Method for predicting residual life of existing asphalt pavement structure

Abstract

The invention discloses a method for predicting the residual life of an existing asphalt pavement structure, particularly relates to the technical field of computer simulation and data processing, and is used for solving the problem that the existing prediction method cannot efficiently represent the non-uniformity of an existing pavement material and the dynamic environment load influence on the premise of ensuring the precision. Existing damage data of a road surface are acquired to construct a simulation model reflecting a real initial state, material fatigue parameters are acquired in combination with cross-scale analysis, a dynamic environment boundary is driven by utilizing climate time sequence data, and then time-frequency coherence analysis is performed on a damage mode and a climate load to identify a high-threat working condition, so that the high-threat working condition is identified. And finally, preferentially operating high-fidelity thermal-mechanical coupling numerical simulation aiming at a high-threat working condition, and mapping damage evolution data output by simulation to a macroscopic performance degradation process so as to determine the residual fatigue life.

Description

Technical Field

[0001] This invention relates to the field of computer simulation and data processing technology, and in particular to a method for predicting the remaining life of existing asphalt pavement structures. Background Technology

[0002] In the field of computer simulation-based engineering prediction, predicting the remaining service life of asphalt pavement structures already in use is an important research direction. Existing prediction methods mainly fall into two categories: one is statistical models based on historical data, whose prediction accuracy is severely limited by the representativeness and completeness of the training data, resulting in insufficient generalization ability. The other is physical models based on numerical simulation. These methods typically employ simplified homogeneous material assumptions and static environmental boundary conditions to construct simulation models, thereby reducing computational complexity. However, when dealing with the specific object of existing asphalt pavement structures, the inherent limitations of these models become prominent: firstly, the material properties of existing pavement structures exhibit complex multi-scale non-homogeneity and initial damage states due to their service history, and existing simulation models lack effective parameterization methods to accurately characterize these initial conditions; secondly, the actual service environment of pavement is dynamically changing, while existing models often use constant or simplified environmental boundaries, leading to distortion in the prediction of long-term performance evolution. This makes it difficult for existing simulation methods to achieve a balance between prediction accuracy, model complexity, and computational efficiency, limiting their reliable application in practical engineering.

[0003] Existing technologies, when using computer simulations to predict the remaining fatigue life of existing asphalt pavement structures, cannot effectively balance prediction accuracy and engineering practicality. This results in a disconnect between the prediction results and the actual service performance evolution. Furthermore, the complexity of the calculations makes it difficult to support efficient evaluation and precise maintenance decisions for large-scale road networks. Summary of the Invention

[0004] This invention addresses the technical problems existing in the prior art by providing a method for predicting the remaining life of existing asphalt pavement structures.

[0005] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: A method for predicting the remaining service life of existing asphalt pavement structures includes: S1. Obtain the existing damage state data of the target asphalt pavement structure, and determine the initial state of the multiphysics coupled numerical simulation model based on the existing damage state data. S2. Obtain fatigue performance parameters of asphalt pavement materials for multi-physics coupled numerical simulation models through cross-scale analysis methods; S3. Obtain time-series climate data for the target area and construct environmental boundary conditions to drive the multiphysics coupled numerical simulation model. S4. Extract damage mode features from existing damage state data, and perform time-frequency analysis on climate time series data to calculate the time-varying coherence between the damage mode features and the climate time series data after time-frequency analysis, and identify high-threat working condition combinations. S5. Load the fatigue performance parameters of asphalt pavement materials and environmental boundary conditions into the multiphysics coupled numerical simulation model, and prioritize the high-threat working condition combination to run high-fidelity numerical simulation to calculate the damage evolution data of the target asphalt pavement structure under the set load spectrum. S6. Based on damage evolution data, determine the performance degradation process according to the correspondence between damage state and service performance, and take the time corresponding to the performance degradation to the allowable limit as the remaining fatigue life.

[0006] Furthermore, S1 includes: The road surface deflection data, interlayer bonding status data, and internal defect data obtained through non-destructive testing and core sampling are used as existing damage status data. The existing damage state data are fused to extract the spatial stiffness attenuation characteristics and initial damage distribution characteristics of the pavement structure. The spatial stiffness attenuation characteristics and initial damage distribution characteristics are transformed into elastic modulus reduction parameters and initial damage field variables of the corresponding material regions in the multiphysics coupled numerical simulation model, so as to determine the initial state of the multiphysics coupled numerical simulation model.

[0007] Furthermore, the multiphysics coupled numerical simulation model is a three-dimensional numerical model of the pavement structure established by the finite element method. It simulates the pavement response under the combined action of temperature, humidity and load by simultaneously solving the heat conduction equation, moisture migration equation and mechanical equilibrium equation. The geometric dimensions and material partitions of the multiphysics coupled numerical simulation model are determined according to the target pavement design drawings and are constructed by assigning material property parameters and boundary conditions.

[0008] Furthermore, S2 includes: The microstructure images of asphalt mixtures are obtained and analyzed to extract the geometric configuration and spatial distribution information of the internal aggregates, asphalt binder and voids. Based on geometric configuration and spatial distribution information, a microscale mechanical simulation calculation is performed to obtain the local stress-strain response at the interface between asphalt mastic and aggregate at the microscale. The local stress-strain response at the microscale is correlated and calibrated with the fatigue test data of macroscopic specimens of asphalt mixtures, forming a cross-scale calibration relationship from microscopic response to macroscopic performance; Based on cross-scale calibration relationships, fatigue performance parameters for characterizing the macroscopic fatigue degradation of materials are calculated and used as fatigue performance parameters of asphalt pavement materials in multi-physics coupled numerical simulation models.

[0009] Furthermore, S3 includes: Historical meteorological monitoring data of the target area is obtained. The historical meteorological monitoring data includes data series of temperature data, humidity data and solar radiation intensity data over time, which constitute climate time series data. The climate time series data are cleaned and interpolated to obtain a continuous and complete time series of climate parameters. The continuous and complete time series of climate parameters are transformed into boundary condition parameters required for the calculation of heat conduction, moisture migration and thermal stress in the multiphysics coupled numerical simulation model, so as to construct the environmental boundary conditions used to drive the multiphysics coupled numerical simulation model.

[0010] Furthermore, S4 includes: Modal analysis was performed on existing damage state data to extract damage modal features that reflect the main damage modes of the pavement structure. Time-frequency transformation is performed on climate time-series data to obtain its energy distribution characteristics in the time-frequency domain; Calculate the time-varying coherence coefficients of damage modal characteristics and energy distribution characteristics in the time-frequency domain; Based on the time-frequency regions where the time-varying coherence coefficient exceeds the preset coherence threshold, the climate environment and time combination that pose a high threat to the evolution of pavement structure damage are identified as high-threat working condition combinations.

