Asphalt pavement maintenance decision-making method and system considering multi-dimensional influence factors

Abstract

The invention relates to an asphalt pavement maintenance decision-making method and system considering multi-dimensional influence factors, belongs to the technical field of pavement maintenance, constructs an asphalt pavement performance multi-source data set, designs an average weight fusion importance analysis method, and aims to solve the problem that association analysis among pavement influence factor multi-source data is insufficient. Feature engineering operation including feature generation, selection and conversion is carried out on the data, a problem domain is represented more accurately, and the model efficiency is improved. After part of the features are subjected to coding and feature dimension reduction, feature importance sorting is performed on a multi-source feature data set by applying multiple statistical and machine learning methods, and a feature set having significant influence on asphalt pavement performance prediction is screened out according to a sorting result; and accurately capturing and predicting the change trend of the pavement performance based on an OOA optimized fusion prediction model. The current performance state and the performance prediction trend are comprehensively considered, and decision-making of the maintenance level and the specific maintenance mode is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of pavement maintenance, and in particular to an asphalt pavement maintenance decision-making method and system taking multi-dimensional influencing factors into consideration. Background Art

[0002] In order to keep the highway in good condition and prevent its quality from deteriorating, highway maintenance is necessary. In recent years, my country's expressway construction has developed rapidly. With the increase in highway mileage, the workload of maintenance, inspection and management after the completion of the highway has also increased sharply.

[0003] Early research on asphalt pavement performance prediction by domestic and foreign scholars was usually limited by the amount of available data, and often relied on empirical regression analysis, probability statistical models, and other classic prediction methods. With the advancement of technology and the accumulation of data, researchers have gradually turned to using machine learning technology and neural network methods for multivariate data analysis to achieve more accurate predictions. The above exploration and research provide important theoretical basis and reference value for this study. However, current research still has problems such as incomplete influencing factors and insufficient prediction accuracy, and more in-depth research is needed on these issues. Summary of the invention

[0004] The present invention aims to solve the above-mentioned technical problems and provide an asphalt pavement maintenance decision-making method and system taking into account multi-dimensional influencing factors.

[0005] In order to solve the above technical problems, the technical solution provided by the present invention is:

[0006] A method for asphalt pavement maintenance decision-making considering multi-dimensional influencing factors includes the following steps:

[0007] S1. Asphalt pavement data integration and data quality improvement:

[0008] Collect and integrate the evaluation indicators and multi-source influencing factors that need to be considered in the performance evaluation of asphalt pavement. The evaluation indicators are PCI, RQI, RDI and SRI. The multi-source influencing factors are analyzed and characterized to lay a data foundation for the construction of data-driven performance prediction models and maintenance decisions.

[0009] S2. Importance analysis of factors affecting asphalt pavement performance based on average weighted multi-method fusion:

[0010] The evaluation indicators and multi-source influencing factors of asphalt pavement performance are screened, and the prediction performance is maintained while streamlining the model input. A large amount of feature data is standardized and feature coding is implemented. The importance of the factors affecting the performance of asphalt pavement is analyzed, and the most predictive feature set is found according to the importance ranking results.

[0011] S3. Establishment of OOA optimization fusion model for asphalt pavement:

[0012] Establish a predictive OOA optimization fusion model to predict and analyze asphalt pavement performance;

[0013] S4. Establishment of multi-source data set of maintenance history:

[0014] Construct a multi-source data set of maintenance history and establish a maintenance decision-making method based on the inverse variance weighted improved random forest model to improve the accuracy of maintenance decisions;

[0015] S401: Analyze the historical data of expressways, the PSSI of structural strength is higher than 90, and there is no need to maintain the structural strength index; therefore, the pavement indexes are selected as PCI, RDI, RQI and SRI; after merging the collected historical maintenance data and the data of unmaintained sections, a maintenance method classification data set is constructed to make decisions on 7 maintenance methods; a maintenance level classification data set is constructed to make decisions on 3 maintenance levels;

[0016] S402: Construct a multi-index maintenance method dataset including historical data and unmaintained road sections. Based on the PCI, RDI, RQI and SRI indicator data, the dataset adds multi-source features including pavement disease data, road basic information data and traffic load data to improve the dataset's ability to represent maintenance decision-making issues;

[0017] S5. Decision-making method:

[0018] Construct a multi-source data set of maintenance history, and establish a maintenance decision-making method based on the inverse variance weighted improved random forest model to improve the accuracy of maintenance decisions, and use the prediction results and current situation to select the maintenance decision method.

[0019] S501: Construct the OOB data of the tree in the random forest method. The OOB data is an independent validation set for each tree that does not participate in the training. During the training process, each tree will use random sampling with replacement to select data from the original training set as training data when it is established. Usually, the number of selected data is the same as the size of the original training set. Due to the sampling with replacement, some data will be repeatedly selected, while others may not be selected; the data that is not selected constitutes the OOB data of this tree;

[0020] S502: The prediction of each decision tree contained in the random forest is given the same weight to calculate the final prediction result; when using the inverse variance weighting method IVW (, a weight is assigned to the prediction of each tree, and the weight of each tree is inversely proportional to its prediction error on the validation set.

[0021] The basic principle of inverse variance weighting is that measurements with high credibility (small error) should have greater influence on the integration results, and vice versa, measurements with large errors should have less influence. Variance reflects the uncertainty of a measurement value and can be considered as the degree of data fluctuation. In the task of maintenance decision-making, the decision error rate on the validation set is used as the basis for weight allocation, and the weight of each tree is the normalized value of the inverse of the decision error rate. The calculation formula is as follows:

[0022]

[0023] Where n represents the number of decision trees in the random forest algorithm; w i represents the weight of the i-th tree; E i Represents the decision error rate of the i-th tree.

[0024] Each tree votes on the decision result, and after inverse variance weighting, the result with the highest vote is taken as the final result of the model to obtain the highway maintenance decision-making method.

[0025] The multi-source influencing factors include climate, traffic flow, road infrastructure information, disease and maintenance history data.

[0026] The importance of factors affecting asphalt pavement performance is analyzed by using PCA and Pearson analysis methods to achieve feature dimensionality reduction. The importance analysis method based on the fusion of Spearman, Permutation and SHAP average weights is used to rank the importance of multi-source factors affecting asphalt pavement performance as the basis for subsequent data selection.

[0027] The aforementioned factors affecting asphalt pavement performance include historical detection data (PCI, PQI, RDI, SRI), basic road information (road age, material, structure), climate data (temperature, precipitation), traffic flow (equivalent, passenger-to-freight ratio, cumulative equivalent), disease data (type, location, area), and maintenance history (time, location, method).

