A method and system for optimizing the proportioning of asphalt mixtures

Abstract

The application relates to the technical field of asphalt mixture, and discloses a proportioning optimization method and system for asphalt mixture, which comprises a collection module, a feature construction module, a comparative analysis module, a judgment and control module and a database module. By collecting mixing response data, compaction response data and volume change data of different proportioning samples, a structural feature set and a response evolution factor set are constructed; differences between a current sample and a historical sample are compared, a proportioning response deviation degree is calculated, and a grade is divided; when the grade is the second grade, a proportioning optimization strategy is generated in combination with the change direction relationship of the structural features and the response factors. The application can realize intelligent judgment of proportioning and output of a control strategy.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of asphalt mixture, in particular to a proportioning optimization method and system for asphalt mixture. BACKGROUND

[0002] As one of the most commonly used paving materials in road engineering, the performance stability of asphalt mixture directly affects the structural strength, service life and construction adaptability of the pavement. In existing engineering practice, the proportioning parameters of asphalt mixture are usually preliminarily set by relying on laboratory trial data or empirical formula, and then adjusted through repeated tests to approach the target performance. This method not only has long cycle and low efficiency, but also lacks systematic correlation between the proportioning structure and the actual response, which makes it difficult to predict the performance evolution trend in the material mixing, compaction and cooling process in advance, affecting the consistency of engineering quality and the real-time of regulation and control.

[0003] Some studies attempt to introduce data-driven methods to model and evaluate the performance of asphalt mixture, but most of them only focus on the mapping relationship between mechanical indicators and static proportioning, ignoring the response evolution process in the construction process, such as viscosity change in the mixing process, density growth trend in the compaction stage and volume shrinkage behavior after cooling. These key factors have a significant impact on the final performance.

[0004] Therefore, it is urgent to build a proportioning optimization system that integrates structural features and response evolution information, accurately identifies the structural differences between the current state and historical samples and the matching degree of response trends, and then realizes intelligent and closed-loop optimization of the proportioning structure. SUMMARY

[0005] In view of this, the present application provides a proportioning optimization method and system for asphalt mixture to solve the above problems.

[0006] In one aspect, the present application provides a proportioning optimization system for asphalt mixture, comprising:

[0007] The acquisition module is configured to obtain mixing response data, compaction response data and volume change data in the cooling process of a plurality of asphalt mixture samples with different proportioning parameters in historical data;

[0008] The feature construction module is configured to construct a structural feature set according to the proportioning parameters, and construct a response evolution factor set according to the mixing response data, compaction response data and volume change data;

[0009] The comparative analysis module is configured to obtain the structural feature set and the response evolution factor of the current sample to be tested, compare the structural feature set and the response evolution factor set of the historical sample, calculate the proportioning response deviation degree, match the proportioning response deviation degree with the set grading standard, and determine the grade; the grade includes a first grade and a second grade with the proportioning response deviation degree increasing in turn;

[0010] The judgment and control module is configured to analyze the matching relationship between the structural feature set change direction and the response evolution factor change direction of the sample to be tested when the grade is the second grade, output the proportioning optimization strategy according to the relationship between the structural feature set change direction and the response evolution factor change direction, and output the proportioning optimization strategy.

[0011] The database module is configured to establish a database including the structural feature set, the response evolution factor and the grade of the calibrated proportioning sample.

[0012] Further, it further includes:

[0013] The path deduction module is configured to compare the sample to be tested with the database according to the structural feature set and the response evolution factor, obtain the adjacent sample, and output the adjustable parameter interval and the structural offset trend risk of the sample to be tested based on the historical response evolution factor change direction and the label grade of the adjacent sample.

[0014] The path deduction module is called in advance before the judgment and control module is started.

[0015] Further, the proportioning parameters include: a coarse and fine aggregate ratio, which is the mass ratio of coarse aggregate to fine aggregate; a mineral powder content, which is the filler content of the asphalt mixture; an asphalt content, which is the asphalt content of the asphalt mixture; and a target air void ratio, which is the air volume fraction of the asphalt mixture.

[0016] The mixing response data is a viscosity change curve in the mixing process; the compaction response data is process data of the density change with the compaction angle or the compaction times in the compaction process; and the volume change data is a volume change curve of the asphalt mixture in the natural cooling process.

[0017] Further, the structural feature set constructed by the feature construction module includes:

[0018] The relative deviation value between the coarse and fine aggregate ratio and the standard gradation curve, the fitting residual value between the mineral powder content and the target air void ratio, and the wrapping factor between the unit aggregate specific surface area and the asphalt content; and the coupling offset between the target air void ratio and the stable density value of the compaction prediction model.

[0019] The response evolution factor set constructed by the feature construction module includes:

[0020] a time period from the start of mixing to the time when the viscosity reaches the maximum value, and a first-order slope value of the viscosity change curve obtained by fitting the viscosity change curve by the least square method; an average density growth rate in the first 25% of the total compaction times, and a density change standard deviation in the last 25% of the total compaction times in the compaction stage; a maximum change rate of the shrinkage rate in the volume change curve and a duration during which the maximum change rate continuously exceeds twice the average change rate in the volume change curve from the unloading time to the time when the temperature drops to room temperature in the cooling stage;

[0021] wherein the unit aggregate specific surface area is the total specific surface area corresponding to unit mass or unit volume of aggregate under the current mix proportion structure.

[0022] Further, the comparative analysis module specifically comprises:

[0023] The vectorized similarity is calculated based on the structure feature set and the response evolution factor set, and the difference degree of the current sample to be tested and the historical sample in the database is calculated by the Euclidean distance, the cosine angle or the Mahalanobis distance.

[0024] The mix proportion response deviation vector is generated according to the similarity, and each index value in the deviation vector is obtained by normalization and weighting to obtain the mix proportion response deviation degree.

[0025] The mix proportion response deviation degree is matched with the set grading standard threshold value, and the first grade or the second grade is respectively given; when the mix proportion response deviation degree exceeds the upper limit of the second grade, it is classified into the second grade.