[0011] Furthermore, the time-varying coherence coefficients of the damage modal features and energy distribution features in the time-frequency domain are calculated using the wavelet coherence method, which includes: performing continuous wavelet transforms on the damage modal features and energy distribution features respectively to obtain their respective wavelet coefficients, calculating the cross-correlation coefficients and autocorrelation coefficients of the two wavelet coefficients at the same time-frequency point, and obtaining the time-varying coherence coefficients.

[0012] Furthermore, S5 includes: The fatigue performance parameters of asphalt pavement materials and environmental boundary conditions are assigned to the material property parameters and boundary condition parameters in the multiphysics coupled numerical simulation model. Based on the climate and environmental conditions and time segments corresponding to the high-threat operating conditions, corresponding simulation tasks are set in the multiphysics coupled numerical simulation model. A multiphysics coupled numerical simulation model was run to perform thermo-mechanical coupled high-fidelity numerical simulation, simulating the mechanical response of the pavement structure under the combined action of a set load spectrum and corresponding climatic environmental conditions. Based on the mechanical response, damage evolution data characterizing the performance degradation and damage propagation process of pavement structure materials are calculated and output.

[0013] Furthermore, S6 includes: Extract a sequence of damage state parameters characterizing the extent of damage propagation within the pavement structure from damage evolution data; Based on a pre-established performance evolution model that describes the quantitative mapping relationship between the damage state parameter sequence and the pavement smoothness index, the performance degradation process curve of the pavement smoothness index over time is calculated. On the performance degradation curve, the time of action corresponding to when the pavement smoothness index drops to the allowable limit specified in the engineering specifications is determined, and the time of action is taken as the remaining fatigue life.

[0014] Furthermore, the performance evolution model is established through regression analysis of historical data, including: collecting measured data of pavement smoothness index under different damage states, and using statistical regression methods to fit the mathematical function relationship between damage state parameters and smoothness index, thereby obtaining a performance evolution model that describes the quantitative mapping relationship.

[0015] The beneficial effects of this invention are: 1. By integrating multiphysics simulation with multi-source heterogeneous data, the accuracy and practicality of predicting the remaining life of existing pavement structures are effectively improved. Material fatigue parameters are obtained through cross-scale analysis, and a simulation model including initial damage is constructed based on field detection data. This enables the model to accurately reflect the heterogeneous material state and spatial damage distribution of existing pavements due to their service history, fundamentally improving the physical realism of the simulation model. Long-term climate data is used to drive dynamic environmental boundaries, and time-frequency coherence analysis of environmental loads and damage modes is performed to identify high-threat conditions. This allows the simulation to focus on the key environmental stages with the strongest driving force for damage evolution, thereby significantly improving the relevance and efficiency of simulation analysis while ensuring high fidelity.

[0016] 2. A balance is achieved between high-precision prediction and computational feasibility. By prioritizing the deployment of high-fidelity simulations on identified high-threat combinations of working conditions, rather than performing equal calculations across all time periods, this method significantly reduces the required simulation computing resources. This makes it possible to perform high-resolution simulations of long-term performance evolution under conventional computing conditions. Based on the damage evolution data output by the refined simulation, the remaining fatigue life, which can be directly used for maintenance decisions, is mapped to an empirically calibrated performance evolution model. This forms a complete, closed-loop digital prediction link from microscopic material response to macroscopic service performance, providing reliable quantitative evidence for pavement maintenance management that combines physical mechanisms and engineering practicality. Attached Figure Description

[0017] Figure 1 This is a flowchart of a method for predicting the remaining life of existing asphalt pavement structures according to the present invention. Figure 2 This is a flowchart illustrating the identification of high-threat combinations of operating conditions in this invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Example: Figure 1 This invention provides a method for predicting the remaining life of existing asphalt pavement structures, comprising: S1. Obtain the existing damage state data of the target asphalt pavement structure, and determine the initial state of the multiphysics coupled numerical simulation model based on the existing damage state data. S2. Obtain fatigue performance parameters of asphalt pavement materials for multi-physics coupled numerical simulation models through cross-scale analysis methods; S3. Obtain time-series climate data for the target area and construct environmental boundary conditions to drive the multiphysics coupled numerical simulation model. S4. Extract damage mode features from existing damage state data, and perform time-frequency analysis on climate time series data to calculate the time-varying coherence between the damage mode features and the climate time series data after time-frequency analysis, and identify high-threat working condition combinations. S5. Load the fatigue performance parameters of asphalt pavement materials and environmental boundary conditions into the multiphysics coupled numerical simulation model, and prioritize the high-threat working condition combination to run high-fidelity numerical simulation to calculate the damage evolution data of the target asphalt pavement structure under the set load spectrum. S6. Based on damage evolution data, determine the performance degradation process according to the correspondence between damage state and service performance, and take the time corresponding to the performance degradation to the allowable limit as the remaining fatigue life.

[0020] S1. Obtain the existing damage state data of the target asphalt pavement structure, and determine the initial state of the multiphysics coupled numerical simulation model based on the existing damage state data. Specifically, this is implemented as follows: To obtain existing damage state data of the target asphalt pavement structure, a falling weight deflectometer is used to measure the deflection at multiple points on the pavement surface. The measurement points are arranged in a regular grid of, for example, 2 meters by 2 meters, along the transverse and longitudinal directions of the lanes. The deflection basin data of each measurement point is obtained, and this deflection basin data is used as the pavement deflection data. Simultaneously, pavement core samples are drilled at representative locations. After the core samples are cut and segmented, a pull-out tester is used to measure the bond strength between the asphalt layer and the base layer. Digital image correlation technology is used to observe the development of interlayer microcracks. The bond strength test results and microcrack observation results are combined as interlayer bonding state data. Furthermore, ground-penetrating radar is used to scan the target pavement area at a center frequency of, for example, 1.6 GHz and a measurement line spacing of 0.5 meters to obtain radar reflection wave images. By processing and interpreting the radar reflection wave images, abnormal areas such as voids, loosening, and cracks within the pavement structure are identified. The location, size, and type information of these abnormal areas are used as internal defect data. The above pavement deflection data, interlayer bonding state data, and internal defect data together constitute the existing damage state data.

[0021] When fusing existing damage state data, the pavement deflection data is first calculated using a deflection basin back-calculation program based on the theory of elastic layered systems. This program iteratively adjusts the modulus values ​​of each structural layer to minimize the error between the theoretical and measured deflection basins, thereby calculating the dynamic resilient modulus values ​​of each pavement structural layer at various spatial points. The degree of attenuation of this dynamic resilient modulus value relative to the initial design modulus of the material is quantified as a modulus attenuation coefficient. The modulus attenuation coefficients at all points constitute a spatial distribution field, which represents the spatial stiffness attenuation characteristics of the pavement structure. Simultaneously, for the fusion of interlayer bond strength data and internal defect data, different fusion weights need to be assigned to different types of data to reflect their relative importance. For example, the weight of interlayer bond strength data can be set to 0.6, and the weight of internal defect data can be set to 0.4. The specific setting of this weight... The threshold for interlayer bond strength failure is determined based on expert experience and statistical analysis of the impact of different damage types on the overall structural performance in historical data. Based on this weighted fusion logic, regions in the interlayer bond strength test data with values ​​below the interlayer bond strength failure threshold, along with all abnormal regions identified from internal defect data, are superimposed and fused in a unified three-dimensional spatial coordinate system to generate a spatial three-dimensional point cloud dataset labeled with different damage types and severity. This dataset describes the spatial location and distribution of damage within the pavement structure, i.e., the initial damage distribution characteristics. The interlayer bond strength failure threshold is determined based on the pavement material design requirements, the standard failure load value from indoor pull-out tests, and statistical analysis of historical test data. For example, by analyzing a large amount of historical test data, the 85th percentile value of the bond strength test value sequence is set as the interlayer bond strength failure threshold.