[0028] The maintenance history multi-source data set includes pre-maintenance performance, maintenance time, maintenance location, maintenance method and post-maintenance performance, and collects each data set, performs importance analysis, reconstructs the data set, and then performs maintenance decision training.

[0029] The importance analysis of factors affecting asphalt pavement performance by fusion of average weighted multiple methods in step 2 includes the following steps:

[0030] S201: The Spearman rank correlation coefficient is often used to measure the dependence of two variables. The characteristics and indicators involved often do not meet the requirements of normal distribution. The formula of the Spearman correlation coefficient is shown in formula (1).

[0031]

[0032] Where n is the number of variables; d i is the rank difference between two variables (i=1,2,3...n). The rank of a value is determined by its relative size in a sequence. If there are identical values ​​in the sequence, their rank will be the average of their positions. The difference between the ranks of two different values ​​is called the rank difference.

[0033] S202: Permutation-based feature importance analysis destroys the correlation between the original features and the target variable by randomly arranging feature values, and then evaluates the importance of the features by comparing the performance differences between the original model and the rearranged model. Its calculation requires training a model, randomly rearranging the data distribution of each feature, disrupting the order of the original feature values, retraining the model using the rearranged feature values ​​and the real target variable, and calculating the change in model performance. If the performance of the rearranged model is significantly reduced, it can be considered that the feature has a greater impact on the performance of the model, that is, the importance of the feature is high;

[0034] S203: SHAP is an explanatory model prediction method based on the Shapley value in game theory. In machine learning, its role is to quantify the contribution of each feature to illustrate their importance to model prediction. The feature importance analysis process of the SHAP method is as follows:

[0035] S20301: Initialize data: select a specific data instance and determine its characteristic value;

[0036] S20302: Define model: Define the machine learning model used, such as a linear regression model;

[0037] S20303: Calculate the baseline value: The baseline value is the predicted value of the model when no features are involved;

[0038] S20304: Calculate marginal contribution: For each feature, consider all possible feature combinations and calculate the change in the predicted value when adding the feature;

[0039] For feature i, its SHAP value is defined as in formula (2).

[0040]

[0041] Where S is the subset of all features except feature i, |S| is the size of S, i.e. the number of features in S, ! represents the factorial, N is the total number of features, and f(S) is the predicted value of the model on the feature subset (S);

[0042] Formula (2) calculates the marginal contribution of feature i under all possible feature subsets and performs weighted summation through the weights in combinatorial mathematics;

[0043] S20305: Assign weights: Assign weights to the marginal contribution of each feature according to the definition of Shapley value; weights are assigned based on the number and feature values ​​of the feature in the combination;

[0044] S204: Use three importance analysis methods to score and rank the factors affecting asphalt pavement performance; due to the differences in the measurement standards and data characteristics of different methods, the final rankings have some commonalities as well as some differences; in order to avoid the limitations and one-sidedness of using only a single method, the Spearman correlation coefficient method, the Permutation method and the SHAP method are integrated; the three importance analysis methods are given the same weight and then combined to ensure the validity of the importance weight analysis results of the factors affecting asphalt pavement. The process is as follows:

[0045] S20401: The dimensions of the results obtained by the three methods are different, so they are unified; for the correlation obtained by the Spearman correlation analysis method, the absolute value needs to be taken before calculation;

[0046] S20402: Process the weight of each method using formula (3);

[0047]

[0048] In the formula, F i is the importance weight of the i-th feature after de-dimensioning; I i is the importance weight of the i-th feature before de-dimensioning; n is the number of features;

[0049] S20403: Take the average of the feature importance weights of the three processed methods to obtain the feature importance weights of each fused indicator, re-sort the fused feature importance weights, and use the result as the basis for feature screening.

[0050] The steps of establishing the OOA optimization fusion model for predicting asphalt pavement performance and performing prediction analysis on asphalt pavement performance in step 3 are as follows:

[0051] S301: Perform matrix modeling according to Publication (4);

[0052] S302: Randomly initialize the position of the osprey in the search space according to Publication (5);

[0053]

[0054] x i,j =lb j+r i,j ·(ub j -lb j )(5);

[0055] Where N is the number of ospreys; m is the number of problem variables; X is the osprey position population matrix; X i is the initial position of the i-th osprey, i=1,2...,n; X i,j is the initial position of the i-th osprey in the j-th problem variable, j = 1, 2, ..., m; lb j ,ub j are the lower and upper bounds of the j-th problem variable; r i,j is a random number, ranging from [0,1];

[0056] S303: Calculate the objective function of each osprey, and the evaluation can be expressed by the vector of formula (6);

[0057]

[0058] Where F is the objective function value; F i is the objective function value of the i-th osprey;

[0059] S304: Phase 1 Update:

[0060] The position of each osprey with a higher objective function value in the search space represents the location of the prey fish school; the fish school position of each osprey is expressed using formula (7);

[0061] FP i ={X k |k∈{1,2,...,N}∧F k <F i}∪{X best}(7);

[0062] Where FP i is the set of fish shoal positions of the i-th osprey; X best Yes F i The optimal corresponding X i value;

[0063] When an osprey attacks a fish randomly, it will move toward the fish's position. The new position of the osprey is calculated using equations (8) and (9). If the new position improves the value of the objective function, it replaces the previous position of the osprey.

[0064]

[0065] In the formula, is the new position of the i-th osprey in the j-th problem variable; SF i,jis the position of the fish selected by the i-th osprey in the j-th problem variable; I i,j The value of is one of {1,2};

[0066] If the osprey improves the value of the objective function in the new position, the new position of the osprey after the update is calculated by formula (10);

[0067]

[0068] In the formula, is the new position of the i-th osprey; yes The objective function value of

[0069] S305: For each member of the population, use equations (11) and (12) to calculate a new random position as a position suitable for eating fish; if the value of the objective function is improved at this new position, replace the previous position of the corresponding osprey according to equation (13);

[0070]

[0071] An asphalt pavement maintenance system considering multi-dimensional influencing factors, comprising:

[0072] Data collection module, used to collect and integrate the evaluation indicators and multi-source influencing factors that need to be considered in the asphalt pavement performance evaluation;

[0073] Data analysis module, used to screen the evaluation indexes and multi-source influencing factors of asphalt pavement performance;

[0074] Asphalt pavement performance prediction module is used to predict and analyze asphalt pavement performance, and verify the performance of neural networks and machine learning algorithms in pavement performance prediction tasks, including grey correlation analysis. In view of the impact of hyperparameter selection on the performance of machine learning models, the OOA algorithm is introduced to optimize the hyperparameters of the machine learning model, and a multi-model fusion prediction method based on Stacking is established. The advantages of multiple models complement each other to achieve high-precision prediction of asphalt pavement performance.

