A method and device for post-evaluation of road maintenance projects

Abstract

The application provides a pavement maintenance project post-evaluation method and device, and relates to the technical field of highway maintenance, which is used to improve the accuracy, comprehensiveness and decision guidance of the post-evaluation of the pavement maintenance project. The method comprises the following steps: determining that a to-be-evaluated pavement maintenance project has evaluation qualifications based on a minimum effect tracking period; the minimum effect tracking period represents the shortest time required to obtain a post-evaluation result; inputting multi-modal data of the to-be-evaluated pavement maintenance project into a post-evaluation quantitative model system to obtain evaluation results of each evaluation index; determining a comprehensive index corresponding to an evaluation mode based on the evaluation mode, the evaluation results of each evaluation index and the weights of each evaluation index; and determining a post-evaluation result corresponding to the evaluation mode according to the comprehensive index.

Description

Technical Field

[0001] This application relates to the field of highway maintenance technology, and in particular to a post-evaluation method and apparatus for road maintenance projects. Background Technology

[0002] Post-construction evaluation of road maintenance projects is a systematic assessment of the effectiveness, benefits, and impacts of completed maintenance work. It is a crucial step in optimizing future maintenance decisions and improving the efficiency of maintenance fund utilization. Depending on the evaluation object, post-construction evaluation of highway maintenance can be divided into four categories: post-construction evaluation of highway maintenance policies, post-construction evaluation of highway maintenance decisions, post-construction evaluation of highway maintenance technologies, and post-construction evaluation of highway maintenance projects. Among these, post-construction evaluation of maintenance technologies assesses the effectiveness of the promotion and application of a specific maintenance technology (such as micro-surfacing or thin-layer overlay), confirming the actual applicability and effectiveness of the technology for different types of road defects and road conditions.

[0003] Existing post-evaluation methods for road maintenance projects, which construct a three-level evaluation index system, have drawbacks such as vague definition of evaluation objects, lack of clear time window requirements, single data source with insufficient processing capacity, strong subjectivity and static invariance in weight determination, incomplete evaluation dimensions, and disconnect between evaluation results and decision-making. Summary of the Invention

[0004] This application provides a method and apparatus for post-evaluation of road maintenance projects, which improves the accuracy, comprehensiveness and decision-making guidance of post-evaluation of road maintenance projects.

[0005] In a first aspect, this application provides a post-evaluation method for road maintenance projects, the method comprising: The minimum effect follow-up period is used to determine whether a road maintenance project is eligible for evaluation; the minimum effect follow-up period represents the shortest time required to obtain the post-evaluation results. The multimodal data of the road maintenance project to be evaluated are input into the post-evaluation quantitative model system to obtain the evaluation results of each evaluation index; The comprehensive index corresponding to the evaluation model is determined based on the evaluation model, the evaluation results of each evaluation indicator, and the weight of each evaluation indicator; the evaluation model includes at least one of the following: road condition evaluation model, economic evaluation model, environmental evaluation model, technology applicability evaluation model, and comprehensive evaluation model. The post-evaluation result corresponding to the evaluation model is determined based on the comprehensive index.

[0006] Optionally, determining whether a road maintenance project qualifies for evaluation based on the minimum effect tracking period includes: The minimum effect tracking period is determined based on the project type, evaluation mode, highway grade, and traffic grade of the road maintenance project to be evaluated. Calculate the difference between the current time and the completion time of the road maintenance project to be evaluated; If the difference is greater than or equal to the minimum effect tracking period, then the road maintenance project to be evaluated is determined to be qualified for evaluation.

[0007] Optionally, before determining the comprehensive index corresponding to the evaluation model based on the evaluation model, the evaluation results of each evaluation indicator, and the weight of each evaluation indicator, the method further includes: The basic weights of each evaluation indicator are determined based on the principle of minimizing total deviation. Time-sensitive weighted multimodal data are input into a random forest-based weighted prediction model to obtain correction coefficients for each evaluation index; the multimodal data includes at least one of the following: historical data before maintenance, pavement maintenance engineering design data, pavement maintenance construction record data, pavement inspection data after maintenance, traffic load data, and meteorological environment data; The weights of each evaluation indicator are obtained based on the basic weights and corresponding correction coefficients of each evaluation indicator.

[0008] Optionally, after determining the post-evaluation result corresponding to the evaluation mode based on the comprehensive index, the method further includes: The post-evaluation results are transformed into entries in the pavement maintenance measures library and / or pavement maintenance scheme library; wherein, each entry in the pavement maintenance measures library includes basic information about the measures, an applicable condition matrix, effect parameters, and sample project statistics; each entry in the pavement maintenance scheme library includes basic information about the scheme, scenario matching rules, scheme content, and expected comprehensive effect; The maintenance decision knowledge graph is updated based on the pavement maintenance measures library and / or pavement maintenance scheme library.

[0009] Optionally, the method further includes: If the post-evaluation result meets any of the triggering conditions, the feedback mechanism is triggered; The triggering conditions include: The applicability assessment level is medium or poor; The effect of the new technology application is improved by the preset percentage compared with the existing data in the road maintenance measures database; The difference in the effect of the same technology under the same conditions reaches a preset difference threshold; The cumulative number of samples has reached a preset threshold. The expected achievement level did not meet the design requirements; The multi-source data consistency index is less than the preset consistency threshold; The feedback mechanism includes at least one of the following: Update the road maintenance measures database; Optimize the road maintenance solution database; Generate technical reports; Push decision-making suggestions; Push the results of the expected achievement analysis; Data quality warning.

[0010] Optionally, the post-evaluation quantification model system includes at least one of the following models: Road surface performance degradation model, before-and-after maintenance effect comparison model, multi-source data correlation analysis model, expected and actual effect comparison model, technology applicability evaluation model, economic assessment model and environmental assessment model.

[0011] Secondly, this application also provides a post-evaluation device for road maintenance projects, the device comprising: The evaluation eligibility determination module is used to determine whether a road maintenance project to be evaluated is eligible for evaluation based on the minimum effect follow-up period; the minimum effect follow-up period represents the shortest time required to obtain the post-evaluation results. The evaluation index assessment module is used to input the multimodal data of the road maintenance project to be evaluated into the post-evaluation quantitative model system to obtain the evaluation results of each evaluation index. The comprehensive index determination module is used to determine the comprehensive index corresponding to the evaluation mode based on the evaluation mode, the evaluation results of each evaluation indicator, and the weight of each evaluation indicator. The evaluation result determination module is used to determine the post-evaluation result corresponding to the evaluation mode based on the comprehensive index.