[0022] When determining the initial state of the multiphysics coupled numerical simulation model, the spatial stiffness attenuation characteristics are input into the preprocessing module of the simulation model. Based on the center point coordinates of each element in the model mesh, the corresponding modulus attenuation coefficient is read at its spatial location. The predefined elastic modulus of the material is reduced according to this modulus attenuation coefficient. The reduction rule is that no reduction occurs when the modulus attenuation coefficient is 1; when the modulus attenuation coefficient is less than 1, the elastic modulus value is adjusted to the original design value multiplied by the coefficient using linear interpolation. This yields the elastic modulus reduction parameter for each material element, which is then assigned a value. Simultaneously, the three-dimensional point cloud dataset from the initial damage distribution characteristics is spatially mapped to the finite element mesh of the simulation model. For elements within the model mesh containing damage points marked by the point cloud data, an initial damage field variable value ranging from 0 to 1 is assigned based on the damage type and density represented by the point cloud data. 0 represents no damage, and 1 represents complete damage. The specific value depends on the number of damage points within the element and the preset damage density. The relationship between damage density and damage field variables is determined by interpolation. The method for establishing this relationship table is as follows: Standard asphalt mixture specimens with different preset defect density levels are prepared in the laboratory. The initial mechanical properties of each specimen are tested through mechanical tests. The properties of the undamaged specimens are normalized to 1, and the properties of the completely damaged specimens are normalized to 0. By fitting the relationship curve between defect density and normalized performance values, a mapping relationship between defect density and damage field variable values ​​is established. This relationship constitutes the relationship table. In this mapping process, when the damage point density calculated in the unit reaches a certain critical value determined according to the above experimental relationship curve, this critical value is the critical damage point density threshold. Reaching the critical damage point density threshold will have a significant impact on macroscopic properties. For example, the corresponding initial damage field variable value can be set to 0.7. By completing the assignment of elastic modulus reduction parameters and initial damage field variables for all relevant material regions, the initial state of the multiphysics coupled numerical simulation model is determined.

[0023] Regarding the establishment of the multiphysics coupled numerical simulation model, firstly, based on the design drawings of the target asphalt pavement structure, the thickness, width, and length of each structural layer of the pavement, as well as the material type used for each layer, are determined. A corresponding three-dimensional solid geometric model is then created in commercial finite element software. Subsequently, the geometric model is divided into different material partitions according to the material type, and each partition is meshed using hexahedral solid elements. The governing equations of the model include the heat conduction equation describing the heat transfer process, the moisture migration equation describing the movement of moisture in the porous medium, and the mechanical equilibrium equation describing the deformation and stress state of the structure under external forces. These three equations are solved simultaneously through the material thermal expansion and contraction effect caused by the temperature field, the material modulus softening effect caused by the humidity field, and the stress and strain generated by the mechanical field. Solving the heat conduction equation requires inputting the material's thermal conductivity and specific heat capacity parameters, while solving the moisture migration equation requires inputting the material's permeability coefficient and moisture diffusion coefficient parameters. Solving the mechanical equilibrium equations requires inputting the material's elastic modulus, Poisson's ratio, and strength parameters. These material property parameters are assigned values ​​based on the pavement design documents and indoor material test reports. The model's boundary conditions include the convective heat transfer boundary and solar radiation heat flow boundary between the model's upper surface and the air, the adiabatic or isothermal boundaries of the model's sides and bottom, the initial temperature and humidity distribution field inside the model, and the mechanical boundary of the model's upper surface bearing vehicle loads. The specific parameter values ​​of these boundary conditions are assigned through environmental boundary conditions constructed in subsequent steps. By completing geometric modeling, material partitioning, mesh generation, setting the governing equations, assigning material property parameters, and defining boundary conditions, the construction of the multiphysics coupled numerical simulation model is completed. This establishment process clarifies the complete technical chain from geometry to physical properties. The governing equations, parameters, and boundary conditions involved are all well-known technical elements in this field used for analyzing pavement structure response. Their specific simultaneous solution is automatically completed by the solver of the selected commercial finite element software based on the set model and conditions.

[0024] S2. Obtain the fatigue performance parameters of asphalt pavement materials for multiphysics coupled numerical simulation models through cross-scale analysis methods. The specific implementation is as follows: To obtain microstructural images of asphalt mixtures, standard specimens drilled from the asphalt mixture were scanned using, for example, an industrial computed tomography (CT) scanner. The image resolution was set to, for example, 15 micrometers per pixel to obtain high-contrast three-dimensional grayscale images that clearly distinguish the aggregate, asphalt binder, and porous phases. The acquired three-dimensional grayscale images were processed. First, median filtering was used to eliminate image noise. Then, a grayscale threshold was set to initially divide the image into solid and porous phases. This grayscale threshold was determined by analyzing the valley grayscale values ​​between the solid and porous peaks on the image's grayscale histogram. For the solid phase... Further segmentation of the bulk phase employs an image segmentation algorithm based on region growing. This algorithm separates aggregate particles from asphalt mortar based on differences in grayscale values ​​and local texture features, labeling them as different material regions. After image segmentation, the geometric configuration and spatial distribution information of each material region are extracted. Geometric configuration information includes the volume, surface area, aspect ratio, and angularity of each aggregate particle; the thickness and continuity of the asphalt mortar region; and the size, shape, and connectivity of voids. Spatial distribution information is characterized by statistically analyzing the coordinate positions of each phase material in three-dimensional space and calculating spatial statistical parameters such as proximity and aggregation degree.

[0025] When acquiring and segmenting the microstructure image of asphalt mixture, the specific method for obtaining and setting the grayscale threshold segmentation threshold is as follows: First, obtain the global grayscale histogram of the three-dimensional grayscale image. This histogram usually exhibits a bimodal shape, corresponding to the lower grayscale region of the porous phase and the higher grayscale region of the solid phase in the image, respectively. Smooth the grayscale histogram using methods such as Gaussian smoothing to eliminate minor fluctuations. Then, find the point with the lowest grayscale value between the two main peaks. The grayscale value of this point is automatically determined as the initial candidate value for the grayscale threshold segmentation. To ensure the robustness of the segmentation, fine-tuning can be performed within a small grayscale range around this initial candidate value, and the visual segmentation effect of the binarized image can be verified. Finally, the grayscale threshold segmentation threshold used to segment the solid phase and the porous phase is determined. For the distinction between aggregates and asphalt mastic within the solid phase, the segmentation is mainly based on the statistical differences in the grayscale value distribution and local texture features of the two types of materials. This process is automatically completed by the image analysis algorithm and does not rely on setting a single global grayscale threshold.

