A Multi-Scenario Asphalt Mixture Design System Based on Comprehensive Performance Scoring Reliability

Abstract

This invention relates to the field of asphalt mixture design technology, and more particularly to a multi-scenario asphalt mixture design system based on the reliability of comprehensive performance scoring. The invention integrates the opinions of multiple experts through a multi-scenario expert evaluation proxy module, and uses artificial intelligence algorithms to construct a mapping model between climate zones and the importance of road performance. It standardizes the differences in evaluations of different experts regarding climate zones such as high temperature, low temperature, and rainfall. This standardization process not only ensures consistency in evaluation standards across different climate zones but also dynamically adapts to the needs of multiple scenarios through the proxy model, significantly improving the universality and reliability of the design scheme. It avoids the scheme conflicts or performance imbalances caused by ambiguous standards in traditional manual evaluation. Finally, it selects the most reliable scheme by calculating and comparing β values. This solves the technical problems of inconsistent standards and low reliability in current methods of manually selecting asphalt mixtures.

Description

Technical Field

[0001] This invention relates to the field of asphalt mixture design technology, and in particular to a multi-scenario asphalt mixture design system based on the reliability of comprehensive performance scoring. Background Technology

[0002] Asphalt mixtures are widely used on roads of all levels due to their excellent road performance and ease of construction. However, the market offers a wide variety of products with varying quality, and different asphalt mixtures have different road performance characteristics. To select the most suitable asphalt mixture solution for a specific construction application environment, it is usually necessary for experienced technical personnel to manually confirm the asphalt mixture design based on the actual construction application environment and the results of indoor road performance tests. This manual selection method relies heavily on human judgment and has a strong subjective element. In practical applications, it has shortcomings such as inconsistent standards and results that are easily influenced by individual cognitive limitations, thus limiting its feasibility and reliability to some extent. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides a multi-scenario asphalt mixture design system based on the reliability of comprehensive performance scoring, which solves the technical problems of inconsistent standards and low reliability in the current method of manually selecting asphalt mixtures.

[0004] To address the aforementioned technical problems, this invention provides the following technical solution: a multi-scenario asphalt mixture design system based on comprehensive performance scoring reliability, the system comprising:

[0005] The custom design parameter determination module is used to obtain all regional samples within the sample area according to the designer's needs, classify the regional samples according to preset rules to set several levels of climate zones, and set the road performance of asphalt mixture for different climate zones.

[0006] A multi-scenario expert evaluation proxy module is used to obtain the importance evaluations of asphalt mixture road performance under different climate zones from several experts, and to establish a multi-scenario expert evaluation proxy model to unify the importance evaluations of road performance under different climate zones.

[0007] The multi-scenario expert scoring agent module is used to obtain the road performance data of asphalt mixture to establish a road performance database, and to evaluate the importance of the unified road performance, using the analytic hierarchy process to obtain the weight of the road performance and calculate the comprehensive performance score of the asphalt mixture.

[0008] The reliability design module is used to establish the type of function function and function function random variable based on the comprehensive performance score, action effect and resistance relationship of asphalt mixture road, and calculate the reliability index of the new asphalt mixture through the reliability index function. Then, the asphalt mixture with the largest reliability index is selected as the optimal design scheme.

[0009] Preferably, the custom design parameter determination module includes:

[0010] Climate zoning unit, which is used to set several levels of climate zoning and division rules, and to divide the regional sample into different climate zoning based on the climate, soil conditions and traffic load of the regional sample.

[0011] A road performance unit is used to define the road performance requirements of asphalt mixtures for different climate zones.

[0012] Preferably, the multi-scenario expert evaluation agent module includes:

[0013] The expert evaluation acquisition unit is used to collect opinions from several experts, obtain the importance evaluation of each expert on road performance under different climate zones, construct the importance vector of road performance in different climate zones, and establish a multi-scenario expert scoring database.

[0014] An evaluation agent model building unit is used to construct a mapping relationship between climate zones and the importance of road performance through artificial intelligence algorithms, so as to establish a multi-scenario expert evaluation agent model that simulates the thinking and evaluation logic of experts.