[0012] Thirdly, this application also provides an electronic 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 post-evaluation method for road maintenance engineering as described in any of the above.

[0013] Fourthly, this application also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the post-evaluation method for road maintenance engineering as described in any of the above.

[0014] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the post-evaluation method for road maintenance engineering as described in any of the above.

[0015] This application provides a post-evaluation method for road maintenance projects. First, based on the minimum effect tracking period, it determines whether the road maintenance project qualifies for evaluation. Then, for qualified projects, a systematic post-evaluation quantitative model is used to perform the evaluation process on the identified evaluation objects. The final post-evaluation result is determined based on the obtained comprehensive index. This method can improve the accuracy, comprehensiveness, and decision-making guidance of post-evaluation of road maintenance projects. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A schematic flowchart illustrating a post-evaluation method for road maintenance engineering provided in this application embodiment; Figure 2 A schematic diagram illustrating the overall process of post-evaluation of road maintenance engineering provided for an embodiment of this application; Figure 3 A schematic diagram of the structure of a post-evaluation device for road maintenance engineering provided in an embodiment of this application; Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] All actions involving the acquisition of signal information or data in this application are carried out in accordance with the relevant data protection laws and policies of the country where the application is located, and with the authorization granted by the owner of the relevant device.

[0020] In the embodiments of this application, "multiple" refers to two or more. Terms such as "first" and "second" are used only for descriptive purposes and should not be construed as indicating or implying relative importance or order.

[0021] Post-construction evaluation of road maintenance projects is a systematic assessment of the effects, benefits, and impacts of completed maintenance work. It is a crucial step in optimizing future maintenance decisions and improving the efficiency of maintenance fund utilization. However, existing post-construction evaluation methods for road maintenance projects still have the following technical shortcomings: (1) The evaluation object is vaguely defined and lacks a clear time window requirement. Current technologies generally fail to clearly define the eligibility criteria for evaluation subjects, and do not specify how long data must be accumulated after the completion of maintenance projects before post-evaluation. This ambiguity leads to the following problems: some evaluations are conducted immediately after maintenance completion, reflecting only short-term effects and failing to assess the long-term durability of the maintenance project; some evaluation data lacks sufficient time span to form a complete "performance decay-recovery-re-decay" curve, resulting in evaluation results that lack scientific rigor and persuasiveness.

[0022] (2) Limited data sources and insufficient processing capacity: Existing methods mainly rely on structured inspection data (such as flatness, damage rate, deflection value, etc.), and make insufficient use of unstructured data such as construction logs, on-site images, and UAV inspection videos. These unstructured data contain rich information about the construction process and clues to potential quality problems, but due to their heterogeneous formats and complex semantics, they are difficult to be effectively used by automated assessment systems, resulting in incomplete assessment information.

[0023] (3) The determination of weights is highly subjective and static: The weights of indicators are mostly determined by expert scoring methods such as the Analytic Hierarchy Process (AHP) and Dematel, which have obvious subjective biases. At the same time, once the weights are determined, they remain unchanged for a long time and cannot be dynamically adjusted according to factors such as engineering characteristics, regional differences, and climate conditions, which leads to deviations between the evaluation results and the actual situation.

[0024] (4) Incomplete assessment dimensions: Existing assessments focus on the immediate recovery effect of technical conditions, lacking a comprehensive consideration of the life cycle cost, long-term performance degradation patterns, and environmental benefits such as carbon emissions, making it difficult to meet the requirements of green transportation and sustainable development in the new era.

[0025] (5) Disconnect between assessment results and decision-making: Assessment conclusions are usually presented in the form of reports, failing to form a structured knowledge base. Post-assessment results cannot be automatically transformed into guidance for future maintenance decisions, resulting in information silos between "assessment" and "decision-making", making it difficult to achieve a closed-loop improvement of "assessment-decision-optimization".

[0026] Figure 1 A schematic flowchart illustrating a post-evaluation method for road maintenance engineering provided in this application embodiment; Figure 2 This application provides a schematic diagram of the overall process for post-evaluation of road maintenance engineering; the following is in conjunction with... Figure 1 and Figure 2 The technical solution of this application will now be introduced.

[0027] like Figure 1 As shown, the method includes the following steps: Step 110: Determine whether the road maintenance project to be evaluated is eligible for evaluation based on the minimum effect tracking period.

[0028] Before conducting a formal assessment, the pavement maintenance project to be assessed must first be qualified to ensure it meets the minimum follow-up period for post-assessment, thereby guaranteeing the scientific validity and effectiveness of the assessment results. The minimum follow-up period represents the shortest time required to obtain post-assessment results; that is, the shortest time span necessary to obtain reliable post-assessment results after the completion of the maintenance project. Determining this time span requires comprehensive consideration of multiple factors, including the type of maintenance project, assessment model, performance degradation patterns, and data collection frequency.

[0029] In one possible implementation, step 110, determining whether the pavement maintenance project to be evaluated qualifies for evaluation based on the minimum effect tracking period, specifically includes the following steps: Step 111: Determine the minimum effect tracking period based on the project type, assessment model, highway grade, and traffic grade of the road maintenance project to be evaluated.

[0030] First, determine the project type, assessment model, highway grade, and traffic grade of the road maintenance project to be assessed.

[0031] Identify the type of road maintenance work to be evaluated. The type of work must include at least preventative maintenance, restorative maintenance, and specialized maintenance. Specifically: (1) Preventive maintenance projects: such as micro-surfacing, slurry seal, fog seal, etc., are mainly used to restore the functional performance of the road surface.

[0032] (2) Restorative maintenance projects: such as milling and repaving, overlay, etc., are mainly used to restore the pavement structure performance.

[0033] (3) Special maintenance projects: such as bridge reinforcement, tunnel repair, and traffic safety facility upgrades.

[0034] Based on the assessment objective and available data, select one or more of the following assessment models: (1) Road condition assessment mode: This mode focuses on the improvement effect of maintenance projects on road surface performance and is a basic assessment mode. This mode has low requirements for visual data and mainly relies on structured detection data and daily inspection data. It is suitable for ordinary national and provincial highways with limited data collection conditions.

[0035] (2) Economic evaluation model: focuses on the input-output benefits and life-cycle cost analysis of maintenance projects. It requires detailed cost data and traffic volume data and is suitable for investment benefit analysis and optimal allocation of maintenance funds.

