Classification evaluation method and system for safety risk of blasting construction of tunnel in cracked zone

Abstract

The invention discloses a cracking zone tunnel blasting construction safety risk grading evaluation method and system, and belongs to the technical field of tunnel engineering safety monitoring. According to the method, dimensionality reduction and conflict resolution are performed on the original indexes by adopting the PCA-EWM method, and the mutually independent composite indexes among the indexes are obtained, so that the independence and effectiveness of the indexes are improved, and the problem that the final evaluation result is inaccurate due to the fact that the information among the original indexes is overlapped and the weight distribution is inevitably caused by repeated distribution is avoided; the weight is dynamically adjusted through the offset of the criterion layer so as to adapt to the complex engineering environment, and it is ensured that the evaluation result can reflect the current risk state; due to the fact that the composite index of tunnel blasting and the construction risk degree are in a nonlinear relation, a normal cloud model is adopted to reconstruct a membership function and process nonlinear risk evolution of the index so as to better adapt to nonlinear characteristics of the tunnel blasting construction data of the cracked zone, the accuracy of risk assessment is improved, and the safety of construction is guaranteed.

Description

Technical Field

[0001] This invention belongs to the field of tunnel engineering safety monitoring technology, specifically relating to a method and system for safety risk classification and evaluation of tunnel blasting construction in fracture zones. Background Technology

[0002] When tunnels traverse fracture zones, the surrounding rock is loose, jointed, and actively seeping, making blasting operations highly susceptible to collapse accidents. Current technologies primarily employ the AHP-FCE (Analog-Hybrid Analysis-FCE) method for risk assessment, but this method suffers from the following drawbacks: Static weights: In traditional methods, indicator weights are set by experts on a one-time basis and cannot be dynamically adjusted based on real-time geological parameters (such as rock mass integrity coefficients), resulting in risk levels lagging behind on-site changes. Redundant and conflicting indicators: Information overlap exists between evaluation indicators (e.g., "blasting vibration velocity" and "drilling deviation" are highly correlated), leading to redundant weight allocation and reduced assessment accuracy. Traditional trapezoidal membership functions exhibit abrupt changes at critical values, failing to characterize the nonlinear evolution of geological parameters in fracture zones. Lack of closed-loop feedback: Evaluation results are only used for post-event analysis and cannot guide real-time adjustments to construction parameters.

[0003] Therefore, there is an urgent need for a risk classification method that can dynamically couple subjective and objective weights, eliminate indicator conflicts, perform nonlinear membership mapping, and achieve real-time closed-loop control. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method and system for safety risk classification and evaluation of tunnel blasting construction in fractured zones, which addresses the shortcomings of the prior art. The system is novel and reasonable in design, simple in structure, highly practical, and easy to promote and use.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A method for safety risk classification and evaluation in tunnel blasting construction in fractured zones is proposed, which constructs a four-layer progressive dynamic risk classification model, specifically including the following steps: S1. Constructing the indicator layer: Collecting raw parameters from four dimensions—geology, construction, environment, and management—as raw indicator data; S2. Constructing a composite index layer: The original indices are reconciled and their dimensions are compressed using the PCA-EWM method to obtain composite indices; S3. Construct a criterion layer, classify composite indicators into four criteria according to engineering logic, construct a judgment matrix using the 1-9 scaling method of AHP, solve for the maximum eigenvalue and corresponding eigenvector using the eigenvalue method, and use the normalized eigenvector as the basic weight after completing the consistency test (CR<0.1); adopt a dynamic weight correction mechanism to nonlinearly correct the basic weight according to the criterion layer offset.

[0006] in, For the first The basic weights of each criterion The correlation coefficient. , It is a non-linear adjustment factor. , It is the hyperbolic tangent function. For the first The criterion layer offset of each criterion. For the first The dynamic weights of each criterion For the criterion number, ; S4. Construct the target layer and reconstruct the membership function using a normal cloud model to handle the nonlinear risk evolution of the indicators; calculate the comprehensive risk index based on the dynamic weights and membership degrees of the criterion layer, using the following formula:

[0007] in, As a comprehensive risk index, For the first The membership degree of each criterion at the criterion level. The total number of criteria; S5. Closed-loop feedback control adjusts construction parameters in real time based on the risk level output from the target layer.