[0026] Further, the judgment and control module, when receiving the second grade, performs the following operations:

[0027] The numerical variation direction of the coarse and fine aggregate ratio, the mineral powder content and the asphalt content in the structure feature set in the current sample to be tested is obtained.

[0028] Meanwhile, the change direction of the time period from the start of mixing to the time when the viscosity reaches the maximum value, the first-order slope value of the viscosity change curve, the average density growth rate, the density change standard deviation, the maximum change rate of the shrinkage rate and the duration in the response evolution factor set are extracted.

[0029] If the change direction of the coarse and fine aggregate ratio, the mineral powder content and the asphalt content in the structure feature set is consistent with the change direction of any index in the response evolution factor set, it is judged that the structure adjustment and the performance response are consistent, and the asphalt content is finely adjusted at this time.

[0030] If the change direction of any parameter in the structure feature set is opposite to the direction of two or more indicators in the response evolution factor set, it is determined that the proportion disturbance is mismatched, the proportion of coarse and fine aggregates and the mineral powder content are re-evaluated, and the asphalt content adjustment is suspended;

[0031] If the maximum change rate of the shrinkage rate and its duration indicator exceed twice the average level of the historical sample at the same time, output the volume stability abnormality prompt, adjust the target void ratio or optimize the cooling process configuration.

[0032] Further, the path deduction module executes the following operations based on the matching results of the structure feature set and the response evolution factor set in the database:

[0033] Select the first five groups of historical samples with the smallest Euclidean distance from the structure feature set of the current mixture sample to be tested as the adjacent mixture samples;

[0034] Extract the corresponding response evolution factor set from these adjacent mixture samples, including the first-order slope value of the viscosity change curve, the average density growth rate, the density change standard deviation, the maximum change rate of the shrinkage rate and the duration;

[0035] According to the numerical variation direction of the response evolution factor set and the corresponding proportion parameter adjustment record, the response path trajectory is constructed;

[0036] Based on the change trend of the response evolution factor set of the mixture sample to be tested under different proportion parameter adjustment conditions, a structure response prediction sequence is formed, and the prediction result is output to the judgment and control module.

[0037] Further, the path deduction module further includes a structure risk analysis unit, which is configured to:

[0038] According to the combined change history of the structure feature set and the response evolution factor set in the adjacent mixture samples, the adjustable parameter interval of the mixture sample to be tested under the current structure condition is calculated;

[0039] According to whether the maximum change rate of the shrinkage rate and the duration are higher than the corresponding 95% confidence interval of the samples in the database, the structure deviation trend risk level is calculated;

[0040] When the maximum change rate of the shrinkage rate and the duration are both in the abnormal interval, a high risk level is output, if only one indicator is high, a medium risk level is output, and if both are in the normal interval, a low risk level is output.

[0041] Further, when the number of adjacent mixture samples in the database that meet the following conditions is less than three groups, a supplementary mechanism is executed:

[0042] The Euclidean distance between the structural feature set and the current sample to be tested is less than or equal to a set threshold; the first-order slope value of the viscosity change curve and the average density growth rate in the response evolution factor set are within a set floating range;

[0043] The supplementary mechanism includes collecting three groups of mixture samples and collecting corresponding mixing response data, compaction response data and volume change data;

[0044] The structural feature set and the response evolution factor set are constructed by the feature construction module and are incorporated into the database module.

[0045] Compared with the prior art, the beneficial effects of the present application are that:

[0046] By constructing structural feature parameters such as the relative deviation of coarse and fine aggregate ratios, the wrapping factor and the coupling offset, and combining dynamic response evolution factors such as the slope of the viscosity curve, the density growth rate and the volume shrinkage rate, an intermediate representation mechanism between structure and performance is established, and the discrimination ability for complex mix proportion structure states is improved.

[0047] The proportion response deviation degree and the grade division mechanism are introduced to enhance the pertinence and gradation of control and judgment

[0048] The system generates a normalized matching response deviation degree by calculating the similarity of the structural feature set and the response evolution factor through vectorization matching, and divides the first level and the second level according to the normalized matching response deviation degree, thereby providing a clear classification basis for subsequent strategy decision-making and avoiding blind adjustment or rash correction. A cooperative judgment path of the structure and the response change direction is constructed to improve the rationality and refinement level of the optimization strategy generation. In the second level state, the system judges whether the structural adjustment leads to performance improvement by analyzing whether the change direction of the coarse and fine aggregate ratio, the mineral powder content and the asphalt content matches the change trend of each response evolution factor. If the directions are consistent, it is clear that the optimization is cooperative, and the asphalt content is adjusted preferentially. If the directions are opposite, it is identified as a disturbance mismatch to prevent false optimization and improve the pertinence of the matching adjustment. The path deduction module and the structure risk analysis mechanism are introduced to realize predictive control and safety boundary prompting. The system constructs a response path trajectory based on historical adjacent samples and outputs a structural response prediction sequence, so that the matching adjustment is no longer dependent on a single point response and has trend foresight. At the same time, the structural deviation risk level is calculated by comparing the maximum shrinkage rate and its duration with the position of the 95% confidence interval in the database, thereby providing quantifiable support for cooling stability evaluation. When the number of adjacent samples is insufficient, the system can automatically trigger a fast sampling mechanism to collect boundary samples and construct their structural feature set and response evolution factor, which are supplemented to the database, thereby effectively solving the reasoning fault caused by sparse samples and maintaining the stability and effectiveness of the control strategy generation. The system can be widely used in the matching initial setting optimization, on-site trial mixing adjustment, quality abnormality diagnosis and construction rhythm control of road engineering, can dynamically perceive the material behavior and provide reasonable strategies, and significantly improves the consistency and intelligent level of the whole process of engineering material design-production-construction.

