A grading optimization method based on performance index evaluation of asphalt mixture

Abstract

The application discloses a grading optimization method based on asphalt mixture performance index evaluation, comprising the following steps: S1, combining the screening result of mineral aggregate, setting a target mix proportion, according to the requirement, trial mixing several groups of candidate target gradings which meet the specification provision mineral aggregate grading range, and determining the optimal asphalt content of each group of grading curve according to the OAC method; S2, molding the trial mixed candidate target gradings into test pieces according to the specification, and performing road performance tests required by the specification, such as rutting, small beam bending and freeze-thaw splitting, 5 parallel test pieces are prepared for each test, and the road performance test results are collected; S3, using the entropy value method and the grey correlation analysis method, the performance index evaluation value of the candidate target grading is determined; S4, combining the local climate, the expert investigation method is used to determine the weight of the conventional and road performance indexes; S5, according to the index evaluation value and the weight, the comprehensive performance evaluation value of each candidate target grading is calculated, and the optimal target mix proportion is optimized.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of engineering material design and research, in particular to a grading optimization evaluation method and implementation process for target mix design of asphalt mixture. BACKGROUND

[0002] Target mix design is one of the key factors affecting the performance of asphalt mixture. In the actual target mix design process, not only is it necessary to make the grading curve of the target mix within the range of the specified aggregate grading, but it is also necessary to make the performance indicators meet the relevant requirements of the specification. In the target mix design process, the screening results of the aggregate can be used to trial-produce multiple groups of grading curves that meet the requirements of the specification, and the performance indicators of these groups also have their own advantages and disadvantages and emphases. According to the performance requirements, the optimization of each grading curve can significantly improve the road performance of the asphalt mixture, and thus improve the road performance and durability of the road.

[0003] In the actual production process, some designers omit the step of grading optimization and only trial-produce a grading curve that meets the requirements of the specification and performance, which is used as the target mix. This is neither scientific nor can it improve the road performance of the mixture. At present, there is no scientific analysis method for optimizing the target mix and grading, and therefore there are many differences in how to determine the optimal target mix.

[0004] The essence of target mix grading optimization is to optimize the target grading curve that best meets the performance expectations, and therefore it is necessary to establish a performance indicator evaluation system for asphalt mixture. The traditional evaluation of the road performance of asphalt mixture only considers the advantages and disadvantages of the experimental values of each performance indicator, but in the actual performance test, the test results of the performance indicators are generally the average values of the parallel test pieces. Although this value can represent the average level of the performance indicators of the mixture, it cannot reflect the stability of the performance of the mixture. Therefore, in the evaluation system of the performance indicators of the mixture, it is also necessary to consider the volatility of the index values of the parallel test pieces, and to select the coefficient of variation to represent the volatility of the index values of the parallel test pieces. This coefficient not only considers the standard deviation of the parallel indicators, but also makes the standard deviation independent of the average value. The quotient of the two (coefficient of variation) can better represent the volatility of the performance indicators.

[0005] The correlation analysis method in the prior art has been widely applied in multiple fields and is widely recognized by the academic circles. Through a series of processes such as index normalization processing, correlation coefficient calculation and weight assignment, a scientific and explicit evaluation system can be established. The entropy value method is an objective weighting method, and the principle is mainly to assign weights according to the dispersion degree of each evaluation index. The greater the dispersion degree of the index, the greater the weight assigned to it. This method is suitable for the demand of no subjective weighting of evaluation indexes and only objective weighting according to the dispersion degree of the index; the expert investigation method collects expert opinions in the form of a questionnaire, sorts the importance of each index, and then obtains the weight of each index, which is a scientific subjective weighting method. SUMMARY

[0006] The purpose of the present application is to overcome the deficiencies in the prior art, to optimize the target mix ratio and gradation that meet the performance expectations, and to provide a gradation optimization method based on asphalt mixture performance index evaluation. The present application takes the performance indicators required by the specification as the benchmark, and through the weighting method combining subjective and objective methods, it starts from the two dimensions of the average level and volatility of the performance indicators to conduct comprehensive evaluation of the performance indicators of asphalt mixture, and optimizes the target mix ratio and the corresponding gradation curve with excellent and stable comprehensive performance. This method can significantly improve the road performance of asphalt mixture, and further improve the road performance and durability of the road.