[0036] (3) Environmental assessment model: focuses on the carbon emissions and energy conservation and emission reduction benefits of maintenance projects, requires data such as material usage, construction technology, and energy consumption, and is suitable for green highway construction and "dual carbon" target assessment.

[0037] (4) Technology Applicability Assessment Mode: This mode focuses on assessing the applicability of specific maintenance technologies to different types of road defects and road conditions, including the technology's service life, unit cost, economic benefits, and optimal application scenarios. This mode is a new core assessment mode added in this invention and is used to construct and update the maintenance measures and solutions library.

[0038] (5) Comprehensive evaluation model: It takes into account all factors such as technology, economy, environment, process, and applicability of technology. It requires the most comprehensive data support and is suitable for demonstration projects, scientific research projects or comprehensive performance evaluation.

[0039] Highway grades include expressways, Class I highways, Class II highways, and Class III and below highways; traffic grades include light traffic, medium traffic, heavy traffic, and extremely heavy traffic.

[0040] Then, based on the project type, evaluation model, highway grade, and traffic grade of the road maintenance project to be evaluated, the minimum effect tracking period is determined. Specifically, the minimum effect tracking period... The calculation formula is: = In the formula, The baseline tracking period is based on the project type; This is an adjustment factor based on the evaluation model; This is an adjustment factor based on highway grade; This is an adjustment factor based on traffic level.

[0041] Different types of road maintenance projects have different performance degradation patterns, requiring differentiated tracking times. Specifically, the baseline tracking period based on the type of maintenance project is shown in Table 1.

[0042] Table 1

[0043] Different evaluation models have significantly different requirements for the data time span. The adjustment coefficients based on the evaluation model are shown in Table 2.

[0044] Table 2

[0045] The adjustment coefficients based on highway grade and traffic grade can be set according to actual needs. For example, the adjustment coefficient for highway grade can be set to 1 for expressways, 1.1 for Class I highways, 1.2 for Class II highways, and 1.3 for Class III and below. The adjustment coefficient based on traffic grade can be set to 1 for light traffic, 1.1 for moderate traffic, 1.2 for heavy traffic, and 1.3 for extremely heavy traffic.

[0046] For example, Table 3 shows the recommended minimum effect tracking period based on project type and evaluation mode.

[0047] Table 3

[0048] Step 112: Calculate the difference between the current time and the completion time of the road maintenance project to be evaluated.

[0049] Current time With completion time The difference is determined by the completion date in the project completion acceptance report.

[0050] Step 113: If the difference is greater than or equal to the minimum effect tracking period, then the road maintenance project to be evaluated is deemed to be qualified for evaluation.

[0051] If the current time With completion time The difference is greater than or equal to the minimum effect tracking period corresponding to the selected mode. If so, the road maintenance project to be evaluated is deemed eligible for evaluation.

[0052] For example, for a preventive maintenance project on a secondary highway completed in June 2022, if the "Road Condition Assessment Mode" is selected for assessment in June 2024, the minimum follow-up period is typically ≥1.5 years. Since the current difference of 2 years meets the requirement, it is deemed qualified for assessment.

[0053] In addition, when determining whether a road maintenance project is eligible for evaluation, besides the minimum effective tracking period, the determination also includes whether the periodic testing data after maintenance meets the frequency requirements corresponding to the highway grade. That is, a road maintenance project that meets the following conditions is considered eligible for evaluation: (1) Current time With completion time The difference is greater than or equal to the minimum effect tracking period corresponding to the selected mode. .

[0054] (2) The periodic testing data after maintenance meet the frequency requirements corresponding to the highway grade.

[0055] If the conditions are not met, the system will prompt "Insufficient data accumulation, not yet meeting the conditions for post-evaluation" and provide a suggested waiting time or supplementary data collection requirements.

[0056] Step 120: Input the multimodal data of the road maintenance project to be evaluated into the post-evaluation quantitative model system to obtain the evaluation results of each evaluation index.

[0057] Before evaluating each indicator, corresponding multimodal raw data is collected according to the evaluation model of the road maintenance project to be evaluated. Then, the collected multimodal raw data undergoes quality cleaning, removing data that does not meet preset quality thresholds. Next, features are extracted from the cleaned unstructured data, converting it into structured feature vectors, and then spatiotemporally aligned and fused with the structured data to form a multimodal fusion dataset oriented towards the evaluation model. Specifically: Specifically, based on the assessment model of the road maintenance project to be evaluated, corresponding multimodal raw data is collected. This multimodal raw data includes six categories: pre-maintenance historical data, road maintenance project design data, road maintenance construction record data, post-maintenance road inspection data, traffic load data, and meteorological environmental data. For example, Table 4 is the data requirement matrix for the assessment model provided in this application, where □ indicates required data; ο indicates optional data that can be collected according to specific circumstances.

[0058] Table 4

[0059] Specifically, the data collection time range requirements are as follows: 1. Historical data prior to maintenance should include at least the following items: (1) Periodic testing data: Testing data for three consecutive years before the commencement of the maintenance project (if applicable).

[0060] (2) Daily inspection records: Daily inspection records for the year prior to the commencement of the maintenance project.

[0061] (3) Minor repair and maintenance records: Repair records for the two years prior to the commencement of the maintenance project.

[0062] 2. Post-maintenance road surface inspection data should include at least the following items: (1) Periodic inspection data: Starting from the first year after completion, data will be collected at the frequency required by the highway grade, for example: Highways / Class I roads: Twice a year (once before and once after the rainy season).

[0063] Secondary highways: once a year.

[0064] Highways of Class III and below: once every 2 years.

[0065] (2) Daily inspection records: at least once a month, record the development of the disease.

[0066] (3) Minor repair and maintenance records: Record the actual occurrences, including the repair location, repair measures, repair time, and repair cost.

[0067] The collected multimodal raw data is cleaned to remove data that does not meet the preset quality threshold, including: 1. Integrity check: (1) Set differentiated integrity thresholds based on different highway grades: For highways / first-class roads: samples with a missing rate of more than 10% for key fields were removed. For secondary highways: Samples with a missing rate of more than 15% for key fields were removed.

[0068] For highways of level 3 and below: Samples with a missing key field rate exceeding 20% ​​will be removed.

[0069] (2) For samples with a missing rate below the threshold, use Lagrange interpolation or linear interpolation to fill in the missing data.