[0008] Furthermore, in step S2, the process of obtaining the composite index is as follows: first, conflict detection is performed, and then a correlation matrix between the original indices is constructed. , This represents the correlation coefficient between the i-th and j-th indicators. When the correlation coefficient between indicators is greater than the correlation threshold, the indicator is placed in a conflict group. PCA is used to merge the indicators in the same conflict group to generate a composite indicator; the correlation threshold is 0.8. The process of merging conflict groups using PCA is as follows: The first principal component is extracted from the conflict groups using PCA as the composite index, expressed as:

[0009] in, For the first The score of each composite indicator, For the first The weights of the original indicators in this composite indicator For the first The mean and standard deviation of each original indicator. For the first Observed values ​​of the original indicators; For the first The number of original indicators within each conflict group The conflict group number. , For local indexes within a conflict group; The information entropy of the composite index is calculated to adjust the weights. The adjustment formula is as follows:

[0010] in, For the first Information entropy of a composite indicator For the first The weights of the composite indicators, This is the sequence number of the composite index. This represents the total number of composite indicators.

[0011] Furthermore, the correlation matrix is ​​the Pearson correlation coefficient matrix, and the expression for the correlation coefficient is:

[0012] in, for covariance, They are respectively standard deviation The first The value of the first indicator and the first The value of each indicator.

[0013] Further, in step S3, the expression for the criterion layer offset is:

[0014] in, For the first Real-time scores of each composite indicator; For the first The preset risk level thresholds for each composite indicator. For the first The total number of composite indicators corresponding to each criterion. For the first A set of composite indicators corresponding to each criterion.

[0015] Furthermore, in step S4, the process of reconstructing the membership function using the normal cloud model is as follows: construct a composite index membership function based on the median of the index threshold, the variation entropy, and the hyperentropy of the normal cloud model; aggregate the composite index membership degree into the criterion layer membership degree function; and update the geological variation entropy using the Bayesian algorithm. The expression for the membership function of the criterion layer is:

[0016] in, For the first The composite index in the first Local weights within each criterion; The expression for the membership function of the composite index is:

[0017] in, For the first The scores of the composite indicators, For the first The median threshold value of a composite indicator. For the first The entropy of variation of a composite index For the first The hyperentropy of a composite index for The range of scores for each composite indicator; The expression for updating the mutation entropy is:

[0018] in, For the updated mutation entropy, For the entropy of historical variation, This represents the standard deviation of the test data for this batch. This represents the standard deviation of historical data.

[0019] Furthermore, in step S5, the comprehensive risk index The mapping relationship with risk level is as follows: Level I risk Level II risk Level III risk It is classified as a Level IV risk.

[0020] Furthermore, the indicator layer includes the following indicators: rock mass integrity coefficient, joint density, permeability coefficient, fault zone width, peak velocity of blasting vibration, blasting frequency, blasting hole spacing, maximum charge per section, borehole deviation, groundwater flow rate, daily rainfall, surface subsidence rate, monitoring frequency, completeness of emergency response plan, and personnel training level.

[0021] Furthermore, the four principles are: geological principles, construction principles, environmental principles, and management principles.

[0022] Furthermore, the composite indicators are: geological integrity and water-rock coupling, blasting energy disturbance and construction accuracy, environmental disturbance, and management response.

[0023] This invention also discloses a safety risk classification and evaluation system for blasting construction in fractured zones, used to implement the above-mentioned safety risk classification and evaluation method for blasting construction in fractured zones. The system includes: The data acquisition module includes a ground-penetrating radar, borehole camera, accelerometer array, pore water pressure gauge, flow meter, and digital inspection system, used to acquire the raw parameters described in step S1; The edge computing module is used to execute the calculation process of steps S2-S4, including a conflict detection unit, a PCA synthesis unit, an entropy weight correction unit, a dynamic weight calculation unit, and a cloud model membership calculation unit. The decision control module includes a PLC controller, which is used to adjust the construction parameters in real time according to the risk level output in step S4. The historical database stores original parameters, intermediate calculation results, and risk level records, providing data support for Bayesian updates.

