A prestressed concrete bridge risk diagnosis and control method, device, equipment and medium

Abstract

The application provides a prestressed concrete bridge risk diagnosis and control method, device, equipment and medium, the method comprises the following steps: constructing a coupling graph, determining the membership relationship and event influence relationship between nodes, and configuring the attribute parameters of each node; the data is converted into a unified evidence tuple corresponding to the coupling graph; the evidence credibility of each evidence tuple is determined based on the preset evaluation index; the candidate risk state node in the coupling graph is taken as the target node, the propagation weight, the contradictory evidence and the missing precondition are introduced to punish and correct the risk score, and the candidate risk score of each candidate risk state node is output through the normalization function; the completeness value of each candidate risk state is evaluated, if it is lower than the preset threshold, the executable detection item is selected from the detection action node set, the optimization model is constructed, the minimum supplementary detection set is solved and determined, and feedback is fed back; if it is not lower than the preset threshold, the feasible treatment scheme is selected in combination with the disease mechanism, the risk level and the existing precondition detection result.

Description

Technical Field

[0001] This invention relates to the field of multimodal information analysis technology, and particularly to the cross-disciplinary fields of bridge engineering, structural safety diagnosis, non-destructive testing, structural health monitoring and engineering knowledge computation. It mainly relates to a method, device, equipment and medium for risk diagnosis and control of prestressed concrete bridges. Background Technology

[0002] Prestressed concrete bridges are widely used in highway and urban bridges, but during long-term service, they often experience problems such as cracks, prestress loss, grouting defects in ducts, anchorage zone defects, bearing anomalies, leakage, concrete spalling, and abnormal alignment. These defects often exhibit location-relatedness, mechanistic coupling, and developmental stages. For example, even with the same term "longitudinal crack," the dominant mechanism and subsequent detection path may be completely different when the crack is located in the web, top slab, anchorage zone, or near the support. Furthermore, beam deflection, abnormal expansion joints, and cable stress attenuation usually have temporal and spatial relationships, and conclusions cannot be drawn based on a single symptom.

[0003] Existing technologies mainly fall into three categories for detection and diagnosis: The first category is diagnostic methods based on human experience and standard provisions. While these methods offer some engineering interpretability, they are weak in their ability to jointly utilize multi-source data and have low efficiency in determining complex disease combinations. The second category is recognition methods based on machine learning or deep learning. These methods are more suitable for image classification, object detection, or single-index prediction, but they often directly output scores or categories, lacking a clear technical chain of "symptom-mechanism-verification detection-treatment action," making it difficult to support engineering decision-making. The third category is knowledge graph or large-model-assisted question-answering solutions. These methods can improve knowledge organization and text interaction capabilities, but most remain at the level of "text retrieval + answer generation," still failing to solve the three most critical problems in bridge engineering: The first is how to simultaneously incorporate site differences, temporal variations, and quantification thresholds into the inference; Secondly, how to automatically provide a minimum and necessary supplementary detection set when there is insufficient evidence, instead of directly outputting unstable conclusions; Third, how to incorporate the prerequisites for maintenance measures, traffic constraints, and construction windows into the same decision-making chain.

[0004] Therefore, it can be seen that existing technologies generally have the following shortcomings: (1) The disease name is used as the main index, but the coupled expression of "component-part-mechanism" is not established, which leads to the mixed use of cases with the same expression but different engineering meanings; (2) There is a lack of credible dissemination and conflict punishment mechanisms for evidence from different sources. Inspection descriptions, detection values ​​and monitoring trends are often disconnected from each other, or there is a lack of clear order of resolution when there is conflict. (3) In the supplementary testing phase, testing items are usually selected manually based on experience, which cannot take into account information gain, traffic occupation and testing intrusion.

[0005] (4) In the treatment recommendation stage, insufficient consideration is given to the pre-test conditions, prohibited construction methods and construction window, which may easily lead to the result that "theoretically feasible but difficult to implement on site".

[0006] Therefore, a more convergent and engineered technical solution is needed so that bridge risk diagnosis no longer relies on simple semantic retrieval or black-box prediction, but can make credible inferences based on the coupling relationship between mechanisms and locations, proactively plan supplementary detections when evidence is insufficient, and output treatment plans that meet the conditions for on-site implementation when evidence is sufficient. Summary of the Invention

[0007] To overcome the problems existing in related technologies, this disclosure provides a method, device, equipment and medium for risk diagnosis and control of prestressed concrete bridges, so as to solve the technical problems in related technologies.

[0008] This specification provides one or more embodiments of a method for risk diagnosis and control of prestressed concrete bridges, including the following steps: 1) Collect multi-dimensional bridge engineering data and standardize the multi-dimensional bridge engineering data to obtain a standardized bridge dataset; 2) Construct a mechanism-location coupling map including components, locations, disease mechanisms, evidence indicators, detection actions, and treatment action nodes, determine the membership relationships and event influence relationships between nodes, and configure the attribute parameters of each node; 3) Transform the data in the standardized bridge dataset into unified evidence tuples of the corresponding coupled graph, including at least the dimensions of components, locations, disease mechanisms, and evidence indicators; 4) Determine the credibility of each evidence tuple based on preset evaluation indicators; 5) Taking the candidate risk state nodes in the coupling graph as target nodes, based on the connection relationship between the evidence tuples and the target nodes, propagation weights, contradictory evidence, and missing preconditions are introduced to penalize and correct the risk scores. The candidate risk scores of each candidate risk state node are then output through a normalization function. 6) Evaluate the completeness value of each candidate risk status based on symptom coverage, key indicator coverage, key site coverage, and evidence consistency indicators; 7) If the completeness value of the candidate risk state is lower than the preset threshold, select executable detection items from the set of detection action nodes, construct an optimization model with the goal of minimizing detection cost, traffic occupation cost, and structural intrusion cost, while constraining the total information gain to be no less than the set threshold, solve to determine the minimum supplementary detection set and feed it back to step 1); if the completeness value of the candidate risk state is not lower than the preset threshold, combine the disease mechanism, risk level, and existing previous detection results to preliminarily screen feasible treatment solutions.