[0070] 2. Outlier Detection and Handling: The 3σ principle is used to identify outliers in structured data: |xi-μ|>3σ Where μ is the mean and σ is the standard deviation. Identified outliers require manual verification or substitution with adjacent values.

[0071] Feature extraction and structuring of cleaned unstructured data includes: 1. Feature extraction from periodic detection data: The periodic detection data from previous years were organized into a time series matrix: = Where i=1,2,...,m are the detection time points, and j=1,2,...,n are the detection indicators (RQI, DR, RD, SFC, etc.).

[0072] 2. Feature extraction from daily inspection data: Extract disease development characteristics from daily inspection records: (1) When the disease first appears t first .

[0073] (2) Disease development rate ,in, This represents the change in the number of defects per unit length.

[0074] (3) Evolution trajectory of disease severity.

[0075] 3. Extract feature data such as maintenance cost characteristics and repair frequency characteristics from minor repair and maintenance records: (1) Annual maintenance frequency .

[0076] (2) Annual maintenance cost per unit mileage .

[0077] (3) The degree of matching between the maintenance location and the type of disease.

[0078] (4) Duration of repair effect.

[0079] In this embodiment of the application, the post-evaluation quantitative model system includes at least one of the following models: pavement performance degradation model, pre- and post-maintenance effect comparison model, multi-source data correlation analysis model, expected and actual effect comparison model, technology applicability evaluation model, economic evaluation model, and environmental evaluation model. Specifically: 1. Road surface performance degradation model: To scientifically evaluate the long-term effects of pavement maintenance projects, a pavement performance degradation model was established. An improved exponential degradation model was adopted: in: for Road surface performance indicators at any given time; These are the initial performance indicators after maintenance; This represents the long-term stable performance value (decay limit). This is the decay rate constant, reflecting how fast the performance degrades; The random error term follows a normal distribution. .

[0080] The model parameters are estimated using the nonlinear least squares method, and the objective function is: decay rate constant The estimated value It is a core indicator for evaluating the long-term effectiveness of road maintenance technology. The smaller the value, the better the performance is maintained.

[0081] 2. Before and after maintenance effect comparison model Establish a multi-dimensional, multi-time-point comparison model of the effects before and after maintenance to systematically evaluate the actual effect of the maintenance project.

[0082] 2.1 Real-time improvement rate of a single indicator For the j-th road condition indicator, define the instant improvement rate. in: These are the values ​​of the indicators from the final test before maintenance. This is the first test result after maintenance (usually within 1-3 months after completion).

[0083] 2.2 Single Indicator Durability Index Considering the detection data at multiple time points during the tracking period, a persistence index is defined. in, The index value is the value of the i-th test after maintenance, and n is the number of tests during the follow-up period.

[0084] 2.3 Comparison of performance degradation curves The degradation curve after maintenance is compared with the historical degradation curve before maintenance to define a degradation improvement index. in: The rate of performance degradation before maintenance. The rate of performance degradation after maintenance. This indicates that the decay has been slowed down and the maintenance is effective; This indicates accelerated decay and poor maintenance.

[0085] 2.4 Overall Maintenance Effectiveness Index Overall Maintenance Effect Index in, , , This serves as a benchmark (the best performance of similar projects can be taken as a reference). , , These are the weighting coefficients. The weights are the indicator weights.

[0086] 3. Multi-source data correlation analysis model Establish a multi-source data correlation analysis model to explore the inherent correlation between detection data, daily inspection records, and minor maintenance records, and achieve cross-verification between data.

[0087] 3.1 Correlation Analysis between Routine Inspection Frequency and Road Condition Deterioration Define the frequency of disease detection during routine inspections. Correlation coefficient with road condition decay rate k: in, For covariance, and The standard deviation is denoted as . This indicates that the more defects found during inspections, the faster the road conditions deteriorate, which aligns with expected patterns and suggests high data reliability; if If not, there may be data quality issues, and a review is required.

[0088] 3.2 Correlation Analysis between Minor Maintenance Frequency and Road Condition Deterioration Define the frequency of disease detection during routine inspections. Correlation coefficient with road condition decay rate k: Road sections with high maintenance frequency show faster road condition deterioration, consistent with the pattern of "maintenance reflecting problems," indicating high data reliability; if If the value is close to 0 or negative, the repair record may be incomplete or the repair effect may be significant, requiring further analysis.

[0089] 3.3 Spatial Consistency Verification of Repair Locations and Defect Analysis Spatial matching is performed between the repair locations recorded in minor maintenance records and the defect locations in the inspection data, and a spatial consistency index S is defined: in, The number of times the repair location is matched with the location of the defect. This represents the total number of repairs. S≥80% indicates good data consistency; S<50% indicates potential data recording issues.

[0090] 3.4 Multi-source data consistency index Multi-source data consistency index in, , , These are the weighting coefficients. The higher the value, the stronger the correlation between multiple data sources, and the higher the data reliability. When the data quality falls below a preset threshold (e.g., 0.5), the system issues a data quality warning, prompting manual review.

[0091] 4. Expected-actual effect comparison model Establish a comparison model of expected and actual results to assess whether the maintenance project meets the design requirements.

[0092] 4.1 Performance indicator achievement For the j-th performance metric, define the achievement degree. in, This represents the actual performance indicators at time t after maintenance. These are the concurrent performance metrics expected in the design documents.

[0093] 4.2 Service life achievement Predicting actual service life based on pavement performance degradation model With respect to the design service life In comparison, define lifespan achievement. Among them, the actual service life is predicted by the decay model when the performance degrades to a critical value. Time: 4.3 Comprehensive Index of Expected and Actual Results Expected-actual effect composite index in, and This is the baseline value (which can be 90% or 100%). and These are the weighting coefficients.

[0094] 5. Applicability Evaluation Model for Road Maintenance Technology Establish a technology applicability evaluation model to systematically evaluate the applicability of road maintenance technologies under different conditions.

[0095] 5.1 Applicability and Matching Degree For a given pavement maintenance technology m and application conditions c (including damage type, traffic level, climate zone, etc.), define applicability matching. in, Let m be the center value of the optimal applicability range of the technique in the k-th conditional dimension; This represents the actual value of condition c in the k-th condition dimension. is the weight coefficient for the k-th conditional dimension.

