Customized pavement preventive maintenance decision-making method

Abstract

The invention discloses a customized road surface preventive maintenance decision-making method, which comprises the following steps of: 1, establishing a PSSI prediction model by adopting a Bayesian optimization convolutional neural network algorithm, and determining a structural performance detection road section by utilizing a sampling inspection mechanism to obtain road surface detection data of the detection road section; 2, preprocessing the road surface detection data, and removing abnormal values; step 3, establishing a maintenance benefit evaluation model based on the pavement use performance index change condition before and after maintenance, and evaluating the maintenance benefit of the preventive maintenance measures; and step 4, making a pavement maintenance decision based on the customized maintenance benefit evaluation model.

Description

Technical Field

[0001] The present invention belongs to the field of highway maintenance, and in particular relates to a customized road surface preventive maintenance decision-making method. Background Art

[0002] At present, the heavy traffic services of high-grade highways all day long make them face preventive maintenance or special maintenance projects every 3 to 5 years on average. However, due to the variety of maintenance measures and materials, the decision-making of existing maintenance management agencies often relies on the engineering experience of managers and lacks scientific basis. Project-level pavement management mainly focuses on the performance of specific sections or specific areas, emphasizes technical decision-making and management support for specific engineering projects, and aims to support management departments to formulate plans for the precise implementation of maintenance measures for the engineering projects. In order to improve the scientific nature of highway system maintenance decisions, the establishment of decision analysis models should be targeted at specific analysis objects, traffic and environmental conditions, which is specifically reflected in the modeling principle of one model for one road and one model for one type of facility. Therefore, it is crucial to establish a customized model for maintenance benefit analysis and evaluation suitable for project decision-making.

[0003] At present, some pavement maintenance decision-making methods are disclosed in the prior art. For example, in patent CN202310750755.2, a highway asphalt pavement maintenance plan decision-making method based on Bayesian multimodal neural network is disclosed, which constructs a highway pavement maintenance information matrix; selects highway pavement maintenance decision indicators; constructs a multimodal neural network model MNN-EAPMD for highway asphalt pavement maintenance plan decision-making; constructs a Bayesian multimodal neural network decision model BMNN-EAPMD for highway asphalt pavement maintenance plan decision-making; constructs a decision optimization plan based on the probability uncertainty of the maintenance plan, etc.; for example, in patent CN202311496412.4, a pavement maintenance plan based on maintenance big data is disclosed. A performance prediction method and system are used to obtain maintenance big data of a target road, wherein the maintenance big data includes basic road data, historical disease data, historical maintenance data, historical pavement performance detection data, climate and environment data, and traffic load data; based on the maintenance big data, a clustering operation is performed on multiple sub-sections in the target road to divide the target road into multiple section sets; for any of the section sets, based on the maintenance big data corresponding to the section set, a preset long short-term memory network is trained to obtain a prediction model corresponding to the section set; for any sub-section in any of the section sets, the data to be predicted of the sub-section is input into the prediction model corresponding to the section set, and a pavement performance prediction result of the sub-section is obtained as output.

[0004] However, in the prediction and decision-making methods disclosed in the above patents: although the pavement maintenance data is fully utilized, maintenance decisions or pavement performance predictions are only made based on the pavement status data, and pavement maintenance decisions based on maintenance benefit evaluation throughout the life cycle of the pavement are not achieved; at the same time, the content disclosed in the above patents does not analyze the original data of the pavement, and the accuracy of the input values ​​of the prediction model or decision model cannot be achieved by relying solely on historical performance test data. Summary of the invention

[0005] The present invention aims to utilize the pavement maintenance history and detection data to achieve preventive pavement decision-making by establishing a customized maintenance benefit evaluation model. The technical problem solved by the present invention is to provide a customized preventive maintenance decision-making method based on maintenance history and detection data for specific analysis objects, traffic and environmental conditions, and realize one model for one road.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0007] The present invention proposes a customized pavement preventive maintenance decision-making method, comprising the following steps:

[0008] Step 1: Use the Bayesian optimization convolutional neural network algorithm to establish a PSSI prediction model, and use the sampling mechanism to determine the structural performance inspection section;

[0009] Step 2: pre-process the road surface detection data and remove abnormal values;

[0010] Step 3: Based on the changes in pavement performance indicators before and after maintenance, a maintenance benefit evaluation model is established to evaluate the maintenance benefits of preventive maintenance measures;

[0011] Step 4: Make pavement maintenance decisions based on the customized maintenance benefit evaluation model.