[0009] This specification provides one or more embodiments of a risk diagnosis and control device for prestressed concrete bridges, including: Multi-source data acquisition and standardization module: used to collect multi-dimensional bridge engineering data and standardize the multi-dimensional bridge engineering data to obtain a standardized bridge dataset; Coupled graph construction module: used to construct a mechanism-part coupling graph including nodes of components, parts, disease mechanisms, evidence indicators, detection actions, and treatment actions, and to determine the membership relationships and event influence relationships between nodes, and configure the attribute parameters of each node; Evidence tuple construction module: used to transform data in the standardized bridge dataset into unified evidence tuples corresponding to the coupled graph, including at least the dimensions of components, locations, disease mechanisms, and evidence indicators; Consistency calculation module: used to determine the credibility of each evidence tuple based on preset evaluation indicators; The credible risk state propagation module is used to take candidate risk state nodes in the coupled graph as target nodes, and to introduce propagation weights, contradictory evidence, and missing preconditions to penalize and correct risk scores based on the connection relationship between evidence tuples and target nodes. The module then outputs the candidate risk scores of each candidate risk state node through a normalization function. The evidence completeness assessment module is used to evaluate the completeness value of each candidate risk status based on symptom coverage, key indicator coverage, key site coverage, and evidence consistency indicators. The supplementary detection set solution module is used to determine whether the completeness value of a candidate risk state is lower than a preset threshold. It then selects executable detection items from the set of detection action nodes, constructs an optimization model that minimizes detection cost, traffic occupancy cost, and structural intrusion cost, while ensuring the total information gain is not lower than a set threshold. The module solves and determines the minimum supplementary detection set, feeding it back to the multi-source data acquisition and standardization module. If the completeness value of a candidate risk state is not lower than the preset threshold, the module feeds it back to the treatment plan determination module. The treatment plan determination module is used to initially screen feasible treatment plans by combining the disease mechanism, risk level, and existing preliminary test results.

[0010] This specification provides one or more embodiments of a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the prestressed concrete bridge risk diagnosis and control method as described above.

[0011] This specification provides one or more embodiments of a computer-readable storage medium storing a computer program that, when executed by a processor, implements the risk diagnosis and control method for prestressed concrete bridges as described above.

[0012] This disclosure provides a method, apparatus, equipment, and medium for risk diagnosis and control of prestressed concrete bridges. Its advantages lie in addressing the shortcomings of existing bridge defect diagnosis methods, such as lack of component modeling, difficulty in unifying multi-source evidence, reliance on manual supplementary inspections, and treatment plans detached from on-site constraints. By standardizing multi-source data and constructing a coupled atlas based on components, locations, and defect mechanisms, the probability of the same defect name being misused in different locations can be significantly reduced. Furthermore, textual, numerical, and temporal information are utilized. This invention, through unified evidence tuples and a consistency calculation mechanism, integrates inspection symptoms, quantitative detection values, and monitoring evolution characteristics into the same risk propagation framework, avoiding independent judgments from multiple sources of evidence. By correcting the risk score of candidate risk state nodes through contradictory evidence penalty items and missing precondition penalty items, it exhibits greater engineering stability than existing schemes that rely solely on similarity ranking. For candidate risk state completeness values ​​below a preset threshold, information gain is the objective, and the minimum supplementary detection set is automatically obtained under constraints of cost, traffic occupancy, and invasiveness, thereby reducing unnecessary detections. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in one or more embodiments of this specification or in the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments recorded in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 A flowchart illustrating a risk diagnosis and control method for prestressed concrete bridges provided for one or more embodiments of this specification; Figure 2 This is a schematic diagram of the coupling spectrum structure provided in one or more embodiments of this specification; Figure 3 This diagram illustrates the calculation of evidence credibility and the propagation of risk status for one or more embodiments of this specification; it shows the relationship between evidence tuple input, mechanism matching, spatial continuity, temporal consistency, threshold compliance, contradiction penalty, and precondition penalty; Figure 4 This diagram illustrates the treatment execution boundary verification and post-test gating update for one or more embodiments of this specification; it shows the pre-test conditions, prohibited construction methods, traffic organization, construction window, retest results, and case write-back conditions. Figure 5 A block diagram of a risk diagnosis and control device for prestressed concrete bridges provided for one or more embodiments of this specification; Figure 6 This is a schematic diagram of the structure of a computer device provided for one or more embodiments of this specification. Detailed Implementation

[0015] To enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the technical solutions in one or more embodiments of this specification 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 this specification, and not all of the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of this invention.

[0016] The present invention will now be described in detail with reference to specific embodiments and accompanying drawings.

[0017] Method Implementation Examples According to embodiments of the present invention, a method for risk diagnosis and control of prestressed concrete bridges is provided, such as... Figure 1 The diagram shown is a flowchart of the risk diagnosis and control method for prestressed concrete bridges provided in this embodiment. The risk diagnosis and control method for prestressed concrete bridges according to this embodiment includes the following steps: S1, Multi-source engineering data acquisition and standardization: Collect multi-dimensional bridge engineering data such as inspection, testing, monitoring, design and construction, maintenance, and traffic restrictions, and standardize the multi-dimensional bridge engineering data to obtain a standardized bridge dataset.

[0018] S2, Mechanism-Site Coupling Map Construction: Construct a mechanism-site coupling map including components, locations, disease mechanisms, evidence indicators, detection actions, and treatment action nodes, determine the membership relationships and event influence relationships between nodes, and configure the attribute parameters of each node.

[0019] S3, Construction of Evidence Tuples: Transform the data in the standardized bridge dataset of step S1 into unified evidence tuples corresponding to the coupling graph of S2, including at least the dimensions of components, locations, disease mechanisms, and evidence indicators; S4, Evidence Consistency Calculation: Determine the credibility of each evidence tuple based on preset evaluation indicators.

[0020] S5, Reliable Propagation of Risk States: Taking the candidate risk state nodes in the coupled graph as target nodes, based on the connection relationship between the evidence tuples and the target nodes, propagation weights, contradictory evidence, and missing preconditions are introduced to penalize and correct the risk scores. The candidate risk scores of each candidate risk state node are output through a normalization function.

