A method and system for joint prediction of road performance degradation based on multi-source data

Abstract

The application belongs to the technical field of road performance prediction, and provides a road performance degradation joint prediction method and system based on multi-source data, which comprises the following steps: acquiring a multi-source database containing road static and dynamic data; constructing a road performance degradation joint prediction model based on the acquired multi-source database; predicting the road performance of a target road section according to the road performance degradation joint prediction model, and completing the road performance degradation joint prediction based on multi-source data.

Description

Technical Field

[0001] This invention belongs to the field of road performance prediction technology, specifically relating to a method and system for joint prediction of road performance degradation based on multi-source data. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] During use, road surface performance gradually deteriorates with the cumulative effect of traffic loads and the passage of time. When the damage reaches a predetermined standard, maintenance or repair measures are required to restore or improve its performance. Road surface performance prediction models are key tools for describing this performance evolution process and are an important foundation for scientific decision-making in road management systems. Accurate prediction models help to rationally formulate maintenance plans, optimize fund allocation, and improve road service levels.

[0004] Currently, there is a wealth of research on the degradation law of pavement performance, and various prediction models have been established, mainly including deterministic models, probabilistic models, grey prediction models, and other models.

[0005] Deterministic models are generally based on explicit mathematical or mechanical relationships and can be divided into three categories: mechanical methods, mechanical-empirical methods, and empirical methods. Mechanical methods, based on elastic or viscoelastic theories, calculate the stress, strain, or displacement response of the pavement under load through structural mechanics analysis. While theoretically rigorous, this method is computationally complex and highly dependent on material parameters and environmental conditions. In practical applications, its high data variability, cumbersome operation, and limited prediction accuracy limit its use. Mechanical-empirical methods, based on mechanical analysis results, combine experimental data to establish regression relationships between pavement performance and mechanical responses such as stress and strain. While balancing theory and experiment, these models often focus on structural damage and fail to comprehensively reflect the overall service performance of the pavement. Furthermore, their complex form makes them difficult to generalize. Empirical methods avoid complex mechanical calculations and directly utilize multiple regression techniques to establish statistical relationships between pavement performance and service time. This method is simple and practical, but its prediction accuracy is highly dependent on data quality and the modeler's understanding of the variable relationships.

[0006] Because pavement performance is affected by multiple factors such as traffic, climate, materials, construction, and maintenance, its degradation process exhibits significant uncertainty. Actual performance does not decline along a single, predictive curve but rather displays a degree of randomness. Therefore, probabilistic prediction models have emerged, capable of reflecting the confidence interval and risk level of prediction results, and better meeting the needs of actual engineering projects.

[0007] Grey prediction models are an effective method for handling "grey systems" where some information is known and some is unknown. By processing the raw data (such as through cumulative generation), randomness is reduced, potential patterns are uncovered, and a differential equation model is established for trend prediction. This method is suitable for situations with limited data and incomplete information, and has certain application value in road performance prediction.

[0008] In recent years, intelligent prediction methods, especially artificial neural networks, have been widely used in pavement performance prediction due to their powerful nonlinear fitting and self-learning capabilities. Neural networks learn the complex relationships between input historical detection data through training, constructing a mapping model between inputs (such as traffic volume, age, and maintenance history) and outputs (such as performance indicators like PCI and RQI). After training, this model can be used for simulation prediction of future performance. Compared to traditional models, neural networks have higher prediction accuracy and adaptability, and can effectively handle nonlinear problems under the coupling of multiple factors.

[0009] Deterministic mechanical models offer clear theory but have limited practicality; empirical models are simple but rely heavily on data; probabilistic models better reflect real-world uncertainties; and intelligent methods offer high accuracy but require substantial data support. The future trend lies in integrating the advantages of multiple models, combining big data and artificial intelligence technologies, to build a more accurate and robust pavement performance prediction system, providing strong support for scientific maintenance decisions. Summary of the Invention

[0010] To address the aforementioned issues, this invention proposes a joint prediction method and system for road performance degradation based on multi-source data. It integrates regression prediction, probability prediction, grey prediction, and artificial intelligence prediction to construct a multi-model collaborative prediction framework, thereby achieving a comprehensive characterization of the road performance degradation process and significantly improving the accuracy, robustness, and reliability of the prediction results.