[0049] In another aspect, the present application provides a matching optimization method for asphalt mixture, comprising:

[0050] S1: obtaining mixing response data, compaction response data and volume change data in the cooling process of a plurality of asphalt mixture samples with different matching parameters in historical data;

[0051] S2: constructing a structural feature set according to the matching parameters, and constructing a response evolution factor set according to the mixing response data, the compaction response data and the volume change data;

[0052] S3: obtaining the structural feature set and the response evolution factor of the current mixture sample to be tested, and comparing them with the structural feature set and the response evolution factor set of the historical samples, calculating the matching response deviation degree, and matching the matching response deviation degree with the set classification standard to determine the level; the level includes a first level and a second level with increasing matching response deviation degrees;

[0053] S4: when the level is the second level, analyzing a matching relationship between a change direction of the set of structural features of the sample of the mixture to be tested and a change direction of the response evolution factor; outputting a proportioning optimization strategy according to the relationship between the change direction of the set of structural features and the change direction of the response evolution factor;

[0054] S5: establishing a database including the set of structural features, the response evolution factor and the level of the calibrated sample of the proportioning.

[0055] It should be noted that the method for proportioning optimization of asphalt mixture and the system thereof have the same beneficial effects, which will not be described here. BRIEF DESCRIPTION OF DRAWINGS

[0056] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of the preferred embodiments and are not intended to limit the scope of the application. Moreover, the same reference numerals are intended to denote the same components throughout the accompanying drawings. In the drawings:

[0057] Figure 1 A functional block diagram of a system for proportioning optimization of asphalt mixture is provided for an embodiment of the application.

[0058] Figure 2 A flowchart of a method for proportioning optimization of asphalt mixture is provided for an embodiment of the application. DETAILED DESCRIPTION

[0059] The exemplary embodiments of the present application will be described hereinafter with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it is understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be accurately conveyed to those skilled in the art. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The present application will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0060] Referring to Figure 1 The embodiment of the present application provides a system for proportioning optimization of asphalt mixture, which comprises:

[0061] The acquisition module is configured to acquire mixing response data, compaction response data and volume change data in a cooling process of a plurality of asphalt mixture samples with different proportioning parameters in historical data.

[0062] The characteristic construction module is configured to construct a structure characteristic set according to the proportioning parameters, and construct a response evolution factor set according to the mixing response data, the compaction response data and the volume change data.

[0063] The comparative analysis module is configured to obtain the structure characteristic set and the response evolution factor of the current to-be-tested mixture sample, compare them with the structure characteristic set and the response evolution factor set of the historical samples, calculate the proportioning response deviation degree, match the proportioning response deviation degree with the set grading standard, and determine the grade; the grade includes a first grade and a second grade with the proportioning response deviation degree increasing in turn.

[0064] The judgment and control module is configured to, when the grade is the second grade, analyze the matching relationship between the structure characteristic set change direction and the response evolution factor change direction of the to-be-tested mixture sample; output the proportioning optimization strategy according to the relationship between the structure characteristic set change direction and the response evolution factor change direction.

[0065] The database module is configured to establish a database including the structure characteristic set, the response evolution factor and the grade of the calibrated proportioning sample.

[0066] It should be noted that the multi-dimensional parameter set of the structure characteristic and the response evolution factor is constructed by collecting the whole-process response data in the mixing, compaction and cooling stages, the quantitative evaluation and grade division of the asphalt mixture proportioning performance are realized, when the second grade with larger deviation is identified, the optimization strategy is further deduced in combination with the relationship between the structure and the response change direction, and the pertinence and scientificity of the control are improved. The system has the functions of intelligent comparison, trend judgment and strategy output, can significantly improve the efficiency and quality of the proportioning adjustment, and enhance the stability and adaptability of the asphalt mixture design.

[0067] In some embodiments of the present application, the path deduction module is further configured to compare the to-be-tested mixture sample with the database according to the structure characteristic set and the response evolution factor, obtain the adjacent mixture sample, and output the adjustable parameter interval of the to-be-tested mixture sample and the structure offset trend risk based on the historical response evolution factor change direction and the label grade of the adjacent mixture sample.

[0068] The path deduction module is called in advance before the judgment and control module is started.

[0069] In the present embodiment, the specific implementation steps of the path deduction module include: 1. sample matching stage: the path deduction module first receives the structure characteristic set and the response evolution factor set of the current to-be-tested mixture sample; then, the database module is called to select the first several groups of adjacent samples with the minimum Euclidean distance between the structure characteristic set and the structure characteristic set of the current sample from the historical samples as the matching sample set.

[0070] 2. Response path extraction stage: for each sample in the matched sample set, extract its corresponding response evolution factor set, and divide it into mixing, compaction, and cooling stages; analyze the change direction of the response evolution factor in each stage (such as the first-order slope increase and decrease, the density increase rate improvement, or the stability fluctuation expansion, etc.) and its corresponding structural feature change history; associate the response evolution factor change direction with the label level (first level or second level) to form the structural-response-level evolution path sample group.

[0071] 3. Structural response trend deduction stage: substitute the structural feature set of the current mixed material sample into the above structural-response path group respectively, simulate its response trend change under the existing historical evolution path; if the response evolution factor trend in the simulation result is highly similar to the corresponding second level sample path in the history, it is determined that there is a structural deviation trend risk, and the risk level is recorded.

[0072] 4. Adjustable parameter interval calculation stage: analyze the adjustment direction and amplitude of the structural parameters (such as the coarse and fine aggregate ratio, the mineral powder content, and the asphalt content) in the path during the optimization process from the second level to the first level; calculate the maximum and minimum boundaries of the current structural parameters in different paths to improve performance, and form the adjustable parameter interval of the current mixed material sample.

[0073] 5. Risk output and module linkage: output the adjustable parameter interval and structural deviation trend risk level in the above deduction result to the judgment and control module; before the judgment and control module is officially started, the system control process automatically calls the path deduction module in advance to complete the above calculation, ensuring that the control strategy generation has a trend prediction basis.