[0075] The visualization interaction module is used to construct a resource integration database based on the fusion data and maintenance detection information, so that users can interact with the resource integration database through a three-dimensional visualization window.

[0076] After adopting the above method, the present invention has the following advantages:

[0077] The present invention constructs a multi-source data set of asphalt pavement performance, designs an importance analysis method of average weight fusion, and extracts monthly and seasonal fine-grained features of rainfall, temperature, and traffic flow data respectively to address the problem of insufficient correlation analysis between multi-source data of pavement influencing factors and incomplete consideration of influencing factors. It also adds historical maintenance data and disease data features to construct a multi-source data set of asphalt pavement performance influencing factors. The multi-source influencing factors of asphalt pavement performance are ranked by importance, and the characteristic factors with the greatest influence on asphalt pavement performance prediction are screened out based on the ranking results.

[0078] The asphalt pavement performance prediction method using OOA optimization fusion model aims at the problem of insufficient accuracy of the pavement performance prediction model. Considering that different models are suitable for the prediction of different performance indicators (PCI, RQI, RDI and SRI), the prediction advantages of RF, XGBoost and CatBoost models are fused through the Stacking method, and OOA is used for parameter optimization to make the model more suitable for the pavement performance prediction task on multi-source feature data sets. It solves the overfitting problem caused by biased learning of data by a single model and achieves high-precision prediction of asphalt pavement performance.

[0079] A multi-source data maintenance decision method based on random forests improved by inverse variance weights was proposed. Aiming at the problem of low accuracy of existing asphalt pavement maintenance decision methods, a multi-source data set of maintenance history was constructed. It has a stronger representation ability for maintenance decision problems than the traditional decision method that only considers performance index data PCI, RQI, RDI and SRI, and improves the accuracy of maintenance decisions. A multi-source data maintenance decision method based on random forests improved by inverse variance weights was designed, which can not only make decisions on maintenance levels, but also make decisions on specific maintenance methods.

[0080] The above summary is for illustrative purposes only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments and features described above, further aspects, embodiments and features of the present invention will be readily apparent by reference to the accompanying drawings and the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0082] Figure 1 It is a technical flow chart of the present invention;

[0083] Figure 2is an analysis diagram of a multi-source data set of asphalt pavement performance according to the present invention;

[0084] Figure 3 It is the annual temperature change diagram of the representative area 1 of the present invention;

[0085] Figure 4 It is the maintenance benefit diagram of the preventive maintenance of the present invention;

[0086] Figure 5 It is a maintenance benefit diagram of the repair and maintenance of the present invention;

[0087] Figure 6 is a flow chart of the present invention;

[0088] Figure 7 It is a comparison chart of the IVW-RF maintenance decision results of the present invention. DETAILED DESCRIPTION

[0089] Specific embodiments of the present invention will now be mentioned in detail. Although the present invention is described in conjunction with these specific embodiments, it should be appreciated that it is not intended to limit the present invention to these specific embodiments. On the contrary, these embodiments are intended to cover substitutions, changes or equivalent embodiments that may be included in the spirit and scope of the invention defined by the claims. In the following description, a large number of specific details are set forth in order to provide a comprehensive understanding of the present invention. The present invention can be implemented without some or all of these specific details. In other cases, in order not to make the present invention unnecessarily obscure, well-known process operations are not described in detail.

[0090] When used in conjunction with "including," "methods comprising," or similar language in this specification and the appended claims, the singular forms "a," "an," and "the" include plural references unless the context clearly dictates otherwise. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0091] The present invention is further described in detail below in conjunction with the full text.

[0092] Combined with Figure 1-Figure 7,The present invention performs feature engineering operations on data, including feature generation, selection and conversion, to more accurately characterize the problem domain and improve model efficiency. After encoding and feature dimension reduction of some features, multiple statistical and machine learning methods are used to sort the feature importance of multi-source feature data sets, and the feature set with significant impact on asphalt pavement performance prediction is screened out according to the sorting results. A fusion prediction model based on OOA optimization is designed, which can accurately capture and predict the changing trend of pavement performance. A multi-source data maintenance decision method based on inverse variance weighted improved random forest is proposed, which comprehensively considers the current performance status and performance prediction trend to realize the decision of maintenance level and specific maintenance method. In the specific implementation of the present invention, PQI comprehensively considers multiple key technical conditions such as pavement damage, pavement flatness, rutting, vehicle jumping, wear, anti-skid performance and structural strength. Each sub-indicator corresponds to a specific performance characteristic of the highway. After scientific weighting and normalization, these sub-indicators form a scale from 0 to 100 to reflect the technical condition of the pavement.

[0093] Different models of the present invention are suitable for the prediction of different performance indicators (PCI, RQI, RDI and SRI). The prediction advantages of RF, XGBoost and CatBoost models are integrated through the Stacking method, and OOA is used for parameter optimization to make the model more suitable for the pavement performance prediction task on multi-source feature data sets. The overfitting problem caused by biased learning of data by a single model is solved, and high-precision prediction of asphalt pavement performance is achieved.

[0094] In order to solve the problem of low accuracy of existing asphalt pavement maintenance decision-making methods, a multi-source data set of maintenance history was constructed. Compared with the traditional decision-making method that only considers performance indicator data PCI, RQI, RDI and SRI, it has a stronger ability to characterize maintenance decision-making problems and improves the accuracy of maintenance decisions.

[0095] The PQI should be calculated according to formula 15:

[0096] PQI=w PCI PCI+w RQI RQI+w RDI RDI+w PBI PBI+w PWI PWI+w SRI SRI+w PSSI PSSI (15);

[0097] In the formula, w PCI 、w RQI 、w RDI 、w PBI 、w PWI 、w SRI 、w PSSI is the weight of each indicator in PQI.

[0098] PCI=100-a0DR a1 (16); DR is the damage rate of the highway pavement; in asphalt pavement, the coefficient a0 is 15.00 and the coefficient a1 is 0.412;

[0099] In equations 17 and 18, the unit of IRI measurement value is m / km; the coefficient a0 is 0.026 for first-class highways and expressways, and 0.0185 for other levels of highways; the coefficient a1 is 0.65 for first-class highways and expressways, and 0.85 for other levels of highways; Z S Z is the absolute vertical movement distance of the vehicle body; u is the absolute vertical movement distance of the tire; L is the distance traveled by the vehicle.

[0100] Among them, in formula 19.

[0101] RD is the rutting depth, in mm; rutting depth parameter RD a Use 10.0; rutting depth parameter RD b Use 40.0; model parameter a0 is 1.0; model parameter a1 is 3.0.