[0012] As a further technical solution, the PSSI prediction model makes a graded prediction of the PSSI status of the target road section based on the road surface performance data; two different combinations of indicator parameters are selected as input variables, and the output variable is the PSSI level, so as to construct a multi-feature input and single output multi-classification model.

[0013] As a further technical solution, the PSSI grading standard is as follows: PSSI>92, no maintenance is considered; 85<PSSI≤92 is for preventive maintenance, and PSSI≤85 is for restorative maintenance.

[0014] As a further technical solution, a set of sections to be inspected is screened out according to the PSSI estimation model; sections with estimated structural strength level 3 in the PSSI estimation model results of combination I input parameters are included in the set of sections to be inspected, and sections with estimated structural strength level 2 and below in the PSSI estimation model results of combination II input parameters are included in the set of sections to be inspected.

[0015] As a further technical solution, when the proportion of the road sections to be inspected in the set is 20% or less of the total road sections, the final PSSI inspection sections are the entire set of road sections to be inspected and some sections that are not included in the set of road sections to be inspected, and the overall proportion is 20%.

[0016] As a further technical solution, when the proportion of sections to be inspected is above 20%, three levels of inspection standards are set; ① Level 1: full inspection of the set of sections to be inspected; ② Level 2: 70% inspection of the set of sections to be inspected; ③ Level 3: select some sections from the set of sections to be inspected for inspection, and the inspection ratio accounts for 30% of the total sections.

[0017] As a further technical solution, in step 2: the detection data preprocessing is based on the maintenance time node t1, according to the time t of the pavement performance decay after maintenance i -t1, set a reasonable range of performance data; according to the current maintenance year and the tth after maintenance i -t1 year performance value interval range, derive the 0~(t i -t1) The upper and lower bounds of the performance data during the period, and any data outside the bounds will be considered as outliers and removed.

[0018] As a further technical solution, when 0≤(t i -t1)<3, the upper limit is PPI=100, and the lower limit is PI=85-5 / 2*(t i -t1);

[0019] When (t i -t1)≥3, the upper limit is PPI=100-10 / 7*[(t i -t1)-3]; the lower bound is PI=85-5 / 2*(t i -t1).

[0020] As a further technical solution, the main evaluation indicators in the maintenance benefit evaluation model are the change in performance indicators and the change in performance decay rate.

[0021] As a further technical solution, the calculation method of the performance index change PVQ is as follows:

[0022] (1) The expression form of the pavement performance decay curve is determined. The specific calculation formula is as follows:

[0023] PPI = 100*{1-exp[-(α / t) β ]} (1)

[0024] In the formula, the initial performance value is set to 100, t represents the road age, α and β are model parameters; α is the life factor, which reflects the length of the road service life; β is the mode factor, which reflects the shape of the decay curve;

[0025] (2) Use the test data from previous years to fit and determine the performance natural decay model equation when there is no maintenance;

[0026] (3) Based on the standard decay curve equation without maintenance and the performance test value P0′ one year before maintenance, the relative road age t0 before maintenance is calculated according to formula (2), so as to determine the specific state of the pavement performance in the decay process before maintenance;

[0027]

[0028] (4) Based on the relative road age t0, the maintenance time node t1 is calculated. t1 is the inspection time one year after t0, so the maintenance time node t1;

[0029] t1=t0+1(4)

[0030] (5) Based on the standard decay curve equation without maintenance and the time node t1 of the maintenance year, the predicted performance value P0 of the current maintenance year is calculated;

[0031] P0=100*{1-exp[-(α / t1) β ]} (5)

[0032] (6) The performance index change PVQ is the difference between the performance P1 and P0 before and after the implementation of the maintenance measures, that is,

[0033] PVQ=P1-P0 (6).