[0024] Compared with the prior art, the present invention has the following advantages: This invention employs the PCA-EWM method to reduce the dimensionality and resolve conflicts of the original indicators, resulting in independent composite indicators. This improves the independence and effectiveness of the indicators and avoids the problem of redundant weight allocation caused by information overlap between the original indicators, which leads to inaccurate final evaluation results. By dynamically adjusting the weights through the criterion layer offset, the risk level is evaluated based on the dynamic weights, which can adapt to complex engineering environments and ensure that the evaluation results reflect the current risk status. Since the composite indicators of tunnel blasting have a non-linear relationship with the degree of construction risk, a normal cloud model is used to reconstruct the membership function to handle the non-linear risk evolution of the indicators, better adapting to the non-linear characteristics of tunnel blasting construction data in fracture zones and improving the accuracy of risk assessment. Finally, based on the risk level output by the target layer, the engineering strategy is adjusted in real time to ensure the safety of construction.

[0025] This invention systematically solves the problems of weight updating, indicator conflict, membership mapping, and real-time decision-making in risk assessment by means of dynamic weight correction, PCA-EWM method, normal cloud model, and closed-loop feedback control. These methods not only improve the accuracy, real-time performance, and adaptability of risk assessment, but also enhance the transparency and interpretability of assessment results, providing a scientific basis and technical support for risk assessment in complex engineering environments.

[0026] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0027] Figure 1 This is a flowchart illustrating an embodiment of the safety risk classification and evaluation method for tunnel blasting construction in fracture zones according to the present invention. Figure 2 This is a structural block diagram of an embodiment of the safety risk classification and evaluation system for tunnel blasting construction in fracture zones according to the present invention. Detailed Implementation

[0028] Example of a safety risk classification and evaluation method for tunnel blasting construction in fracture zones: like Figure 1 As shown, the safety risk classification and evaluation method for tunnel blasting construction in fracture zones constructs a four-layer progressive dynamic risk classification model, which includes the following steps: S1. Constructing the Indicator Layer: This involves collecting raw parameters from four dimensions: geology, construction, environment, and management, as the initial indicator data. The indicator layer includes the following indicators: rock mass integrity coefficient, joint density, permeability coefficient, fault zone width, peak velocity of blasting vibration, blasting frequency, blasting hole spacing, maximum charge per section, borehole deviation, groundwater flow rate, daily rainfall, surface subsidence rate, monitoring frequency, emergency response plan completeness, and personnel training level. Specific indicator classifications are shown in Table 1 below.

[0029] Table 1

[0030] Among them, Kv, Jd, and Kw are derived from real-time calculations of ground-penetrating radar and borehole camera data; PPV and Fm are derived from an accelerometer array; Gw is derived from a pore water pressure gauge and flow meter; and Ep and Tl are derived from a digital inspection system.

[0031] S2. Constructing a composite index layer: The original indices are reconciled and their dimensions compressed using the PCA-EWM method (Principal Component Analysis-Entropy Weight Fusion Method) to obtain composite indices.

[0032] The process of obtaining the composite index is as follows: first, conflict detection is performed, and then a correlation matrix between the original indices is constructed. , This represents the correlation coefficient between the i-th and j-th indicators. When the correlation coefficient between indicators is greater than the correlation threshold, the indicator is placed in a conflict group. PCA is used to merge the indicators in the same conflict group to generate a composite indicator. The correlation threshold is 0.8.

[0033] The process of merging conflict groups using PCA is as follows: The first principal component is extracted from the conflict groups using PCA as the composite index, expressed as:

[0034] in, For the first The score of each composite indicator, For the first The weights of the original indicators in this composite indicator For the first The mean and standard deviation of each original indicator. For the first Observed values ​​of the original indicators; For the first The number of original indicators within each conflict group The conflict group number. , This is a local index within the conflict group. , The information entropy of the composite index is calculated to adjust the weights. The adjustment formula is as follows:

[0035] in, For the first Information entropy of a composite indicator For the first The weights of the composite indicators, The total number of composite indicators, , The total number of composite indicators. .

[0036] The correlation matrix is ​​the Pearson correlation coefficient matrix, and the expression for the correlation coefficient is:

[0037] in, for covariance, They are respectively standard deviation The first The value of the first indicator and the first The value of each indicator.