[0021] S6, Evidence completeness assessment: The completeness value of each candidate risk status is evaluated based on the indicators of symptom coverage, key indicator coverage, key site coverage, and evidence consistency. S7, Information Gain-Driven Minimum Supplement Detection Set Solution: If the completeness value of a candidate risk state is lower than a preset threshold, select executable detection items from a preset set of detection action nodes according to rules that they are related to the current candidate risk state, can supplement missing key indicators, meet on-site implementation conditions, and do not violate dependency or mutual exclusion relationships. Construct an optimization model with the goal of minimizing detection cost, traffic occupation cost, and structural intrusion cost, while constraining the total information gain to not be lower than a set threshold, and solve to obtain the minimum supplement detection set; after the minimum supplement detection set is executed, the newly added detection results are fed back as new bridge data to... Step S1: Regenerate the evidence tuple and re-enter the risk reasoning process until the completeness reaches a preset threshold. If the completeness value of the candidate risk state is not lower than the preset threshold, proceed to step S8. The preset set of detection action nodes is established based on the detection action nodes in step S2, forming an optional supplementary inspection item library. This library includes attributes such as inspection item name, applicable components and locations, verifiable disease mechanisms, supplementary evidence indicators, inspection costs, traffic occupancy costs, structural intrusion costs, implementation conditions, dependencies, and mutual exclusion relationships. These attributes provide quantifiable supplementary evidence sources for candidate risk states with insufficient completeness. Step S8: Determine the treatment plan: Combining the disease mechanism, risk level, and existing preliminary detection results, initially screen feasible treatment plans.

[0022] The risk diagnosis and control method for prestressed concrete bridges provided in this embodiment addresses the shortcomings of existing bridge defect diagnosis methods, such as lack of component modeling, difficulty in unifying multi-source evidence, reliance on manual supplementary inspections, and treatment plans being detached from on-site constraints. By standardizing multi-source data and constructing a coupled atlas based on components, locations, and defect mechanisms, the method significantly reduces the probability of the same defect name being misused in different locations. It also utilizes textual, numerical, and temporal information. This invention, through unified evidence tuples and a consistency calculation mechanism, integrates inspection symptoms, quantitative detection values, and monitoring evolution characteristics into the same risk propagation framework, avoiding independent judgments from multiple sources of evidence. By correcting the risk score of candidate risk state nodes through contradictory evidence penalty items and missing precondition penalty items, it exhibits greater engineering stability than existing schemes that only rank based on similarity. For candidate risk state completeness values ​​below a preset threshold, it automatically calculates the minimum supplementary detection set under constraints of cost, traffic occupancy, and invasiveness, based on information gain, thereby reducing unnecessary detections.

[0023] In this embodiment, step S1, multi-source engineering data acquisition and standardization, specifically includes the following steps: Step 11: Collect bridge inspection records, special inspection reports, structural health monitoring data, design drawings, construction records, maintenance files, existing maintenance cases, and traffic organization restriction information; Step 12: Standardize the format, terminology, units, time, location, and source of the collected data to form a standardized bridge dataset.

[0024] In this embodiment, to resolve the confusion caused by cases with the same name but different engineering meanings, which would lead to unclear correspondence between disease symptoms and mechanisms and inaccurate detection schemes, step S2 is used to construct six types of nodes: components, locations, disease mechanisms, evidence indicators, detection actions, and treatment actions. Attribute parameters for each node are established to clearly define the location of disease symptoms and mechanisms. Step S2 specifically includes the following steps: Step 21: Construct a coupled graph consisting of component nodes, location nodes, disease mechanism nodes, evidence index nodes, detection action nodes, and treatment action nodes. At least seven types of node relationships are established, including location belonging to component, disease mechanism acting on location, evidence index characterizing mechanism, detection action verifying mechanism, treatment action reducing risk, treatment action depending on pre-detection, and treatment action having an incompatible relationship. Step 22: Configure threshold range, applicable conditions, disabled conditions, time validity and source reliability attributes for each node or edge.

[0025] Step 23 involves establishing a set of detection action nodes based on the detection action nodes. This set of detection action nodes originates from the detection action nodes in the coupling graph constructed in Step 21 and serves as a candidate supplementary detection item library. Each detection action node is configured with at least the following attributes: detection item name, applicable components, applicable locations, verifiable disease mechanism, supplementary evidence indicators, detection cost, traffic occupancy cost, structural intrusion cost, implementation conditions, dependency relationships, and mutual exclusion relationships. Furthermore, other nodes or edges are configured with attributes such as threshold range, applicable operating conditions, prohibited operating conditions, time validity, and source reliability.

[0026] By setting up step S2 above, components, parts, and disease mechanisms are modeled together, without using disease names as a single index. This allows for the differentiation of bridge diseases with the same name but different meanings, significantly reducing the probability of the same disease name being misused in different parts, and enabling more accurate search and configuration of detection schemes.

[0027] In this embodiment, step S3, constructing the evidence tuple, specifically includes the following steps: Step S31: Convert each data point in the standardized bridge dataset into a unified evidence tuple based on the nodes of the coupling graph. ,in, Indicates components, Indicates location. Indicates the mechanism of candidate diseases. This represents an observation vector composed of symptom descriptions, numerical indicators, and time-series features. Indicates the observation time. Indicate the source category; and assign an initial reliability based on the source type. . refer to Figure 3 The diagram shown illustrates the calculation of evidence credibility and the propagation of risk status provided in this embodiment. This embodiment uses a unified evidence tuple and consistency calculation mechanism to bring inspection symptoms, quantitative detection values, and monitoring evolution characteristics into the same risk propagation framework, avoiding independent judgment of multiple sources of evidence.

[0028] In this embodiment, step S4 determines the credibility of each evidence tuple based on preset evaluation indicators, such as... Figure 3 The preset evaluation indicators include source reliability, mechanism matching degree, spatial continuity, temporal consistency, and threshold compliance, which solve the problem of the inability to quantify the credibility of multi-source evidence. The specific calculation is as follows: For each tuple of evidence, the computer's matching degree Spatial continuity Time consistency and threshold compliance Thus, the credibility of the evidence is determined: ; in, ω1 to ω5 For preset weights, and ω1 + ω2 + ω3 + ω4 + ω5 =1, i This represents an evidence tuple.