[0102] In formula 20, SFC is the lateral force coefficient; SRI is min For calibration parameters, 35.0 is used; a0 and a1 are model coefficients, which are 28.6 and -0.105 respectively.

[0103] like Figure 2 As shown in the figure, multi-source data including climate, traffic flow, basic road information, disease and maintenance history data are integrated. The pavement performance of a certain RC highway is analyzed, and the multi-source influencing factors are analyzed and feature constructed, so as to prepare data for feature importance analysis, asphalt pavement performance prediction model construction and maintenance decision-making. The basic information data of the road include: material quality, thickness and road age, which will directly affect the performance decay process of the pavement. For example, a large pavement thickness can slow down the damage rate of the pavement to a certain extent and extend the service life of the pavement. Therefore, the basic information data of the road plays an important role in evaluating the decay process.

[0104] Traffic flow is also an important factor affecting the degradation of pavement performance. High-intensity traffic on a road section will impose a large load on the road surface, thereby aggravating road fatigue and deformation. In particular, the frequency and weight of heavy vehicles passing by will further affect the speed of pavement degradation.

[0105] In addition, climate conditions also have a significant impact on the performance decay of pavement. For example, temperature may cause the pavement material to shrink and expand, aggravating pavement cracks and wear. Precipitation and snow accumulation can also deteriorate road conditions and have an adverse effect on pavement performance.

[0106] At the same time, during the use of roads, various types and degrees of diseases will appear. Typical road surface diseases include potholes, repairs and cracks. These diseases have a negative effect on the performance of the road, accelerate the degradation of road performance, and shorten the overall life of the road.

[0107] The complex and diverse influencing factors increase the difficulty of data collection and collation. The selection of unreasonable influencing factors may also have a negative impact on the prediction results. The data collected in this paper include expressway regular inspection data, road basic information data, climate data, traffic flow data, disease data and maintenance data. In addition, some data that are difficult to apply directly need to be feature constructed.

[0108] Take a certain RC highway as an example:

[0109] 1. Road basic information data processing

[0110] The road basic data of a certain RC highway is mainly divided into two sections, with road ages of 23 years and 20 years respectively. The corresponding start and end pile numbers are shown in Table 1.

[0111] Table 1 Highway section age information

[0112]

[0113]

[0114] The material information of a RC highway pavement structure is shown in Table 2

[0115] Table 2 Material information of a RC highway pavement structure

[0116]

[0117] As can be seen from Table 2, Section 1 (except the bridge section) and Section 2 (except the bridge section) used the same surface and middle and lower layer materials, including 4 cm AK-16 asphalt concrete overlay, 5 cm AC-20 asphalt concrete middle surface layer and 6 cm AC-25 asphalt concrete lower layer. The base materials of the two are different. Section 1 uses 20, 30 (33) cm lime fly ash stabilized soil, and Section 2 uses 20 cm of the same material.

[0118] 2. Climate data processing

[0119] The climate data obtained along a certain RC highway mainly includes temperature data and precipitation data. Considering that the climate data of similar sections are basically the same, the climate data of five counties along the road are used as a substitute. The official meteorological data of each region are used, and the temperature and precipitation of each region are processed and integrated separately.

[0120] 3. Climate data processing

[0121] The climate data obtained along a certain RC highway mainly includes temperature data and precipitation data. Considering that the climate data of similar sections are basically the same, the climate data of five counties along the road are used as a substitute. The official meteorological data of each region are used, and the temperature and precipitation of each region are processed and integrated separately.

[0122] like Figure 3 As shown in the figure, temperature data has a strong periodicity. In the feature construction process, representative data can be used instead of all data to reduce data redundancy, which is conducive to the model's learning of data. At the same time, pavement performance is significantly affected by temperature fluctuations, especially when experiencing high temperature cycles and freeze-thaw cycles. Seasonal data can better capture these periodic changes.

[0123] The time points within a year are divided into four different categories based on the characteristics of the data, rather than simply dividing them into the traditional four seasons of spring, summer, autumn and winter. The climate data is clustered into four seasons using the K-means algorithm.

[0124] 1) Initialization: Randomly select k samples from the data as the initial cluster centers.

[0125] 2) Cluster assignment: Calculate the Euclidean distance between each data sample and each cluster center, and assign each sample to the nearest cluster center. The Euclidean distance formula 21 is as follows.

[0126]

[0127] In the formula, x i is the i-th data point in the data set, c j is the jth cluster center, D is the dimension or number of features of the data, Represents data point x i The value in the dth dimension, Represents the cluster center c j The value in the dth dimension.

[0128] (3) Center update: Based on the newly formed clusters, recalculate the center of each cluster, generally taking the average value of all cluster points. The cluster center update formula, Formula 22, is as follows.

[0129]

[0130] In the formula, c j ′ is the new cluster center of the jth cluster, S j is the set of data points assigned to the jth cluster, |S j | is the set S j The size is the number of data points, x i is the data point assigned to the jth cluster.

[0131] The temperature difference between seasons is obvious, which can show the characteristics of high temperature cycle and freeze-thaw cycle. The temperature between different regions also has certain differences, which can be used as the characteristics of asphalt pavement performance prediction model.

[0132] Seasonal characteristics are shown in Table 3.

[0133] Table 3 Seasonal characteristics display

[0134]

[0135]

[0136] The seasonal characteristics of the temperature data after clustering are shown in Table 4. Season 2 has the lowest average temperature, and the lowest temperature is mostly below 0°C; Season 1 has moderate temperature, with the highest temperature mostly between 13-22°C and the lowest temperature between 4-10°C; Season 4 has a highest temperature of around 25°C and a lowest temperature of around 15°C; Season 3 is the traditional summer, with the highest temperature above 30°C. The seasonal characteristics of each region are shown in Table 3.

[0137] Extreme temperatures can also affect pavement performance. When the temperature is high, the pavement is more sensitive to the heavy pressure of vehicles, which may cause deformation or rutting. High temperatures can also accelerate the aging process of asphalt pavement and reduce its durability. When the temperature is low, especially below freezing, the pavement material shrinks and becomes more fragile. This can cause cracks to form and worsen, and freeze-thaw cycles can also damage the pavement structure, resulting in potholes and other damage. Therefore, the number of extreme high temperature days (maximum temperature greater than 35°C) and extreme low temperature days (minimum temperature less than 0°C) each year are counted as influencing factors in pavement performance prediction. The data characteristics are shown in Table 4.