[0034] As a further technical solution, the calculation method of the performance decay rate change ΔK is as follows:

[0035] (1) Based on the standard decay curve equation without maintenance, the slope k before maintenance is calculated by taking the derivative at the maintenance time node t1. a

[0036]

[0037] (2) Use the performance value P1 of the year after maintenance and the effective value P1 of the performance test in the following 1 to 3 years. i , the slope k after curing is obtained by linear fitting calculation b

[0038]

[0039] (3) According to k before and after maintenance a and k b The difference in decay rates is calculated

[0040] △K=k b -k a (9).

[0041] The beneficial effects of the present invention are as follows:

[0042] In the maintenance decision-making method proposed in the present invention, a random inspection mechanism is used to carry out structural random inspections to improve the scientific nature of the inspections; at the same time, the maintenance benefit results are calculated based on the maintenance history and inspection data; and a customized evaluation model for the maintenance benefit is established using the maintenance benefit results and maintenance status grouping modeling, which can achieve one model for one path and improve the accuracy of maintenance decisions.

[0043] The present invention predicts the pavement structure strength index (PSSI) based on the historical performance test data, and conducts random inspections on the pavement according to the prediction results to achieve the mastery of the pavement structure performance. The maintenance plan is determined by PSS I classification, which can improve the scientific nature of the test; the maintenance benefit evaluation model is different from the pavement performance prediction model in the prior art (patent CN 202311496412.4) in that the maintenance benefit model can analyze the pavement maintenance benefit and evaluate different maintenance measures. While making full use of pavement test data, the patent of the present invention proposes a random inspection method to master the pavement structure performance and determine the pavement maintenance plan. Furthermore, by establishing a pavement maintenance benefit evaluation model, the maintenance benefits of different maintenance measures are determined, and then preventive maintenance decisions for the pavement are made. The proposal of this decision-making method can significantly improve the scientific nature of pavement decisions, reduce the maintenance cost of the pavement throughout its life cycle, and achieve one model for one road. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a schematic diagram of the sampling inspection process of the present invention;

[0045] Figure 2 Schematic diagram of the changes in road performance before and after maintenance.

[0046] In the figure: a: the test point in the year before the maintenance measures were implemented; a': the test point in the most recent year before the maintenance measures were implemented; b: the test point in the year after the maintenance measures were implemented; c: the test point after the maintenance measures were implemented; P0: the performance value in the year before the maintenance measures were implemented; P0': the performance value in the most recent year before the maintenance measures were implemented; P1: the performance value in the year after the maintenance measures were implemented; P i : Performance value after maintenance measures are implemented; t0: Time node in the most recent year before maintenance implementation; t1: Time node in the year of maintenance implementation; t i: Time node after maintenance implementation; k a : The slope of the original curve at t1 before maintenance; k b : The slope of the curve at t1 after maintenance; PVQ: The change in performance index; △K: The change in decay rate. DETAILED DESCRIPTION

[0047] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used in the present invention have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0048] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless otherwise explicitly stated in the present invention, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "include" and / or "include" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof;

[0049] Glossary: ​​PSSI in the present invention refers to pavement structural strength; PCI refers to pavement damage; RDI refers to pavement rutting depth index; PBI refers to pavement bounce index; SRI refers to pavement skid resistance index;

[0050] This embodiment provides a customized preventive maintenance decision-making method based on maintenance history and detection data for specific analysis objects, traffic and environmental conditions, and implements one model for one road. The specific solution is as follows: Figure 1 As shown, the details are as follows:

[0051] Step 1: Use Bayesian optimization convolutional neural network algorithm to establish PSSI prediction model, and use sampling mechanism to determine the section for structural performance testing. The purpose of establishing PSSI prediction model is to determine the sampling section, and the PSSI classification and related steps based on the results of PSSI prediction model are used to determine the sampling section. The "Highway Technical Condition Assessment Standard" (JTG 5210-2018) stipulates that the pavement structure strength is a sampling detection index, and the sampling detection route or section should be determined according to the needs of pavement maintenance management. The minimum sampling ratio shall not be less than 20% of the highway network maintenance mileage. However, the standard does not provide a specific method for determining the sampling section. The present invention provides a method for determining the sampling section: using the PSSI prediction model, the sections are graded, and the set of sections to be sampled is determined according to the grading results. Finally, the sampling section is determined by the method of the present invention.