[0029] In this embodiment, step S5 sets a dual penalty item to address the lack of a clear resolution order when there is evidence conflict, thereby improving the stability of engineering diagnosis. The specific candidate risk scores for each candidate risk state node are calculated as follows: Among them, candidate risk state nodes are determined based on the mechanism-location coupling map in step S2. Specifically, for the current evidence tuple Search along the association edges of "evidence indicators - disease mechanism - location - component", and select the disease risk status that has a direct or indirect relationship with the current component, location, and evidence indicators, and whose edge weight or matching degree reaches a preset threshold as candidate nodes to be evaluated; the candidate risk status can be represented as a combination node of "component / location - disease mechanism - development stage or risk level".

[0030] ; Specifically, Represents evidence tuples The support strength for the k-th candidate risk state is determined by the path weights, location matching degree, and mechanism matching degree from the evidence index node to the risk state node in the coupling graph. This represents the penalty for contradictory evidence, used to quantify the reduction in risk score caused by evidence that is contrary to or inconsistent with the k-th risk state. This indicates a penalty for missing prerequisites, used to quantify the deduction of risk score from key test results, key evidence, or necessary prerequisites that must be present but have not yet been obtained for this risk status; λ1 and λ2 are the weights of the corresponding penalty items.

[0031] In this embodiment, the completeness value for evaluating each candidate risk status based on symptom coverage, key indicator coverage, key site coverage, and evidence consistency is calculated as follows: The aforementioned indicators are obtained by matching the risk status evidence template pre-configured in the coupling graph in step S2 with the current evidence tuple generated in step S3. The risk status evidence template includes a set of typical symptoms, a set of essential detection indicators, a set of key locations, and contradictory evidence rules; after comparing the current evidence tuple with each set item by item, the corresponding coverage ratio and consistency score are obtained.

[0032] In this embodiment, the completeness value of each candidate risk status is calculated based on symptom coverage, key indicator coverage, key site coverage, and evidence consistency indicators as follows:

[0033] in, to To preset weights, This indicates symptom coverage, which is the proportion of the typical symptoms of the disease covered by existing evidence. This indicates the coverage of key indicators, that is, the completeness of the essential testing indicators; This indicates the coverage of key areas, i.e., whether the evidence related to key areas of the disease is complete; The score represents the consistency among evidence, i.e. the degree of non-conflict and logical self-consistency among multiple pieces of evidence.

[0034] In one implementation, symptom coverage Key indicator coverage Coverage of key areas ;in , , These represent the set of typical symptoms, the set of essential detection indicators, and the set of key body parts corresponding to the k-th candidate risk state. , , These represent the sets of symptoms, indicators, and locations already covered by the current tuple of evidence. The consistency score between pieces of evidence is also considered. This can be obtained by subtracting the weighting percentage of contradictory evidence from 1. In other words, the more consistent multiple pieces of evidence are in spatial location, temporal order, and threshold direction, the better. The closer it is to 1.

[0035] In this embodiment, if the completeness value of a candidate risk state is lower than a preset threshold, it indicates that the existing evidence is insufficient and supplementary testing planning is required. Step S7 first generates candidate supplementary testing items from the set of testing action nodes. The screening conditions include: there is an association edge between the testing action node and the disease mechanism node corresponding to the current candidate risk state, which is "testing action to verify mechanism"; the testing action can supplement the key evidence indicators that are missing or lack credibility in the current evidence tuple; the testing action meets the requirements of site accessibility, traffic organization, testing equipment, construction window and safety conditions; the testing action does not violate dependency and mutual exclusion relationships and is not covered by the results of completed testing. For candidate supplementary testing items that meet the conditions, a 0-1 variable is used to represent whether to select them, and an optimization model is constructed with the goal of minimizing testing cost, traffic occupation cost and structural intrusion cost, while constraining the total information gain to be no less than a set threshold, as follows: ; ; in, The information gain brought by the j-th detection action to the current risk state is determined comprehensively based on the degree of information supplementation, the correlation strength with candidate risk states, and the ability of the detection result to distinguish different candidate states. To reduce testing costs, For the cost of traffic occupation, Γ represents the structural invasiveness cost, and Γ represents the target information gain threshold.

[0036] In this embodiment, the minimum supplementary detection set is obtained by solving the 0-1 optimization model described above. During the solution process, the dependencies, mutual exclusions, and on-site implementation boundaries between detection items are considered simultaneously. After the minimum supplementary detection set is executed, the newly obtained detection results are returned as new bridge data to step S1 for re-collection, standardization, and generation of evidence tuples, and then re-enter the candidate risk state reasoning and completeness evaluation process. If the completeness still does not reach the preset threshold, supplementary inspection planning continues until the completeness reaches the standard, and then step S8 is executed. Based on the disease mechanism, risk level, and existing preliminary detection results, feasible treatment solutions are initially screened.

[0037] In the supplementary testing phase, testing items are typically selected manually based on experience, which fails to balance information gain, traffic occupancy, and testing intrusion. Furthermore, the treatment recommendation phase lacks sufficient consideration of pre-testing conditions, prohibited construction methods, and construction windows, easily leading to results that are "theoretically feasible but difficult to implement on-site." Therefore, how to complete the boundary verification of pre-testing conditions, prohibited construction methods, and construction windows and traffic organization before treatment output is a problem that this embodiment needs to address. Thus, this embodiment considers jointly verifying pre-testing conditions, prohibited construction methods, and the prohibited relationship between traffic and construction windows to obtain a treatment plan, as detailed below: refer to Figure 4 The diagram shown illustrates the treatment execution boundary verification and post-hoc gating update provided in this embodiment. When the evidence completeness of a candidate risk state reaches or exceeds a preset threshold, a treatment candidate set is generated based on the disease mechanism, risk level, pre-detection results, incompatible relationships, traffic organization conditions, and construction window conditions. An execution score is then calculated for each treatment candidate to obtain the optimal and alternative treatment solutions. The specific calculation formula for each treatment candidate score is as follows: ; in, Δ represents the risk reduction of the p-th treatment plan, used to characterize the percentage decrease in the target risk status score before and after the implementation of the plan. Specifically, it can be calculated as follows: Calculate, R k The risk score is the value of the k-th candidate risk state before treatment. , p This represents the predicted residual risk score after adopting the p-th treatment plan; The accessibility of the p-th treatment plan is used to characterize whether the p-th treatment plan can be successfully implemented under the site conditions. It is obtained by normalized weighting of the degree of satisfaction of the working space, the conditions for equipment entry, the conditions for material supply, the degree of matching of the construction window, and the conditions for safety protection. The durability gain of the p-th treatment scheme is used to characterize the degree of improvement of the p-th treatment scheme on the structural service performance and the ability to suppress disease recurrence. It is obtained by normalized weighting of the expected life extension ratio, the reduction ratio of disease recurrence probability, waterproof and anti-corrosion effect and subsequent maintenance convenience. The traffic impact penalty for the p-th treatment plan is used to characterize the adverse effects of the treatment construction on the bridge's traffic capacity and traffic organization. It is obtained by normalizing and weighting the number of lanes occupied, the duration of lane occupation, whether it affects peak traffic, the increase in detour distance, and the traffic control level. Dis represents the construction disturbance penalty for the p-th treatment scheme, used to characterize the degree of disturbance to the existing structure and surrounding environment during construction. It is obtained by normalizing and weighting the excavation or drilling range, vibration and impact intensity, temporary support requirements, impact on the existing prestressed system, and secondary damage risk. All the above indicators are normalized to the interval [0,1], where ΔR p Fea p and Dur p The larger the value, the better the treatment effect. p and Dis p The larger the value, the higher the implementation cost.