[0015] The importance proxy unit is used to re-input each road performance importance vector from the multi-scenario expert scoring database into the model, and obtain the importance value of road performance through the multi-scenario expert evaluation proxy model.

[0016] Preferably, the evaluation agent model building unit uses a bootstrap method to take 80% of the samples in the multi-scenario expert scoring database as the training set and the remaining 20% ​​as the test set. The climate zone combinations in the training set are used as the input to the artificial intelligence machine learning model, and the road performance importance vector corresponding to the climate zone is used as the output of the artificial intelligence machine learning model. After training, a multi-scenario expert evaluation agent model is obtained. The model output layer uses a linear activation function to output the importance score of road performance. The model loss function uses the mean squared error loss function, and the MSE loss function formula is:

[0017]

[0018] In the above formula, LOSS is the mean squared error loss value, n represents the number of samples in the multi-scenario expert scoring database, C is the number of road performance parameters to be considered, and y ic The actual expert score for the i-th road performance item. The score is the importance score of the i-th road performance item predicted by the multi-scenario expert evaluation agent model, and the score reflects the importance of various road performance items.

[0019] Preferably, the multi-scenario expert scoring proxy module includes:

[0020] A road performance acquisition unit is used to acquire road performance data of asphalt mixtures and integrate them into a road performance database.

[0021] The expert scoring calculation unit is used to obtain the weights of road performance based on the experts' evaluation of the importance of road performance, and to calculate the comprehensive performance score of different asphalt mixtures based on the obtained weights and the samples in the road performance database.

[0022] The scoring proxy model establishment unit is used to construct a mapping relationship from the importance of road performance to the comprehensive performance score through artificial intelligence algorithms, so as to establish a multi-scenario comprehensive performance scoring model.

[0023] Preferably, the formula for calculating the comprehensive performance score of the asphalt mixture is:

[0024]

[0025] In the above formula, P j For the comprehensive performance score of any asphalt mixture, α i Let α be the value of the i-th indoor test index of the road performance of the asphalt mixture. i，max This represents the maximum value of the i-th indoor test index for road performance in the database.

[0026] Preferably, the reliability design module includes:

[0027] A random variable determination unit is used to determine the type of random variable of the function based on the comprehensive performance score;

[0028] A function determination unit is used to establish a function based on the relationship between the effect and the resistance.

[0029] A reliability index calculation unit is used to establish a reliability index function for the comprehensive performance of asphalt mixtures using the first second moment method.

[0030] An asphalt mixture design unit is used to conduct road performance tests on new asphalt mixtures, calculate reliability indices based on reliability index functions, and select asphalt mixtures with high reliability indices as the optimal design scheme.

[0031] By employing the above technical solution, the present invention provides a multi-scenario asphalt mixture design system based on comprehensive performance score reliability, which has at least the following beneficial effects:

[0032] 1. This invention integrates the opinions of multiple experts through a multi-scenario expert evaluation proxy module, and uses artificial intelligence algorithms to construct a mapping model between the importance of climate zones and road performance. It standardizes the differences in evaluations of different experts on climate zones such as high temperature, low temperature, and rainfall. This standardization process not only ensures the consistency of evaluation standards under different climate zones, but also dynamically adapts to the needs of multiple scenarios through the proxy model, significantly improving the universality and credibility of the design scheme, and avoiding the problems of scheme conflict or performance imbalance caused by the ambiguity of standards in traditional manual evaluation.

[0033] 2. This invention is based on the analytic hierarchy process and a road performance database. It combines expert weights and test data to calculate a comprehensive performance score, ensuring the comparability of the score results. Furthermore, it establishes a nonlinear mapping relationship between road performance and the score through an AI model, dynamically corrects weight biases, and reduces subjective human intervention. Compared with traditional experience-based evaluation, this method significantly improves the objectivity and engineering applicability of the score through data-driven quantitative analysis, providing accurate input for subsequent reliability design.