[0007] The purpose of the present application is achieved by the following technical solutions:

[0008] A gradation optimization method based on asphalt mixture performance index evaluation, the steps are as follows:

[0009] S1. Combine the screening results of the mineral aggregate to design the target mix ratio. According to the demand, a plurality of candidate target gradations that meet the specification requirements of the mineral aggregate gradation range are trial-produced, and the optimal asphalt content of the gradation curve of each group is determined according to the OAC method.

[0010] S2. The trial-produced candidate target gradation is molded into test pieces according to the specification, and the rutting, small beam bending, freeze-thaw splitting and other road performance tests required by the specification are carried out. Five parallel test pieces are made for each test, and the road performance test results are collected.

[0011] S3. The entropy value method and the grey correlation analysis method are used to determine the performance index evaluation value of the candidate target gradation. Specifically as follows:

[0012] S301. Calculate the average value and the coefficient of variation of the test results of each parallel test piece. The formula for calculating the coefficient of variation is:

[0013]

[0014] Wherein, σ is the standard deviation of the experimental results of the performance index of the parallel test piece, μ is the average value of the experimental results of the performance index of the parallel test piece;

[0015] S302. The average value and the coefficient of variation of the performance index of the parallel test piece are normalized, and the actual value is converted to the range of [0, 1]. The normalization formula is

[0016]

[0017] d jk is the average value, or the normalized value of the coefficient of variation; a jk is the average value and the coefficient of variation of each group of candidate target gradation performance index; maxa j is the maximum value of the average value and the coefficient of variation; mina jk is the minimum value of the average value and the coefficient of variation; j = 1, 2, j = 1 represents the average value, and j = 2 represents the coefficient of variation;

[0018] The correlation analysis between the reference data and the comparison data is carried out on the basis of the analysis of a single index, and the correlation degree of a single index is the correlation coefficient. The optimal value d j is the reference value, D j = [d1, d2, …, d n ] T is the compared data. For a certain performance index, j is the average value or the coefficient of variation, and the correlation coefficient ξ jk of the average value or the coefficient of variation of the kth gradation and the optimal value is calculated according to the following formula:

[0019]

[0020] d j is the optimal value of the average value or the coefficient of variation, d jk is the normalized value of the average value or the coefficient of variation of each group of candidate target gradation, and p is the resolution coefficient, which is in the range of 0 < p < 1, and is taken as 0.5. k is the number of candidate target gradations;

[0021] The correlation coefficient ξ jk between each index is obtained, and a judgment matrix is formed

[0022] S303. The weight of the average value and the coefficient of variation is determined by using the entropy method. First, the index value x jk of the average value and the coefficient of variation of each group of candidate target gradation is calculated according to the following formula:

[0023]

[0024] Then, the entropy value e jThe decision information of each index can be represented as:

[0025]

[0026] wherein q is the number of candidate gradations.

[0027] The utility value f of each index is calculated j That is, the deviation degree of the index, the greater the deviation degree, the greater the value of the index, and the greater the weight.

[0028] f j = |1-e j |, j = 1, 2 (6)

[0029] e j is the entropy value of each index

[0030] The weight factor is calculated to obtain the weight. The weight factor of the average value or the coefficient of variation is expressed as:

[0031]

[0032] Thus, the average value and the coefficient of variation weight matrix is formed

[0033] S304. Determine the evaluation value of each performance index.

[0034] According to the correlation coefficient and the weight of the average value and the coefficient of variation of each performance index obtained above, the evaluation matrix of each performance index of the candidate target gradation is calculated.

[0035] M = W T F (8)

[0036] F is the evaluation matrix.

[0037] S4. Combine the local climate, and use the expert survey method to determine the weight of the conventional and road performance index.