[0038] This embodiment of the method involves ten steps: coupling graph construction, unified expression of evidence tuples, weighted calculation of evidence credibility, reliable risk propagation, evidence completeness determination, information gain-driven solution of the minimum supplementary detection set, treatment boundary verification and scoring, structured output, and posterior gating update. It is equipped with a dedicated system, electronic equipment, and storage media to achieve differentiated identification of diseases with the same name, multi-source evidence fusion to resolve contradictions, intelligent planning of the minimum supplementary inspection items, and output of treatment plans that fit construction and traffic constraints. This enables bridge risk diagnosis to move beyond simple semantic retrieval or black-box prediction, and instead make reliable inferences based on the coupling relationship between mechanisms and locations. It can also proactively plan supplementary inspections when evidence is insufficient and output treatment plans that meet on-site implementation conditions when evidence is sufficient.

[0039] In this embodiment, structured result output is also included, which mainly outputs the risk status, dominant mechanism, component location, key evidence, risk level, supplementary inspection / treatment plan and applicable boundaries.

[0040] This embodiment avoids continuous pollution caused by unverified results directly entering the knowledge base. A verification feedback mechanism ensures secure updates to the knowledge base, significantly improving engineering practicality and diagnostic reliability. Specifically, it also includes the following steps: S9, Verification Feedback Gated Update: The retest effect of the treatment plan after implementation is verified. When the retest result meets the risk score reduction standard, the key disease indicator improvement standard, the monitoring trend stability standard, and the evidence consistency standard, it is judged as meeting the standard, and the corresponding case is written back to the case library. At the same time, the restricted step size update mechanism—location coupling spectrum parameters is updated.

[0041] In one specific embodiment, the results of the retest, monitoring of recovery, and treatment effectiveness are verified post-implementation. The compliance standards include: First, the percentage decrease in the target risk status score after treatment compared to before treatment shall not be less than the preset risk reduction threshold. Second, the key indicators corresponding to the dominant disease mechanism meet the standard limits or reach the preset improvement ratio compared with the original treatment. For example, indicators such as crack width, deflection increment, cable force deviation, support displacement or leakage range meet the preset control requirements. Third, the continuous monitoring data did not show a continuous deterioration trend within the preset observation period, and the fluctuation amplitude was lower than the preset stability threshold; Fourth, there is no new strongly contradictory evidence between the retested evidence and the original diagnostic conclusion, and the evidence consistency score is not lower than the preset consistency threshold.

[0042] Only when all of the above conditions are met will the corresponding case be written back to the historical case database, and the propagation weight, threshold parameter, and treatment effect parameter in the coupling graph will be updated with a limited step size; if any condition is not met, only the implementation record and retest record will be retained, and the core graph parameters will not be updated.

[0043] The implementation process of the method described in this application will be illustrated below through a specific case.

[0044] Step 1. Data Preparation: Collect inspection records, crack observation data, bearing inspection data, cable tension test data, main beam deflection monitoring curves, design drawings, and existing maintenance data for the prestressed concrete continuous box girder bridge. Standardize component names, disease terminology, and location codes for text records; standardize dimensions and align time for test data; extract long-term slope, abrupt change points, fluctuation amplitude, and out-of-limit states from monitoring curves.

[0045] Step 2. Coupled Atlas Construction: Nodes are set for components and locations such as main beams, diaphragms, anchorage zones, webs, top slabs, support zones, and expansion joint zones; nodes are set for disease mechanisms such as prestress loss, shear cracks, temperature shrinkage cracks, incomplete grouting, support failure, and structural alignment anomalies; nodes are set for evidence indicators such as crack width, crack direction, crack location, deflection increment, abnormal opening and closing of expansion joints, cable force attenuation rate, and water seepage; nodes are set for detection actions such as crack re-measurement, cable force verification, endoscopic inspection, ultrasonic testing, and support displacement detection; nodes are set for treatment actions such as sealing repair, secondary grouting, support replacement, local reinforcement, load limiting, and continuous monitoring; further, applicable conditions, threshold ranges, dependency order, and prohibited relationships are configured for each node and edge.

[0046] Step 3. Evidence Tuple Generation: Standardized bridge data includes: longitudinal crack width of 0.32 mm in the web, located near the mid-span; main girder deflection monitoring shows a continuous increasing trend; no obvious shear cracking at the supports; downstream expansion joints are closed, while upstream joint width is increasing. The bridge data is converted into evidence tuples, where the main component is the main girder, and the locations include the web, mid-span, and expansion joint area. Candidate mechanisms include prestress loss, alignment anomalies, and local crack development, sourced from inspection, special testing, and monitoring data, respectively.