[0034] 3. This invention, based on function and random variable analysis, transforms comprehensive performance scores into reliability indicators, quantitatively assesses the failure probability of asphalt mixtures, constructs a function using the first second-moment method, and calculates the mean and standard deviation of the function by combining log-normally distributed random variables of road performance, ultimately deriving the reliability indicator β value. By comparing β values, highly reliable solutions are selected. This method overcomes the limitations of empirical limits in traditional specifications, quantifies the redundancy of material performance from a probabilistic perspective, provides a scientific basis for the design of asphalt mixtures under extreme climate or complex load conditions, and significantly improves the durability and safety reserve of road structures.

[0035] 4. This invention divides the climate into multiple levels based on multi-dimensional parameters such as climate, soil, and traffic load through a custom design parameter module. It also dynamically configures the road performance priority by combining secondary zones such as salt erosion and traffic volume. By establishing a mapping rule between climate zones and performance requirements, the system can automatically generate targeted design objectives. For example, it can enhance corrosion resistance in salt erosion zones and focus on water stability in humid zones. This refined zoning strategy ensures that asphalt mixtures can achieve optimal performance configuration in different scenarios such as high temperature and rain, severe cold and snow accumulation, and salt and alkali corrosion, solving the problem of material waste or performance shortcomings caused by traditional "one-size-fits-all" design. Attached Figure Description

[0036] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0037] Figure 1 This is a structural block diagram of the multi-scenario asphalt mixture design system based on comprehensive performance scoring reliability according to the present invention.

[0038] In the diagram: 1. Custom design parameter determination module; 11. Climate zoning unit; 12. Road performance unit; 2. Multi-scenario expert evaluation proxy module; 21. Expert evaluation acquisition unit; 22. Evaluation proxy model establishment unit; 23. Importance proxy unit; 3. Multi-scenario expert scoring proxy module; 31. Road performance acquisition unit; 32. Expert scoring calculation unit; 33. Scoring proxy model establishment unit; 4. Reliability design module; 41. Random variable determination unit; 42. Functional function determination unit; 43. Reliability index calculation unit; 44. Asphalt mixture design unit. Detailed Implementation

[0039] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. This will allow for a full understanding of how the present application uses technical means to solve technical problems and achieve technical effects, and to facilitate its implementation.

[0040] Reliability theory is a reliable quantitative calculation method that is now widely used in the field of civil engineering, especially in the design of engineering structures and components, where it plays a key role. However, unlike engineering structures and components with relatively clear load effects, the effects of asphalt mixtures on road structures are difficult to determine due to their viscoelastic properties. Unlike definite mechanical indicators such as compressive strength, which can be directly used as the basis for road structure design using reliability theory, many road performance indicators do not have clear design values. Instead, the minimum values ​​of empirical indoor test indicators for road performance are often used to standardize the design of asphalt mixtures. This uniform minimum limit makes it difficult to apply reliability theory in the design of asphalt mixtures. In fact, after modification with various admixtures, the indoor test values ​​of asphalt mixture road performance indicators often far exceed the minimum values ​​specified in the standards.

[0041] Therefore, to address the technical problems of inconsistent standards and low reliability in current methods of manually selecting asphalt mixtures, this invention provides a multi-scenario asphalt mixture design system based on the reliability of comprehensive performance scoring. This system can quantitatively calculate the reliability of comprehensive performance scoring to obtain the probability that asphalt mixtures meet design expectations under different road service conditions, enabling a more direct and reliable selection of the asphalt mixture with the best comprehensive performance. The system includes:

[0042] Module 1, used to determine custom design parameters for all regional samples within a sample area based on designer needs, classifies regional samples according to preset rules to set several levels of climate zones, and sets the pavement performance of asphalt mixtures for different climate zones. For example, if China is set as the sample area, it obtains possible combinations of climate zones for all regions within China (i.e., regional samples), and then sets pavement performance that may affect the performance of asphalt mixtures according to the climate zones. Module 2, used to obtain the importance evaluations of asphalt mixture pavement performance under different climate zones from several experts, establishes a multi-scenario expert evaluation proxy model to unify the importance evaluations of pavement performance under different climate zones, and eliminates differences in opinions among different experts. The system employs a multi-scenario expert scoring agent module 3 to acquire pavement performance data of asphalt mixtures and establish a pavement performance database. Based on the unified evaluation of the importance of pavement performance, it uses the analytic hierarchy process (AHP) to obtain the weights of pavement performance and calculate the comprehensive performance score of asphalt mixtures. A reliability design module 4 is used to establish the type of function function and function function random variable based on the relationship between the comprehensive performance score of asphalt mixture pavement, the effect, and the resistance. It calculates the reliability index of the new asphalt mixture using a reliability index function, and then selects the asphalt mixture with the highest reliability index as the optimal design scheme. Through the improved reliability index calculation method applicable to asphalt mixtures, a quantitative evaluation of different asphalt mixtures is completed, enabling a more objective and scientific selection of the asphalt mixture with the best comprehensive performance.