[0026] Based on geometric configuration and spatial distribution information, a microscale mechanical simulation is performed. Specifically, the segmented and labeled 3D image data is directly converted into a finite element mesh model. Each image voxel corresponds to a hexahedral element. The material properties of the elements are assigned according to their labeled type; for example, aggregate elements are assigned the elastic modulus and Poisson's ratio of rock, asphalt mastic elements are assigned viscoelastic material parameters based on the generalized Maxwell model, and void elements are assigned near-zero stiffness values ​​to simulate cavities. The boundary conditions for the microscale mechanical simulation are set based on the macroscopic stress state. A uniformly distributed pressure is applied to the top of the model to simulate wheel loads. A fixed constraint is applied to the bottom of the model, and periodic boundary conditions or symmetric constraints are applied to the sides of the model to simulate an infinite mass. Under the set boundary conditions, the stress and strain fields of the model under load are calculated by solving the linear elastic or viscoelastic mechanical control equations, with particular attention paid to the stress and strain response of the asphalt mortar-aggregate interface region. The maximum principal stress, equivalent stress, and maximum shear stress of the interface elements are extracted as local stress response indices, and the maximum principal strain and equivalent strain of the interface elements are extracted as local strain response indices. This simulation calculation is completed on commercial finite element software or a dedicated micromechanical calculation platform, and the process from image to stress and strain results is automated through parametric scripts.

[0027] To correlate and calibrate the local stress-strain response at the microscale with fatigue test data of macroscopic asphalt mixture specimens, macroscopic fatigue tests must first be conducted. These tests employ a four-point bending fatigue test mode with controlled stress. Standard beam specimens are repeatedly loaded under constant stress levels and loading frequencies, and the number of loading cycles until the specimen stiffness decreases to a specific proportion of its initial stiffness, such as 50%, is recorded as the macroscopic fatigue life. To establish the correlation, a representative stress-strain parameter for all asphalt mastic-aggregate interface elements under simulated loads is statistically calculated from the micromechanical simulation results. For example, the volume-weighted average of the equivalent stress of all interface elements is used as a representative mechanical response index at the microscale. The above micromechanical simulation is then repeated under different macroscopic test stress levels. The simulation process yields a set of data pairs between microscopic representative mechanical response indices and corresponding macroscopic fatigue life. Using nonlinear curve fitting methods, such as power functions, these two sets of data are fitted to determine the mathematical functional relationship between the microscopic representative mechanical response indices and the macroscopic fatigue life. This functional relationship constitutes the cross-scale calibration relationship from microscopic response to macroscopic performance. A goodness-of-fit threshold is used to judge the reliability of the calibration relationship. This threshold is typically set based on statistical significance requirements; for example, the minimum acceptable value of the coefficient of determination is set to 0.85. When the coefficient of determination is lower than this goodness-of-fit threshold, the validity of the material parameter assignments in the microscopic model or the macroscopic experimental data needs to be checked and recalibrated.

[0028] When establishing cross-scale calibration relationships, a goodness-of-fit threshold is used to ensure that the established mathematical relationship has sufficient predictive reliability. This goodness-of-fit threshold is set based on the general requirements for model prediction accuracy in engineering practice and statistical significance criteria. Specifically, after fitting the microscopic representative mechanical response indexes and macroscopic fatigue life data using models such as power functions, the coefficient of determination is calculated. In the field of road material fatigue analysis, to ensure the model has sufficient explanatory power and predictive stability, this coefficient of determination is usually required to be no less than a pre-set critical value. This critical value is the goodness-of-fit threshold, and its specific value is determined with reference to the industry's general standards for the reliability of material performance models and the statistical results of historical data from a large number of similar calibration experiments. For example, based on experience, when the coefficient of determination reaches 0.85 or higher, the calibration relationship is considered reliable and acceptable; therefore, the goodness-of-fit threshold can be set to 0.85. If the fitting result does not reach this threshold, it indicates that there may be anomalies in the microscopic simulation parameter settings or macroscopic experimental data. It is necessary to check and adjust the relevant parameters and recalculate until the goodness-of-fit threshold is met.

[0029] Based on cross-scale calibration relationships, fatigue performance parameters characterizing the macroscopic fatigue degradation of materials are calculated. The specific process involves transforming the mathematical expression of the cross-scale calibration relationship to align it with the form of the standard macroscopic fatigue equation. The standard macroscopic fatigue equation typically represents a power-law relationship between macroscopic stress level and fatigue life. By comparing the transformed calibration relationship with the standard fatigue equation, material-specific coefficients, i.e., fatigue performance parameters, in the standard fatigue equation can be directly derived. For example, if the standard macroscopic fatigue equation expresses a relationship where macroscopic stress is inversely proportional to the Nth power of fatigue life, then this Nth power can be analytically derived through the cross-scale calibration relationship. The specific value of the power is a key fatigue performance parameter; another fatigue performance parameter may be a proportionality coefficient related to the initial strength of the material, which can also be obtained by comparing the constant terms in the equation; the final set of fatigue performance parameters can be directly input into the multiphysics coupled numerical simulation model to predict the damage accumulation and performance degradation process of asphalt pavement materials under cyclic loading on a macroscopic scale; the entire process from obtaining microscopic images to calculating macroscopic parameters constitutes a complete cross-scale analysis framework, ensuring that physically meaningful and experimentally calibrated material constitutive parameters are provided for subsequent numerical simulations.

[0030] S3. Obtain time-series climate data for the target area and construct environmental boundary conditions to drive the multiphysics coupled numerical simulation model. The specific implementation is as follows: When acquiring time-series climate data for the target area, historical meteorological monitoring data is collected from meteorological monitoring stations in and around the target area. This historical meteorological monitoring data includes hourly or daily air temperature data, relative humidity data, and solar radiation intensity data. These data constitute the raw time-series climate data indexed by time. To ensure data quality and spatial representativeness, spatial consistency checks are performed on data from different stations during the collection process. For example, temperature readings at adjacent stations at the same time point are compared. If the difference exceeds a pre-set spatial consistency difference threshold, the data at that time point is marked as data to be verified. This spatial consistency difference threshold is not a fixed value. Its setting needs to comprehensively consider the climate type, topographic complexity, and average distance between meteorological stations in the target area. It is determined by statistically analyzing the distribution of differences in data from multiple stations during the same historical period. For example, the standard deviation of temperature differences between adjacent stations in multi-year historical data can be calculated, and the spatial consistency difference threshold can be set to twice this standard deviation. The collected raw time-series climate data may contain missing values ​​due to sensor failure or transmission interruption, as well as obvious outliers caused by extreme weather or instrument malfunctions.