[0047] Step 4. Calculation of Evidence Credibility: For evidence of "longitudinal web cracks," calculate the theoretical matching degree based on their location and direction; determine spatial continuity based on the crack distribution at adjacent measuring points; determine temporal consistency based on the results of previous remeasurements; determine threshold compliance based on the crack width threshold; and finally, combine this with the source reliability to obtain... The same method was used to calculate the corresponding values ​​for both the evidence of "continuously increasing deflection" and the evidence of "abnormal expansion joints". .

[0048] Step 5. Risk Status Communication: ... The risk status is propagated to candidate nodes such as "prestress loss risk status," "support anomaly risk status," and "local non-continuous crack risk status." If a candidate status requires cable stress testing as a prerequisite but is currently missing, a prerequisite missing penalty term is introduced; if a status is significantly inconsistent with existing evidence in spatial location, a contradiction penalty term is introduced. The final ranking of each risk status is then obtained.

[0049] Step 6. Evidence Completeness Determination and Supplementary Inspection: If the evidence completeness of the "prestress loss risk status" is lower than the threshold, for example, although there are abnormal deflection and expansion joint abnormalities, but there is still a lack of measured cable force and internal duct information, the system will not directly output the conclusion of support replacement or steel strand treatment. Instead, it will prioritize cable force verification and local endoscopic inspection based on the information gain model. If traffic occupancy restrictions are strict, the system will prioritize the retention of inspection items with short road occupancy time and high information gain among multiple optional inspection actions to form a minimum supplementary inspection set.

[0050] Step 7. Treatment Plan Output: If the main mechanism is confirmed to be prestress loss accompanied by local grouting defects after re-inspection, the treatment plan with the highest execution score is selected from the options such as secondary grouting, local reinforcement, load limitation and continuous monitoring, based on the preconditions, prohibited construction methods and construction window constraints. The output includes the main mechanism, key evidence, risk level, recommended plan, construction methods that are prohibited from being directly used, traffic organization requirements and re-inspection requirements.

[0051] Step 8. Post-hoc gating update: After treatment, if the cable tension retest, deflection retest, and crack propagation rate meet the preset improvement threshold, this case is allowed to be written back to the historical case library, and the relevant propagation weights are updated in small steps; if the improvement is not obvious, only the implementation record is retained, and the core parameters are not updated.

[0052] Device Examples According to embodiments of the present invention, a risk diagnosis and control device for prestressed concrete bridges is provided, such as... Figure 5 The diagram shown is a block diagram of a test device for section evacuation based on subway emergency faults provided in this embodiment. The device for monitoring, displaying, and optimizing interaction in complex scenarios according to this embodiment includes: Multi-source data acquisition and standardization module 10: Used to collect multi-dimensional bridge engineering data and standardize the multi-dimensional bridge engineering data to obtain a standardized bridge dataset.

[0053] Coupled Graph Construction Module 20: Used to construct a mechanism-part coupling graph including nodes of components, parts, disease mechanisms, evidence indicators, detection actions, and treatment actions, and to determine the membership relationships and event influence relationships between nodes, and configure the attribute parameters of each node.

[0054] Evidence tuple construction module 30: used to transform the data in the standardized bridge dataset into unified evidence tuples corresponding to the coupling graph, including at least the dimensions of components, parts, disease mechanisms, and evidence indicators; Consistency calculation module 40: used to determine the credibility of each evidence tuple based on preset evaluation indicators.

[0055] Risk State Credible Propagation Module 50: This module uses candidate risk state nodes in the coupled graph as target nodes. Based on the connection relationship between the evidence tuple and the target node, it introduces propagation weights, contradictory evidence, and missing preconditions to penalize and correct the risk scores. The module then outputs the candidate risk scores of each candidate risk state node through a normalization function.

[0056] Evidence completeness assessment module 60: Used to evaluate the completeness value of each candidate risk status based on symptom coverage, key indicator coverage, key site coverage and evidence consistency indicators. The supplementary detection set solution module 70 is used to determine whether the completeness value of a candidate risk state is lower than a preset threshold. Based on the correlation between detection action nodes and candidate risk states and missing evidence indicators, and combined with on-site accessibility, dependency, and mutual exclusion, it selects candidate supplementary inspection items from the set of detection action nodes. It constructs an optimization model with the goal of minimizing detection cost, traffic occupation cost, and structural invasiveness cost, while constraining the total information gain to be no less than a set threshold. The model solves for the minimum supplementary detection set, and after execution of the minimum supplementary detection set, the newly added detection results are fed back to the multi-source data acquisition and standardization module 10. If the completeness value of a candidate risk state is not lower than the preset threshold, it is fed back to the treatment plan determination module 80. The treatment plan determination module 80 is used to preliminarily screen feasible treatment plans based on the disease mechanism, risk level, and existing preliminary detection results.

[0057] The prestressed concrete bridge risk diagnosis and control device provided in this embodiment addresses the shortcomings of existing bridge defect diagnosis methods, such as lack of component modeling, difficulty in unifying multi-source evidence, reliance on manual supplementary inspections, and treatment plans being detached from on-site constraints. By standardizing multi-source data and constructing a coupled atlas based on components, locations, and defect mechanisms, it significantly reduces the probability of the same defect name being misused in different locations. It also utilizes textual, numerical, and temporal information. This invention, through unified evidence tuples and a consistency calculation mechanism, integrates inspection symptoms, quantitative detection values, and monitoring evolution characteristics into the same risk propagation framework, avoiding independent judgments from multiple sources of evidence. By correcting the risk score of candidate risk state nodes through contradictory evidence penalty items and missing precondition penalty items, it exhibits greater engineering stability than existing schemes that only rank based on similarity. For candidate risk state completeness values ​​below a preset threshold, it automatically calculates the minimum supplementary detection set under constraints of cost, traffic occupancy, and invasiveness, based on information gain, thereby reducing unnecessary detections.

[0058] The coupling graph construction module 20 includes a coupling graph construction submodule and a configuration submodule; The coupling graph construction submodule is used to construct a coupling graph consisting of component nodes, location nodes, disease mechanism nodes, evidence indicator nodes, detection action nodes, and treatment action nodes. It shall establish at least seven types of node relationships, including location belonging to component, disease mechanism acting on location, evidence indicator characterizing mechanism, detection action verifying mechanism, treatment action reducing risk, treatment action depending on pre-detection, and treatment action having a prohibitive relationship. The configuration submodule is used to configure threshold ranges, applicable operating conditions, disabled operating conditions, time validity, and source reliability attributes for each node or edge.