[0138] Table 4 Extreme temperature characteristics of each area of ​​a RC highway

[0139]

[0140]

[0141] Rainfall data feature construction,

[0142] The accumulated rainfall data of each region are shown in Table 5. These data reflect the climate characteristics and precipitation distribution of each region and can be used as factors affecting the degradation of pavement performance.

[0143] Table 5 Cumulative rainfall characteristics of each area on a certain RC highway

[0144] Rainfall in the past year Rainfall in the past two years Rainfall in the past three years Rainfall in the past 4 years Region 1 638.39 1699.92 2472.27 3245.48 Area 2 788.31 2026.74 3014.6 3963.48 Area 3 777.93 2015.97 3050.31 4076.38 Region 4 616.85 1676.46 2498.67 3351.98 Area 5 458.24 1340.27 1921.15 2563.99

[0145] Traffic flow data processing,

[0146] Taking the section from K0+000 to K4+960 as an example, the data are shown in Table 6.

[0147] Table 6 Traffic flow data sample display of a RC highway section from K0+000 to K4+960

[0148]

[0149]

[0150] The original data sources are different, the data formats are different, and the data and technical indicators of multiple types of vehicles are not completely consistent. For data availability, we extract equivalent data that can reflect the overall characteristics of traffic volume. In addition, passenger cars and trucks have different impacts on the road surface, so we extract cargo-passenger ratio data that can reflect the specific distribution of traffic volume.

[0151] As can be seen from the table, there are large differences in traffic volume in different sections and at different times. Fine-grained monthly data including monthly equivalents and passenger-to-freight ratio data can be used as important features for predicting asphalt pavement performance. At the same time, when integrating and constructing features of the data, the cumulative equivalents of the past four years, which have an important impact on pavement performance, are selected to represent the cumulative effect of traffic flow, which can also represent the impact of traffic flow on the pavement at the annual time granularity.

[0152] Disease data processing,

[0153] The road surface damage reflects the decline of road performance, and the further development of the damage will affect the service life of the road. The road surface damage rate DR reflects 11 types of damage on the road surface. The distribution and statistics of various damages are obtained by analyzing the original data of the road inspection vehicle. When integrating the damage data, the main 7 indicators are initially selected, including longitudinal cracks, transverse cracks, strip repairs, block cracks, block repairs, potholes and cracks.

[0154] When predicting the performance of asphalt pavement, ensuring data quality is a key step. Time and space scale alignment refers to the integration and alignment of different types of collected data according to a unified spatial location and time standard for accurate analysis and prediction. Using the 100-meter stakes on the highway to mark the location, all data are positioned according to this reference system.

[0155] The surface materials and thickness of the road surface and base layers in the road basic information data are the attributes of the road itself and do not change over time. They are aligned on the 100-meter stakes through location information. For climate data and traffic flow data, representative historical data and features are selected and then aligned to the 100-meter stakes through location information. Disease data and maintenance history data are aligned in the same way as pavement performance indicators on a spatial scale.

[0156] For climate data, we divided the data into regions, matched the start and end pile numbers of the road sections in each region according to the actual location, and matched the data according to the 100-meter piles. For traffic flow data and maintenance history data, we aligned the data according to the start and end pile numbers of each monitored road section. The disease data counted the cumulative disease distribution on each 100-meter road section. We selected some representative samples of the integrated data for display, and the results are shown in Table 7-14.

[0157] Table 7 Characteristics of representative samples of seasonal temperature data of a certain RC highway (unit: °C)

[0158]

[0159] Table 8 Characteristics of representative samples of monthly rainfall on a certain RC highway (unit: mm)

[0160]

[0161]

[0162] Table 9 Characteristics of representative samples of cumulative rainfall on a certain RC highway (unit: mm)

[0163] Starting pile number Rainfall in the past year Rainfall in the past two years Rainfall in the past three years Rainfall in the past 4 years K0+400 638.39 1699.92 2472.27 3245.48 K17+300 788.31 2026.74 3014.6 3963.48 K46+900 616.85 1676.46 2498.67 3351.98 K42+200 777.93 2015.97 3050.31 4076.38 K50+700 458.24 1340.27 1921.15 2563.99 ...

[0164] Table 10 Characteristics of representative samples of RC highway disease data (unit: m 2 )

[0165] direction Starting pile number Vertical cracks Horizontal cracks Strip repair Block cracks Block repair Potholes Cracking Up K11+300 0.00 0.22 0.00 0.00 0.06 0.00 0.00 Up K14+700 0.19 0.00 0.00 0.32 0.56 0.35 0.00 Downside K38+400 0.20 0.50 0.00 0.00 0.00 0.20 0.00 Downside K45+200 0.13 0.00 0.53 0.00 0.54 0.00 0.60 Downside K79+800 0.00 0.00 0.00 0.52 0.40 0 0 ...

[0166] Table 11 Characteristics of representative samples of a certain RC highway monthly freight-passenger ratio

[0167]

[0168]

[0169] Table 12 Characteristics of representative samples of a certain RC high-speed monthly vehicle equivalent (unit: ten thousand)

[0170] Starting pile number January February March April May June July August September October November December K0+000 81 174 163 149 184 181 192 195 189 135 117 258 K20+200 99 245 260 252 305 284 292 209 307 194 166 293 K33+400 131 58 212 73 261 262 284 98 95 68 98 162 K54+700 83 178 228 203 215 201 200 189 190 128 97 153 K61+800 140 277 291 268 286 179 130 190 269 188 166 186 ...

[0171] Table 13 Characteristics of representative samples of a certain RC high-speed cumulative equivalent

[0172] Starting pile number Cumulative equivalent in the past year Cumulative equivalent in the past two years Cumulative equivalent in the past three years Cumulative equivalent in the past four years K0+000 20198024 36265656 51065598 67574548 K20+200 29054842 62415879 70668447 81098687 K33+400 18035862 44687498 56386688 72006133 K54+700 20641427 57279625 74093665 94735145 K61+800 25692382 55866798 71773158 90217703 ...

[0173] Table 14 Characteristics of sample maintenance history data of a certain RC highway

[0174]

[0175]

[0176] In summary, the evaluation indicators and multi-source influencing factors that need to be considered in the asphalt pavement performance evaluation are integrated for multi-source data including climate, traffic flow, road basic information, disease and maintenance history data. The pavement performance of a certain RC highway is analyzed, and the multi-source influencing factors are analyzed and feature constructed, which prepares data for feature importance analysis, asphalt pavement performance prediction model construction and maintenance decision-making.

[0177] Before the feature dimension reduction and importance analysis of multi-source feature influencing factors, the non-numerical features are firstly encoded, and then the data is standardized to avoid the influence of data format and numerical value on the analysis results. Three feature importance analysis methods are used for average weight fusion to screen out the important characteristic factors of pavement performance evaluation indicators and eliminate the influencing factors of collinearity, so as to prepare for the subsequent prediction of pavement performance evaluation indicators.

