Quantitative evaluation method and system for safety level of tmb tunneling in migmatite tunnel

Abstract

This invention provides a quantitative evaluation method and system for the safety level of TBM excavation in mixed rock tunnels, relating to the field of tunnel excavation technology. By combining the analytic hierarchy process (AHP) and entropy weighting method to quantify the weights of the evaluation index system, the support vector machine evaluation model can accurately and in real time evaluate the monitoring data and output the safety level. The monitoring data includes TBM parameters, geological parameters, and disaster risks. This invention solves the problem of lagging risk response in existing TBM safety evaluations and is applicable to TBM excavation in mixed rock tunnels.

Description

Technical Field

[0001] This invention relates to the field of tunnel excavation technology, and in particular to a method and system for quantitatively evaluating the safety level of TBM excavation in mixed rock tunnels. Background Technology

[0002] TBMs (Tunnel Boring Machines) are highly mechanized tunnel excavation machines, offering significant advantages, particularly in deep-buried, long-distance, and large-diameter tunnel projects. They achieve rapid excavation through continuous mechanized operation, while shield structures and synchronous support significantly enhance construction safety. Laser guidance systems ensure tunnel forming accuracy, and blasting vibrations, noise, and ecological damage are reduced, significantly improving the efficiency and quality of underground tunnel construction.

[0003] As the scale of TBM construction continues to expand, the safety risks faced by TBM construction are becoming increasingly prominent. In high-stress mixed rock tunnels, TBM excavation mainly faces prominent problems such as complex geological conditions, high ground stress hazards, and low construction efficiency. Mixed rock in deeply buried tunnels (such as interbedded granite and schist, and heterogeneous soft and hard rocks) causes abrupt changes in cutterhead stress, abnormal cutter wear, and frequent jamming. Under high ground stress conditions, it is prone to rock bursts, collapses, and TBM jamming. Downtime due to geological risks accounts for as much as 30% to 50% of the total downtime, seriously restricting the construction period.

[0004] Existing TBM safety assessment methods mostly rely on empirical formulas or single indicators (such as Q system, RMR classification), lacking a dynamic quantitative assessment system that integrates geological, machine, and construction parameters, resulting in a lag in risk response. Summary of the Invention

[0005] The technical problem solved by this invention: This invention provides a quantitative evaluation method and system for the safety level of TBM tunneling in mixed rock tunnels, solving the problem of lagging risk response in existing TBM safety evaluations.

[0006] The technical solution adopted by this invention to solve the above-mentioned technical problems is: a quantitative evaluation method for the safety level of TBM tunneling in mixed rock tunnels, comprising the following steps:

[0007] S1. Construct an evaluation index system that includes TBM parameters, geological parameters, and disaster risks;

[0008] S2. Determine subjective weights by constructing an expert judgment matrix using the analytic hierarchy process, obtain objective weights by analyzing the dispersion of monitoring data using the entropy weight method, and determine the combined weights of indicators based on subjective and objective weights.

[0009] S3. Real-time acquisition of geological parameters, TBM operating parameters, and environmental data, followed by standardization processing to construct feature vectors;

[0010] S4. Use combined weights to weight the feature vectors to obtain the input feature vectors, and use the support vector machine to evaluate the model and output the security level.

[0011] Furthermore, in S1, the TBM parameters include cutterhead thrust, cutterhead torque, cutterhead rotation speed, cutterhead tunneling speed, and cutterhead opening ratio; the geological parameters include uniaxial compressive strength, integrity coefficient, wear resistance index, ground stress level, and tunnel depth; the disaster risks include fault fracture zones, soft rock jamming deformation, composite strata, sudden water inrush, and surface subsidence.

[0012] Furthermore, in S2, the subjective weights are determined by constructing an expert judgment matrix using the analytic hierarchy process (AHP), including the following steps:

[0013] S21. Based on the evaluation index system, create a hierarchical structure;

[0014] S22. Construct the judgment matrix for each layer and verify the accuracy of the judgment matrix through consistency.

[0015] S23. Use the eigenvector method to solve the judgment matrix and obtain the weight of the index of this layer relative to the previous layer.