[0011] According to some embodiments, the first aspect of the present invention provides a method for joint prediction of road performance degradation based on multi-source data, employing the following technical solution: A method for joint prediction of road performance degradation based on multi-source data includes: Obtain a multi-source database containing static and dynamic road data; Based on the acquired multi-source database, a joint prediction model for road performance degradation was constructed. The road performance of the target road segment is predicted based on the joint prediction model of road performance degradation, and the joint prediction of road performance degradation based on multi-source data is completed.

[0012] As a further technical limitation, before obtaining a multi-source database of road performance, multi-source data of road performance should be collected. The collected multi-source data should at least include basic information data of each route code, route name, direction, lane, starting station number and ending station number in a certain road network, including static attribute data of environmental location, service life, lane type, traffic load level and pavement structure type, as well as dynamic data including pavement performance test data and maintenance history records over the years.

[0013] Furthermore, using route code, direction, lane, and station interval as unique indexes for road segment units, valid road segment units that have undergone pavement performance testing in previous years are selected. Their corresponding static attribute information, as well as pavement performance testing data and maintenance history records from previous years, are integrated to construct a structured and standardized multi-source database.

[0014] Furthermore, the multi-source databases are classified and data preprocessing is performed before model construction. The data preprocessing includes at least maintenance interference removal and abnormal data cleaning.

[0015] As a further technical limitation, a joint prediction model for road performance degradation is constructed for various types of data, including at least a deterministic prediction model, an artificial intelligence prediction model, a probabilistic prediction model, and a grey prediction model.

[0016] As a further technical constraint, based on the static attribute parameters of the road segment to be predicted, a joint prediction model for road performance degradation of the corresponding type is matched, and prediction is carried out in conjunction with historical data of the road segment to be predicted. First, the prediction results for the next year (t0+1 year) of the deterministic prediction model, the artificial intelligence prediction model, and the grey prediction model are obtained, denoted as follows: , , .

[0017] Furthermore, based on the mapping rules used in the probabilistic prediction model, the current (year t0) performance detection data of the road segment to be predicted is... y Mapped to road condition status The prediction results of the three types of models , , Mapped to road condition status respectively , , Furthermore, the probabilistic prediction model is used to analyze the... Migrate to , , The corresponding transition probabilities are denoted as follows: , , ; Calculate the joint prediction results That is:

[0018] According to some embodiments, the second aspect of the present invention provides a road performance degradation joint prediction system based on multi-source data, which adopts the following technical solution: A joint prediction system for road performance degradation based on multi-source data includes: The acquisition module is configured to acquire a multi-source database containing static and dynamic road data; The building module is configured to construct a joint prediction model for road performance degradation based on the acquired multi-source database. The prediction module is configured to predict the road performance of the target road segment based on the joint prediction model for road performance degradation, and to complete the joint prediction of road performance degradation based on multi-source data.

[0019] According to some embodiments, a third aspect of the present invention provides a computer-readable storage medium, employing the following technical solution: A computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps in the road performance degradation joint prediction method based on multi-source data as described in the first aspect of the present invention.

[0020] According to some embodiments, the fourth aspect of the present invention provides an electronic device, which adopts the following technical solution: An electronic device includes a memory, a processor, and a program stored in the memory and running on the processor, wherein the processor executes the program to implement the steps in the road performance degradation joint prediction method based on multi-source data as described in the first aspect of the present invention.

[0021] According to some embodiments, the fifth aspect of the present invention provides a computer program product, which adopts the following technical solution: A computer program product includes software code, wherein the program in the software code performs the steps of the road performance degradation joint prediction method based on multi-source data as described in the first aspect of the present invention.