[0074] In some embodiments of the present application, the proportioning parameters include: the coarse and fine aggregate ratio, which is the mass ratio of coarse aggregate to fine aggregate; the mineral powder content, which is the filler content of the asphalt mixture; the asphalt content, which is the asphalt content of the asphalt mixture; the target air void, which is the air volume fraction of the asphalt mixture; the mixing response data is the viscosity change curve in the mixing process; the compaction response data is the process data of the density change with the compaction angle or the compaction times in the compaction process; and the volume change data is the volume change curve of the asphalt mixture in the natural cooling process.

[0075] It should be noted that the coarse-fine aggregate ratio refers to the mass ratio or volume ratio between coarse aggregate (such as gravel) and fine aggregate (such as sand) for controlling the grading curve of the mixture. The filler content refers to the proportion of filler (such as limestone powder) added to the unit mixture, which has a significant impact on the filling and bonding properties. The asphalt content refers to the mass proportion of asphalt in the unit mixture, which is a key parameter for controlling the bonding performance, durability and anti-rutting performance. The target air void ratio refers to the desired air volume fraction of the total volume inside the formed mixture under design requirements, which plays a decisive role in compactness, durability and the like.

[0076] In some embodiments of the present application, the structural feature set constructed by the feature construction module includes: the relative deviation value between the coarse-fine aggregate ratio and the standard grading curve, the fitting residual value between the filler content and the target air void ratio, and the wrapping factor between the unit aggregate specific surface area and the asphalt content; the coupling offset between the target air void ratio and the stable density value of the compaction prediction model; the response evolution factor set constructed by the feature construction module includes: in the mixing stage, the time period experienced from the start of mixing to the maximum viscosity, and the first-order slope value obtained by fitting the viscosity change curve in this time period by the least square method; in the compaction stage, the average density growth rate in the first 25% interval of the total compaction times, and the density change standard deviation in the last 25% interval of the total compaction times; in the cooling stage, the maximum change rate of the shrinkage rate in the volume change curve during the period from unloading to the temperature dropping to room temperature, and the duration during which the maximum change rate continuously exceeds twice the average change rate in the volume change curve; wherein the unit aggregate specific surface area is the total specific surface area corresponding to unit mass or unit volume of aggregate under the current proportioning structure.

[0077] It should be noted that the construction steps of the structural feature set: obtain the proportioning parameters of the current mixture sample: including coarse-fine aggregate ratio, filler content, asphalt content, target air void ratio. Calculate the relative deviation value between the coarse-fine aggregate ratio and the standard grading curve.

[0078] Fit and compare the particle size grading data of the current coarse-fine aggregate with the target standard grading curve; calculate the difference value at each particle size point, take the absolute value average after normalization as the relative deviation value output.

[0079] Calculate the fitting residual value between the filler content and the target air void ratio, use the existing statistical regression model of filler content and target air void ratio for prediction; compare the actual measured air void ratio with the model predicted value, take the residual value as the feature output.

[0080] Calculate the wrapping factor between the specific surface area of unit aggregate and the asphalt content: obtain the particle size composition and density of all aggregates in the current mix; estimate the specific surface area per unit mass or volume according to the aggregate particle size and shape; model the ratio of specific surface area to asphalt content, and output the wrapping factor representing the wrapping capacity of asphalt.

[0081] Calculate the coupling offset between the target air void and the stable density value of the compaction prediction model: call the compaction prediction model established in the database, input the current structure parameters, and predict the final compaction value; the difference between the predicted density and the target air void is taken as the coupling offset output.

[0082] Mixing stage index extraction: extract the time period from the start of mixing to the maximum viscosity, locate the starting point and peak value of the viscosity curve from the collected mixing response data; calculate the time interval between the two, as the viscosity rising time period output.

[0083] Fitting the viscosity curve and extracting the first order slope value: least squares linear fitting is performed on the viscosity data in the above time period; the first order derivative value of the linear function is taken, which is the first order slope value of the viscosity growth.

[0084] Compaction stage index extraction: calculate the average density growth rate in the first 25% compaction interval, extract the total compaction number from the compaction response data; intercept the first 25% compaction process data, calculate the ratio of density growth to compaction number, and output the average growth rate.

[0085] Calculate the standard deviation of density change in the last 25% interval: extract the data in the last 25% compaction stage, calculate the standard deviation of density change in this interval, which is used to reflect the volatility of later compaction.

[0086] Cooling stage index extraction: calculate the maximum change rate of shrinkage rate in the volume change curve, take the derivative of the volume change curve to get the volume change rate per unit time; get the maximum change rate from the derivative curve, which is the maximum negative slope point.

[0087] Calculate the average value of the derivative of the volume change curve; identify the time length of the derivative value greater than twice the average value in the continuous time interval, and output the duration.

[0088] Unit aggregate specific surface area acquisition method: for each particle size of aggregate, set its shape coefficient and surface area estimation coefficient (determined by experience or pre-test); obtain the total specific surface area of the current sample by weighted average according to the mass or volume proportion; output the unit aggregate specific surface area as one of the structure characteristic indexes in unit mass (g / m²) or volume (cm² / cm³).

[0089] It should be noted that the establishment process of the compaction prediction model includes the following steps: first, collecting the mixing ratio parameters and corresponding compaction response data of historical asphalt mixture samples, wherein the mixing ratio parameters include the coarse and fine aggregate ratio, the mineral powder content, the asphalt content and the target air void ratio, and the compaction response data includes the complete curve of the density change with the compaction times. Second, based on the compaction curve of each sample, the average density of the stable interval at the end of the section is extracted as the stable density value label, and the relative deviation value between the coarse and fine aggregate ratio in the mixing ratio parameter and the standard gradation curve, the fitting residual value between the mineral powder content and the target air void ratio, the wrapping factor between the unit aggregate specific surface area and the asphalt content, and the coupling offset between the target air void ratio and the stable density value of the compaction prediction model are used as input variables to establish the input-output mapping relationship. Subsequently, a nonlinear regression model or a regression model based on gradient boosting algorithm is used for training, and the model performance is optimized by the least mean square error or the average absolute error, and the appropriate model structure and parameters are selected by cross-validation. After the model training is completed, the offset between the predicted stable density value and the theoretical density value obtained by converting the sample target air void ratio is calculated, which is used as the coupling offset indicator in the structure feature set for subsequent mixing ratio coordination judgment and path deduction. The model can be continuously trained with the database update to improve the prediction accuracy of the compaction trend under different mixing ratio structures.