[0096] 5.2 Comprehensive Index of Technical Effectiveness The comprehensive technical effectiveness index takes into account service life, performance retention rate, and cost-effectiveness. in, This refers to the actual service life. Based on the service life; This represents the actual decay rate. The baseline decay rate; For the actual cost-benefit ratio, Based on the cost-benefit ratio; , , Let be the weighting coefficient, satisfying .

[0097] 5.3 Technical Reliability Index Based on multi-project statistics, a technical reliability index is defined. in, Let m be the average of the performance indicators of technology across multiple projects. The standard deviation is denoted as . The larger the value, the more stable the technical effect.

[0098] 6. Economic Evaluation Model Establish an economic evaluation model to calculate the economic benefits of road maintenance projects.

[0099] 6.1 Total life cycle cost Total life cycle cost : in: Initial maintenance costs; denoted as t, the maintenance cost in year t (obtained from minor repair and maintenance records); r is the discount rate (usually 4%-6%); T is the analysis period (10-15 years); SV is the residual value.

[0100] 6.2 Cost-benefit ratio Cost-effectiveness The calculation formula is: in, The benefits in year t include user benefits (fuel savings, time savings) and maintenance department benefits (reduced subsequent maintenance costs).

[0101] 7. Environmental Assessment Model Establish an environmental assessment model to quantify the carbon emissions and energy conservation and emission reduction benefits of road maintenance projects.

[0102] 7.1 Carbon emissions during the construction period Carbon emissions during construction for: in: Let k be the amount of material used. Let k be the carbon emission factor of the k-th material; This represents the consumption of the l-th type of energy. The carbon emission factor for the lth energy source.

[0103] 7.2 Energy saving and emission reduction benefits during operation Energy saving and emission reduction benefits during operation The calculation formula is: in, and The figures are fuel consumption per unit mileage (liters / km) before and after maintenance. Let be the traffic volume in year t; The length of the road section to be maintained; It is a carbon emission factor for fuel.

[0104] Step 130: Determine the comprehensive index corresponding to the evaluation model based on the evaluation model, the evaluation results of each evaluation indicator, and the weight of each evaluation indicator.

[0105] The method for determining the weights of each evaluation indicator includes the following steps: Step 1301: Determine the basic weights of each evaluation index based on the principle of minimizing total deviation.

[0106] (1) Construct the objective dominance matrix For m evaluation objects and n evaluation indicators, construct an objective dominance matrix. , where represents the overall advantage of evaluation object i relative to evaluation object k, and is calculated using the following formula: in, To evaluate the normalized value of object i on index j, Based on the weights.

[0107] (2) Construct the subjective preference matrix Construct a subjective preference matrix based on expert experience. ,in This indicates that experts believe they have a high degree of preference for evaluation object i relative to evaluation object k, satisfying the following conditions: ∈[-1,1] and .

[0108] (3) Construct the optimization model Define the scaling difference function The goal is to minimize the sum of scale differences in pairwise comparisons of all evaluated objects, and an optimization model is constructed to solve for the basic weights. : Constraints: Step 1302: Input the time-sensitive weighted multimodal data into the weight prediction model based on random forest to obtain the correction coefficients of each evaluation index.

[0109] (1) Timeliness empowerment To differentiate the impact of the timeliness of different data on weight adjustment, timeliness-based weighting is applied to the input data: For static baseline data exceeding 12 months, assign a timeliness weight τ=0.5. For real-time detection data within the past 3 months, a timeliness weight τ=1.0 is assigned. For data falling between these two extremes, timeliness weights are calculated using linear interpolation: in, This represents the number of months since the data was created.

[0110] (2) Constructing a weighted prediction model A weighted prediction model is constructed using the random forest algorithm. The model input is the feature vector X. include: 1) Engineering characteristics: highway grade, traffic grade, climate zone, pavement type 2) Performance characteristics: current performance index values, performance degradation rate 3) Data characteristics: data integrity, data timeliness, and multi-source data consistency index Q. The model output is a vector of dynamic adjustment coefficients for the weights of each indicator. ,in (3) Mathematical expression of the random sensen model A random forest consists of B decision trees. For a new input X, the model output is the average of the outputs of all decision trees. in, This is the output of the b-th decision tree.

[0111] Step 1303: Based on the basic weights and corresponding correction coefficients of each evaluation indicator, obtain the weights of each evaluation indicator.

[0112] In calculating the weights of each evaluation indicator, this application incorporates subjective preference information and objective test data, and calculates the basic weights based on the principle of minimizing total deviation. At the same time, it constructs a weight prediction model based on random forest, inputs real-time multimodal data with time-sensitive weighting, and outputs dynamic correction coefficients. The basic weights and dynamic correction coefficients are then integrated to obtain the final dynamic comprehensive weights of each evaluation indicator.

[0113] Step 140: Determine the post-evaluation results corresponding to the evaluation model based on the comprehensive index.

[0114] Based on the selected evaluation model, the corresponding quantitative model is applied to calculate the comprehensive index, which is then compared with a preset level threshold to generate an evaluation level. Specifically: 1. Calculation of Comprehensive Index for Road Condition Assessment Model The road condition assessment model's comprehensive index takes into account maintenance effectiveness, durability, and the degree to which expectations are met. in: The comprehensive index of maintenance effect is calculated based on the comprehensive index of maintenance effect in the above-mentioned before-and-after maintenance effect comparison model. calculate; The durability index is the decay improvement index in the above-mentioned pre- and post-maintenance effect comparison model. calculate; The expected-actual effect composite index is calculated based on the expected-actual effect comparison model described above. calculate; , , It is a dynamic comprehensive weight.

[0115] 2. Calculation of the comprehensive index of the technology suitability assessment model in: The actual service life is predicted using a performance degradation model; The actual decay rate is estimated using the above-mentioned model comparing the effects before and after maintenance. The actual cost-benefit ratio is calculated from the cost-benefit ratio (BCR) in the economic evaluation model. The technical reliability index is derived from the technical reliability index in the pavement maintenance technology applicability evaluation model. calculate; , , , Let be the weighting coefficient, satisfying 3. Calculation of the composite index of the economic evaluation model in, , It is a dynamic comprehensive weight.

[0116] 4. Calculation of the comprehensive index of the environmental assessment model in, , It is a dynamic comprehensive weight.

[0117] 5. Calculation of the comprehensive index of the all-factor comprehensive evaluation model in, , , , For dynamic comprehensive weighting, satisfying .