[0022] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention integrates multiple algorithms, including regression prediction, probabilistic prediction, grey prediction, and artificial intelligence prediction, to construct a multi-model collaborative prediction framework. By leveraging the complementary advantages of various algorithms in trend fitting, uncertainty quantification, small sample adaptation, and nonlinear modeling, it achieves a comprehensive characterization of pavement performance degradation processes, significantly improving the accuracy, robustness, and reliability of prediction results.

[0023] At the data level, this invention integrates historical detection data from multiple regions and types of roads, covering multi-dimensional information such as different environmental locations, road structure types, and traffic load levels, to achieve multi-source fusion and refined classification of data. By dividing the data into scenarios and features, it differentiates the road performance degradation patterns under different usage conditions, and the constructed model library has good adaptability to complex application scenarios.

[0024] At the model level, this invention constructs a "multi-model joint prediction" architecture, which integrates the advantages of deterministic prediction models, probabilistic prediction models and intelligent learning models (grey prediction, neural networks, etc.). It adopts an integrated learning and weighted fusion strategy to collaboratively optimize and complement the output results of each sub-model. It can not only capture the overall trend of performance degradation, but also effectively reflect the random fluctuations and uncertainties in the degradation process, and the distribution of prediction results is more consistent with reality. Attached Figure Description

[0025] The accompanying drawings, which form part of this embodiment, are used to provide a further understanding of this embodiment. The illustrative embodiments and their descriptions are used to explain this embodiment and do not constitute an improper limitation of this embodiment.

[0026] Figure 1 This is a flowchart of the road performance degradation joint prediction method based on multi-source data in Embodiment 1 of the present invention; Figure 2 This is an architecture diagram of the road performance degradation joint prediction method based on multi-source data in Embodiment 1 of the present invention; Figure 3 This is a schematic diagram of the road segment attribute classification rule system in Embodiment 1 of the present invention; Figure 4 This is a schematic diagram of the training data input format for the artificial intelligence prediction model in Embodiment 1 of the present invention; Figure 5 This is a schematic diagram of the road condition status marking of a certain road segment unit in a certain year in Embodiment 1 of the present invention; Figure 6 This is a schematic diagram of the distribution cloud of a certain probability transition matrix P in Embodiment 1 of the present invention; Figure 7 This is a flowchart illustrating the use of the road performance degradation joint prediction model in Embodiment 1 of the present invention. Figure 8 This is a schematic diagram of the rolling forecasting process for the next few years in Embodiment 1 of the present invention; Figure 9 This is a structural block diagram of the road performance degradation joint prediction system based on multi-source data in Embodiment 2 of the present invention. Detailed Implementation

[0027] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0028] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, 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 pertains.

[0029] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0030] In this invention, terms such as "upper," "lower," "left," "right," "front," "back," "vertical," "horizontal," "side," and "bottom" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used only to facilitate the description of the structural relationships of the various components or elements of this invention and do not specifically refer to any component or element in this invention. They should not be construed as limiting the invention.

[0031] In this invention, terms such as "fixed connection," "connected," and "linked" should be interpreted broadly, indicating a fixed connection, an integral connection, or a detachable connection; a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can determine the specific meaning of these terms in this invention based on the specific circumstances, and they should not be construed as limitations on the invention.

[0032] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0033] Terminology Explanation: In this embodiment, road performance refers to the pavement technical condition. The construction and prediction of performance degradation models for other indicators can refer to this method.

[0034] Environmental location refers to the overall description of the environmental characteristics of a specific road segment or geographical unit within its spatial distribution. This includes, but is not limited to, natural elements such as climate conditions (temperature, precipitation, and wind), topography, hydrological characteristics, and soil type, as well as differences in the surrounding environment affected by human activities (such as urbanization level, pollution level, and ecological sensitivity). Its core lies in emphasizing the environmental differences between different road segments and the potential impact of these differences on road performance evolution. The classification dimensions can be selected according to actual needs, such as by climate (temperate, tropical regions, etc.); by natural geographical features (mountainous, plain, coastal, etc.); or by political and economic factors (core area, important area, general area, etc.).