[0038] PCA synthesis generates six composite indicators: geological integrity and water-rock coupling, blasting energy disturbance and construction accuracy, environmental disturbance, and management response. Specifically, geological integrity Z1 is synthesized from Kv and Jd; blasting energy disturbance Z2 from PPV, Fm, and Qmax; construction accuracy Z3 from Hs and Dd; water-rock coupling Z4 from Kw and Gw; environmental disturbance Z5 from Rf and Sr; and management response Z6 from Mf, Ep, and Tl. It is important to note that the symbols used for Z1, Z2, etc., differ from those used for the composite indicators mentioned earlier; this is explained to avoid confusion.

[0039] S3. Construct a criterion layer, categorizing composite indicators into four main criteria based on engineering logic. These four criteria are: geological criteria, construction criteria, environmental criteria, and management criteria. The specific correspondence between the geological criteria layer and the composite indicators is shown in the table below: Table 2

[0040] The judgment matrix is ​​constructed using the 1-9 scaling method of AHP.

[0041] Judgment Matrix for:

[0042] The largest eigenvalue and its corresponding eigenvector are obtained using the eigenvalue method. After a consistency check, the basic weights are determined. The expression for the basic weights is as follows: The consistency test is CR < 0.1.

[0043] A dynamic weight correction mechanism is adopted to nonlinearly correct the basic weights based on the criterion layer offset. Specifically, the correction formula is as follows:

[0044] in, For the first The basic weights of each criterion The correlation coefficient. , It is a non-linear adjustment factor. , It is the hyperbolic tangent function. For the first The criterion layer offset of each criterion. For the first The dynamic weights of each criterion For the criterion number, This implementation addresses the issue of excessive subjectivity in existing weighting systems, which rely solely on expert scoring. By combining subjective and objective methods, it ensures the accuracy of the weights.

[0045] The expression for the criterion layer offset is:

[0046] in, For the first Real-time scores of each composite indicator; For the first The preset risk level thresholds for each composite indicator. For the first The total number of composite indicators corresponding to each criterion. For the first A set of composite indices corresponding to a criterion. For example, the set of geological criteria is {Z1, Z2}; corresponding to It equals 2. These correspond to geology, construction, environment, and management guidelines, respectively.

[0047] S4. Construct the target layer and reconstruct the membership function using a normal cloud model to handle the nonlinear risk evolution of the indicators; calculate the comprehensive risk index based on the dynamic weights of the criterion layer and the membership degrees of the cloud model. The target is the blasting collapse risk level, and the formula for calculating the comprehensive risk index is:

[0048] in, As a comprehensive risk index, For the first The membership degree of each criterion at the criterion level. The total number of criteria.

[0049] The normal cloud model is an uncertainty modeling method based on fuzzy mathematics and probability statistics. It includes key parameter indices, such as the median threshold and geological variability entropy. The median threshold is the center point of the normal cloud model, and the geological variability entropy measures the uncertainty of the cloud model. The process of reconstructing the membership function using the normal cloud model is as follows: A composite index membership function is constructed based on the median threshold, variability entropy, and hyperentropy of the normal cloud model; the composite index membership degrees are aggregated into a criterion-level membership function; and the geological variability entropy is updated using a Bayesian algorithm. The expression for the membership function of the criterion layer is:

[0050] in, For the first The composite index in the first Local weights within each criterion; only the geological criterion and the construction criterion have weight allocation issues, and the local weights for these two criteria are both set to 0.5; the local weights can also be adaptively changed according to the actual situation of the engineering logic classification; the other two criteria each have only one composite index, and the corresponding weight is set to 1.

[0051] The expression for the membership function of the composite index is:

[0052] in, For the first The scores of the composite indicators, For the first The median threshold value of a composite indicator. For the first The entropy of variation of a composite index For the first The hyperentropy of a composite index , for The range of scores for a composite indicator is the difference between the maximum and minimum values ​​of that composite indicator.

[0053] The expression for updating the mutation entropy is:

[0054] in, For the updated mutation entropy, For the entropy of historical variation, This represents the standard deviation of the test data for this batch. This represents the standard deviation of historical data.

[0055] Historical geological variation entropy refers to the geological variation entropy obtained from offline training of the previous batch or the original batch. It is mainly used as prior information in Bayesian updates to reflect the degree of fluctuation in historical data.

[0056] S5. Closed-loop feedback control adjusts construction parameters in real time based on the risk level output from the target layer.