[0043] The custom design parameter determination module 1 involves the division of climate zones. The following section provides a more detailed explanation of the climate zone division methods. This module includes:

[0044] This is used to set up several levels of climate zones and division rules, and to divide regional samples into different climate zone units 11 based on the climate, soil conditions, and traffic load of the regional samples. For example, the climate zones are divided into primary zones and secondary zones. The primary zones include high temperature zones, low temperature zones, rainfall zones, salt erosion zones, and traffic volume zones. Further, the secondary zones will further subdivide the primary zones. The high temperature zones are divided into hot summer zones, hot summer zones, and cool summer zones; the low temperature zones are divided into severe winter zones, cold winter zones, and temperate winter zones; the rainfall zones are divided into humid zones, wet zones, and arid zones; the salt erosion zones are divided into severely salted zones, moderately salted zones, and no salted zones; and the traffic volume zones are divided into overloaded traffic zones, heavily loaded traffic zones, and moderately light traffic zones. The designers select the primary zones to be considered according to their wishes, and arrange and combine the secondary zones, retaining feasible climate zone combinations within China. After removing other climate zone combinations that would not actually occur in the region, the secondary zones can be defined using the following criteria: High Temperature Zone: A maximum average temperature greater than 35℃ in the hottest month is considered a hot summer zone, 20–35℃ is a hot summer zone, and less than 25℃ is a cool summer zone; Low Temperature Zone: An extreme minimum temperature less than -21.5℃ is considered a severe winter zone, -21.5–-9℃ is a cold winter zone, and greater than -9℃ is a temperate winter zone; Rainfall Zone: An annual rainfall greater than 1 000mm is a humid zone, 250-1000mm is a moist zone, and less than 250mm is a dry zone; salt erosion zoning: soil salt content greater than 0.4% is a severely salted zone, 0.1%-0.4% is a moderately salted zone, and less than 0.1% is a non-salted zone; traffic volume zoning: cumulative equivalent axle trips greater than 10 million are overloaded traffic zones, 1 million to 10 million are heavy traffic zones, and less than 1 million are medium-light traffic zones.

[0045] It also includes a road performance unit 12 for setting the road performance requirements of asphalt mixtures for different climate zones.

[0046] The following section provides a detailed explanation of the functions of the multi-scenario expert evaluation agent module 2, specifically including:

[0047] The expert evaluation acquisition unit 21 is used to collect opinions from several experts and obtain their evaluations of the importance of road performance under different climate zones. This is used to construct a road performance importance vector for different climate zones and establish a multi-scenario expert scoring database. Specifically, based on the designer's wishes, all possible combinations of climate zones within China and the road performance that needs to be considered can be obtained. Using expert surveys, emails, questionnaires, and other methods, multiple experts can be consulted to obtain their scores on the importance of various road performance characteristics under different combinations of secondary climate zones. For example, in Fujian, a region with high summer temperatures and heavy rainfall, high temperatures and rainfall have a significant impact on asphalt mixtures. Furthermore, Fujian has a high traffic volume, resulting in repeated vehicle loads. However, sub-zero temperatures are rare throughout the year. Taking the high-temperature performance, low-temperature performance, water stability, and fatigue performance of asphalt mixtures as an example, it would be considered that high-temperature performance and water stability are equally important in asphalt mixture design, but more important than fatigue performance, and significantly more important than low-temperature performance. Thus, they can be scored from 9 to 1 point according to their importance, with higher scores indicating greater importance. Experts might ultimately score both high-temperature and low-temperature performance at 9 points, fatigue performance at 6 points, and low-temperature performance at 1 point, thus constructing an importance vector X = (9, 1, 6, 9). Higher importance results in a larger score. The road performance importance vector X = (x1, x2, ..., x...) can be used. i ), where x i The road performance importance score represents the i-th road performance importance score. The road performance importance vector represents the various road performance importance scores under all possible climate zone combinations within China. The climate zone combinations and the corresponding expert road performance importance scores form a multi-scenario expert scoring database.