[0178] The data of pavement performance influencing factors from multiple sources and multiple domains are standardized and feature encoding is implemented. Then, feature dimension reduction is achieved through various techniques such as PCA and Pearson analysis. An importance analysis method based on the fusion of Spearman, Permutation and SHAP average weights is proposed to rank the importance of multi-source influencing factors of asphalt pavement performance, and the most influential important characteristic factors for asphalt pavement performance prediction are screened out based on the ranking results.

[0179] By using these important influencing factors and considering the historical situation of various road condition detection indicators, an asphalt pavement performance prediction model is established. The performance of gray correlation analysis, neural network and machine learning algorithms in the pavement performance prediction task is verified. In view of the impact of hyperparameter selection on the performance of the machine learning model, the OOA algorithm is introduced to optimize the hyperparameters of the machine learning model. A multi-model fusion prediction method based on the Stacking method is established, and the advantages of multiple models complement each other to achieve high-precision prediction of asphalt pavement performance.

[0180] A single grey prediction model, deep learning models such as TabNet and LSTM-FCNN, and machine learning models such as random forest, XGBoost, and CatBoost were established to predict asphalt pavement performance. The introduction of the OOA optimization algorithm significantly improved the selection process of model hyperparameters, and the effect was better than the Bayesian optimization algorithm. Finally, the Stacking method of multi-model fusion was used to achieve mutual complementation of models and more comprehensively capture the changing laws of pavement performance. Through experimental comparison, the superiority and reliability of the proposed model in predicting asphalt pavement performance were established.

[0181] Effect analysis based on the maintenance method dataset. The accuracy of all models on the maintenance level classification dataset is higher than their performance on the maintenance method classification dataset. Combined with the confusion matrix given in Table 15, this is because the distribution of data in the maintenance level classification dataset is more uniform, and the model is more likely to make correct predictions.

[0182] Table 15 Confusion matrix of IVW-RF maintenance method decision

[0183] No.1 No.2 No.3 No.4 No.5 No.6 No.7 precision recall F1-score No.1 1 0 2 4 0 1 0 0.5 0.12 0.2 No.2 0 4 6 1 0 3 0 0.44 0.29 0.35 No.3 0 1 148 15 1 22 10 0.56 0.75 0.64 No.4 0 3 19 505 0 14 0 0.87 0.93 0.9 No.5 0 0 13 2 1 12 2 0.14 0.03 0.05 No.6 0 1 41 39 3 71 8 0.49 0.44 0.46 No.7 1 0 36 15 2 22 9 0.31 0.11 0.16

[0184] No.1 to No.7 represent synchronous fiber wearing layer, ultra-thin wearing layer, micro-surfacing, no maintenance, milling upper middle surface layer resurfacing, milling upper surface layer resurfacing and milling full surface layer resurfacing respectively.

[0185] The confusion matrix in Table 20 is a display of the decision results using IVW-Random Forest, where each row represents the true category and each column represents the predicted category. The values ​​on the diagonal represent the number of correct decisions, that is, the actual solution is the same as the decision solution. Precision refers to the proportion of true positive samples among the positive samples identified by the classifier. Recall refers to the proportion of samples that the classifier correctly identifies as positive samples among all positive samples. F1-score is the harmonic mean of precision and recall.

[0186] From the confusion matrix in Table 20, we can see that the accuracy of no maintenance is very high, with a precision of 0.87, a recall of 0.93, and an F1-score of 0.90. It performs best among all categories, and the prediction results are highly consistent with the actual situation. The accuracy of micro-surfacing is relatively high, with a precision of 0.56, a recall of 0.75, and an F1-score of 0.64, which shows that the prediction of this category has a relatively high consistency compared with the actual situation. However, the decision-making performance for the milling and resurfacing method is poor, and the scores of multiple indicators are very low, indicating that the model has poor prediction effect on these categories and the probability of correct judgment is low.

[0187] In addition, the number of samples of synchronous fiber wear layer and ultra-thin wear layer is small, and many samples are misclassified into other categories because the number of samples of these two categories in the dataset is not enough for the model to learn enough features for accurate decision-making.

[0188] (2) Effect analysis based on maintenance level dataset

[0189] The data distribution of the maintenance level dataset is relatively even, and the data of various decision results are relatively balanced. Table 16 shows the performance of IVW-RF on this dataset.

[0190] Table 16 Confusion matrix for IVW-RF maintenance level decision

[0191] Repair and maintenance No maintenance Preventive maintenance precision recall F1-score Repair and maintenance 154 54 70 0.75 0.55 0.64 No maintenance 8 532 1 0.87 0.98 0.92 Preventive maintenance 44 24 151 0.68 0.69 0.68

[0192] First, for the no-maintenance class, the model performs best, with the highest precision of 0.87, recall of 0.98, and F1-score of 0.92. This means that the model is very accurate and reliable in identifying situations where no maintenance is required. Almost all no-maintenance instances are correctly identified, and only 9 of the 541 samples are misclassified as other labels. At the same time, there are very few instances that are misclassified as no-maintenance.

[0193] For the preventive maintenance class, the precision is slightly lower at 0.68, while the recall and F1-score are 0.69 and 0.68 respectively. This shows that the model balances the error when predicting preventive maintenance, but the performance is slightly lower than the prediction of no maintenance. There are 151 instances that are correctly classified, while 44 are misclassified as repair maintenance and 24 are misclassified as no maintenance.

[0194] The precision of the repair and maintenance category is 0.75, which is higher than preventive maintenance but lower than no maintenance. However, its recall is 0.55, which is low, and its F1-score is 0.64, which is at a medium level. This means that the model is relatively good at identifying instances that need repair and maintenance, but a large proportion of instances are still misclassified.

[0195] (3) Effect analysis based on multi-index maintenance method dataset

[0196] It is difficult to make accurate decisions on some sections of roads based solely on performance indicators, and more features that can assist in decision-making need to be added. Since the data comes from the real historical data of a certain RC highway, data collection is limited. The multi-source feature dataset only contains repair and maintenance data. The performance of IVW-RF on the maintenance method decision dataset containing multi-source features is shown in Table 17.

[0197] The prediction of no maintenance is very accurate, with the highest precision of 0.96 and recall of 0.98, and the highest F1-score of 0.97, showing that the classification model is highly accurate in identifying this category. The recall value of the milling top layer resurfacing category is high at 0.86, indicating that the model is more effective in detecting this category, but its precision is low at 0.77, which means that there is a certain error rate in predicting this category. Some data points in the other three categories were misclassified as milling top layer resurfacing, especially the misclassification of milling full layer resurfacing as milling top layer resurfacing, which is more common, with 26 misclassifications.