[0057] In step S5, the comprehensive risk index The mapping relationship with risk level is as follows: Level I risk Level II risk Level III risk The risk level is classified as Level IV. Specifically, the correspondence between risk level and decision-making is as follows: Level I corresponds to normal construction, with a monitoring frequency of once per hour; Level II corresponds to an early warning state, with a monitoring frequency of three times per hour; Level III corresponds to "automatic triggering of reducing explosives by 20% and performing grouting reinforcement"; Level IV corresponds to immediate work stoppage and activation of the emergency plan. The classification of these levels can be customized according to actual needs; the corresponding risk levels and decision-making relationships can also be set according to the actual situation.

[0058] This invention constructs a four-layer progressive dynamic risk classification model to achieve dynamic weight correction, index conflict resolution, nonlinear membership function reconstruction, and closed-loop feedback and real-time control. It can dynamically assess the collapse risk of tunnel blasting construction in fracture zones in real time and adaptively adjust construction parameters, effectively improving construction safety and efficiency.

[0059] Example of a safety risk classification and evaluation system for tunnel blasting construction in fracture zones: like Figure 2 As shown, the safety risk classification and evaluation system for blasting construction in fractured zones is used to implement the aforementioned safety risk classification and evaluation method for blasting construction in fractured zones. Specifically, it includes: The data acquisition module includes a ground-penetrating radar, borehole camera, accelerometer array, pore water pressure gauge, flow meter, and digital inspection system, used to collect the raw parameters described in step S1 above.

[0060] The edge computing module is used to perform the calculation process of steps S1-S4 above, including a conflict detection unit, a PCA synthesis unit, an entropy weight correction unit, a dynamic weight calculation unit, and a cloud model membership calculation unit.

[0061] The decision control module includes a PLC controller, which is used to adjust the construction parameters in real time based on the risk level output in step S4 above.

[0062] The historical database stores original parameters, intermediate calculation results, and risk level records, providing data support for Bayesian updates.

[0063] This embodiment is implemented with reference to the above-described embodiment of the method for safety risk classification and evaluation of tunnel blasting construction in fractured zones, and will not be repeated here.

[0064] The above description is merely a preferred embodiment of the present invention and does not constitute any limitation on the present invention. Any simple modifications, alterations, or equivalent structural changes made to the above embodiments based on the technical essence of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for safety risk classification and evaluation in tunnel blasting construction in fractured zones, characterized in that: The construction of a four-layer progressive dynamic risk classification model includes the following steps: S1. Constructing the indicator layer: Collecting raw parameters from four dimensions—geology, construction, environment, and management—as raw indicator data; S2. Constructing a composite index layer: The original indices are reconciled and their dimensions are compressed using the PCA-EWM method to obtain composite indices; S3. Construct a criterion layer, classifying composite indicators into four criteria according to engineering logic. Use the AHP 1-9 scaling method to construct a judgment matrix, and solve for the maximum eigenvalue and corresponding eigenvector using the eigenvalue method. After completing the consistency test (CR < 0.1), use the normalized eigenvectors as the basic weights. Employ a dynamic weight correction mechanism to non-linearly correct the basic weights based on the criterion layer offset. The correction formula is as follows: in, For the first The basic weights of each criterion The correlation coefficient. , It is a non-linear adjustment factor. , It is the hyperbolic tangent function. For the first The criterion layer offset of each criterion. For the first The dynamic weights of each criterion For the criterion number, ; S4. Construct the target layer and reconstruct the membership function using a normal cloud model to handle the nonlinear risk evolution of the indicators; calculate the comprehensive risk index based on the dynamic weights and membership degrees of the criterion layer, using the following formula: in, As a comprehensive risk index, For the first The membership degree of each criterion at the criterion level. The total number of criteria; S5. Closed-loop feedback control adjusts construction parameters in real time based on the risk level output from the target layer.

2. The method for safety risk classification and evaluation of tunnel blasting construction in fracture zones according to claim 1, characterized in that: In step S2, the process of obtaining the composite index is as follows: first, conflict detection is performed, and then a correlation matrix between the original indices is constructed. , This represents the correlation coefficient between the i-th and j-th indicators. When the correlation coefficient between indicators is greater than the correlation threshold, the indicator is placed in a conflict group. PCA is used to merge the indicators in the same conflict group to generate a composite indicator; the correlation threshold is 0.