[0052] Furthermore, the PSSI prediction model estimates the PSSI status of the target road section based on the road surface performance data. Two different combinations of indicator parameters are selected as input variables, and the output variable is the PSSI level, so as to construct a multi-feature input and single output multi-classification model;

[0053] The input and output variables of the model are shown in Table 1, the PSSI classification standards are shown in Table 2, and the specific steps of the random inspection mechanism for structural performance detection sections are as follows:

[0054] Table 1 Description of input and output variables of PSSI estimation model

[0055]

[0056]

[0057] Table 2 PSS I grading standards

[0058] PSSI Level Level judgment criteria Specific implementation Level 1 PSSI>92 No consideration of maintenance Second level 85<PSSI≤92 Preventive maintenance Level 3 PSSI≤85 Repair and maintenance

[0059] 1) According to the PSS I estimation model, a set of sections to be sampled is selected. The sections with estimated structural strength level 3 in the PSS I estimation model results of combination I input parameters in Table 1 are included in the set of sections to be sampled, and the sections with estimated structural strength level 2 and below in the PSS I estimation model results of combination II input parameters in Table 1 are included in the set of sections to be sampled.

[0060] 2) When the proportion of the road sections to be inspected is 20% or less of the total road sections, the minimum PSS I inspection standard stipulated in the "Highway Technical Condition Assessment Standard" (JTG5210-2018) is 20%. The final PSS I inspection sections are the entire set of road sections to be inspected and some sections that are not included in the set to be inspected, with a total proportion of 20%.

[0061] 3) When the proportion of road sections to be sampled is more than 20%, three levels of sampling standards are set, as follows:

[0062] ① Level 1: Full inspection of the set of sections to be inspected (when the set of sections to be inspected is less than 30%, some sections that are not included in the set to be inspected will be included so that the final inspection ratio is not less than 30%);

[0063] ② Level 2: 70% of the road sections to be inspected are inspected (when the road sections to be inspected are less than 30%, some road sections that are not included in the road sections to be inspected are included so that the final inspection ratio is not less than 30%);

[0064] ③Level 3: Select some sections from the set of sections to be inspected for random inspection, and the inspection ratio accounts for 30% of the total sections.

[0065] Step 2: preprocess the road surface detection data obtained in step 1 to remove abnormal values;

[0066] Furthermore, the detection data preprocessing is based on the maintenance time node t1, and the pavement performance decay time t i -t1, set a reasonable range of performance data; according to the current maintenance year and the tth after maintenance i -t1 year performance value interval range, derive the 0~(t i -t1) The upper and lower bounds of the performance data during the period, and any data outside the bounds will be considered as outliers and removed.

[0067] For example: Specifically, the maintenance of the current year (ie, t i -t1 = 0 years) performance value is specified in the range of 85 to 100, and 10 years after maintenance (i.e. t i -t1=10) is specified to be in the range of 60~90. Based on the performance value range of the current year of maintenance and the 10th year after maintenance, the upper and lower bound formulas for the performance data between 0 and 10 years after maintenance are derived. For data outside this range, it is considered as an outlier and removed. The upper and lower bounds of the performance test data after maintenance are set as shown in Table 3.

[0068] Table 3 PPI upper and lower limits of performance test data after curing

[0069]

[0070] Note: i -t1: the decay time of pavement performance after maintenance calculated based on the maintenance time node t1;

[0071] PPI_0: The corresponding limit is at t i -Performance value at t1=0;

[0072] PPI_10: The corresponding limit is at t i -Performance value when t1=10.

[0073] Step 3: Based on the changes in pavement performance indicators before and after maintenance, a maintenance benefit evaluation model is established to evaluate the maintenance benefits of preventive maintenance measures.

[0074] Furthermore, the main evaluation indicators in the maintenance benefit evaluation model are the change in performance indicators and the change in performance decay rate, specifically:

[0075] 1) Performance Vary Quantity (PVQ)

[0076] The change in pavement performance can directly reflect the effect of maintenance measures on the improvement of pavement performance. The larger the change, the more obvious the effect of the maintenance measure. The maintenance effect can be evaluated by comparing the improvement value of pavement performance before and after maintenance.

[0077] 2) Performance decay rate change △K

[0078] The most intuitive manifestation of the change in the decay curve of pavement performance is the slowdown in the rate of pavement performance attenuation. The maintenance benefit is evaluated by the change in the decay rate of the performance curve before and after maintenance, △K.