[0090] In some embodiments of the present application, the comparative analysis module specifically includes: calculating the vectorized similarity based on the structure feature set and the response evolution factor set, calculating the difference between the current mixture sample to be tested and the historical sample in the database by Euclidean distance, cosine angle or Mahalanobis distance; generating a mixing ratio response deviation vector according to the similarity, and obtaining the mixing ratio response deviation degree by normalizing and weighting each index value in the deviation vector; matching the mixing ratio response deviation degree with the set grading standard threshold, and respectively assigning the first grade or the second grade; when the mixing ratio response deviation degree exceeds the upper limit of the second grade, it is classified as the second grade.

[0091] In the present embodiment, the specific implementation steps of the comparative analysis module include the following contents: first, obtaining the structure feature set and the response evolution factor set of the current mixture sample to be tested. The structure feature set includes: the relative deviation value between the coarse and fine aggregate ratio and the standard gradation curve, the fitting residual value between the mineral powder content and the target air void ratio, the wrapping factor between the unit aggregate specific surface area and the asphalt content, and the coupling offset between the target air void ratio and the stable density value of the compaction prediction model; the response evolution factor set includes: the viscosity growth time period and the first-order slope value in the mixing stage, the average density growth rate and the density change standard deviation in the compaction stage, and the maximum change rate and the duration of the cooling stage.

[0092] Then, the above structural feature set and response evolution factor set are subjected to numerical normalization processing and converted into a uniform dimension vector form. Next, all calibrated historical samples in the database are selected, and the corresponding structural feature set and response evolution factor set are extracted and also normalized into vectors.

[0093] After the vector data is prepared, the similarity between the current sample to be tested and each historical sample in the structural feature dimension and the response factor dimension is calculated in the form of Euclidean distance, cosine angle or Mahalanobis distance, etc. The similarity indexes in each dimension are combined into a matching response deviation vector.

[0094] Subsequently, normalization and weighting calculation are applied to each index in the deviation vector to obtain a comprehensive matching response deviation degree value. The larger the value, the more significant the overall deviation of the current matching and the historical database sample in structure and response.

[0095] Finally, the calculated matching response deviation degree is compared with the system set grading standard: if it is within the first grade threshold range, it is assigned a first grade label; if it is within the second grade threshold range or exceeds the upper threshold, it is uniformly assigned a second grade label, but the system will simultaneously issue a structure adjustment strategy to assist the subsequent analysis of the judgment and control module. This process can be combined with a dynamic threshold updating mechanism to automatically adjust the grade determination standard according to the continuous expansion of the database samples, in order to improve the classification accuracy.

[0096] In some embodiments of the present application, when the judgment and control module receives the second grade, it performs the following operations: obtaining the numerical variation direction of the coarse and fine aggregate ratio, mineral powder content and asphalt content in the structural feature set in the current sample to be tested; simultaneously extracting the time period experienced from the start of mixing to the maximum viscosity, the first order slope value obtained by fitting the viscosity curve, the average density growth rate, the density change standard deviation, the maximum change rate of shrinkage rate and the change direction of duration in the response evolution factor set; if the change direction of the coarse and fine aggregate ratio, mineral powder content and asphalt content in the structural feature set is consistent with the change direction of any index in the response evolution factor set, it is judged that the structure adjustment and performance response are consistent, and the asphalt content is adjusted finely; if the change direction of any parameter in the structural feature set is opposite to the direction of two or more indexes in the response evolution factor set, it is judged that the matching of the mixture disturbance is mismatched, and the coarse and fine aggregate ratio and the mineral powder content are reevaluated, and the adjustment of the asphalt content is suspended; if the maximum change rate of shrinkage rate and its duration index both exceed twice the average level of historical samples, an abnormal volume stability prompt is output, and the target air void or the cooling process configuration is adjusted.

[0097] Specifically, the structural parameter variation direction extraction: first, extract the mix proportion parameters from the current sample, i.e. the proportion of coarse and fine aggregates, the mineral powder content and the asphalt content; compare them with the mean value of the structural feature set in the historical sample or the reference sample to determine the variation direction of each structural parameter in the current sample (increase, decrease or basically unchanged).

[0098] Next, from the response evolution factor set of the sample, extract: in the mixing stage, the time period from the start of mixing to the maximum viscosity; the first-order slope value obtained by least squares fitting of the viscosity change curve in this period; the average density growth rate in the compaction stage; the density change standard deviation in the late compaction stage.

[0099] The maximum change rate of the shrinkage rate of the volume change curve in the cooling stage and the duration during which the change rate is continuously higher than twice the average change rate.

[0100] Compare these indicators with historical data to determine the trend of their values (up, down or stable).

[0101] If any of the coarse and fine aggregate proportion, mineral powder content and asphalt content has the same variation direction as any of the response evolution factors (e.g. increase in aggregate proportion accompanied by increase in viscosity growth slope), it is inferred that the structural adjustment in performance response is a synergistic enhancement relationship, and the fine adjustment of asphalt content is performed (e.g. fine tuning within ±0.1%).

[0102] If any of the three structural parameters has the opposite variation direction as two or more response evolution factors (e.g. increase in mineral powder content but decrease in average density growth rate and increase in density fluctuation), it is judged that the current mix proportion structure leads to performance mismatch, and it is determined as a disturbance mismatch type abnormality. At this time, the asphalt content adjustment is suspended, and the coarse and fine aggregate proportion and mineral powder content are re-evaluated.