[0048] The evaluation agent model building unit 22, used to construct the mapping relationship between climate zones and the importance of road performance through artificial intelligence algorithms, establishes a multi-scenario expert evaluation agent model that simulates the thinking and evaluation logic of experts. The evaluation agent model building unit 22 uses a bootstrap method, taking 80% of the samples from the obtained multi-scenario expert scoring database as the training set and the remaining 20% ​​as the test set. The climate zone combinations in the training set are used as the input to the artificial intelligence machine learning model, and the road performance importance vectors corresponding to the climate zones are used as the output of the artificial intelligence machine learning model. After training, the multi-scenario expert evaluation agent model is obtained. The model output layer uses a linear activation function to output the importance score of road performance. The model loss function uses the mean squared error (MSE) loss function, and the MSE loss function formula is:

[0049]

[0050] In the above formula, LOSS is the mean squared error loss value, n represents the number of samples in the multi-scenario expert scoring database, C is the number of road performance parameters to be considered, and y ic The actual expert score for the i-th road performance item. The score is the importance score of the i-th road performance item predicted by the multi-scenario expert evaluation agent model, and the score reflects the importance of various road performance items.

[0051] The importance proxy unit 23 is used to re-input each road performance importance vector from the multi-scenario expert scoring database into the importance proxy unit 23 of the model. The importance value of road performance is obtained through the multi-scenario expert evaluation proxy model. The underlying logic of the data is fully learned by the artificial intelligence machine learning algorithm, and the thinking of the experts is understood, so that the opinions of the experts are effectively unified, eliminating the evaluation differences caused by the differences in the prior knowledge of the experts, and unifying the evaluation standard of the importance of road performance.

[0052] The following details the functions of the multi-scenario expert scoring agent module 3, specifically including:

[0053] The road performance acquisition unit 31 is used to acquire road performance data of asphalt mixtures and integrate it into a road performance database. It can collect and import road performance data samples from publicly available datasets, laboratory road performance test data samples, and road performance test data samples from papers or publicly available texts. After data preprocessing to remove missing and outlier values ​​and normalizing the data, road performance data is obtained. The expert scoring calculation unit 32 is used to calculate the weights of road performance based on expert evaluations of its importance. It uses the analytic hierarchy process (AHP) to obtain the weights of road performance and calculates the comprehensive performance score of different asphalt mixtures based on the obtained weights and samples from the road performance database. For example, based on the multi-scenario expert evaluation proxy model and the reconstructed evaluation of the importance of road performance, a road performance judgment matrix for different scenarios is established using the 1-9 scale method. The weights of each road performance are calculated using the hierarchical single-ranking method, followed by a consistency check. The multi-scenario expert evaluation proxy model objectifies the subjective expert scoring weights in the AHP, making it more universal and accurate. The formula for calculating the comprehensive performance score of asphalt mixtures is:

[0054]

[0055] In the above formula, P j For the comprehensive performance score of any asphalt mixture, α i Let α be the value of the i-th indoor test index of the road performance of the asphalt mixture. i，max This represents the maximum value of the i-th indoor test index for road performance in the database.

[0056] The scoring proxy model building unit 33, used to construct a mapping relationship from the importance of road performance to the comprehensive performance score through artificial intelligence algorithms, and to establish a multi-scenario comprehensive performance scoring model, specifically establishes a multi-scenario road performance-scoring database based on the road performance database, the reconstructed evaluation of the importance of road performance, and the comprehensive performance scores of the aforementioned samples under different road performance importance levels. Using a bootstrap method, 80% of the samples in the multi-scenario road performance-scoring database are used as the training set, and the remaining 20% ​​as the test set. The reconstructed evaluation of the importance of road performance and the road performance data of asphalt mixtures in the training set are used as input to the artificial intelligence machine learning model, and the comprehensive performance score of asphalt mixtures is used as the output. The multi-scenario comprehensive performance scoring model is obtained by training the model using the training set, and the model's prediction accuracy is verified and hyperparameters are adjusted using the test set, thereby obtaining the optimal multi-scenario comprehensive performance scoring model.

