3 < ≤0.5 indicates critical failure, 0.5 < ≤0.7 indicates significant decline, 0.7 < ≤0.9 indicates slight degradation, 0.9 < ≤1.0 indicates intact; S33: Based on dynamic monitoring data from a total station track inspection vehicle and referring to current specifications, the track smoothness deterioration level is divided into five levels: excellent, good, acceptable, poor, and very poor. Then, the influence index is determined. Impact Index Scoring is done using a range of 0 to 1, where 0 ≤ 1. ≤0.40 is the range, 0.40 < ≤0.60 is considered poor, 0.60 < ≤0.75 is considered acceptable; 0.75 < ≤0.90 is considered good, 0.90 < ≤1.00 is considered excellent.

5. The method for assessing the risk of water hazard in subway tunnels under extreme rainfall based on multi-source data according to claim 1, characterized in that, Step S4 specifically includes: S41: Using the Analytic Hierarchy Process (AHP) to analyze the influence index ~ Perform weight calculation; S42: Using the entropy weight method to evaluate the influence index ~ Perform weight calculation; S43: Impact index calculated based on step S41 ~ The weights and the influence index calculated in step S42 ~ Weight calculation affects the index ~ The dynamic weights are then adjusted to correct any abnormal dynamic weights, resulting in the influence index. ~ The final weight; S44: Based on the final weights calculated in step S43, calculate the comprehensive risk score for subway flooding, map the comprehensive score to a five-level risk system, and match graded prevention and control measures. Through the dynamic judgment of parameter risk level and the matching mechanism of construction methods, optimize and adjust the prevention and control measures in real time.

6. The method for assessing the risk of water hazard in subway tunnels under extreme rainfall based on multi-source data according to claim 5, characterized in that, Step S41 specifically includes: S411: Using a nine-level scaling method to measure the influence index ~ Assign pairwise importance scales, construct a judgment matrix, and normalize the column vectors of the judgment matrix. S412: Calculate the influence index using the eigenvector method. ~ The weights; S413: Check the consistency of the matrix through a consistency test. If the test passes, proceed to the influence index calculated in step S412. ~ If the weights are reasonable, and the test fails, repeat steps S411 to S413.

7. The method for assessing the risk of water hazard in subway tunnels under extreme rainfall based on multi-source data according to claim 5, characterized in that, Step S42 specifically includes: S421: Impact Elimination Index ~ The difference in the dimensions of the source data is used to generate a dimensionless dataset; S422: Statistical Impact Index ~ The proportion of gravity; S423: Calculate the impact index ~ The weight.

8. The method for assessing the risk of water hazard in subway tunnels under extreme rainfall based on multi-source data according to claim 5, characterized in that, Step S43 specifically includes: S431: Calculating the influence index based on weights calculated using the analytic hierarchy process (AHP) and the entropy weight method. ~ Dynamic weights; S432: Will affect the index ~ The dynamic weights are compared with the standard weights in the historical engineering database, and the influence index is calculated accordingly. ~ The offset is used to correct the dynamic weights of the influence index for offsets that do not meet the requirements, thus obtaining the influence index. ~ The final dynamic weight; S433: Impact index in step S432 ~ The final dynamic weights are normalized to obtain the influence index. ~ The final weight.

9. The method for assessing the risk of water hazard in subway tunnels under extreme rainfall based on multi-source data according to claim 5, characterized in that, Step S44 specifically includes: S441: Based on the calculated final weights, calculate the comprehensive risk score for subway flooding, map the comprehensive score to a five-level risk system, and match it with graded prevention and control measures; S442: Set independent risk level thresholds for individual parameters. When the combination of multiple parameters reaches the preset risk level, trigger the superimposed prevention and control measures and adjust the weight allocation to update the comprehensive score and adjust the prevention and control measures. S443: Based on historical engineering cases and real-time monitoring data, establish and update a database of emergency construction methods with strong parameter correlations, screen out high-frequency associated water hazard parameter combinations, establish a direct correspondence between parameter anomalies and emergency construction methods through manual experience and data statistics, form a basic construction method association list, and then form a construction method matching degree list based on historical engineering cases, and divide the execution priority according to the matching degree.

Images