[0016] Furthermore, in S2, the entropy weight method is used to analyze the dispersion of monitoring data and obtain objective weights, including the following steps:

[0017] S201. Construct the initial matrix: ,in, , , Indicates the number of samples. Indicates the number of indicators. Denotes the initial matrix. Indicates the first The first sample The value of each indicator;

[0018] S202. Standardize the sample indicator values;

[0019] S203. Based on the standardized sample index values, calculate the entropy weight of the index. The calculation formula is as follows: ,in This represents the entropy weight of the j-th index. , , , The first standardization The j-th index value of a sample.

[0020] Furthermore, in S2, the formula used to determine the combined weight of the indicators based on subjective and objective weights is as follows: ,in, Indicates the portfolio weight. This represents the weight of the j-th index obtained by the analytic hierarchy process relative to the previous level. This represents the entropy weight of the j-th index.

[0021] Furthermore, in S4, when a rockburst precursor is detected, the combined weight of the ground stress level is increased by 20% to 30%, and when a large deformation of soft rock is detected, the combined weight of the soft rock jamming deformation is increased by 20% to 30%.

[0022] Furthermore, the method also includes: issuing corresponding early warnings based on the security level.

[0023] Furthermore, the training process of the support vector machine evaluation model includes the following steps:

[0024] S41. Use the input feature vector and the corresponding evaluation results as sample data;

[0025] S42. Select the Gaussian kernel function and set the penalty factor to 1.5 and the kernel parameter to 0.8;

[0026] S43. Use sample data to train the model and obtain the support vector machine evaluation model.

[0027] Furthermore, the regression function for the support vector machine evaluation model is: ,in, Indicates the first The first Lagrange multiplier of each sample data point Indicates the first The second Lagrange multiplier of each sample data, Indicates the first The input feature vector in each sample data, This represents the total number of sample data. Indicates a bias value. This represents the independent variable of the regression function. This represents the dependent variable in the regression function. and The solution formula is as follows:

[0028] ,

[0029] ,

[0030] Find a group and , so that the objective function The largest, of which, This represents the total number of sample data. Indicates the first Evaluation results of individual sample data Indicates the first The input feature vector in each sample data, Indicates insensitivity to parameters. Indicates the first The first Lagrange multiplier of each sample data point Indicates the first The second Lagrange multiplier of each sample data, Indicates the first The input feature vector of each sample data, This represents the penalty factor.

[0031] The present invention also provides a quantitative evaluation system for the safety level of TBM tunneling in mixed rock tunnels, which implements the method described above. The system includes a data acquisition module, a data preprocessing module, a combined weighting module, and an analysis and processing module.

[0032] The data acquisition module is used to collect geological parameters, TBM operating parameters, and environmental data in real time.

[0033] The data preprocessing module is used to standardize the real-time collected geological parameters, TBM operating parameters, and environmental data, construct feature vectors, and call combined weights to weight the feature vectors to obtain input feature vectors.

[0034] The combined weighting module is used to construct an expert judgment matrix using the analytic hierarchy process to determine subjective weights, analyze the dispersion of monitoring data using the entropy weight method to obtain objective weights, and determine the combined weights of indicators based on subjective and objective weights.

[0035] Analysis and processing module: It evaluates the model using support vector machines and outputs the security level based on the input feature vectors.

[0036] The beneficial effects of this invention are as follows: This invention provides a quantitative evaluation method and system for the safety level of TBM tunneling in mixed rock tunnels. By combining the analytic hierarchy process and the entropy weight method to quantify the weights of the evaluation index system, the support vector machine evaluation model can accurately and in real time evaluate the monitoring data and output the safety level, thus solving the problem of lagging risk response in existing TBM safety evaluations. Attached Figure Description

[0037] Figure 1 This is a flowchart illustrating a quantitative evaluation method for the safety level of TBM tunneling in mixed rock tunnels provided by the present invention.

[0038] Figure 2 This is a diagram of the evaluation index system in the quantitative evaluation method for the safety level of TBM tunneling in mixed rock tunnels provided by the present invention. Detailed Implementation

[0039] This invention addresses the problem of delayed risk response in existing TBM safety assessments by providing a quantitative evaluation method for the safety level of TBM excavation in mixed rock tunnels, such as... Figure 1 As shown, it includes the following steps:

[0040] S1. Construct an evaluation index system that includes TBM parameters, geological parameters, and disaster risks.