[0103] When the maximum change rate of the shrinkage rate and its abnormal duration in the cooling stage both exceed twice the average value of the database history (based on statistical indicators), the system determines that there is a volume stability problem; at this time, the system outputs a volume stability abnormality prompt, recommending the user to correct the target air void or adjust the process parameters of the cooling stage (such as extending the natural cooling time, optimizing the unloading time point, etc.) to reduce the late volume strain.

[0104] In some embodiments of the present application, the path deduction module performs the following operations based on the matching results of the structure feature set and the response evolution factor set in the database: selecting the first five groups of historical samples with the smallest Euclidean distance from the structure feature set of the current sample to be tested as the adjacent mixture samples; extracting the corresponding response evolution factor set of these adjacent mixture samples, including the first-order slope value of the viscosity change curve, the average density growth rate, the density change standard deviation, the maximum change rate and the duration of the shrinkage rate; constructing the response path trajectory according to the numerical variation direction of the response evolution factor set and the corresponding mix proportion parameter adjustment record; forming a structure response prediction sequence based on the change trend of the response evolution factor set of the sample to be tested under different mix proportion parameter adjustment conditions, and outputting the prediction result to the judgment and control module.

[0105] Specifically, the path deduction module is based on the matching results of the structure feature set and the response evolution factor set in the database, and the specific implementation is as follows: first, the system vectorizes the structure feature set of the current sample to be tested, and calculates the Euclidean distance between it and the structure feature set of all calibrated historical samples in the database. The system automatically selects the first five groups of historical samples with the smallest Euclidean distance as the adjacent mixture samples, ensuring that the structural similarity is higher than the set matching threshold, so as to enhance the reliability of the response trend deduction.

[0106] Subsequently, the response evolution factor set corresponding to the above-mentioned adjacent mixture samples is extracted, including the first-order slope value of the viscosity change curve, the average density growth rate and the density change standard deviation in the compaction process, and the maximum change rate of the shrinkage rate and the duration of the change rate in the cooling stage, etc. Key parameters.

[0107] Next, the system combines the historical adjustment records of each mix proportion parameter in the adjacent samples, analyzes the numerical variation direction of the response evolution factor, constructs multiple response path trajectories, forms a mapping relationship diagram of mix proportion disturbance-performance response, and serves as a subsequent prediction basis.

[0108] Finally, the system performs path extrapolation simulation on the current sample to be tested within the adjustable mix proportion parameter range, calculates the trend change of the response evolution factor set, generates a structure response prediction sequence, and submits the prediction sequence together with the response trend stability index to the judgment and control module for subsequent mix proportion optimization strategy reference and adjustment decision basis.

[0109] In some embodiments of the present application, the path deduction module further comprises a structural risk analysis unit, which is configured to: calculate the adjustable parameter interval of the to-be-tested mixture sample under the current structural condition according to the combined variation history of the structural feature set and the response evolution factor set in the adjacent mixture sample; calculate the structural offset trend risk level according to whether the maximum change rate and the duration of the shrinkage rate are higher than the corresponding 95% confidence interval of the sample in the database; output a high risk level when the maximum change rate and the duration of the shrinkage rate are both in the abnormal interval, output a medium risk level if only one index is high, and output a low risk level when both indexes are in the normal interval.

[0110] Specifically, first, a plurality of adjacent samples with high similarity to the structural feature set of the current to-be-tested mixture sample are screened from the database (such as the first several groups with the minimum Euclidean distance).

[0111] The relevant data in the corresponding structural feature set and response evolution factor set of these adjacent samples are extracted, and the combined variation history trajectory of “structure-response” is constructed.

[0112] The historical adjustment records of the above-mentioned adjacent samples on each structural feature parameter (such as the proportion of coarse and fine aggregates, the mineral powder content, and the asphalt content) and the improvement trend of the corresponding response evolution factor are analyzed to identify which parameter adjustment has led to performance improvement (such as reduced density fluctuation and reduced shrinkage rate).

[0113] The parameter value range of these successful adjustments is integrated to form the adjustable parameter interval of the to-be-tested mixture under the current structural condition. For example, if the historical adjacent sample adjusts the asphalt content from 5.3% to 5.0% to significantly reduce the shrinkage rate, the system can set 5.0-5.2% as the recommended asphalt adjustable range of the current sample.

[0114] Risk index statistical analysis: from the volume change data of the current mixture sample in the cooling stage, the maximum change rate of the shrinkage rate and the duration of the change rate continuously exceeding twice the average change rate are extracted. Correspondingly, the distribution of these two indexes is statistically analyzed from all the calibrated samples in the database, and the 95% confidence interval is calculated (for example, the upper limit of the maximum change rate is 0.012 mm / min, and the upper limit of the duration is 180s).

[0115] If the maximum change rate and the duration of the shrinkage rate of the current sample are both higher than the upper limit of the above-mentioned 95% confidence interval, that is, they fall into the abnormal interval, it is judged as a high risk level; if only one of the indexes exceeds, while the other is within the normal interval, it is judged as a medium risk level; if both indexes do not exceed, it is judged as a low risk level.

[0116] The system automatically outputs the risk level of the structural deviation trend according to the analysis results, and can generate a strategy for the judgment and control module in combination with the adjustable parameter interval, prompting whether to allow the execution of the structural adjustment operation or whether to prioritize the control of the cooling process.

[0117] In some embodiments of the present application, when the number of adjacent mixture samples in the database that meet the following conditions is less than three groups, a replenishment mechanism is performed: the Euclidean distance between the structural feature set and the current mixture sample to be tested is less than or equal to a set threshold; the first-order slope value of the viscosity change curve and the average density growth rate in the response evolution factor set are within a set floating range; the replenishment mechanism includes collecting three groups of mixture samples and collecting corresponding mixing response data, compaction response data, and volume change data; the structural feature set and the response evolution factor set are constructed by the feature construction module and are included in the database module.