[0118] After obtaining the comprehensive index, the post-evaluation results corresponding to the evaluation model are determined based on the comprehensive index. For example, the evaluation results are divided into multiple levels, and the evaluation level of the road maintenance project to be evaluated is determined based on the comprehensive index of the road maintenance project to be evaluated and the range of each level.

[0119] Furthermore, after obtaining the composite index, its credibility can be corrected based on the multi-source data consistency index Q: in, is the original composite index, and Q is the multi-source data consistency index. When... When the value is less than 0.5, the system issues a data quality warning, prompting manual review.

[0120] According to the revised composite index The value is used to classify the post-project evaluation results of road maintenance into multiple levels, for example: Furthermore, after determining the post-evaluation results corresponding to the evaluation model based on the comprehensive index, the process also includes: converting the post-evaluation results into entries in the pavement maintenance measures library and / or pavement maintenance scheme library; wherein each entry in the pavement maintenance measures library includes basic information about the measures, an applicable condition matrix, effect parameters, and sample project statistics; each entry in the pavement maintenance scheme library includes basic scheme information, scenario matching rules, scheme content, and expected comprehensive effect; and updating the maintenance decision knowledge graph based on the pavement maintenance measures library and / or pavement maintenance scheme library.

[0121] Based on the above evaluation results, a road maintenance measures database and a road maintenance scheme database are constructed and dynamically updated to form a closed-loop mechanism of "evaluation-accumulation-optimization-application".

[0122] 1. Construction of a road maintenance measures database The results of the technology suitability assessment model are transformed into structured entries for a pavement maintenance measure library. Each measure library entry includes the following information: (1) Basic information of the measures Measure ID: Unique Identifier Measures name: such as "micro-surfacing", "thin-layer coating", etc. Measures can be categorized as: preventative maintenance, restorative maintenance, and specialized maintenance. Technical standards: The standard or specification number on which it is based (2) Applicable Condition Matrix

[0123] (3) Effect parameter library Expected service life: decay rate constant: Immediate improvement rate: Sustained effect index: Unit cost: Yuan / square meter Results-to-Benefit Ratio: Technical reliability index: Expected achievement rate: (4) Sample project statistics Number of sample projects N Effect distribution histogram parameters Mean of multi-source data consistency index 2. Construction of a road maintenance solution database (1) Basic information of the plan Solution ID: Unique Identifier Scheme Name: For example, "Preventive Maintenance Scheme for Asphalt Pavement of Secondary Highways" Applicable Scenarios Description Solution Components: Included maintenance measures and their combinations (2) Scene matching rules Highway classification criteria: Expressway / Class I Highway / Class II Highway / Class III and below Road surface type: Asphalt concrete / Cement concrete Traffic volume range: [Q] min Q max ] Disease type combinations: [Disease type code list] Road Condition Index Range: [PQI] min PQI max ] (3) Content of the plan

[0124] (4) Expected overall effect Expected comprehensive benefit index: Expected service life: Year Total life cycle cost: Yuan / square meter / year Expected achievement rate: 3. Dynamic updates of the measure base based on knowledge graphs A knowledge graph for maintenance decision-making is constructed, with a pavement maintenance measures database and a maintenance plan database as its core components. Entity node V: includes disease type, maintenance measures, material type, applicable conditions, effect parameters, expected achievement, etc.; Relation edge E: includes relations such as "applies to", "treats", "has", "is superior to", and "does not apply to". Knowledge reasoning is performed using Graph Convolutional Networks (GCNs). The inter-layer propagation rules of GCNs are as follows: Once the newly evaluated engineering data is entered, the system automatically: (1) Compare the effects of new technology applications with existing data in the measures database; (2) When the confidence level of a new sample reaches the threshold, update the effect parameters (mean, standard deviation) of the measure library. (3) When new applicable conditions or abnormal effects are discovered, an early warning is generated and a manual review is requested; (4) Based on the accumulated sample data, dynamically adjust the priority of recommended measures; (5) Feed back the results of the expected achievement analysis to the design department to optimize the design requirements of future projects.

[0125] In one possible implementation, a feedback mechanism is triggered if the post-evaluation result meets any of the triggering conditions, wherein the triggering conditions include: (1) The technical suitability assessment level is medium or poor.

[0126] (2) The effect of the new technology application is improved by a preset percentage compared with the existing data in the road maintenance measures database, for example, the effect improvement is >20%.

[0127] (3) The difference in effect of the same technology under the same conditions reaches the preset difference threshold, for example, the coefficient of variation > 0.3.

[0128] (4) The cumulative number of samples reaches the preset threshold, for example, from N<10 to N≥30.

[0129] (5) The expected achievement rate did not meet the design requirements. For example, the expected achievement rate A j <80% or A T <80% indicates that the design requirements have not been met.

[0130] (6) The consistency index of multi-source data is less than the preset consistency threshold. For example, if the consistency index of multi-source data Q < 0.5, it indicates that there may be quality problems with the data.

[0131] The feedback mechanism includes at least one of the following: (1) Update the pavement maintenance measures library, for example, recalculate the mean and standard deviation of the effect parameters.

[0132] (2) Optimize the road maintenance scheme library, for example, by adjusting the priority and applicable condition threshold of the scheme recommendations.

[0133] (3) Generate technical reports, for example, summarize the effects of technology application and propose improvement suggestions.

[0134] (4) Push decision-making suggestions, for example, push optimized scheme recommendation rules to the maintenance decision-making system.

[0135] (5) Push the results of the expected achievement analysis. For example, push the results of the expected achievement analysis to the design department to optimize the design standards and requirements of future projects.

[0136] (6) Data quality warning, for example, when Q<0.5, prompt the data collection department to review the relevant data.

[0137] Example 1: Post-evaluation of a preventive maintenance project involving micro-surfacing of asphalt pavement on a secondary highway (road condition assessment model) This embodiment takes a micro-surfacing preventive maintenance project on a secondary highway from K100+000 to K110+000 as an example and uses a road condition assessment model for evaluation.

[0138] S0: Qualification of Evaluation Subjects Project type: Microsurfacing preventative maintenance Highway grade: Class II highway Completion date: June 2022 Assessment date: June 2024; follow-up period: 2 years. Minimum effect tracking period calculation: The preventive maintenance road condition assessment model requires ≥1.5 years, and the actual tracking period is 2 years, which meets the requirements.

[0139] Design requirements: Design service life of 5 years, expected RQI ≥ 85 S1: Data Acquisition and Preprocessing Collect testing data, construction logs, daily inspection records, minor repair and maintenance records from 2021 to 2024, and perform quality cleaning and feature extraction.