[0057] The following details the functions of Reliability Design Module 4, including:

[0058] The random variable determination unit 41, used to determine the type of random variable of the function based on the comprehensive performance score, specifically treats the evaluation of road performance and its importance as random variables. The road performance random variables are all considered to satisfy a log-normal distribution, with their mean and standard deviation obtained from experimental data. Several parallel experimental data points for a particular road performance are considered as independent samples used to calculate the statistical parameters of the road performance random variable. The random variable evaluating the importance of road performance is considered to have a constant mean and a zero standard deviation, with its mean taken as the input value of the importance of road performance in the multi-scenario comprehensive performance scoring model. i Let N represent the random variable of the i-th road performance indoor test index. i Let X represent the random variable that scores the importance of the i-th road performance. Then, the multi-scenario comprehensive performance scoring model can be regarded as a resistance nonlinear function R(X1,…,X) composed of random variables. i ,…,X C ,N1,…,N i ,…,N C ).

[0059] Unit 42, used to determine the function function based on the relationship between the effect and resistance, specifically selects the score value P that the designer needs to meet the actual engineering requirements. Then, the effect function is S = P. Based on the effect function, the resistance nonlinear function, and the composition of the function function, the function function for the comprehensive performance evaluation of asphalt mixture is established. The expression of the function function is as follows:

[0060] Z = g(X1,…,X) i ,…,XC ,N1,…,N i ,…,N C )=R(X1,…,X i ,…,X C ,N1,…,N i ,…,N C )

[0061] In the above formula, Z is the function of function, g is the function name of function of function, R is the function name of resistance nonlinear function, and S is the effect function.

[0062] The reliable index calculation unit 43 is used to establish a reliable index function for the comprehensive performance of asphalt mixtures using the first second moment method. Specifically, the function Z for the comprehensive performance score of asphalt mixtures is expanded by Taylor series at the mean of each random variable according to the mean second moment method, and only the linear term is retained. The approximate expression of the function function is as follows:

[0063]

[0064] In the above formula, Z' is an approximate expression for the function. Represented as random variable X i The mean, The partial derivative of the function with respect to the random variable is the value at the mean. Assignment at the location, Let N be a random variable i The mean; The partial derivative of the function with respect to the random variable is the value at the mean. The assignment at that location.

[0065] The expression for the mean of the function is as follows:

[0066]

[0067] The expression for the standard deviation of the function is as follows:

[0068]

[0069] In the above formula, μ z σ is the mean of the function; z The standard deviation of the function. Let X be a random variable i Standard deviation Let N be a random variable i The standard deviation.

[0070] The asphalt mixture design unit 44 is used to conduct road performance tests on new asphalt mixtures. It calculates the reliability index based on the reliability index function and selects the asphalt mixture with the highest reliability index as the optimal design scheme. Specifically, based on the mean and standard deviation of the function obtained in the preceding steps, the reliability index β of different asphalt mixtures is calculated. The higher the reliability index β, the more reliable the asphalt mixture. The reliability index β can be calculated using the following formula:

[0071]

[0072] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, this application 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.