[0118] Specifically, when the number of adjacent mixture samples in the database that meet the similarity condition is less than three groups, the system automatically triggers the replenishment mechanism, which has the following specific implementation: adjacent sample screening judgment: first, the system performs comparison and retrieval of the current mixture sample to be tested in the following two dimensions: in the structural feature set dimension, the Euclidean distance between all historical samples and the current sample is calculated, and samples with a distance less than or equal to a set threshold are selected; in the response evolution factor set dimension, further select samples in which the first-order slope value of the viscosity change curve and the average density growth rate are both within a set floating range.

[0119] If the number of samples that finally meet the above double conditions is less than three groups, the replenishment mechanism process is entered.

[0120] Based on the boundary conditions of the current mix proportion parameters and the abnormal performance of their response evolution factors, the structural configuration of the three groups of replenishment samples to be collected should preferentially cover the parameter combinations with the largest numerical changes or the highest uncertainties in the current structural feature set. For example: if the coarse and fine aggregate ratio deviates significantly from the standard gradation curve, a group of samples with adjusted coarse and fine aggregate ratio is preferentially collected; if the mineral powder content and the target air voids fitting residual are high, a group of mineral powder content adjustment samples is supplemented; if the asphalt content corresponding to the wrapping factor is significantly abnormal, a group of samples with optimized asphalt content is supplemented.

[0121] Real-time response data collection: for the designed three groups of replenishment structural samples, standardize the process flow operation, and record the following data respectively: viscosity change curve data during mixing; density change data during compaction; volume change curve during natural cooling.

[0122] All data collection needs to be synchronized with time stamps and process control parameters to ensure data consistency and comparability.

[0123] Feature and factor construction: the system call feature construction module parses the raw data of the three groups of supplementary samples to generate the structural feature set and the response evolution factor set. The calculation content should include: coarse and fine aggregate proportion deviation, mineral powder void ratio residual error, wrapping factor, coupling offset; viscosity curve first-order slope value, density growth rate, density standard deviation, shrinkage rate change rate and duration.

[0124] Database supplement and regression call: supplement the structural feature set and the response evolution factor set constructed above into the database module, update the database index, and automatically re-run the path deduction module for structural response matching and optimization strategy output to ensure the sustainable operation of the subsequent regulation module and the accuracy of the strategy generation.

[0125] Referring to Figure 2 The embodiment of the present application provides a proportioning optimization method for asphalt mixture, comprising:

[0126] S1: Obtain the mixing response data, compaction response data and volume change data in the cooling process of a plurality of asphalt mixture samples with different proportioning parameters in historical data.

[0127] S2: Construct a structural feature set according to the proportioning parameters, and construct a response evolution factor set according to the mixing response data, compaction response data and volume change data.

[0128] S3: Obtain the structural feature set and response evolution factor of the current mixture sample to be tested, and compare them with the structural feature set and response evolution factor set of the historical sample, calculate the proportioning response deviation degree, match the proportioning response deviation degree with the set grading standard, and determine the grade; the grade includes a first grade and a second grade with increasing proportioning response deviation degrees.

[0129] S4: When the grade is the second grade, analyze the matching relationship between the change direction of the structural feature set and the change direction of the response evolution factor of the mixture sample to be tested; according to the relationship between the change direction of the structural feature set and the change direction of the response evolution factor, output the proportioning optimization strategy.

[0130] S5: Establish a database including the structural feature set, response evolution factor and grade of the calibrated proportioning sample.

[0131] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, but not to limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that the specific embodiments of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the present application. Any modification or equivalent replacement without departing from the spirit and scope of the present application should be covered in the protection scope of the claims of the present application.

Claims

1. A system for optimizing the ratio of asphalt mixture, characterized in that: include: an acquisition module configured to obtain mixing response data, compaction response data, and volume change data during cooling of a plurality of asphalt mixture samples with different mix parameters in historical data; a feature construction module configured to construct a set of structural features based on the mix ratio parameters, and to construct a set of response evolution factors based on the mixing response data, the compaction response data, and the volume change data; a comparative analysis module configured to obtain a set of structural features and a response evolution factor of a current mixture sample to be tested, compare the set of structural features and the set of response evolution factors with those of historical samples, calculate a degree of deviation in the proportion response, and match the degree of deviation in the proportion response with a set grading standard to determine a grade; the grade includes a first grade and a second grade with increasing degrees of deviation in the proportion response; The judgment and control module is configured to, when the level is the second level, analyze the matching relationship between the change direction of the structural feature set of the tested mixture sample and the change direction of the response evolution factor; and output a ratio optimization strategy based on the relationship between the change direction of the structural feature set and the change direction of the response evolution factor; A database module is configured to establish a database including a set of structural features, response evolution factors and levels of calibrated proportioned samples; When receiving the second level, the judgment and control module performs the following operations: Obtain the numerical change direction of the coarse and fine aggregate ratio, mineral powder content and asphalt content in the current mixture sample to be tested in the structural feature set; At the same time, the following factors are extracted from the set of response evolution factors: the time period from the start of mixing to the maximum viscosity, the first-order slope value obtained by fitting the viscosity change curve, the average density growth rate, the standard deviation of density change, the maximum change rate of shrinkage rate and the change direction of duration; If the change direction of the coarse-fine aggregate ratio, mineral powder content, and asphalt content in the structural feature set is consistent with the change direction of any indicator in the response evolution factor set, it is determined that the structural adjustment is consistent with the performance response, and the asphalt content is finely adjusted. If the direction of change of any parameter in the structural feature set is opposite to the direction of two or more indicators in the response evolution factor set, it is judged as a mix disturbance mismatch, and the coarse and fine aggregate ratio and mineral powder content are re-evaluated, and the asphalt content adjustment is temporarily suspended; If the maximum change rate of the shrinkage rate and its duration index are both more than twice the average level of historical samples, a volume stability abnormality prompt is output, and the target void ratio is adjusted or the cooling process configuration is optimized.