8. The process of merging conflict groups using PCA is as follows: The first principal component is extracted from the conflict groups using PCA as the composite index, expressed as: in, For the first The score of each composite indicator, For the first The weights of the original indicators in this composite indicator For the first The mean and standard deviation of each original indicator. For the first Observed values ​​of the original indicators; For the first The number of original indicators within each conflict group The conflict group number. , For local indexes within a conflict group; The information entropy of the composite index is calculated to adjust the weights. The adjustment formula is as follows: in, For the first Information entropy of a composite indicator For the first The weights of the composite indicators, This is the sequence number of the composite index. This represents the total number of composite indicators.

3. The method for safety risk classification and evaluation of tunnel blasting construction in fracture zones according to claim 2, characterized in that: The correlation matrix is ​​the Pearson correlation coefficient matrix, and the expression for the correlation coefficient is: in, for covariance, They are respectively standard deviation The first The value of the first indicator and the first The value of each indicator.

4. The method for safety risk classification and evaluation of tunnel blasting construction in fracture zones according to claim 1, characterized in that: In step S3, the expression for the criterion layer offset is: in, For the first Real-time scores of each composite indicator; For the first The preset risk level thresholds for each composite indicator. For the first The total number of composite indicators corresponding to each criterion. For the first A set of composite indicators corresponding to each criterion.

5. The method for safety risk classification and evaluation of tunnel blasting construction in fractured zones according to claim 4, characterized in that: In step S4, the process of reconstructing the membership function using the normal cloud model is as follows: construct a composite index membership function based on the median of the index threshold, the entropy of variation, and the hyperentropy of the normal cloud model; aggregate the composite index membership degree into the criterion layer membership degree function; and update the geological variation entropy using the Bayesian algorithm. The expression for the membership function of the criterion layer is: in, For the first The composite index in the first Local weights within each criterion; The expression for the membership function of the composite index is: in, For the first The scores of the composite indicators, For the first The median threshold value of a composite indicator. For the first The entropy of variation of a composite index For the first The hyperentropy of a composite index for The range of scores for each composite indicator; The expression for updating the mutation entropy is: in, For the updated mutation entropy, For the entropy of historical variation, This represents the standard deviation of the test data for this batch. This represents the standard deviation of historical data.

6. The method for safety risk classification and evaluation of tunnel blasting construction in fractured zones according to claim 1, characterized in that: In step S5, the comprehensive risk index The mapping relationship with risk level is as follows: Level I risk Level II risk Level III risk It is classified as a Level IV risk.

7. The method for safety risk classification and evaluation of tunnel blasting construction in fractured zones according to claim 1, characterized in that: The indicator layer includes the following indicators: rock mass integrity coefficient, joint density, permeability coefficient, fault zone width, peak velocity of blasting vibration, blasting frequency, blasting hole spacing, maximum charge per section, borehole deviation, groundwater flow rate, daily rainfall, surface subsidence rate, monitoring frequency, completeness of emergency response plan, and personnel training level.

8. The method for safety risk classification and evaluation of tunnel blasting construction in fractured zones according to claim 1, characterized in that: The four principles are: geological principles, construction principles, environmental principles, and management principles.

9. A method for safety risk classification and evaluation of tunnel blasting construction in fracture zones according to claim 1, characterized in that: The composite indicators are: geological integrity, water-rock coupling, blasting energy disturbance, construction accuracy, environmental disturbance, and management response.

10. A safety risk classification and evaluation system for tunnel blasting construction in fractured zones, characterized in that, The system for implementing the safety risk classification and evaluation method for tunnel blasting construction in fractured zones as described in any one of claims 1-9 comprises: The data acquisition module includes a ground-penetrating radar, borehole camera, accelerometer array, pore water pressure gauge, flow meter, and digital inspection system, used to acquire the raw parameters described in step S1; The edge computing module is used to execute the calculation process of steps S2-S4, including a conflict detection unit, a PCA synthesis unit, an entropy weight correction unit, a dynamic weight calculation unit, and a cloud model membership calculation unit. The decision control module includes a PLC controller, which is used to adjust the construction parameters in real time according to the risk level output in step S4. The historical database stores original parameters, intermediate calculation results, and risk level records, providing data support for Bayesian updates.

Images