[0031] When cleaning and interpolating climate time-series data, the first step is data cleaning to identify and handle outliers and missing values ​​in the raw data. For outlier identification, a statistical distribution-based method is used, such as calculating the mean and standard deviation of each climate parameter's time series within a 24-hour rolling time window. Data points continuously exceeding the rolling mean ± 3 times the standard deviation are initially identified as outliers. For these initially identified outliers, a review is needed based on the data trends of adjacent time periods. If the outlier shows a discontinuous abrupt change with the preceding and following data points and cannot be explained by known meteorological events, it is confirmed as invalid data and marked. The handling of missing values ​​and other confirmed outliers follows a similar process. To impute invalid outlier data, time series interpolation methods are employed. Specifically, different interpolation strategies are used depending on the length of the missing data segment. For shorter missing segments, such as those with fewer than six consecutive time points, linear interpolation is used, filling the missing segment linearly based on the values ​​of valid data points before and after it. For longer missing segments or data that cannot be reliably filled using simple interpolation, an interpolation method based on the average of historical data from the same period is used. This involves extracting climate parameter data from the same date and time over the past few years from the historical database and calculating their arithmetic mean as the filling value for the missing time. Through the above cleaning and interpolation processes, a temporally continuous and complete time series of climate parameters is obtained.

[0032] The continuous and complete time series of climate parameters are transformed into boundary condition parameters required for calculating heat conduction, moisture migration, and thermal stress in a multiphysics coupled numerical simulation model. For heat conduction calculations, the air temperature time series and solar radiation intensity time series need to be transformed into convective heat transfer boundary conditions and heat flux boundary conditions on the model's surface. Specifically, the convective heat transfer boundary conditions require calculating the convective heat transfer coefficient at each time point. This coefficient is calculated based on air temperature and wind speed data using an empirical relationship in the form of Newton's law of cooling; that is, the convective heat transfer coefficient is equal to a coefficient related to wind speed multiplied by a power of the wind speed, where the wind speed-related coefficient is obtained experimentally or from literature. Simultaneously, the air temperature time series is directly used as the reference ambient temperature for convective heat transfer. The solar radiation intensity time series is then converted based on the heat absorption rate of the road surface material. The incident heat flux density acting on the upper surface of the model is converted to the relationship that heat flux density equals solar radiation intensity multiplied by the material's thermal absorptivity. For moisture migration calculations, the relative humidity time series needs to be converted into humidity boundary conditions for the near-surface layer of the model. This is achieved by querying the isothermal hygroscopic curve of the material. The relative humidity value at each time point is input into the curve to obtain the corresponding equilibrium moisture content of the material, thereby constructing a boundary condition sequence for the change of surface moisture content over time. For thermal stress calculations, the boundary conditions are directly determined by the temperature field output from the heat conduction calculation. Finally, the convective heat transfer coefficient time series, ambient temperature time series, surface heat flux density time series, and surface moisture content time series obtained from the above calculations together constitute a time-synchronized and physically defined set of environmental boundary condition parameters, which is used to drive the multiphysics coupled numerical simulation model.

[0033] Figure 2 A flowchart for identifying high-threat operating condition combinations according to the present invention is provided. S4: Extract damage mode features based on existing damage state data, and perform time-frequency analysis on climate time-series data to calculate the time-varying coherence between the damage mode features and the climate time-series data after time-frequency analysis, thereby identifying high-threat operating condition combinations. The specific implementation is as follows: When performing modal analysis on existing damage state data to extract damage modal features, based on the spatial three-dimensional point cloud dataset and spatial stiffness attenuation feature field that represent the initial damage distribution characteristics of the pavement structure obtained and fused in step S1, the damage points at different spatial locations in the spatial three-dimensional point cloud dataset are first quantized and assigned values ​​according to their damage type and severity, forming a three-dimensional damage scalar field with spatial coordinates as variables. Simultaneously, the modulus attenuation coefficient in the spatial stiffness attenuation feature field is also considered as a three-dimensional scalar field with spatial coordinates as variables. After spatially aligning these two three-dimensional scalar field data, for example, this... Modal analysis is performed using the orthogonal decomposition method, which solves for the eigenvalues ​​and eigenvectors of the covariance matrix constructed from the spatial data field. The calculated eigenvectors are the spatial modes, and each spatial mode corresponds to an eigenvalue. The magnitude of the eigenvalue represents the contribution rate of the mode to the total variance of the original data field. By retaining the top few modes with the largest contribution rates, such as the top three modes with a cumulative contribution rate of 85% of the total variance, the damage mode features that best represent the main distribution pattern of existing damage in the pavement structure can be extracted. These features are specifically represented by a set of spatial distribution functions and associated initial mode coordinate values.

[0034] When performing time-frequency transformation on climate time-series data to obtain energy distribution characteristics, based on the continuous and complete climate parameter time series obtained after cleaning and interpolation in step S3, the temperature time series is used as an example. A continuous wavelet transform method is used to perform time-frequency analysis on the temperature time series. The continuous wavelet transform requires selecting a wavelet basis function, such as the Morlet wavelet, whose shape is controlled by the center frequency parameter and bandwidth parameter. By translating the selected wavelet basis function on the time axis and scaling it on the frequency scale, a family of wavelet functions is constructed. The temperature time series is then multiplied by each wavelet function to obtain the wavelet coefficients of the signal in the two-dimensional time-frequency plane. The square of the modulus of these wavelet coefficients represents the joint distribution of signal energy in the time and frequency dimensions. This two-dimensional distribution is the energy distribution characteristic of the climate time-series data in the time-frequency domain. In specific calculations, the range of the frequency scale is set according to the actual physical cycle of temperature changes, for example, from the high frequency corresponding to a 1-hour cycle to the low frequency corresponding to a 1-year cycle. The resolution of the time-frequency transformation is controlled by adjusting the step size of the scale parameter.

[0035] The calculation of the time-varying coherence coefficients of damage modal features and energy distribution features in the time-frequency domain is specifically achieved through the wavelet coherence method. First, the modal coordinate time series corresponding to the extracted main damage modal features are synchronized with the aforementioned climate temperature time series. Next, continuous wavelet transforms are performed on these two time series with parameters consistent with those in the previous steps to obtain their respective wavelet coefficient matrices. Then, at each grid point in the two-dimensional time-frequency plane, the cross-correlation coefficients of these two wavelet coefficients, i.e., their cross-power spectral density, are calculated. Simultaneously, the autocorrelation coefficients of their respective wavelet coefficients, i.e., their respective auto-power spectral density, are calculated. Finally, the cross-correlation coefficient at each time-frequency point is divided by the square root of the product of the two autocorrelation coefficients at that point. The result is the time-varying coherence coefficient at that time-frequency point, with a value ranging from 0 to 1.