[0059] The detection action node set construction submodule is used to build a detection action node set based on the detection action nodes. The detection action node set originates from the detection action nodes in the constructed coupling graph and is used as a candidate supplementary detection item library. Each detection action node is configured with at least the following attributes: detection item name, applicable components, applicable parts, verifiable disease mechanism, supplementary evidence indicators, detection cost, traffic occupation cost, structural intrusion cost, implementation conditions, dependency relationship, and mutual exclusion relationship; and it is configured with threshold range, applicable working conditions, prohibited working conditions, time validity, and source reliability attributes for other nodes or edges.

[0060] This embodiment models components, locations, and disease mechanisms together, rather than using disease names as a single index. This allows for the differentiation of bridge diseases with the same name but different meanings, significantly reducing the probability of the same disease name being misused in different locations, and enabling more accurate search and configuration of detection schemes.

[0061] In one embodiment, the evaluation indicators set by the consistency calculation module 40 include source reliability, mechanism matching degree, spatial continuity, temporal consistency and threshold compliance indicators, which solves the problem that the credibility of multi-source evidence cannot be quantified.

[0062] In one embodiment, the supplementary detection set solving module 70 constructs an optimization model that aims to minimize detection cost, traffic occupancy cost, and structural intrusion cost, while constraining the total information gain to be no less than a set threshold, as follows: ; ; in, Let be the information gain brought by the j-th detection action in terms of its ability to distinguish the current risk state. To reduce testing costs, For the cost of traffic occupation, Γ represents the intrusive cost, and Γ represents the target information gain threshold.

[0063] In this embodiment, the minimum supplementary detection set is obtained by solving the 0-1 optimization model. During the solution process, the dependencies and mutual exclusion relationships between detection items are considered simultaneously to obtain the optimal minimum supplementary detection set. The optimal minimum supplementary detection set returns to S1 to regenerate the evidence tuple and re-enter the reasoning process until the completeness meets the standard. Then, the feasible treatment plan is initially screened by combining the disease mechanism, risk level, and existing preliminary detection results.

[0064] In this embodiment, the treatment plan determination module 80 is specifically configured to: when the evidence completeness of a candidate risk status reaches or exceeds a preset threshold, generate a set of treatment candidates based on the disease mechanism, risk level, pre-detection results, incompatible relationships, traffic organization conditions, and construction window conditions; then calculate and execute a score for each treatment candidate to obtain the optimal and alternative treatment plans. The specific calculation of the score for each treatment candidate is as follows: ; Wherein, ΔR p This represents the risk reduction margin of the p-th treatment plan, used to characterize the percentage decrease in the target risk status score before and after the implementation of the plan. Specifically, it can be represented by... Calculate, R k The risk score is the value of the k-th candidate risk state before treatment. , p Fea is the predicted residual risk score after adopting the p-th treatment plan. p The accessibility indicator, representing the p-th treatment plan, is used to characterize whether it can be successfully implemented under site conditions. It is obtained by a normalized weighted average of the degree of availability of workspace, equipment access conditions, material supply conditions, construction window matching degree, and safety protection conditions. p The durability gain, representing the degree to which the p-th treatment scheme improves the structural service performance and disease recurrence suppression capability, is obtained by normalizing and weighting the expected life extension ratio, the disease recurrence probability reduction ratio, waterproofing and corrosion protection effect, and subsequent maintenance convenience; Tra p The traffic impact penalty, used to characterize the adverse effects of construction work on bridge capacity and traffic organization, is calculated by a normalized weighted average of the number of lanes occupied, duration of lane occupancy, impact on peak traffic, increase in detour distance, and traffic control level. p The construction disturbance penalty is used to characterize the degree of disturbance to the existing structure and surrounding environment caused by the construction process. It is obtained by normalized weighting of the excavation or drilling range, vibration and impact intensity, temporary support requirements, impact on the existing prestressed system, and secondary damage risk.

[0065] All the above indicators are normalized to the [0,1] interval, where ΔR p Fea p and Dur pThe larger the value, the better the treatment effect. p and Dis p The larger the value, the higher the implementation cost.

[0066] This indicates the degree of risk reduction, construction accessibility, durability gain, traffic impact penalty, and construction disturbance penalty.

[0067] The embodiments of the present invention are device embodiments corresponding to the above method embodiments. The specific operations of each module processing step can be understood with reference to the description of the method embodiments, and will not be repeated here.

[0068] like Figure 6 As shown, the present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the risk diagnosis and control method for prestressed concrete bridges described in the above embodiments.

[0069] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the risk diagnosis and control method for prestressed concrete bridges described in the above embodiments.

[0070] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0071] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for apparatus or system embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. The apparatus and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0072] Furthermore, the functional modules in the various embodiments of this disclosure can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0073] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features. These modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and the contents not described in detail in the specification of the present invention are well known to those skilled in the art.

Claims

1. A method for risk diagnosis and control of prestressed concrete bridges, characterized in that, Includes the following steps: 1) Collect multi-dimensional bridge engineering data and standardize the multi-dimensional bridge engineering data to obtain a standardized bridge dataset; 2) Construct a mechanism-location coupling map including components, locations, disease mechanisms, evidence indicators, detection actions, and treatment action nodes, determine the membership relationships and event influence relationships between nodes, and configure the attribute parameters of each node; 3) Transform the data in the standardized bridge dataset into unified evidence tuples of the corresponding coupled graph, including at least the dimensions of components, locations, disease mechanisms, and evidence indicators; 4) Determine the credibility of each evidence tuple based on preset evaluation indicators; 5) Taking the candidate risk state nodes in the coupling graph as target nodes, based on the connection relationship between the evidence tuples and the target nodes, propagation weights, contradictory evidence, and missing preconditions are introduced to penalize and correct the risk scores. The candidate risk scores of each candidate risk state node are then output through a normalization function. 6) Evaluate the completeness value of each candidate risk status based on symptom coverage, key indicator coverage, key site coverage, and evidence consistency indicators; 7) If the completeness value of the candidate risk state is lower than the preset threshold, select executable detection items from the set of detection action nodes, construct an optimization model with the goal of minimizing detection cost, traffic occupation cost, and structural intrusion cost, while constraining the total information gain to be no less than the set threshold, solve to determine the minimum supplementary detection set and feed it back to step 1); if the completeness value of the candidate risk state is not lower than the preset threshold, combine the disease mechanism, risk level, and existing previous detection results to preliminarily screen feasible treatment solutions.