[0198] Table 17 Confusion matrix of maintenance method decision under IVW-RF multi-source characteristics

[0199]

[0200] Based on the experimental results of the three data sets, the data sets are imperfect because they are collected from real historical data. Data defects make it impossible to fully characterize the decision-making ability of the model in various situations. However, by comprehensively considering the performance on the three data sets, we can analyze the performance of the IVW-RF model and the problems that may be encountered in the data-based maintenance decision-making method.

[0201] The IVW-RF model performs better than other models and outperforms other models on multiple data sets. Analyzing the decision results, the data with a smaller sample size has a lower decision accuracy, while the accuracy of the no maintenance category is the highest because the no maintenance case has more obvious characteristics and the data volume is sufficient for the model to mine the characteristics of this category.

[0202] In other categories, significant improvements were achieved when using a maintenance grade dataset with better sample balance, which shows that the IVW-RF model has certain potential in maintenance decision-making and may achieve better performance on a highly balanced dataset.

[0203] If some categories are very similar in features, the model may have difficulty distinguishing them. For example, milling full-layer resurfacing and milling upper layer resurfacing may have high overlap in features, causing model confusion. The accuracy of model decisions can be improved by adding more multi-source features that are helpful for maintenance decisions to assist in distinguishing similar categories.

[0204] Preventive maintenance is a maintenance measure that is carried out before the road surface has slightly worn out or when minor diseases just begin to appear. The core concept is to slow down the development and degradation of road diseases through early intervention. In the absence of obvious functional loss, potential problems can be identified in advance through regular inspection and evaluation of the road surface, and maintenance work can be implemented. Preventive maintenance usually includes filling and encapsulating cracks, fog sealing, thin layer paving, road surface cleaning and drainage system maintenance. These measures are relatively low-cost, have less interference with traffic, can effectively protect the pavement structure, extend the actual service life of the road surface, and reduce the need for future major repairs.

[0205] Unlike preventive maintenance, repair maintenance focuses more on obvious damage and functional degradation that have already occurred. It is a repair measure taken when pavement damage develops to the point where it affects traffic safety and driving comfort. The goal of repair maintenance is to restore the structural integrity and service performance of the pavement, which usually involves clearing, filling, reinforcing or rebuilding the damaged parts. It includes local deep repairs, large-scale pavement milling and resurfacing, and pavement structure reinforcement. Repair maintenance work is expensive and has a greater impact on traffic, but it can prevent further damage to the pavement and delay the overall decline of the pavement structure.

Claims

1. A decision-making method for asphalt pavement maintenance considering multi-dimensional influencing factors, characterized in that: The following steps are involved: S1. Asphalt pavement data integration and data quality improvement: Collect and integrate the evaluation indicators and multi-source influencing factors that need to be considered in the performance evaluation of asphalt pavement, analyze and construct the characteristics of multi-source influencing factors, and lay a data foundation for the construction of data-driven performance prediction models and maintenance decisions. S2. Importance analysis of factors affecting asphalt pavement performance based on average weighted multi-method fusion: The evaluation indicators and multi-source influencing factors of asphalt pavement performance are screened, and the prediction performance is maintained while streamlining the model input. A large amount of feature data is standardized and feature coding is implemented. The importance of the factors affecting the performance of asphalt pavement is analyzed, and the most predictive feature set is found according to the importance ranking results. S3. Establishment of OOA optimization fusion model for asphalt pavement: Establish a predictive OOA optimization fusion model to predict and analyze asphalt pavement performance; S4. Establishment of multi-source data set of maintenance history: Construct a multi-source data set of maintenance history and establish a maintenance decision-making method based on the inverse variance weighted improved random forest model to improve the accuracy of maintenance decisions; S401: Analyzing the historical data of expressways, the PSSI of structural strength is higher than 90, and no maintenance is required for structural strength indicators; therefore, the pavement indicators selected are PCI, RDI, RQI and SRI; After merging the collected maintenance history data and the data of the road sections without maintenance, a maintenance method classification data set was constructed to make decisions on seven maintenance methods; a maintenance level classification data set was constructed to make decisions on three maintenance levels; S402: Construct a multi-index maintenance method dataset including historical data and unmaintained road sections. Based on the PCI, RDI, RQI and SRI indicator data, the dataset adds multi-source features including pavement disease data, road basic information data and traffic load data to improve the dataset's ability to represent maintenance decision-making issues; S5. Pavement maintenance decision-making method: Construct a multi-source data set of maintenance history, and establish a maintenance decision-making method based on the inverse variance weighted improved random forest model to improve the accuracy of maintenance decisions, and use the prediction results and current situation to select the maintenance decision-making method; S501: Construct the OOB data of the trees in the random forest method. The OOB data is an independent validation set for each tree that does not participate in the training. During the training process, each tree will randomly select data from the original training set using replacement sampling as training data when it is established. Usually, the number of selected data is the same as the size of the original training set. Due to the replacement sampling, some data will be repeatedly selected, while others may not be selected. The data that is not selected constitutes the OOB data of this tree. S502: The prediction of each decision tree contained in the random forest is given the same weight to calculate the final prediction result; when using the inverse variance weighting method IVW (, a weight is assigned to the prediction of each tree, and the weight of each tree is inversely proportional to its prediction error on the validation set.

2. The asphalt pavement maintenance decision-making method considering multi-dimensional influencing factors according to claim 1 is characterized by: The multi-source influencing factors include climate, traffic flow, road infrastructure information, disease and maintenance history data.

3. The asphalt pavement maintenance decision-making method considering multi-dimensional influencing factors according to claim 1 is characterized in that: The importance of factors affecting asphalt pavement performance is analyzed by using PCA and Pearson analysis methods to achieve feature dimensionality reduction. The importance analysis method based on the fusion of Spearman, Permutation and SHAP average weights is used to rank the importance of multi-source factors affecting asphalt pavement performance as the basis for subsequent data selection.

4. The asphalt pavement maintenance decision-making method considering multi-dimensional influencing factors according to claim 1 is characterized by: The aforementioned factors affecting asphalt pavement performance include historical detection data (PCI, PQI, RDI, SRI), basic road information (road age, material, structure), climate data (temperature, precipitation), traffic flow (equivalent, passenger-to-freight ratio, cumulative equivalent), disease data (type, location, area), and maintenance history (time, location, method).