2. The asphalt mixture ratio optimization system according to claim 1, characterized in that: Also includes: a path deduction module configured to compare the mixture sample to be tested with the database based on the set of structural features and the response evolution factor, obtain adjacent mixture samples, and output the adjustable parameter range and structural deviation trend risk of the mixture sample to be tested based on the historical response evolution factor change direction and label level of the adjacent mixture samples; Before the judgment and control module is started, the path deduction module is called in advance.

3. The asphalt mixture ratio optimization system according to claim 2, characterized in that: The mixing parameters include: coarse-fine aggregate ratio, which is the mass ratio of coarse aggregate to fine aggregate; mineral powder content, which is the filler content of the asphalt mixture; asphalt content, which is the asphalt content of the asphalt mixture; target void ratio, which is the air volume fraction of the asphalt mixture; The mixing response data is the viscosity change curve during the mixing process; the compaction response data is the process data of the density changing with the compaction angle or the number of compactions during the compaction process; and the volume change data is the volume change curve of the asphalt mixture during the natural cooling process.

4. The asphalt mixture ratio optimization system according to claim 3, characterized in that: The structural feature set constructed by the feature construction module includes: The relative deviation between the coarse and fine aggregate ratio and the standard gradation curve, the fitting residual between the mineral powder content and the target void ratio, and the wrapping factor between the unit aggregate specific surface area and the asphalt content; the coupling offset between the target void ratio and the stable density value of the compaction prediction model; The response evolution factor set constructed by the feature construction module includes: During the mixing stage, the time period from the start of mixing to the time when the viscosity reaches its maximum value, and the first-order slope value obtained by least squares fitting of the viscosity change curve during this time period; during the compaction stage, the average density growth rate within the first 25% of the total compaction times, and the standard deviation of the density change within the last 25% of the total compaction times; during the cooling stage, the maximum rate of change of the shrinkage rate in the volume change curve from the time of unloading to the time when the temperature drops to room temperature, and the duration during which the maximum rate of change continuously exceeds twice the average rate of change in the volume change curve; The unit aggregate specific surface area is the total specific surface area corresponding to unit mass or unit volume of aggregate under the current proportioning structure.

5. The asphalt mixture ratio optimization system according to claim 4, characterized in that: The comparative analysis module specifically includes: The vectorized similarity is calculated based on the structural feature set and the response evolution factor set, and the difference between the current mixture sample to be tested and the historical samples in the database is calculated by Euclidean distance, cosine angle or Mahalanobis distance; Generate a matching response deviation vector based on the similarity, and obtain the matching response deviation degree by normalizing and weighting each indicator value in the deviation vector; The degree of the ratio response deviation is matched with the set grading standard threshold and assigned to the first grade or the second grade respectively; when the degree of the ratio response deviation exceeds the upper limit of the second grade, it is classified as the second grade.

6. The asphalt mixture ratio optimization system according to claim 5, characterized in that: The path deduction module performs the following operations based on the matching results of the structural feature set and the response evolution factor set in the database: Select the top five historical samples with the smallest Euclidean distance to the structural feature set of the current mixture sample to be tested as the neighboring mixture samples; Extract the corresponding response evolution factor set from these adjacent mixture samples, including the first-order slope value of the viscosity change curve, the average density growth rate, the standard deviation of density change, the maximum change rate and duration of the shrinkage rate; Constructing a response path trajectory according to the numerical change direction of the response evolution factor set and the corresponding ratio parameter adjustment record; Based on the changing trend of the response evolution factor set of the tested mixture sample under different ratio parameter adjustment conditions, a structural response prediction sequence is formed, and the prediction results are output to the judgment and control module.

7. The asphalt mixture ratio optimization system according to claim 6, characterized in that: The path deduction module further includes a structural risk analysis unit, which is configured to: Calculating the adjustable parameter range of the tested mixture sample under the current structural conditions according to the combined change history of the structural feature set and the response evolution factor set in the adjacent mixture samples; The structural deviation trend risk level is calculated based on whether the maximum change rate and duration of the shrinkage rate are higher than the 95% confidence interval corresponding to the samples in the database; When the maximum change rate and duration of the contraction rate are both in the abnormal range, a high risk level is output. If only one indicator is high, a medium risk level is output. When both are in the normal range, a low risk level is output.

8. The asphalt mixture ratio optimization system according to claim 7, characterized in that: When the number of neighboring mixture samples in the database that meet the following conditions is less than three groups, the supplement mechanism is executed: The Euclidean distance between the structural feature set and the current mixture sample to be tested is less than or equal to a set threshold; the response evolution factor set includes the first-order slope value of the viscosity change curve and the average density growth rate within a set floating range; The supplementary mechanism includes collecting three groups of mixture samples and collecting corresponding mixing response data, compaction response data and volume change data; The feature construction module constructs a set of structural features and a set of response evolution factors, and incorporates them into the database module.

9. A method for optimizing the ratio of asphalt mixture, characterized in that: Applied to the system according to any one of claims 1 to 8, the method comprises: S1: Obtain mixing response data, compaction response data, and volume change data during cooling of several asphalt mixture samples with different mix parameters in historical data; S2: constructing a structural feature set according to the mix ratio parameters, and constructing a response evolution factor set according to the mixing response data, compaction response data and volume change data; S3: Obtaining a set of structural features and a response evolution factor of the current mixture sample to be tested, and comparing them with the set of structural features and response evolution factor sets of historical samples, calculating a degree of ratio response deviation, and matching the degree of ratio response deviation with a set grading standard to determine a grade; the grades include a first grade and a second grade with increasing degrees of ratio response deviation; S4: When the level is the second level, analyzing the matching relationship between the change direction of the structural feature set of the tested mixture sample and the change direction of the response evolution factor; outputting a ratio optimization strategy based on the relationship between the change direction of the structural feature set and the change direction of the response evolution factor; S5: Establish a database including a set of structural characteristics, response evolution factors and levels of calibrated ratio samples.

Images