[0073] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. Since the above embodiments are substantially similar to the method embodiments, their descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0074] The above embodiments provide a detailed description of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A multi-scenario asphalt mixture design system based on comprehensive performance scoring reliability, characterized in that, The system includes: Custom design parameter determination module (1) is used to obtain all regional samples within the sample area according to the designer's needs, classify the regional samples according to preset rules to set several levels of climate zones, and set the road performance of asphalt mixture for different climate zones. A multi-scenario expert evaluation proxy module (2) is used to obtain the importance evaluations of asphalt mixture road performance under different climate zones from several experts, and to establish a multi-scenario expert evaluation proxy model to unify the importance evaluations of road performance under different climate zones; the multi-scenario expert evaluation proxy module (2) includes: The expert evaluation acquisition unit (21) is used to collect opinions from several experts, obtain the importance evaluation of each expert on road performance under different climate zones, construct the importance vector of road performance in different climate zones, and establish a multi-scenario expert scoring database. Evaluation agent model building unit (22) is used to construct a mapping relationship between climate zones and the importance of road performance through artificial intelligence algorithms, so as to establish a multi-scenario expert evaluation agent model that simulates the thinking and evaluation logic of experts. The sample in the obtained multi-scenario expert scoring database is divided into training set and test set by using the bootstrap method. The combination of climate zones in the training set is used as the input of the artificial intelligence machine learning model, and the road performance importance vector corresponding to the climate zone is used as the output of the artificial intelligence machine learning model. After training, the multi-scenario expert evaluation agent model is obtained. The model output layer adopts a linear activation function to output the importance score of road performance. The model loss function adopts the mean squared error loss function. The MSE loss function formula is: In the above formula, This is the mean squared error loss value. C represents the number of samples in the multi-scenario expert scoring database, and C is the number of road performance parameters that need to be considered. The actual expert score for the i-th road performance item. The score is the importance score of the i-th road performance item predicted by the multi-scenario expert evaluation agent model, and the score reflects the importance of various road performance items. Importance proxy unit (23), the importance proxy unit (23) is used to re-input each road performance importance vector in the multi-scenario expert scoring database into the model, and obtain the importance value of road performance through the multi-scenario expert evaluation proxy model; Multi-scenario expert scoring agent module (3) is used to obtain the road performance data of asphalt mixture to establish a road performance database, and to evaluate the importance of the unified road performance, and to obtain the weight of the road performance and calculate the comprehensive performance score of asphalt mixture by using the analytic hierarchy process. The reliability design module (4) is used to establish the type of function function and function function random variable based on the comprehensive performance score, action effect and resistance relationship of the asphalt mixture road, and calculate the reliability index of the new asphalt mixture through the reliability index function. Then, the asphalt mixture with the largest reliability index is selected as the optimal design scheme.

2. The asphalt mixture design system according to claim 1, characterized in that, The custom design parameter determination module (1) includes: Climate zoning unit (11), the climate zoning unit (11) is used to set several levels of climate zoning and division rules, and to divide the regional sample into different climate zoning according to the climate, soil conditions and traffic load of the regional sample; Road performance unit (12), which is used to set the road performance required for asphalt mixtures in different climate zones.

3. The asphalt mixture design system according to claim 1, characterized in that, The multi-scenario expert scoring agent module (3) includes: The road performance acquisition unit (31) is used to acquire the road performance data of asphalt mixture and integrate it into a road performance database. The expert scoring calculation unit (32) is used to obtain the weight of road performance based on the evaluation of the importance of road performance by experts, and to calculate the comprehensive performance score of different asphalt mixtures based on the obtained weights and the samples of the road performance database. The scoring agent model establishment unit (33) is used to construct a mapping relationship from the importance of road performance and road performance to comprehensive performance score through artificial intelligence algorithm, so as to establish a multi-scenario comprehensive performance score model.

4. The asphalt mixture design system according to claim 3, characterized in that, The formula for calculating the comprehensive performance score of the asphalt mixture is as follows: In the above formula, The comprehensive performance score for any asphalt mixture. Let be the value of the i-th indoor test index of the road performance of the asphalt mixture. This represents the maximum value of the i-th indoor test index for road performance in the database.

5. The asphalt mixture design system according to claim 1, characterized in that, The reliability design module (4) includes: The random variable determination unit (41) is used to determine the type of random variable of the function based on the comprehensive performance score; Functional function determination unit (42) is used to establish a functional function based on the relationship between the effect and the resistance; The reliable index calculation unit (43) is used to establish the reliable index function of the comprehensive performance of asphalt mixture through the first second moment method; Asphalt mixture design unit (44) is used to conduct road performance tests on new asphalt mixtures, calculate reliability indexes based on reliability index functions, and select asphalt mixtures with large reliability indexes as the optimal design scheme.

Images