[0036] When identifying high-threat operating condition combinations based on time-varying coherence coefficients, a preset coherence threshold needs to be set as a judgment criterion. This preset coherence threshold is set based on statistical significance testing. The specific method is as follows: First, a large number of random noise sequence pairs with the same length and spectral characteristics as the original time series are generated. Using the same continuous wavelet transform and wavelet coherence calculation process, the distribution of time-varying coherence coefficients among these random sequence pairs is calculated. The critical values ​​of this distribution at different significance levels are statistically analyzed. For example, at a 95% confidence level, the critical coherence coefficient value corresponding to the 95th quantile can be found. This critical value is set as the preset coherence threshold, meaning that when the actually calculated time-varying coherence coefficient exceeds this threshold, there is a 95% confidence that the coherence is present. The properties are not caused by random noise and are statistically significant. After obtaining the two-dimensional distribution map of the time-varying coherence coefficient, all time-frequency regions are judged, and all time-frequency regions with time-varying coherence coefficient values ​​exceeding the preset coherence threshold are identified. These time-frequency regions will be retained for subsequent analysis. At the same time, time-frequency regions with time-varying coherence coefficient values ​​not exceeding the preset coherence threshold will be ignored and not considered as candidates for high-threat conditions. By mapping each retained time-frequency region back to the original time domain and climate parameter domain, its corresponding time period and the dominant climate condition characteristics within that time period can be resolved. Each such combination of time segment and climate conditions is identified as a high-threat condition combination that poses a high threat to the evolution of pavement structure damage.

[0037] S5. Load the fatigue performance parameters of asphalt pavement materials and environmental boundary conditions into a multiphysics coupled numerical simulation model, and prioritize high-fidelity numerical simulation for high-threat working conditions to calculate the damage evolution data of the target asphalt pavement structure under a set load spectrum. The specific implementation is as follows: When loading the fatigue performance parameters of asphalt pavement materials and environmental boundary conditions into a multiphysics coupled numerical simulation model, the fatigue performance parameters obtained through cross-scale analysis in step S2 are first input into the material property definition module of the simulation model in the form of material constitutive model parameters. Specifically, if the fatigue performance parameters include parameters describing the stiffness degradation law of the material under cyclic loading, these parameters are assigned to the corresponding degradation law variables in the viscoelastic or elastoplastic constitutive model of asphalt mixture defined in the model. At the same time, the set of environmental boundary condition parameters constructed in step S3, namely the convective heat transfer coefficient time series, ambient temperature time series, surface heat flux density time series, and surface moisture content time series, are assigned to the convective heat transfer boundary, temperature boundary, heat flux boundary, and humidity boundary on the upper surface of the simulation model according to their corresponding time nodes. This completes the comprehensive update of the model's material properties and boundary conditions, enabling it to reflect the fatigue characteristics of specific materials and operate under specific climatic conditions.

[0038] When setting up simulation tasks based on the climate and time segments corresponding to high-threat operating condition combinations, each high-threat operating condition combination identified in step S4 explicitly defines a specific time segment and the dominant climate characteristics within that time segment. In the preprocessing settings of the simulation model, firstly, based on this time segment, the corresponding data segment is extracted from the loaded complete set of environmental boundary condition parameters. For example, if a high-threat operating condition combination corresponds to the high-temperature period from 13:00 to 17:00 in the summer afternoon, only all boundary condition time series data within this 4-hour period are extracted. Then, in the simulation analysis step settings, the physical time range of the simulation is set to the length of this time segment, and the corresponding time integration step size is configured. For example, the simulation with a total duration of 4 hours is divided into an increment step of one minute, and the step size setting must meet the numerical stability condition. At the same time, based on the climate characteristics emphasized by the operating condition combination, such as high temperature and high solar radiation, the corresponding physical field coupling relationship is activated in the simulation to ensure that the bidirectional coupling effect of heat conduction and mechanical field is fully considered, thereby generating an independent simulation task with a clear time range and climate driving conditions for this specific high-threat operating condition.

[0039] When performing high-fidelity numerical simulations of thermo-mechanical coupling using a multiphysics coupled numerical simulation model, the solver is activated for the pre-defined simulation task. The solver proceeds sequentially step-by-step according to time increments. Within each increment, it first solves the heat conduction equation based on the current boundary conditions, calculating the three-dimensional temperature field distribution within the pavement structure at that moment. Subsequently, the calculated temperature field is used as a thermal load, while also considering the softening effect of humidity on the material modulus, and substituted into the mechanical equilibrium equations to calculate the stress, strain, and displacement fields within the pavement structure under the set vehicle load spectrum at that moment. The set vehicle load spectrum is determined based on traffic survey data of the target road, defining different vehicle types and axle loads. The dynamic moving load generated on the surface of the road surface model by tire contact pressure and traffic frequency is applied in the form of distributed pressure that varies with time and space. The entire solution process is iterated in each time step until the solution of the thermal field and force field meets the pre-set coupling convergence tolerance. This convergence tolerance is a minimum value set according to the numerical calculation accuracy requirements and computational resource constraints. For example, it can be required that the absolute value of temperature change is less than 0.1 degrees Celsius in two consecutive iterations and the absolute value of stress change is less than 1 kPa. Through time-step solution, the complete mechanical response time series data of the road structure under the combined action of changing climatic boundary conditions and cyclic vehicle loads are finally obtained in the entire high-threat working condition time segment.

[0040] When calculating and outputting damage evolution data based on mechanical response, the mechanical response results calculated at each simulation time step, especially the stress and strain history of key areas, are used in conjunction with the fatigue performance parameters assigned to the material. The evolution of the material's damage state is calculated through a cumulative damage model. Specifically, stress-based fatigue damage laws, such as Miner's linear cumulative damage law, or strain energy-based nonlinear damage evolution equations are used to accumulate the damage generated by each load cycle. After each simulation time step, the damage state variable of each material integration point in the model is updated according to the selected damage evolution equation. This variable accumulates from its initial value, with a value range between 0 and 1, where 0 represents no damage and 1 represents complete failure. As the simulation progresses, the curves of the damage state variables of all material points changing over time are recorded, as well as the expansion of macroscopic damage indicators at the overall structural level, such as the area or volume of the damaged region. The final output damage evolution data includes the damage time history curves of all material integration points, the sequence of damage distribution cloud maps of each structural layer changing over time, and the sequence of macroscopic damage indicators reflecting the overall damage level development. These data fully characterize the degradation of pavement structure material performance and the expansion of damage under high-threat conditions.

[0041] S6. Based on damage evolution data, determine the performance degradation process according to the correspondence between damage state and service performance, and take the time corresponding to the performance degradation to the allowable limit as the remaining fatigue life. The specific implementation is as follows: When determining the performance degradation process based on damage evolution data, the damage state parameter sequence, which characterizes the degree of damage expansion within the pavement structure, is first extracted from the damage evolution data output in step S5. This damage evolution data includes the damage time history curves of all material integration points and the macroscopic damage index sequence. The method for extracting the damage state parameter sequence is to select one or more indicators that can comprehensively reflect the overall damage state of the structure. For example, the volume-weighted average of the damage state variables of all units in the pavement asphalt layer is calculated as the macroscopic damage index, or the percentage of the total volume of units whose damage state variable values ​​exceed a certain critical damage threshold is used as the damage area ratio. As the simulation progresses, the macroscopic damage index or damage area ratio corresponding to each output time point is calculated, thus forming a discrete data sequence with time as the horizontal axis and damage state parameters as the vertical axis, i.e., the damage state parameter sequence.