[0038] Due to the performance requirements of asphalt mixture, which is greatly affected by the climate conditions in each place, in order to make the performance of asphalt mixture better match the characteristics of the local climate environment, the expert survey method is used to determine the weight of the performance index of the mixture. This method is a subjective weighting method, through each expert combining the climate characteristics, the importance of each performance index is sorted, so as to determine the weight of each index.

[0039] S401. Collect expert opinions and establish an index importance sequence array.

[0040] According to the traditional Delphi expert investigation method, the questionnaire is sent to the experts with high academic authority according to the requirements, and the opinions of the experts are sorted and scored for the secondary indicators under the common quantitative indicators and the performance indicators. The scoring rules are as follows: if an expert thinks that a secondary indicator is the "first consideration", the score is 1; if an expert thinks that a secondary indicator is the "second choice", the score is 2, and so on, until each indicator is scored.

[0041] If n experts are invited to conduct a questionnaire survey and the number of indicators is j, the following table results are obtained:

[0042] Expert questionnaire survey results

[0043]

[0044] The indicators for evaluation are j * , so they can be converted into an n x j * matrix to obtain the indicator importance sequence matrix A.

[0045]

[0046] S402. Calculate the membership degree

[0047] The membership function for converting the expert score value is

[0048]

[0049] where G = A nj* , n is the number of experts, l is the score value of each indicator given by the expert, l = {1, 2, 3…j *}, j * is the total number of indicators, and m = j * + 2.

[0050] δ(G) is a variable, and the closer δ(G) of an indicator is to 1, the more important the indicator is.

[0051] Therefore, the membership degrees of the common quantitative indicators and the road performance indicators are calculated according to the formula, and let c ii* = δ(G) be the membership degree of A ij* , the following membership degree matrix can be obtained.

[0052]

[0053] S403. "Blindness analysis" and calculation of overall understanding degree:

[0054] First, define the average understanding degree c j*

[0055] c j* = (c1j* +c 2j* +c 3j* +…+c nj* ) / n (10)

[0056] Then define a "blindness" B of an expert due to the uncertainty caused by cognition j* ,

[0057] B j ={[max(c 1j* ,c 2j* ,c 3j* ,…,c nj* )-c j* ]+[min(c 1j* ,c 2j* ,c 3j* ,…,c nj* )-c j* ]} / 2(11)

[0058] Finally, calculate the overall understanding x of n experts on two groups of indicators j* ,

[0059] x j* =c j* (1-B j* ) (12)

[0060] S404. Normalization processing:

[0061] After normalization processing on the overall understanding x j* , the following normalization formula is obtained:

[0062] α k* =x k* / (x1+x2+x3+…+x j* ), k*=1,2,…,j*; (13) Finally, the weight matrix of each performance indicator is obtained

[0063] Q=[α1 α2 … α j* ]

[0064] S5. According to the evaluation value and weight of each indicator, the comprehensive performance evaluation value of each candidate target grading is calculated, and the optimal target mixing ratio is selected.

[0065] From the above calculation, the evaluation matrix M of each performance indicator of the candidate target grading is obtained, and the weight matrix Q of each performance indicator, and the comprehensive evaluation matrix of the candidate target grading of asphalt mixture is obtained from the following formula:

[0066] Z=MQ T (14)

[0067] The optimal gradation can be obtained by the comprehensive evaluation matrix Z, and the corresponding mix ratio is the target mix ratio.

[0068] Compared with the prior art, the technical scheme of the present application has the beneficial effects that:

[0069] 1. The method of the present application screens the candidate target gradation through target mix ratio design, collects the test results of parallel test pieces through conventional and road performance tests, calculates the evaluation values of each performance index using the entropy method and grey correlation analysis, calculates the weight of each performance index through expert investigation method, and finally calculates the comprehensive performance evaluation value of the candidate target gradation.