[0035] Example 1 Embodiment 1 of this invention introduces a joint prediction method for road performance degradation based on multi-source data.

[0036] Current road performance prediction models are mostly built based on data from local areas or specific road sections, with relatively simple application scenarios. They generally lack systematic consideration of the differences in key factors such as different environmental locations, road structure types, and traffic load levels. This results in insufficient generalization ability and adaptability of the models in complex scenarios of cross-regional and cross-type roads, making it difficult to meet the diverse needs of road network management. The evolution of pavement performance is affected by a variety of random factors such as climate change, traffic growth, and maintenance intervention. Actual performance degradation often shows a non-stationary and fluctuating downward trend. Most traditional prediction models still simplify it to a linear or fixed pattern, ignoring the randomness and variability in the degradation process and lacking a quantitative expression of the confidence probability of the prediction results, which limits its reliability in actual decision-making. Intelligent prediction models based on big data and machine learning possess powerful nonlinear fitting capabilities and adaptive learning characteristics. Their performance heavily relies on sufficient, high-quality, and multi-dimensional historical data. In reality, road detection data often faces challenges such as small sample sizes, missing data, outlier interference, and inconsistent data collection standards. These issues severely restrict the model's training effectiveness and generalization ability, impacting prediction stability. Different pavement performance prediction methods have their own advantages and limitations in terms of time scale applicability, data requirements, prediction accuracy and engineering operability. Existing research and practice mostly use a single model for prediction, lacking the collaborative integration and fusion of multiple algorithms. The integration mechanism of multi-model output results is not yet sound, and the complementary characteristics of various models have not been effectively utilized. The robustness, stability and accuracy of prediction results still have room for improvement.

[0037] Therefore, this embodiment proposes a method such as Figure 1 and Figure 2 The road performance degradation joint prediction method based on multi-source data shown includes: Obtain a multi-source database containing static and dynamic road data; Based on the acquired multi-source database, a joint prediction model for road performance degradation was constructed. The road performance of the target road segment is predicted based on the joint prediction model of road performance degradation, and the joint prediction of road performance degradation based on multi-source data is completed.

[0038] As one or more implementation methods, this embodiment uses a pavement maintenance management system to acquire multi-year, multi-dimensional data from various expressways within the provincial road network. Basic information data includes route codes, route names, directions, lanes, starting and ending chainages, etc. Static attribute data includes environmental location, service life, lane type, traffic load level, and pavement structure type. Dynamic data includes historical PQI, PCI, RQI, RDI, SRI, and other pavement performance testing data, as well as maintenance history records, providing solid data support for subsequent analysis. This embodiment uses route codes, directions, lanes, and chainage intervals as unique indexes for road segment units, filtering out valid road segment units that have undergone pavement performance testing in previous years. It then integrates their corresponding static attribute information, along with historical pavement performance testing data and maintenance history records, to construct a structured and standardized multi-source database. An example of the data structure for a specific road segment unit is shown in Table 1.

[0039] Table 1. Database Data Structure Diagram

[0040] In this embodiment, various static attribute indicators are graded, and a systematic road segment attribute classification rule system is established through the permutation and combination of different indicator levels. Based on this rule, all road segment units in the database are classified and divided, constructing subsets oriented towards different attribute conditions. During the classification process, the data distribution balance of each category must be considered to ensure that the sample size of each subset is relatively reasonable; if significant imbalance occurs, the grading strategy of some indicators should be dynamically adjusted to optimize the subset distribution; thus, the following is obtained: Figure 3 The classification rule system shown.