[0042] The specific method for setting and obtaining the critical damage threshold is as follows. This threshold is determined based on the physical mechanism relating the microscopic damage state of the material to its macroscopic mechanical properties. First, asphalt mixture specimens with different initial damage levels are prepared in the laboratory. The initial damage level is quantified by controlling the number of fatigue preloading cycles or introducing different proportions of microcracks. Subsequently, uniaxial compression or indirect tensile tests are conducted on these specimens to obtain their macroscopic mechanical property parameters, such as dynamic modulus or tensile strength. Simultaneously, microscopic observation data on the internal damage evolution of the specimens during the test are obtained using digital image correlation technology or acoustic emission monitoring. The statistical characteristic values ​​of the corresponding microscopic damage observation data, such as the average value of damage variables across all observed units, are analyzed when macroscopic mechanical properties show a sharp drop or inflection point. This statistical characteristic value of microscopic damage that causes a significant deterioration in macroscopic properties is set as the critical damage threshold. For example, when the dynamic modulus of the specimen is observed to drop to 70% of its initial value, the corresponding average value of the microscopic damage variables is approximately 0.5, and the critical damage threshold can be initially calibrated as 0.5. This calibration process needs to be repeated on multiple specimens, and the final critical damage threshold is the statistical average of all valid calibration results to ensure its representativeness.

[0043] The performance evolution model was established through regression analysis of historical data. The specific construction process is as follows: A large amount of on-site pavement inspection data from different road sections and service stages was collected. This data must appear in pairs, including pavement internal damage state assessment values ​​indirectly obtained through non-destructive testing and mechanical inversion, and pavement smoothness indices measured at the same location using a laser smoothness meter or inertial profiler. After data collection, preprocessing was required, including removing obvious outliers and normalizing damage state assessment values ​​from different sources to ensure comparability. The processed damage state data was then analyzed. The damage state assessment value is used as the independent variable, and the corresponding pavement smoothness index is used as the dependent variable to form a sample dataset for regression analysis. Statistical regression methods, such as multinomial regression or nonlinear least squares, are used to fit the mathematical function relationship between the damage state assessment value and the pavement smoothness index. During the fitting process, the model order or parameters are adjusted to make the coefficient of determination of the fit reach a preset model fit threshold to ensure that the model has sufficient interpretability. The final obtained and verified mathematical function relationship constitutes the performance evolution model describing the quantitative mapping relationship from damage state parameters to pavement smoothness index.

[0044] The specific methods for setting and obtaining the model fit threshold are as follows. This threshold is determined to objectively evaluate the reliability of the mathematical relationship obtained from statistical regression when building a performance evolution model, ensuring it has sufficient predictive power. Its setting is not a single fixed value, but rather a combination of conventional statistical requirements and the accuracy needs of specific engineering application scenarios. First, in statistics, goodness of fit is usually quantified by the coefficient of determination; the closer the coefficient of determination is to 1, the stronger the model's explanatory power for data variation. In engineering applications, models used for prediction and decision-making typically require high interpretability. Therefore, the initial reference range for the model fit threshold can be determined based on extensive engineering modeling experience. For example, in many pavement performance prediction models, the acceptable minimum coefficient of determination is usually set in the range of 0.7 to 0.9. Second, the specific value of the threshold needs to consider the quantity and quality of the historical data samples on which the model is based. When the sample size is sufficient and the data noise is low, a higher threshold can be set, such as 0.85, to pursue a more accurate model. When the sample size is limited or the data dispersion is large, the threshold requirement can be appropriately relaxed, for example, set to 0.75, to prioritize obtaining usable trend relationships. In practice, an initial value can be preset within the above empirical range, and then adjusted through the model validation step. For example, historical data can be randomly divided into training and validation sets, and the prediction error of the selected model on the validation set under different threshold requirements can be observed. Finally, a threshold that can stabilize the prediction error within an acceptable range can be selected as the final model fit threshold.

[0045] When calculating the performance degradation process based on the performance evolution model, the extracted damage state parameter sequence is input into the established performance evolution model. Using the mathematical function relationship described by the model, the damage state parameter value corresponding to each time point is converted into the estimated pavement smoothness index value. This conversion process is performed sequentially at all time points, thereby generating an estimated data sequence of pavement smoothness index changing with time, corresponding one-to-one with the time points of the damage state parameter sequence. The discrete data sequence is smoothed using curve fitting methods such as moving average or spline interpolation to obtain a continuous performance degradation process curve of pavement smoothness index changing with time. This curve intuitively shows the degradation trajectory of pavement service performance over time under high-threat conditions and load spectrum.

[0046] When determining the remaining fatigue life, a permissible limit for the pavement smoothness index needs to be defined based on the performance requirements of the pavement in engineering practice. On the performance degradation process curve that has been plotted, the time point corresponding to the first drop of the pavement smoothness index value to the permissible limit is found along the positive time axis. The difference between this time point and the simulation start time point represents the estimated time that the pavement will take from its current state to no longer meet the smoothness requirements under the working conditions. This estimated time is determined as the remaining fatigue life of the pavement structure under the current damage state and the coupling effect of a specific high-threat environment and load.

[0047] The specific methods for obtaining and setting the permissible limits for road surface smoothness index are as follows. Determining these limits involves a decision-making process that integrates engineering standards, historical data, and functional requirements. First, the design specifications and maintenance procedures applicable to the target road grade are investigated. These documents typically specify the range of standard maintenance values ​​for road surface smoothness. Second, a database of historical maintenance records for roads of the same grade within the jurisdiction is collected, extracting the measured values ​​of the road surface smoothness index recorded during each preventative or corrective maintenance decision. By statistically analyzing these data that triggered maintenance actions, the high-frequency intervals or specific quantiles of their distribution, such as the 80th percentile, are calculated to obtain a limit reference based on actual maintenance experience. Simultaneously, considering the requirements for driving comfort and safety, the correlation between different smoothness index levels and passenger subjective comfort scores or vehicle vibration acceleration can be established through vehicle dynamics simulation or real-vehicle driving evaluation. The smoothness index corresponding to the lower limit of the acceptable comfort score is used as another reference. Finally, by combining standard values ​​from regulations, empirical values ​​from historical data, and comfort limits from functional requirements, a final permissible limit for the pavement smoothness index for life assessment is determined through expert review or a weighted average method.

[0048] All calculations involved in the embodiments are dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to the actual situation.

[0049] It should be noted that this invention can be deployed on the device itself to realize embedded applications, or it can run on a PC or other terminal with a user interface, thereby meeting various hardware environments and usage requirements.