[0041] Specifically, the TBM parameters include cutterhead thrust, cutterhead torque, cutterhead rotation speed, cutterhead tunneling speed, and cutterhead opening ratio; the geological parameters include uniaxial compressive strength, integrity coefficient, abrasion resistance index, ground stress level, and tunnel depth; and the disaster risks include fault fracture zones, soft rock jamming deformation, composite strata, sudden water inrush, and surface subsidence. The evaluation index system is as follows: Figure 2 As shown.

[0042] S2. Subjective weights are determined by constructing an expert judgment matrix using the analytic hierarchy process (AHP), and objective weights are obtained by analyzing the dispersion of monitoring data using the entropy weight method. The combined weights of the indicators are then determined based on the subjective and objective weights.

[0043] Specifically, the subjective weights are determined by constructing an expert judgment matrix using the analytic hierarchy process (AHP), which includes the following steps:

[0044] S21. Based on the evaluation index system, a hierarchical structure is created; the hierarchy includes a target layer, a criterion layer, and an index layer. The target layer includes the TBM tunneling safety level; the criterion layer includes TBM parameters, geological parameters, and disaster risks; and the index layer includes 15 indicators, corresponding to TBM parameters, geological parameters, and disaster risks, such as cutterhead thrust, cutterhead torque, cutterhead rotation speed, cutterhead tunneling speed, cutterhead opening ratio, uniaxial compressive strength, integrity coefficient, wear resistance index, ground stress level, tunnel depth, fault fracture zone, soft rock jamming deformation, composite strata, sudden water inrush, and surface subsidence, which correspond to TBM parameters, geological parameters, and disaster risks.

[0045] S22. Construct the judgment matrix for each layer and verify the accuracy of the judgment matrix through consistency.

[0046] Specifically, if the consistency index ratio is less than 0.1, the judgment matrix is ​​considered reasonable; if the consistency index ratio is not less than 0.1, the judgment matrix is ​​re-assigned until the consistency verification is passed. The formula for calculating the consistency index ratio is: , ,in, This indicates the consistency index ratio. Indicators of consistency This represents the random consistency index. This represents the largest eigenvalue of the judgment matrix. This indicates the order of the judgment matrix.

[0047] S23. Use the eigenvector method to solve the judgment matrix to obtain the weight of this layer relative to the previous layer.

[0048] For example, the judgment matrix of the criterion layer relative to the target layer that passes the consistency verification is: The largest eigenvalue Corresponding feature vector: [0.105, 0.258, 0.637]; CR=0.033<0.1 (verified), final weight results: TBM parameter: 0.105; geological parameter: 0.258; disaster risk: 0.637.

[0049] The entropy weight method is used to analyze the dispersion of monitoring data and obtain objective weights, including the following steps:

[0050] S201. Construct the initial matrix: ,in, , , Indicates the number of samples. Indicates the number of indicators. Denotes the initial matrix. Indicates the first The first sample The value of each indicator;

[0051] S202. Standardize the sample indicator values. Specifically, data standardization includes eliminating the influence of order of magnitude and normalizing the indicator values ​​to a positive direction.

[0052] S203. Based on the standardized sample index values, calculate the entropy weight of the index. The calculation formula is as follows: ,in This represents the entropy weight of the j-th index. , , , The first standardization The j-th index value of a sample.

[0053] The formula used to determine the combined weight of the indicators based on subjective and objective weights is as follows: ,in, Indicates the portfolio weight. This represents the weight of the j-th index obtained by the analytic hierarchy process relative to the previous level. This represents the entropy weight of the j-th index.

[0054] S3. Real-time acquisition of geological parameters, TBM operating parameters, and environmental data, followed by standardization processing to construct feature vectors.

[0055] Specifically, the standardization process includes using wavelet transform to eliminate noise interference.

[0056] S4. Use combined weights to weight the feature vectors to obtain the input feature vectors, and use the support vector machine to evaluate the model and output the security level.