[0070] 2. The method of the present application takes the average value and the coefficient of variation of the parallel test piece experimental values of the candidate target gradation as the evaluation objects, so that the performance evaluation of the candidate target gradation not only considers the test values of its conventional and road performance, but also considers the volatility of the performance index, making the gradation optimization evaluation method comprehensive and scientific.

[0071] 3. The method of the present application uses the grey correlation analysis method to construct the evaluation system, evaluates the road performance of each gradation by analyzing the correlation degree between the index data column of the candidate target gradation and the optimal value data column, and scientifically optimizes. The weight determination method combining subjective and objective methods is used to determine the weight of each index, ensuring the authority of the index weight determination.

[0072] 4. The method of the present application considers the optimal asphalt content of the asphalt mixture and various road performance indexes in combination with the asphalt pavement construction technical specification, simulates the complexity in engineering, and provides a more scientific and practical evaluation idea for the optimization of the target mix ratio gradation of the asphalt mixture.

[0073] 5. The method of the present application scientifically optimizes the target mix ratio, solves the differences in selecting and determining the optimal target mix ratio in actual engineering, and the optimal target mix ratio obtained based on the performance comprehensive evaluation can effectively ensure the comprehensive road performance of the asphalt mixture.

[0074] 6. The weight determination method involved in the method of the present application considers the climate characteristics of the engineering location, and allocates the weight of each road performance index according to local conditions, to determine the target mix ratio of the asphalt mixture most suitable for the local climate characteristics. BRIEF DESCRIPTION OF DRAWINGS

[0075] Figure 1 is a flowchart of the method of the present application.

[0076] Figure 2 is a curve graph of three groups of candidate target gradations.

[0077] Figure 3 is a weight diagram of each index.

[0078] Figure 4 is the comprehensive evaluation result of each candidate target gradation. DETAILED DESCRIPTION

[0079] The application will be further described below in conjunction with the accompanying drawings and specific examples. It should be understood that the specific examples described herein are only used to explain the application and do not limit the application.

[0080] The embodiment provides a gradation optimization method based on performance index evaluation of asphalt mixture, first target mixture ratio design is performed, a plurality of groups of candidate target mixture ratios and gradations are obtained through standard trial blending in combination with the specification, the best asphalt content of each candidate target gradation is determined, and road performance tests are performed. Each test contains five parallel test pieces, test results of the parallel test pieces are collected, evaluation values of each performance index are determined through an entropy value method and a grey correlation analysis method, the best asphalt content and the weights of each road performance index are determined through an expert investigation method, and finally the comprehensive performance evaluation values of each candidate target gradation are calculated from the evaluation values and the weights of the indexes, the best target mixture ratio and gradation are optimized as the final target mixture ratio design result.

[0081] Specifically, the embodiment takes AC-20 asphalt concrete as an example.

[0082] Step one: design a target mixture ratio, and trial blend a plurality of groups of candidate target gradations.

[0083] In this embodiment, AC-20 asphalt mixture is selected, and the screening results of each specification of mineral aggregate are shown in the following table through a screening experiment:

[0084] Table 2 Screening results of each specification of mineral aggregate

[0085]

[0086] Continued Table

[0087]

[0088] The screening results are based on the target mix design according to the Technical Specification for Construction of Asphalt Pavement. Three different mineral aggregate gradations are designed as candidates within the range of the engineering design gradation. The trial method is used for the mix design in this case. It is assumed that the particles of a certain size in the mixture are composed of a certain dominant size of aggregate, and there is no such particle in other aggregates. Then the approximate proportion of each aggregate is calculated according to the main particle size, and the proportion is adjusted to meet the specification requirements. Finally, the three candidate synthetic gradations analyzed in this case are designed. At the same time, the technical indicators of the three synthetic gradations are tested to meet the specification requirements. The mix ratio that does not meet the specification is corrected until it meets the requirements. The screening is a conventional test, and the aggregate sieve is used for screening, which is a common method to obtain the passing rate of each size aggregate. The specific method is described in the Highway Engineering Aggregate Test Procedure.