[0041] Data preprocessing for each subset of data mainly includes: (1) Eliminate the impact of maintenance interference. By comparing the spatial location of each road segment unit with the implementation scope of maintenance projects over the years, road segment units that have not been subject to maintenance intervention are selected, effectively eliminating the interference of maintenance measures on the road surface technical condition, thereby more realistically reflecting the evolution law of road surface performance under natural use conditions.

[0042] (2) Abnormal data cleaning. The box plot method is used to statistically analyze the road surface technical condition indicators over the years, identify and remove outliers that deviate significantly from the main distribution; at the same time, for unreasonable data that violates the engineering rules during the decay process (such as large fluctuation data, continuous full score data, etc.), systematic cleaning and correction are carried out to ensure the rationality and reliability of the data sequence.

[0043] Predictive models are trained on the preprocessed subsets, including deterministic prediction models, artificial intelligence prediction models, and probabilistic prediction models. In this embodiment, the deterministic prediction model, for each subset, uses service life as the independent variable and various technical condition indicators as the dependent variable, and performs regression analysis using various functional forms. The functional expressions are shown in Table 2. By comparing the key evaluation indicators such as RMSE, SSE, and R² of different regression models, the model with the best regression performance is selected as the optimal deterministic prediction model for that subset. An example of the comparison process is shown in Table 3. The above modeling process is executed sequentially for all subsets, ultimately constructing a set of deterministic prediction models corresponding one-to-one with each subset, forming a deterministic prediction model library; an example of a deterministic prediction model library for PCI of a certain province's expressway is shown in Table 4.

[0044] Table 2 Deterministic Prediction Model Library

[0045] Table 3. Comparison and selection process of the optimal deterministic prediction model Model Name RMSE SSE <![CDATA[R 2 ]]> Is it the optimal model? Negative exponential 0.42298 1.2524 0.7666 no Power type 0.6741 3.1809 0.7664 no linear model 0.4234 1.2546 0.7662 no Two-parameter model 0.42294 1.25213 0.7767 yes Table 4. Examples of PCI deterministic prediction models for highways in a certain province

[0046] In this embodiment, the artificial intelligence prediction model constructs an input data format suitable for the artificial intelligence model based on the historical technical status data of each subset of datasets. For example... Figure 4 As shown, the input data for each subset consists of time-series data of all road segment units under that category, organized as a three-dimensional tensor structure with a shape of (batch, seq, feature). Here, batch represents the number of road segment units; seq represents the length of the detection history time series, usually no less than 5 years to ensure the effectiveness of trend learning; and feature represents the detection indicators over the years, generally including 5 core indicators such as PQI, PCI, RQI, RDI, and SRI.

[0047] A feedback neural network (such as LSTM), which has advantages in temporal modeling, is used to learn and train the above-mentioned structured time-series data. By systematically adjusting key hyperparameters such as the number of hidden layer memory units and the number of network layers, and combining this with validation set performance evaluation, the model structure is optimized to improve prediction accuracy. Finally, the trained artificial intelligence prediction model is saved and exported.

[0048] The above modeling process is executed sequentially for all subsets of data, ultimately constructing a set of artificial intelligence prediction models that correspond one-to-one with each subset, forming an artificial intelligence prediction model library. An example of an artificial intelligence prediction model library for highways in a certain province is shown in Table 5.

[0049] Table 5. Examples of Artificial Intelligence Prediction Models for Expressways in a Certain Province

[0050] The probabilistic prediction model in this embodiment uses a Markov model to perform probabilistic analysis and prediction of the road performance evolution process: (1) Divide various technical condition indicators (such as PQI, PCI, RQI, RDI, SRI) into several levels according to performance level, and construct a complete road condition status characterization system. The combination of detection data (PQI, PCI, RQI, RDI, SRI) for each road segment unit in a certain year can be transformed into a road condition status composed of level labels for each indicator through this road condition status characterization system. An example of a road condition status label is as follows: Figure 5 As shown.