[0057] Specifically, when a rockburst precursor is detected, i.e. a stress change exceeding 30%, the combined weight of the ground stress level is increased by 20% to 30%. When large deformation of soft rock is detected, i.e. a convergence rate greater than 2 mm / h, the combined weight of soft rock jamming deformation is increased by 20% to 30%. This is to achieve the purpose of dynamically adjusting the combined weight and ensure the accuracy of the assessment under special circumstances.

[0058] The training process of a support vector machine evaluation model includes the following steps:

[0059] S41. Use the input feature vector and the corresponding evaluation results as sample data;

[0060] S42. Select the Gaussian kernel function and set the penalty factor to 1.5 and the kernel parameter to 0.8;

[0061] S43. Use sample data to train the model and obtain the support vector machine evaluation model.

[0062] The regression function for the support vector machine evaluation model is: ,in, Indicates the first The first Lagrange multiplier of each sample data point Indicates the first The second Lagrange multiplier of each sample data, Indicates the first The input feature vector in each sample data, This represents the total number of sample data. Indicates a bias value. This represents the independent variable of the regression function. This represents the dependent variable in the regression function. and The solution formula is as follows:

[0063] ,

[0064] ,

[0065] Find a group and , so that the objective function The largest, of which, This represents the total number of sample data. Indicates the first Evaluation results of individual sample data Indicates the first The input feature vector in each sample data, Indicates insensitivity to parameters. Indicates the first The first Lagrange multiplier of each sample data point Indicates the first The second Lagrange multiplier of each sample data, Indicates the first The input feature vector of each sample data, This represents the penalty factor.

[0066] This invention also provides a quantitative evaluation system for the safety level of TBM tunneling in mixed rock tunnels, realizing the quantitative evaluation method for the safety level of TBM tunneling in mixed rock tunnels as described above. The system includes a data acquisition module, a data preprocessing module, a combined weighting module, and an analysis and processing module. The data acquisition module is used to collect geological parameters, TBM operating parameters, and environmental data in real time. The data preprocessing module is used to standardize the real-time collected geological parameters, TBM operating parameters, and environmental data, construct feature vectors, and use combined weights to weight the feature vectors to obtain input feature vectors. The combined weighting module is used to construct an expert judgment matrix using the analytic hierarchy process to determine subjective weights, analyze the dispersion of monitoring data using the entropy weighting method to obtain objective weights, and determine the combined weights of the indicators based on the subjective and objective weights. The analysis and processing module outputs the safety level based on the input feature vectors using a support vector machine evaluation model.

[0067] Specifically, the security levels can be divided into five categories, as shown in Table 1.

[0068] Table 1 Safety Level Table

[0069]

[0070] Safety levels can be dynamically marked with five color codes to indicate risk areas, from green (Level I) to red (Level V). When a Level III risk is detected, a multimodal alarm is triggered, such as a 120dB directional sound wave and an LED strobe array. Simultaneously, handling suggestions are pushed to the responsible person's mobile device, such as immediately reducing the cutterhead vibration speed to 2.5 rpm if it exceeds the threshold by 15%. When a Level IV risk is detected, a support plan is automatically generated from the case library, along with construction drawings with three-dimensional coordinates. Support plans may include adding W-shaped steel strips at the arch foot, with spacing increased to 0.5m. When a Level V risk is detected, a shutdown command is directly sent to the TBM main control system with a response time of less than 200ms, and the emergency response plan is activated. All handling records are automatically archived, forming a closed-loop digital twin archive.