[0140] S2: Application of Quantization Models S2.1 Fitting of pavement performance degradation model Using the S2.1 model, the decay rate constant k = 0.16, the initial performance P0 = 92.5, and the decay limit P∞ = 75.0 were obtained by fitting.

[0141] S2.2 Comparison of effects before and after maintenance Immediate improvement rate: RQI increased from 82.5 to 92.5, an improvement rate of 12.1%. Sustained Effectiveness Index: The average RQI during the tracking period was 90.2, an increase of 9.3% compared to before the maintenance. Decay Improvement Index: Decay rate k before maintenance 前 =0.22k, k after curing 后 After reaching 0.16k, the decay rate slowed by 27.3%. S2.3 Correlation Analysis of Multi-Source Data Correlation coefficient ρ between routine inspection frequency and decay rate F,k =0.72 The correlation coefficient ρ between the frequency of minor maintenance and the decay rate M,k =0.68 The spatial consistency index between maintenance location and disease distribution is S=85%. The comprehensive consistency index of multi-source data is Q=0.75. S2.4 Comparison of Expected and Actual Results Two years after maintenance, the actual RQI was 88.5, the expected RQI was 87.0, and the achievement rate was 101.7%. The predicted actual service life is 5.2 years, while the designed service life is 5 years, achieving 104% of the target life. S3: Dynamic Weight Calculation The basic weights are calculated based on the principle of minimizing total deviation, and the dynamic comprehensive weights are obtained by combining time-sensitive weighting.

[0142] S4: Comprehensive Index Calculation and Rating Overall maintenance effect index E 效果 =0.86 Durability index D = 0.82 Expected-actual effect composite index E 预期 =0.92 Road condition assessment composite index E 路况 =0.86 Data credibility correction E 修正=0.86×(0.7+0.3×0.75)=0.86×0.925=0.80 Assessment grade: Good S5: Closed-loop feedback The evaluation results show that the micro-surfacing technology was applied effectively in this project, with high consistency among multi-source data and an expected achievement rate exceeding 100%. The system will incorporate the evaluation results into the measure library and update the effectiveness parameters of the micro-surfacing technology.

[0143] Example 2: Verification Case of Correlation Analysis of Multi-Source Data This example demonstrates the application effect of a multi-source data correlation analysis model.

[0144] Engineering Background A highway asphalt pavement restoration and maintenance project was tracked for 3 years, and the following data were collected: Periodic testing data: twice a year, for a total of 6 testing data. Routine inspection records: once a month, totaling 36 inspection records. Minor repair and maintenance records: Recorded based on actual occurrences, totaling 24 repair records. Correlation analysis results

[0145] Model calculation The correlation coefficient ρ between inspection frequency and decay rate F,k =0.89 The correlation coefficient ρ between maintenance frequency and decay rate M,k =0.91 The comprehensive consistency index of multi-source data is Q=0.87. in conclusion High correlation indicates good consistency among multi-source data and high data reliability. In particular, the K5-K10 and K15-K20 sections show high maintenance frequency and rapid data decay, consistent with the pattern of "maintenance reflecting problems"; while the K0-K5 and K10-K15 sections show low maintenance frequency and slow data decay, consistent with the pattern of "good maintenance". This analysis verifies the reliability of the evaluation model.

[0146] Example 3: Comparison of Expected and Actual Results and Design Feedback Case Study This example demonstrates the application of the expected-actual effect comparison model and the feedback to the design department.

[0147] Engineering Background In the past three years, a certain province has implemented five thin-layer overlay maintenance projects, all of which are designed to have a service life of 6 years and an expected RQI of ≥80 after 5 years.

[0148] Summary of evaluation results

[0149] Multi-source data correlation analysis Multi-source data correlation analysis of Project #03 revealed: The project is located on a high-traffic section of road, and the actual traffic volume exceeded the design expectations by 30%. Correlation coefficient ρ between routine inspection frequency and decay rate F,k =0.85 The correlation coefficient ρ between maintenance frequency and decay rate M,k =0.88 The comprehensive consistency index of multi-source data is Q=0.82. Closed-loop feedback system analysis revealed that the main reason Project #03 failed to meet design requirements was that actual traffic volume exceeded design expectations. System-triggered feedback mechanism: Update the measures library: Add a "traffic volume correction factor" to the parameters for thin-layer overlay effect. Optimize the solution library: Add traffic volume matching requirements to the solution recommendation rules. Feedback to the design department: It is recommended to raise design standards for high-traffic sections or adopt higher-level maintenance measures. Based on feedback, the design department revised the "Design Guidelines for Maintenance Engineering of Heavy Traffic Sections", adjusting the expected service life of heavy traffic sections to 5 years and the design index to RQI≥78.

[0150] This application provides a post-evaluation method for pavement maintenance projects. By establishing a scientific definition mechanism for the minimum effect tracking period, it fundamentally solves the problems of vague definition of evaluation objects and only reflecting short-term effects in existing technologies. By constructing a systematic quantitative evaluation system covering pavement performance degradation models, multi-time-point comparison models before and after maintenance, evaluation models of expected and actual effect achievement, and multi-source data correlation analysis models, it achieves a leap from single-point comparison to full life cycle trend analysis, from subjective judgment to data-driven approaches, and from isolated detection data to cross-verification of multi-source data such as daily inspection records and minor maintenance frequency. Among them, the positive correlation analysis between daily maintenance frequency and pavement condition degradation rate can effectively verify the reliability of the evaluation results. Through the pioneering technology applicability evaluation model, the post-evaluation results are precipitated into a structured maintenance knowledge base containing basic information of measures, an applicability condition matrix, and an effect parameter library. Combined with knowledge graph reasoning, it realizes intelligent recommendation of maintenance measures, dynamic optimization of the scheme library, and closed-loop feedback from the design department, ultimately forming a complete closed loop of "evaluation-precipitation-optimization-application-feedback", which significantly improves the scientificity, accuracy, and decision-guiding value of pavement maintenance post-evaluation.

[0151] The following describes a post-evaluation device for road maintenance engineering provided by the present invention. The post-evaluation device for road maintenance engineering described below can be referred to in correspondence with the post-evaluation method for road maintenance engineering described above.