[0051] (2) For the historical technical condition data of each subset, identify the specific road condition status of each road segment unit in each year, and then count the frequency of the transition of a certain road condition status to the next year's road condition status, and finally construct the probability transition matrix P corresponding to the subset; the distribution cloud map of a certain probability transition matrix P is shown in Figure 1. Figure 6 As shown.

[0052] The probability transition matrix provides a quantitative basis for predicting the uncertainty of pavement performance, and the following relationship exists: ; (2) The above modeling process is executed sequentially for all subsets of data to finally construct a set of probability prediction models that correspond one-to-one with the subsets of data, forming a probability prediction model library; an example of the probability prediction model library for a certain province's expressway is shown in Table 6.

[0053] Table 6. Examples of Probabilistic Prediction Models for Expressways in a Certain Province

[0054] As one or more implementation methods, when performing road performance prediction on a target road segment (which may contain multiple road segment units), this embodiment requires the following preparation of input data: (1) Data collection and integration: Read the complete information of the target road segment, including static attribute data such as environmental location, service life, lane type, traffic load level, and current pavement structure, as well as the technical condition detection data of each road segment unit such as PQI, PCI, RQI, RDI, and SRI over the years, to ensure the integrity and temporal continuity of the data.

[0055] (2) Model matching: Based on the established road segment attribute classification rule system, all road segment units within the target road segment are classified; then, from each prediction model library, the corresponding deterministic prediction model, artificial intelligence prediction model and probabilistic prediction model are accurately matched to ensure the consistency between the prediction model and the road segment characteristics.

[0056] (3) Input Formatting: According to the defined three-dimensional data structure, organize the historical detection data of each unit of the target road segment into a three-dimensional tensor with shape (batch, seq, feature); where batch is the number of units contained in the road segment, seq is the length of the historical sequence, and feature is the dimension of the detection index. Special attention should be paid to the fact that the historical detection duration of the target road segment should not be less than the minimum sequence length required by the model (usually ≥5 years), otherwise the artificial intelligence prediction model cannot be effectively invoked for inference.

[0057] The process of constructing the grey prediction model in this embodiment is as follows: (1) Perform a level comparison test on the historical data series of each technical condition index of the target road section. Let the original historical data series be... Then its series ratio sequence is If the series ratio is arbitrary All meet If the result is positive, the grade ratio test passes; otherwise, it fails.

[0058] (2) For datasets that pass the grade ratio test and meet the requirements for grey system modeling, the GM(1,1) model is used for prediction modeling.

[0059] First, the original technical condition sequence is accumulated once. Based on the generated sequence, a first-order univariate grey differential equation is established, which has the following form: ; The development coefficient 'a' and the grey action quantity 'b' are solved using the least squares method.

[0060] (3) The predicted values ​​of performance over the years can be calculated using the following formula: ; This embodiment uses, as follows: Figure 7 The multi-model joint prediction shown is implemented as follows: (1) By using a deterministic prediction model and an artificial intelligence prediction model matched with the target road segment, and combining them with a grey prediction model constructed from historical data of this road segment, a multi-model joint prediction can be carried out, resulting in three performance prediction results, denoted as... , , ; (2) Through the road condition status representation system, the values ​​of each performance index are mapped to the road condition status. The current road condition status is mapped to... The three performance prediction results are mapped to three different road condition states, denoted as follows: , , ;Analyze the probabilistic prediction model matched with the model. Migrate to years , , The corresponding transition probabilities are denoted as follows: , , ; (3) Calculate the joint prediction results based on this:

[0061] like Figure 8 As shown, in this embodiment, the first The prediction results for the current year are used as new input information, embedded into the original sequence, and used for recursive predictions of subsequent years. Specifically, the prediction results for the current year are used as new input information, embedded into the original sequence, and used for recursive predictions of subsequent years. The predicted value for the year is appended to the end of the constructed original input data sequence, while the oldest historical data in the sequence is removed to ensure that the length of the input sequence remains constant at seq, forming the model input for year t+1. Based on the updated input data, the grey model update and multi-model joint prediction are repeated sequentially to obtain the predicted value for the year. The combined forecast results for the year. By analogy, the development of road performance in the coming years can be predicted.