Claims

1. A method for quantitatively evaluating the safety level of TBM tunneling in a mixed rock tunnel, characterized in that, The method comprises the following steps: S1, constructing an evaluation index system comprising TBM parameters, geological parameters and disaster risks; S2, determining subjective weights by constructing an expert judgment matrix through the analytic hierarchy process, obtaining objective weights by analyzing the discrete degree of monitoring data using the entropy weight method, and determining the combined weights of the indexes based on the subjective weights and the objective weights; The subjective weights are determined by constructing an expert judgment matrix through the analytic hierarchy process, comprising the following steps: S21, creating a hierarchical structure according to the evaluation index system; S22, constructing a judgment matrix for each layer and verifying the accuracy of the judgment matrix through consistency; S23, solving the judgment matrix using the eigenvector method to obtain the weights of the indexes in the current layer relative to the previous layer; The objective weights are obtained by analyzing the discrete degree of monitoring data using the entropy weight method, comprising the following steps: S201. Construct the initial matrix: ,in, , , Indicates the number of samples. Indicates the number of indicators. Denotes the initial matrix. Indicates the first The first sample The value of each indicator; S202, standardizing the sample index values; S203、According to the standardized sample index value, the entropy weight of the index is calculated, and the calculation formula is: wherein represents the entropy weight of the jth index, , , , represents the jth index value of the standardized jth sample; and the jth index value of the standardized jth sample. The formula for determining the combination weight of the indexes based on the subjective weight and the objective weight is: wherein, represents the combination weight, represents the weight of the jth index relative to the previous layer obtained by the analytic hierarchy process, represents the entropy weight of the jth index; S3, obtaining geological parameters, TBM operation parameters and environmental data in real time and performing standardization processing to construct a feature vector; S4, weighting the feature vector using the combined weights to obtain an input feature vector, and outputting a safety level using a support vector machine evaluation model; the training process of the support vector machine evaluation model comprises the following steps: S41, using the input feature vector and the corresponding evaluation results as sample data; S42, selecting a Gaussian kernel function and setting a penalty factor to 1.5 and a kernel parameter to 0.8; S43, training the model using the sample data to obtain the support vector machine evaluation model.

2. The method according to claim 1, wherein, In S1, the TBM parameters include cutterhead thrust, cutterhead torque, cutterhead speed, cutterhead excavation speed and cutterhead opening rate; the geological parameters include uniaxial compressive strength, integrity coefficient, wear resistance index, ground stress level and tunnel depth; and the disaster risks include fault fracture zone, soft rock jamming deformation, composite stratum, gushing water and surface subsidence.

3. The method according to claim 2, wherein the method is characterized by, In S4, when a rock burst precursor is detected, the value of the combined weight of the ground stress level is increased by 20% to 30%, and when a soft rock large deformation is detected, the value of the combined weight of the soft rock jamming deformation is increased by 20% to 30%.

4. The method according to claim 1, wherein, The method further comprises giving a corresponding warning according to the safety level.

5. The method according to claim 1, wherein, The regression function for the support vector machine evaluation model is: ,in, Indicates the first The first Lagrange multiplier of each sample data point Indicates the first The second Lagrange multiplier of each sample data, Indicates the first The input feature vector in each sample data, This represents the total number of sample data. Indicates a bias value. This represents the independent variable of the regression function. This represents the dependent variable in the regression function. and The solution formula is as follows: ， ， find a set and such that the objective function is maximized, where denotes the total number of sample data, denotes the evaluation result of the th sample data, denotes the input feature vector in the th sample data, denotes the insensitive band parameter, denotes the first Lagrange multiplier of the th sample data, denotes the second Lagrange multiplier of the th sample data, denotes the input feature vector of the th sample data, denotes the penalty factor.

6. The quantitative evaluation system for the safety level of TBM tunneling in mixed rock, which realizes the quantitative evaluation method for the safety level of TBM tunneling in mixed rock as claimed in claim 1, characterized in that, The system comprises a data acquisition module, a data preprocessing module, a combined weight module and an analysis processing module; The data acquisition module is used to acquire geological parameters, TBM operation parameters and environmental data in real time; The data preprocessing module; The data preprocessing module is used to standardize the real-time acquired geological parameters, TBM operation parameters and environmental data, construct a feature vector, and call the combined weight module to weight the feature vector to obtain an input feature vector; The combined weight module is used to determine subjective weights by constructing an expert judgment matrix through the analytic hierarchy process, obtain objective weights by analyzing the discrete degree of monitoring data using the entropy weight method, and determine the combined weights of the indexes based on the subjective weights and the objective weights; The analysis processing module; The analysis processing module outputs a safety level based on the input feature vector through a support vector machine evaluation model.

Images