[0079] Furthermore, the specific calculation method of the maintenance benefit evaluation index is as follows:

[0080] 1) Calculation method of performance index change PVQ

[0081] (1) Determine the expression form of the pavement performance decay curve. Use the pavement performance standard equation, which conforms to the pavement performance decay law and has a clear physical meaning. The specific calculation formula is shown in formula (1).

[0082] PPI = 100*{1-exp[-(α / t) β ]} (1)

[0083] In the formula, the initial performance value is set to 100, t represents the road age, and α and β are model parameters. α is the life factor, which reflects the length of the road service life; β is the mode factor, which reflects the shape of the decay curve.

[0084] (2) Use the test data from previous years to fit and determine the natural decay model equation of performance without maintenance.

[0085] (3) Based on the standard decay curve equation without maintenance and the performance test value P0′ one year before maintenance, the relative road age t0 before maintenance is calculated according to formula (2), so as to determine the specific state of the pavement performance in the decay process before maintenance.

[0086]

[0087] (4) Based on the relative road age t0, the maintenance time node t1 is calculated. t1 is the inspection time one year after t0, so the maintenance time node t1 is used.

[0088] t1=t0+1(4)

[0089] (5) Based on the standard decay curve equation without maintenance and the time node t1 of the maintenance year, the predicted performance value P0 of the current maintenance year is calculated.

[0090] P0=100*{1-exp[-(α / t1) β ]} (5)

[0091] (6) The performance index change PVQ is the difference between the performance P1 and P0 before and after the implementation of the maintenance measures, that is,

[0092] PVQ=P1-P0 (6)

[0093] 2) Calculation method of performance decay rate change △K

[0094] (1) Based on the standard decay curve equation without maintenance, the slope k before maintenance is calculated by taking the derivative at the maintenance time node t1. a

[0095]

[0096] (2) Use the performance value P1 of the year after maintenance and the effective value P1 of the performance test in the following 1 to 3 years. i , the slope k after curing is obtained by linear fitting calculation b .

[0097]

[0098] (3) According to k before and after maintenance a and k b The difference in decay rates is calculated

[0099] △K=k b -k a (9)

[0100] Step 4: Make pavement maintenance decisions based on the customized maintenance benefit evaluation model.

[0101] Based on the maintenance history and test data, the maintenance benefit evaluation model can be used to determine the performance index change PVQ and the performance decay rate change △K of the performance in the year after the implementation of different maintenance measures, forming the maintenance benefit evaluation result of a certain highway. According to the current test results, combined with the performance index change PVQ and the performance decay rate change △K, appropriate maintenance measures can be selected.

[0102] The customized preventive maintenance decision-making method based on maintenance history and detection data proposed in this embodiment establishes a customized evaluation model for maintenance benefits, which is used to analyze the applicability of different maintenance measures to a certain highway, thereby realizing one model for each highway.

[0103] The preventive maintenance benefit evaluation of the present invention is further described in detail below through specific examples.

[0104] Taking the project indicator PCI as an example, the maintenance benefit analysis is carried out, and the benefit evaluation results of four preventive maintenance measures under different pavement performance conditions are displayed by statistical distribution characteristics. The following table shows the corresponding maintenance benefit evaluation results.

[0105] Table 4 PCI maintenance benefit evaluation results

[0106]

[0107]

Claims

1. A customized pavement preventive maintenance decision-making method, characterized in that: The following steps are involved: Step 1: Use the Bayesian optimization convolutional neural network algorithm to establish a PSSI prediction model, use the sampling mechanism to determine the structural performance test section, and obtain the road surface test data of the test section; Step 2: pre-process the road surface detection data and remove abnormal values; Step 3: Based on the changes in pavement performance indicators before and after maintenance, a customized maintenance benefit evaluation model is established to evaluate the maintenance benefits of preventive maintenance measures; Step 4: Make pavement maintenance decisions based on the customized maintenance benefit evaluation model.

2. The customized road preventive maintenance decision-making method according to claim 1, characterized in that: The PSSI prediction model makes a graded prediction of the PSSI status of the target road section based on the road surface performance data. Two different combinations of indicator parameters are selected as input variables, and the output variable is the PSSI level, thereby constructing a multi-feature input and single-output multi-classification model.