[0062] It should be noted that in some cases, the historical detection time of the road segment to be predicted may be less than the minimum sequence length required for the artificial intelligence prediction model or the construction of the grey prediction model (usually ≥ 5 years). In this case, it may not be possible to simultaneously enable the deterministic prediction model, the artificial intelligence prediction model, and the grey prediction model. When only two of the models are enabled, the following joint prediction process can still be carried out: (1) By using two types of models matched with the target road segment for prediction, two performance prediction results can be obtained, denoted as follows: , ; (2) Through the road condition status representation system, the values ​​of each performance index are mapped to the road condition status. The current road condition status is mapped to... The two performance prediction results are mapped to two different road condition states, denoted as... , ;Analyze the probabilistic prediction model matched with the model. Migrate to years , The corresponding transition probabilities are denoted as follows: , ; (3) Calculate the joint prediction results based on this:

[0063] When only one type of model is used, the prediction result of that model is the final result. At this point, the transition probability given by the probabilistic prediction model... This represents the confidence level of the result.

[0064] This embodiment integrates multiple algorithms such as regression prediction, probability prediction, grey prediction, and artificial intelligence prediction to construct a multi-model collaborative prediction framework. By leveraging the complementary advantages of various algorithms in trend fitting, uncertainty quantification, small sample adaptation, and nonlinear modeling, it achieves a comprehensive characterization of the road performance degradation process, significantly improving the accuracy, robustness, and reliability of the prediction results.

[0065] At the data level, this embodiment integrates historical detection data from multiple regions and types of roads, covering multi-dimensional information such as different environmental locations, road structure types, and traffic load levels, to achieve multi-source fusion and refined classification of data; by dividing the data into scenarios and features, it differentiates the road performance degradation patterns under different usage conditions, and the constructed model library has good adaptability to complex application scenarios.

[0066] At the model level, this embodiment constructs a "multi-model joint prediction" architecture, which integrates the advantages of deterministic models, probabilistic models and intelligent learning models (grey prediction, neural networks, etc.). It adopts an ensemble learning and weighted fusion strategy to collaboratively optimize and complement the output results of each sub-model. It can not only capture the overall trend of performance degradation, but also effectively reflect the random fluctuations and uncertainties in the degradation process, and the distribution of prediction results is more consistent with reality.

[0067] Example 2 Embodiment 2 of the present invention introduces a joint prediction system for road performance degradation based on multi-source data.

[0068] like Figure 9 The system shown is a joint prediction system for road performance degradation based on multi-source data, comprising: The acquisition module is configured to acquire a multi-source database containing static and dynamic road data; The building module is configured to construct a joint prediction model for road performance degradation based on the acquired multi-source database. The prediction module is configured to predict the road performance of the target road segment based on the joint prediction model for road performance degradation, and to complete the joint prediction of road performance degradation based on multi-source data.

[0069] The detailed steps are the same as those of the road performance degradation joint prediction method based on multi-source data provided in Example 1, and will not be repeated here.

[0070] Example 3 Embodiment 3 of the present invention provides a computer-readable storage medium.

[0071] A computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps in the road performance degradation joint prediction method based on multi-source data as described in Embodiment 1 of the present invention.

[0072] The detailed steps are the same as those of the road performance degradation joint prediction method based on multi-source data provided in Example 1, and will not be repeated here.

[0073] Example 4 Embodiment 4 of the present invention provides an electronic device.

[0074] An electronic device includes a memory, a processor, and a program stored in the memory and running on the processor. When the processor executes the program, it implements the steps in the road performance degradation joint prediction method based on multi-source data as described in Embodiment 1 of the present invention.

[0075] The detailed steps are the same as those of the road performance degradation joint prediction method based on multi-source data provided in Example 1, and will not be repeated here.