[0050] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions according to the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wireless or wired transmission; wired transmission methods include optical fiber, twisted pair, coaxial cable, etc.; wireless transmission includes infrared, microwave, etc. Computer-readable storage media can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more sets of available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0051] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0052] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0053] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0054] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0055] If a function is implemented as a software module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0056] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0057] In conclusion, the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for predicting the remaining service life of existing asphalt pavement structures, characterized in that, include: S1. Obtain the existing damage state data of the target asphalt pavement structure, and determine the initial state of the multiphysics coupled numerical simulation model based on the existing damage state data. S2. Obtain fatigue performance parameters of asphalt pavement materials for multi-physics coupled numerical simulation models through cross-scale analysis methods; S3. Obtain time-series climate data for the target area and construct environmental boundary conditions to drive the multiphysics coupled numerical simulation model. S4. Extract damage mode features from existing damage state data, and perform time-frequency analysis on climate time series data to calculate the time-varying coherence between the damage mode features and the climate time series data after time-frequency analysis, and identify high-threat working condition combinations. S5. Load the fatigue performance parameters of asphalt pavement materials and environmental boundary conditions into the multiphysics coupled numerical simulation model, and prioritize the high-threat working condition combination to run high-fidelity numerical simulation to calculate the damage evolution data of the target asphalt pavement structure under the set load spectrum. S6. Based on damage evolution data, determine the performance degradation process according to the correspondence between damage state and service performance, and take the time corresponding to the performance degradation to the allowable limit as the remaining fatigue life.

2. The method for predicting the remaining service life of existing asphalt pavement structures according to claim 1, characterized in that, S1 includes: The road surface deflection data, interlayer bonding status data, and internal defect data obtained through non-destructive testing and core sampling are used as existing damage status data. The existing damage state data are fused to extract the spatial stiffness attenuation characteristics and initial damage distribution characteristics of the pavement structure. The spatial stiffness attenuation characteristics and initial damage distribution characteristics are transformed into elastic modulus reduction parameters and initial damage field variables of the corresponding material regions in the multiphysics coupled numerical simulation model, so as to determine the initial state of the multiphysics coupled numerical simulation model.

3. The method for predicting the remaining service life of existing asphalt pavement structures according to claim 2, characterized in that, The multiphysics coupled numerical simulation model is a three-dimensional numerical model of the road structure established by the finite element method. It simulates the road response under the combined action of temperature, humidity and load by solving the heat conduction equation, moisture migration equation and mechanical equilibrium equation simultaneously. The geometric dimensions and material partitions of the multiphysics coupled numerical simulation model are determined based on the target road surface design drawings, and the model is constructed by assigning material property parameters and boundary conditions.

4. The method for predicting the remaining service life of existing asphalt pavement structures according to claim 1, characterized in that, S2 include: The microstructure images of asphalt mixtures are obtained and analyzed to extract the geometric configuration and spatial distribution information of the internal aggregates, asphalt binder and voids. Based on geometric configuration and spatial distribution information, a microscale mechanical simulation calculation is performed to obtain the local stress-strain response at the interface between asphalt mastic and aggregate at the microscale. The local stress-strain response at the microscale is correlated and calibrated with the fatigue test data of macroscopic specimens of asphalt mixtures, forming a cross-scale calibration relationship from microscopic response to macroscopic performance; Based on cross-scale calibration relationships, fatigue performance parameters for characterizing the macroscopic fatigue degradation of materials are calculated and used as fatigue performance parameters of asphalt pavement materials in multi-physics coupled numerical simulation models.

5. The method for predicting the remaining service life of existing asphalt pavement structures according to claim 1, characterized in that, S3 include: Historical meteorological monitoring data of the target area is obtained. The historical meteorological monitoring data includes data series of temperature data, humidity data and solar radiation intensity data over time, which constitute climate time series data. The climate time series data are cleaned and interpolated to obtain a continuous and complete time series of climate parameters. The continuous and complete time series of climate parameters are transformed into boundary condition parameters required for the calculation of heat conduction, moisture migration and thermal stress in the multiphysics coupled numerical simulation model, so as to construct the environmental boundary conditions used to drive the multiphysics coupled numerical simulation model.

6. The method for predicting the remaining service life of existing asphalt pavement structures according to claim 1, characterized in that, S4 include: Modal analysis was performed on existing damage state data to extract damage modal features that reflect the main damage modes of the pavement structure. Time-frequency transformation is performed on climate time-series data to obtain its energy distribution characteristics in the time-frequency domain; Calculate the time-varying coherence coefficients of damage modal characteristics and energy distribution characteristics in the time-frequency domain; Based on the time-frequency regions where the time-varying coherence coefficient exceeds the preset coherence threshold, the climate environment and time combination that pose a high threat to the evolution of pavement structure damage are identified as high-threat working condition combinations.

7. The method for predicting the remaining service life of existing asphalt pavement structures according to claim 6, characterized in that, The time-varying coherence coefficients of damage modal features and energy distribution features in the time-frequency domain are calculated using the wavelet coherence method, which includes: performing continuous wavelet transforms on the damage modal features and energy distribution features respectively to obtain their respective wavelet coefficients; calculating the cross-correlation coefficients and autocorrelation coefficients of the two wavelet coefficients at the same time-frequency point to obtain the time-varying coherence coefficients.

8. The method for predicting the remaining service life of existing asphalt pavement structures according to claim 1, characterized in that, S5 include: The fatigue performance parameters of asphalt pavement materials and environmental boundary conditions are assigned to the material property parameters and boundary condition parameters in the multiphysics coupled numerical simulation model. Based on the climate and environmental conditions and time segments corresponding to the high-threat operating conditions, corresponding simulation tasks are set in the multiphysics coupled numerical simulation model. A multiphysics coupled numerical simulation model was run to perform thermo-mechanical coupled high-fidelity numerical simulation, simulating the mechanical response of the pavement structure under the combined action of a set load spectrum and corresponding climatic environmental conditions. Based on the mechanical response, damage evolution data characterizing the performance degradation and damage propagation process of pavement structure materials are calculated and output.

9. The method for predicting the remaining service life of existing asphalt pavement structures according to claim 1, characterized in that, S6 include: Extract a sequence of damage state parameters characterizing the extent of damage propagation within the pavement structure from damage evolution data; Based on a pre-established performance evolution model that describes the quantitative mapping relationship between the damage state parameter sequence and the pavement smoothness index, the performance degradation process curve of the pavement smoothness index over time is calculated. On the performance degradation curve, the time of action corresponding to when the pavement smoothness index drops to the allowable limit specified in the engineering specifications is determined, and the time of action is taken as the remaining fatigue life.

10. The method for predicting the remaining service life of existing asphalt pavement structures according to claim 9, characterized in that, The performance evolution model is established through regression analysis of historical data, including: collecting measured data of pavement smoothness index under different damage states, and using statistical regression methods to fit the mathematical function relationship between damage state parameters and smoothness index, thereby obtaining a performance evolution model that describes the quantitative mapping relationship.

Images