3. The customized road preventive maintenance decision-making method according to claim 1, characterized in that: The PSSI grading standard is as follows: PSSI>92, no maintenance is considered; 85<PSSI≤92 is for preventive maintenance, and PSSI≤85 is for restorative maintenance.

4. The customized road preventive maintenance decision-making method according to claim 1, characterized in that: The road sections to be inspected are selected based on the PSSI estimation model; It is stipulated that the third-level road sections estimated in the PSSI estimation model results of combination I input parameters are included in the set of sections to be sampled, and the second-level and below road sections estimated in the PSSI estimation model results of combination II input parameters are included in the set of sections to be sampled. When the proportion of the road sections to be inspected in the total road sections is 20% or less, the sections that PSSI finally inspects are the entire road sections to be inspected and some sections that are not included in the road sections to be inspected, and their overall proportion is 20%.

5. The customized road preventive maintenance decision-making method according to claim 2, characterized in that: When the proportion of sections to be inspected is above 20%, three levels of inspection standards are set, namely: ① Level 1: full inspection of the set of sections to be inspected; ② Level 2: 70% inspection of the set of sections to be inspected; ③ Level 3: select some sections from the set of sections to be inspected for inspection, and the inspection ratio accounts for 30% of the total sections.

6. The customized road preventive maintenance decision-making method according to claim 2, characterized in that: In the step 2, the detection data preprocessing is based on the maintenance time node t1, and the road performance decay time t i -t1, set a reasonable range of performance data; according to the current maintenance year and the tth after maintenance i -t1 year performance value interval range, derive the 0~(t i -t1) The upper and lower bounds of the performance data during the period, and any data outside the bounds will be considered as outliers and removed.

7. The customized road preventive maintenance decision-making method according to claim 5, characterized in that: When 0≤(t i -t1)<3, the upper limit is PPI=100, and the lower limit is PI=85-5 / 2*(t i -t1); When (t i -t1)≥3, the upper limit is PPI=100-10 / 7*[(t i -t1)-3]; the lower bound is PI=85-5 / 2*(t i -t1).

8. The customized road preventive maintenance decision-making method according to claim 1, wherein: The main evaluation indicators in the maintenance benefit evaluation model are the change in performance indicators and the change in performance decay rate.

9. The customized road preventive maintenance decision-making method according to claim 8, characterized in that: The calculation method of the performance index change PVQ is as follows: (1) The expression form of the pavement performance decay curve is determined. The specific calculation formula is as follows: PPI=100*{1-exp[-(α / t) β ]} (1) In the formula, the initial performance value is set to 100, t represents the road age, α and β are model parameters; α is the life factor, which reflects the length of the road service life; β is the mode factor, which reflects the shape of the decay curve; (2) Use the test data from previous years to fit and determine the performance natural decay model equation when there is no maintenance; (3) Based on the standard decay curve equation without maintenance and the performance test value P0′ one year before maintenance, the relative road age t0 before maintenance is calculated according to formula (2), so as to determine the specific state of the pavement performance in the decay process before maintenance; (4) Based on the relative road age t0, the maintenance time node t1 is calculated. t1 is the inspection time one year after t0, so the maintenance time node t1; t0=t0+1(4) (5) Based on the standard decay curve equation without maintenance and the time node t1 of the maintenance year, the predicted performance value P0 of the current maintenance year is calculated; P0=100*{1-exp[-(α / t1) β ]} (5) (6) The performance index change PVQ is the difference between the performance P1 and P0 before and after the implementation of the maintenance measures, that is, PVQ = P1-P0 (6).

10. The customized road preventive maintenance decision-making method according to claim 8, wherein: The calculation method of the performance decay rate change ΔK is as follows: (1) Based on the standard decay curve equation without maintenance, the slope k before maintenance is calculated by taking the derivative at the maintenance time node t1. a (2) Use the performance value P1 of the year after maintenance and the effective value P1 of the performance test in the following 1 to 3 years. i , the slope k after curing is obtained by linear fitting calculation b (3) According to k before and after maintenance a and k b The difference in decay rates is calculated △K=k b -k a (9)。

Images