[0076] Example 5 Embodiment 5 of the present invention provides a computer program product.

[0077] A computer program product includes software code, wherein the program in the software code performs the steps of the road performance degradation joint prediction method based on multi-source data as described in Embodiment 1 of the present invention.

[0078] The detailed steps are the same as those of the road performance degradation joint prediction method based on multi-source data provided in Example 1, and will not be repeated here.

[0079] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention can be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0080] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0081] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0082] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0083] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0084] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

[0085] The above description is merely a preferred embodiment of this practice and is not intended to limit the scope of this practice. Various modifications and variations can be made to this practice by those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of this practice should be included within the protection scope of this practice.

Claims

1. A method for joint prediction of road performance degradation based on multi-source data, characterized in that, include: Obtain a multi-source database containing static and dynamic road data; Based on the acquired multi-source database, a joint prediction model for road performance degradation was constructed. The road performance of the target road segment is predicted based on the joint prediction model of road performance degradation, and the joint prediction of road performance degradation based on multi-source data is completed.

2. The method for joint prediction of road performance degradation based on multi-source data as described in claim 1, characterized in that, Before obtaining a multi-source database of road performance, multi-source data on road performance should be collected. The collected multi-source data should at least include basic information data on each route code, route name, direction, lane, starting station number, and ending station number in a road network, including static attribute data such as environmental location, service life, lane type, traffic load level, and pavement structure type, as well as dynamic data including pavement performance test data and maintenance history records over the years.

3. The method for joint prediction of road performance degradation based on multi-source data as described in claim 2, characterized in that, Using route code, direction, lane, and station interval as unique indexes for road segment units, valid road segment units that have undergone pavement performance testing in previous years are selected. Their corresponding static attribute information, as well as pavement performance testing data and maintenance history records from previous years, are integrated to construct a structured and standardized multi-source database.

4. The method for joint prediction of road performance degradation based on multi-source data as described in claim 1, characterized in that, The multi-source database is classified and data preprocessing is performed before model building. The data preprocessing includes at least maintenance interference removal and abnormal data cleaning.

5. The method for joint prediction of road performance degradation based on multi-source data as described in claim 1, characterized in that, A joint prediction model for road performance degradation is constructed for various types of data, including at least a deterministic prediction model, an artificial intelligence prediction model, a probabilistic prediction model, and a grey prediction model.

6. The method for joint prediction of road performance degradation based on multi-source data as described in claim 5, characterized in that, Based on the static attribute parameters of the road segment to be predicted, a joint prediction model for road performance degradation of the corresponding type is matched, and prediction is carried out in conjunction with historical data of the road segment to be predicted. First, the prediction results of the deterministic prediction model, the artificial intelligence prediction model, and the grey prediction model for the next year (t0+1 year) are obtained, and are denoted as follows: , , ; Based on the mapping rules used in the probabilistic prediction model, the current (year t0) performance detection data of the road segment to be predicted is used. y Mapped to road condition status The prediction results of the three types of models , , Mapped to road condition status respectively , , Furthermore, the probabilistic prediction model is used to analyze the... Migrate to , , The corresponding transition probabilities are denoted as follows: , , ; Calculate the joint prediction results That is: 。 7. A joint prediction system for road performance degradation based on multi-source data, characterized in that, include: The acquisition module is configured to acquire a multi-source database containing static and dynamic road data; The building module is configured to construct a joint prediction model for road performance degradation based on the acquired multi-source database. The prediction module is configured to predict the road performance of the target road segment based on the joint prediction model for road performance degradation, and to complete the joint prediction of road performance degradation based on multi-source data.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the road performance degradation joint prediction method based on multi-source data as described in any one of claims 1-6.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the program, it implements the steps of the road performance degradation joint prediction method based on multi-source data as described in any one of claims 1-6.

10. A computer program product, comprising software code, characterized in that, The program in the software code executes the steps of the road performance degradation joint prediction method based on multi-source data as described in any one of claims 1-6.

Images