[0152] Figure 3 A schematic diagram of the structure of a post-evaluation device for road maintenance engineering provided in this application embodiment is shown below. Figure 3 As shown, the device 300 includes: The evaluation eligibility determination module 310 is used to determine whether the road maintenance project to be evaluated is eligible for evaluation based on the minimum effect follow-up period; the minimum effect follow-up period represents the shortest time required to obtain the post-evaluation results. The evaluation index assessment module 320 is used to input the multimodal data of the road maintenance project to be evaluated into the post-evaluation quantitative model system to obtain the evaluation results of each evaluation index. The comprehensive index determination module 330 is used to determine the comprehensive index corresponding to the evaluation mode based on the evaluation mode, the evaluation results of each evaluation indicator, and the weight of each evaluation indicator. The evaluation result determination module 340 is used to determine the post-evaluation result corresponding to the evaluation mode based on the comprehensive index.

[0153] It should be noted that the post-evaluation device for road maintenance engineering provided in this application embodiment can realize all the method steps implemented in the above-mentioned post-evaluation method embodiment for road maintenance engineering, and can achieve the same technical effect. Here, the parts that are the same as those in the method embodiment and the beneficial effects will not be described in detail.

[0154] Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4 As shown, the electronic device may include a processor 410, a communications interface 420, a memory 430, and a communication bus 440. The processor 410, communications interface 420, and memory 430 communicate with each other via the communication bus 440. The processor 410 can call logical instructions from the memory 430 to execute post-evaluation methods for road maintenance projects.

[0155] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0156] On the other hand, this application also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to perform the post-evaluation method for road maintenance engineering provided by the above methods.

[0157] In another aspect, this application also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform the post-evaluation methods for road maintenance projects provided by the methods described above.

[0158] The device 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 any creative effort.

[0159] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the post-evaluation method of road maintenance engineering described in various embodiments or some parts of embodiments.

[0160] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application 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 of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A post-evaluation method for road maintenance projects, characterized in that, The method includes: The minimum effect follow-up period is used to determine whether the road maintenance project to be evaluated is eligible for evaluation; the minimum effect follow-up period represents the shortest time required to obtain the post-evaluation results. The multimodal data of the road maintenance project to be evaluated are input into the post-evaluation quantitative model system to obtain the evaluation results of each evaluation index; The comprehensive index corresponding to the evaluation model is determined based on the evaluation model, the evaluation results of each evaluation indicator, and the weight of each evaluation indicator; the evaluation model includes at least one of the following: road condition evaluation model, economic evaluation model, environmental evaluation model, technology applicability evaluation model, and comprehensive evaluation model. The post-evaluation result corresponding to the evaluation model is determined based on the comprehensive index.

2. The post-evaluation method for road maintenance projects according to claim 1, characterized in that, The determination of whether a road maintenance project qualifies for evaluation based on the minimum effect tracking period includes: The minimum effect tracking period is determined based on the project type, evaluation mode, highway grade, and traffic grade of the road maintenance project to be evaluated. Calculate the difference between the current time and the completion time of the road maintenance project to be evaluated; If the difference is greater than or equal to the minimum effect tracking period, then the road maintenance project to be evaluated is determined to be qualified for evaluation.

3. The post-evaluation method for road maintenance projects according to claim 1, characterized in that, Before determining the comprehensive index corresponding to the evaluation model based on the evaluation model, the evaluation results of each evaluation indicator, and the weights of each evaluation indicator, the method further includes: The basic weights of each evaluation indicator are determined based on the principle of minimizing total deviation. Time-sensitive weighted multimodal data are input into a random forest-based weighted prediction model to obtain correction coefficients for each evaluation index; the multimodal data includes at least one of the following: historical data before maintenance, pavement maintenance engineering design data, pavement maintenance construction record data, pavement inspection data after maintenance, traffic load data, and meteorological environment data; The weights of each evaluation indicator are obtained based on the basic weights and corresponding correction coefficients of each evaluation indicator.

4. The post-evaluation method for road maintenance projects according to claim 1, characterized in that, After determining the post-evaluation result corresponding to the evaluation mode based on the comprehensive index, the method further includes: The post-evaluation results are transformed into entries in the pavement maintenance measures library and / or pavement maintenance scheme library; wherein, each entry in the pavement maintenance measures library includes basic information about the measures, an applicable condition matrix, effect parameters, and sample project statistics; each entry in the pavement maintenance scheme library includes basic information about the scheme, scenario matching rules, scheme content, and expected comprehensive effect; The maintenance decision knowledge graph is updated based on the pavement maintenance measures library and / or pavement maintenance scheme library.

5. The post-evaluation method for road maintenance projects according to claim 1, characterized in that, The method further includes: If the post-evaluation result meets any of the triggering conditions, the feedback mechanism is triggered; The triggering conditions include: The applicability assessment level is medium or poor; The effect of the new technology application is improved by the preset percentage compared with the existing data in the road maintenance measures database; The difference in the effect of the same technology under the same conditions reaches a preset difference threshold; The cumulative number of samples has reached a preset threshold. The expected achievement level did not meet the design requirements; The multi-source data consistency index is less than the preset consistency threshold; The feedback mechanism includes at least one of the following: Update the road maintenance measures database; Optimize the road maintenance solution database; Generate technical reports; Push decision-making suggestions; Push the results of the expected achievement analysis; Data quality warning.

6. The post-evaluation method for road maintenance projects according to claim 1, characterized in that, The post-evaluation quantitative model system includes at least one of the following models: Road surface performance degradation model, before-and-after maintenance effect comparison model, multi-source data correlation analysis model, expected and actual effect comparison model, technology applicability evaluation model, economic assessment model and environmental assessment model.

7. A post-evaluation device for road maintenance projects, characterized in that, The device includes: The evaluation eligibility determination module is used to determine whether a road maintenance project to be evaluated is eligible for evaluation based on the minimum effect follow-up period; the minimum effect follow-up period represents the shortest time required to obtain the post-evaluation results. The evaluation index assessment module is used to input the multimodal data of the road maintenance project to be evaluated into the post-evaluation quantitative model system to obtain the evaluation results of each evaluation index. The comprehensive index determination module is used to determine the comprehensive index corresponding to the evaluation mode based on the evaluation mode, the evaluation results of each evaluation indicator, and the weight of each evaluation indicator. The evaluation result determination module is used to determine the post-evaluation result corresponding to the evaluation mode based on the comprehensive index.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the post-evaluation method for road maintenance projects as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the post-evaluation method for road maintenance projects as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the post-evaluation method for road maintenance projects as described in any one of claims 1 to 6.

Images