5. The asphalt pavement maintenance decision-making method considering multi-dimensional influencing factors according to claim 1 is characterized by: The maintenance history multi-source data set includes pre-maintenance performance, maintenance time, maintenance location, maintenance method and post-maintenance performance, and collects each data set, performs importance analysis, reconstructs the data set, and then performs maintenance decision training.

6. The asphalt pavement maintenance decision-making method considering multi-dimensional influencing factors according to claim 1 is characterized by: The importance analysis of factors affecting asphalt pavement performance by fusion of average weighted multiple methods in step 2 includes the following steps: S201: The Spearman rank correlation coefficient is often used to measure the dependence of two variables. The characteristics and indicators involved often do not meet the requirements of normal distribution. The formula of the Spearman correlation coefficient is shown in formula (1). Where n is the number of variables; d i is the rank difference between two variables (i=1,2,3...n). The rank of a value is determined by its relative size in a sequence. If there are identical values ​​in the sequence, their rank will be the average of their positions. The difference between the ranks of two different values ​​is called the rank difference. S202: Permutation-based feature importance analysis destroys the correlation between the original features and the target variable by randomly arranging feature values, and then evaluates the importance of the features by comparing the performance differences between the original model and the rearranged model. Its calculation requires training a model, randomly rearranging the data distribution of each feature, disrupting the order of the original feature values, retraining the model using the rearranged feature values ​​and the real target variable, and calculating the change in model performance. If the performance of the rearranged model is significantly reduced, it can be considered that the feature has a greater impact on the performance of the model, that is, the importance of the feature is high; S203: SHAP is an explanatory model prediction method based on the Shapley value in game theory. In machine learning, its role is to quantify the contribution of each feature to illustrate their importance to model prediction. The feature importance analysis process of the SHAP method is as follows: S20301: Initialize data: select a specific data instance and determine its characteristic value; S20302: Define model: Define the machine learning model used, such as a linear regression model; S20303: Calculate the baseline value: The baseline value is the predicted value of the model when no features are involved; S20304: Calculate marginal contribution: For each feature, consider all possible feature combinations and calculate the change in the predicted value when adding the feature; For feature i, its SHAP value is defined as formula (2); Where S is the subset of all features except feature i, |S| is the size of S, i.e. the number of features in S, ! represents the factorial, N is the total number of features, and f(S) is the predicted value of the model on the feature subset (S); Formula (2) calculates the marginal contribution of feature i under all possible feature subsets and performs weighted summation through the weights in combinatorial mathematics; S20305: Assign weights: Assign weights to the marginal contribution of each feature according to the definition of Shapley value; weights are assigned based on the number and feature values ​​of the feature in the combination; S204: Use three importance analysis methods to score and rank the factors affecting asphalt pavement performance; due to the differences in the measurement standards and data characteristics of different methods, the final rankings have some commonalities as well as some differences; in order to avoid the limitations and one-sidedness of using only a single method, the Spearman correlation coefficient method, the Permutation method and the SHAP method are integrated; the three importance analysis methods are given the same weight and then combined to ensure the validity of the importance weight analysis results of the factors affecting asphalt pavement. The process is as follows: S20401: The dimensions of the results obtained by the three methods are different, so they are unified; for the correlation obtained by the Spearman correlation analysis method, the absolute value needs to be taken before calculation; S20402: Process the weight of each method using formula (3); In the formula, F i is the importance weight of the i-th feature after de-dimensioning; I i is the importance weight of the i-th feature before de-dimensioning; n is the number of features; S20403: Take the average of the feature importance weights of the three processed methods to obtain the feature importance weights of each fused indicator, re-sort the fused feature importance weights, and use the result as the basis for feature screening.

7. The asphalt pavement maintenance decision-making method considering multi-dimensional influencing factors according to claim 1 is characterized by: The steps of establishing the OOA optimization fusion model for predicting asphalt pavement performance and performing prediction analysis on asphalt pavement performance in step 3 are as follows: S301: Perform matrix modeling according to Publication (4); S302: Randomly initialize the position of the osprey in the search space according to Publication (5); x i,j =lb j +r i,j ·(ub j -lb j )(5); Where N is the number of ospreys; m is the number of problem variables; X is the osprey position population matrix; X i is the initial position of the i-th osprey, i=1,2...,n; X i,j is the initial position of the i-th osprey in the j-th problem variable, j = 1, 2, ..., m; lb j ,ub j are the lower and upper bounds of the j-th problem variable; r i,j is a random number, ranging from [0,1]; S303: Calculate the objective function of each osprey, and the evaluation can be expressed by the vector of formula (6); Where F is the objective function value; F i is the objective function value of the i-th osprey; S304: Phase 1 Update: The position of each osprey with a higher objective function value in the search space represents the location of the prey fish school; the fish school position of each osprey is expressed using formula (7); FP i ={X k |k∈{1,2,...,N}∧F k <F i }∪{X best }(7); Where FP i is the set of fish school positions of the i-th osprey; X best Yes F i The optimal corresponding X i value; When an osprey attacks a fish randomly, it will move toward the fish's position. The new position of the osprey is calculated using equations (8) and (9). If the new position improves the value of the objective function, it replaces the previous position of the osprey. In the formula, is the new position of the i-th osprey in the j-th problem variable; SF i,j is the position of the fish selected by the i-th osprey in the j-th problem variable; I i,j The value of is one of {1,2}; If the osprey improves the value of the objective function in the new position, the new position of the osprey after the update is calculated by formula (10); In the formula, is the new position of the ith osprey; F i P1 yes The objective function value of S305: For each member of the population, use equations (11) and (12) to calculate a new random position as a position suitable for eating fish; if the value of the objective function is improved at this new position, replace the previous position of the corresponding osprey according to equation (13); 8. An asphalt pavement maintenance system considering multi-dimensional influencing factors, characterized by: include: Data integration and improvement module, which is used to integrate the evaluation indicators and multi-source influencing factors that need to be considered in the asphalt pavement performance evaluation and improve the data quality; Data analysis module, used to screen the evaluation indexes and multi-source influencing factors of asphalt pavement performance; Asphalt pavement performance prediction module is used to predict and analyze asphalt pavement performance, and verify the performance of neural networks and machine learning algorithms in pavement performance prediction tasks, including grey correlation analysis. In view of the impact of hyperparameter selection on the performance of machine learning models, the OOA algorithm is introduced to optimize the hyperparameters of the machine learning model, and a multi-model fusion prediction method based on Stacking is established. The advantages of multiple models complement each other to achieve high-precision prediction of asphalt pavement performance. The visualization interaction module is used to construct a resource integration database based on the fusion data and maintenance detection information, so that users can interact with the resource integration database through a three-dimensional visualization window.

Images