2. The risk diagnosis and control method for prestressed concrete bridges as described in claim 1, characterized in that, Step 2) specifically includes the following steps: Construct a coupled graph consisting of component nodes, location nodes, disease mechanism nodes, evidence index nodes, detection action nodes, and treatment action nodes. It should at least include the following: location belongs to component, disease mechanism acts on location, evidence index represents mechanism, detection action verifies mechanism, treatment action reduces risk, and treatment action depends on pre-detection and treatment action have an incompatible relationship. A set of detection action nodes is established based on the detection action nodes to serve as a candidate supplementary detection item library. Each detection action node is configured with at least the following attributes: detection item name, applicable components, applicable parts, verifiable disease mechanism, supplementary evidence indicators, detection cost, traffic occupation cost, structural intrusion cost, implementation conditions, dependency relationship, and mutual exclusion relationship. Threshold range, applicable working conditions, prohibited working conditions, time validity, and source reliability attributes are also configured for other nodes or edges.

3. The risk diagnosis and control method for prestressed concrete bridges as described in claim 1, characterized in that, The preset evaluation indicators include source reliability, mechanism matching degree, spatial continuity, temporal consistency, and threshold compliance.

4. The risk diagnosis and control method for prestressed concrete bridges as described in claim 1, characterized in that, Determining whether the completeness value of a candidate risk state is not lower than a preset threshold includes the following steps: Based on the disease mechanism, risk level, preliminary test results, incompatible relationships, traffic organization conditions and construction window conditions, a set of treatment candidates is generated. Then, an execution score is calculated for each treatment candidate to obtain the optimal and alternative treatment solutions.

5. The risk diagnosis and control method for prestressed concrete bridges as described in claim 4, characterized in that, The specific calculation of the score for each treatment candidate is as follows: ； in, This represents the risk reduction margin of the p-th treatment plan, used to characterize the percentage decrease in the target risk status score before and after the implementation of the plan. This indicates the accessibility of the p-th treatment plan. This represents the durability gain of the p-th treatment option. This represents the traffic impact penalty for the p-th treatment plan. This represents the construction disturbance penalty for the p-th treatment plan.

6. The risk diagnosis and control method for prestressed concrete bridges as described in claim 1, characterized in that, It also includes the following steps: The retesting effect of the treatment plan after implementation is verified. When the standard is met, the case library is written back and the mechanism-site coupling map parameters are updated with a limited step size.

7. A risk diagnosis and control device for prestressed concrete bridges, characterized in that, include: Multi-source data acquisition and standardization module: used to collect multi-dimensional bridge engineering data and standardize the multi-dimensional bridge engineering data to obtain a standardized bridge dataset; Coupled graph construction module: used to construct a mechanism-part coupling graph including nodes of components, parts, disease mechanisms, evidence indicators, detection actions, and treatment actions, and to determine the membership relationships and event influence relationships between nodes, and configure the attribute parameters of each node; Evidence tuple construction module: used to transform data in the standardized bridge dataset into unified evidence tuples corresponding to the coupling graph, including at least components, locations, disease mechanisms, and evidence index dimensions; Consistency calculation module: used to determine the credibility of each evidence tuple based on preset evaluation indicators; The credible risk state propagation module is used to take candidate risk state nodes in the coupled graph as target nodes, and to introduce propagation weights, contradictory evidence, and missing preconditions to penalize and correct risk scores based on the connection relationship between evidence tuples and target nodes. The module then outputs the candidate risk scores of each candidate risk state node through a normalization function. The evidence completeness assessment module is used to evaluate the completeness value of each candidate risk status based on symptom coverage, key indicator coverage, key site coverage, and evidence consistency indicators. The supplementary detection set solution module is used to determine if the completeness value of the candidate risk state is lower than the preset threshold. It selects executable detection items from the set of detection action nodes, constructs an optimization model with the goal of minimizing detection cost, traffic occupation cost, and structural intrusion cost, while constraining the total information gain to be no less than the set threshold, solves and determines the minimum supplementary detection set, and feeds it back to the multi-source data acquisition and standardization module. If the completeness value of the candidate risk status is not lower than the preset threshold, feedback is sent to the treatment plan determination module. The treatment plan determination module is used to initially screen feasible treatment plans by combining the disease mechanism, risk level, and existing preliminary test results.

8. The risk diagnosis and control device for prestressed concrete bridges as described in claim 7, characterized in that, The coupling graph construction module includes a coupling graph construction submodule and a configuration submodule; The coupling graph construction submodule is used to construct a coupling graph consisting of component nodes, location nodes, disease mechanism nodes, evidence indicator nodes, detection action nodes, and treatment action nodes. It shall establish at least seven types of node relationships, including location belonging to component, disease mechanism acting on location, evidence indicator characterizing mechanism, detection action verifying mechanism, treatment action reducing risk, treatment action depending on pre-detection, and treatment action having a prohibitive relationship. The configuration submodule is used to configure threshold ranges, applicable operating conditions, disabled operating conditions, time validity, and source reliability attributes for nodes or edges. The detection action node set construction submodule is used to build a detection action node set based on the detection action nodes. The detection action node set originates from the detection action nodes in the constructed coupling graph and is used as a candidate supplementary detection item library. Each detection action node is configured with at least the following attributes: detection item name, applicable components, applicable parts, verifiable disease mechanism, supplementary evidence indicators, detection cost, traffic occupation cost, structural intrusion cost, implementation conditions, dependency relationship, and mutual exclusion relationship; and it is configured with threshold range, applicable working conditions, prohibited working conditions, time validity, and source reliability attributes for other nodes or edges.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the risk diagnosis and control method for prestressed concrete bridges as described in any one of claims 1 to 6.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the risk diagnosis and control method for prestressed concrete bridges as described in any one of claims 1 to 6.

Images