[0089] The candidate target mix design results and the passing rate of the synthetic gradation are shown in the following table:

[0090] Table 3 Candidate target mix design results

[0091]

[0092] Table 4 Passing rate of synthetic gradation

[0093]

[0094]

[0095] The curves of the three candidate target gradations are shown in Figure 2 .

[0096] Based on the three candidate synthetic gradations, the OAC method is used to calculate the optimum asphalt content. The optimum asphalt contents corresponding to the candidate target gradations 1, 2 and 3 are 4.9%, 4.7% and 4.8% respectively.

[0097] Step 2: Form test pieces for each candidate target gradation, and conduct road performance tests to collect test results.

[0098] According to the Highway Engineering Asphalt and Asphalt Mixture Test Procedure, road performance tests are conducted on the three target gradations. Five parallel test pieces are used for each test. The test results are shown in the following table:

[0099] Table 5 Test results of candidate target gradation parallel test pieces

[0100]

[0101] Table 1 (continued)

[0102]

[0103] Table 2 (continued)

[0104]

[0105]

[0106] Step three: using entropy method and grey correlation analysis method, the performance index evaluation value of candidate target gradation is determined.

[0107] 301. The average value and variation coefficient of each parallel test specimen test result are calculated.

[0108] The variation coefficient is calculated according to formula (1), and the average value and variation coefficient of each performance index of the candidate target gradation are shown in Table 6 and Table 7, and the optimal asphalt content is taken as the average value.

[0109] Table 6 Average value of performance index of candidate target gradation

[0110]

[0111] Table 7 Variation coefficient of performance index of candidate target gradation

[0112]

[0113] 302. The performance index is normalized, the correlation coefficient is calculated, and the judgment matrix is formed.

[0114] According to formula (2), the average value and variation coefficient of the performance index of the asphalt mixture are normalized, and the optimal value is selected. For the index with the larger value, the larger the average value, the better the benefit, and the optimal value is the maximum value. For the index with the smaller value, the smaller the average value, the better the benefit, and the optimal value is the minimum value. The minimum value of the variation coefficient is the optimal value.

[0115] Then, the correlation coefficient of each performance index average value and variation coefficient and the optimal value is calculated by formula (3). Thus, the judgment matrix F of the optimal asphalt content and each performance index is formed, as shown in the following matrix:

[0116] Optimal asphalt content: F 0 = [0.333, 1.000, 0.500]

[0117] Marshall stability: Dynamic stability:

[0118] Low temperature bending failure strain: Residual stability:

[0119] Freeze-thaw splitting strength ratio: Water permeability coefficient:

[0120] 303. The average value and the coefficient of variation weight are calculated by the entropy method

[0121] The weight of the average value and the coefficient of variation is determined by the entropy method. The weight factor w of the average value and the coefficient of variation of each test index is calculated by formula (4)-(7) j , forming the average value and the coefficient of variation weight matrix

[0122] Marshall stability: Dynamic stability:

[0123] Low-temperature bending failure strain: Residual stability:

[0124] Freeze-thaw splitting strength ratio: Water permeation coefficient:

[0125] 304. Determine the evaluation value of each performance index.

[0126] The evaluation value of each performance index of the candidate target gradation is obtained by formula (8), and the correlation coefficient is used as the evaluation value of the optimal asphalt content. The index evaluation values are shown in the following table:

[0127] Optimal asphalt content: M 0 = [0.333 1.000 0.500]

[0128] Marshall stability: M 1 = [0.644 0.790 0.403] Dynamic stability: M 2 = [0.333 1.000 0.824]

[0129] Low-temperature bending strain: M 3 = [0.575 0.633 0.729] Residual stability: M 4 = [50.702 0.631 0.660]

[0130] Freeze-thaw splitting strength ratio: M 5 = [0.586 0.333 1.000] Water permeation coefficient: M 6 = [1.000 0.427 0.626]

[0131] Step four: combine the local climate and use the expert survey method to determine the weight of the conventional and road performance indexes.

[0132] 401. Collect expert opinions and establish an index importance sequence matrix.

[0133] The area where the example is located is in the south, and it is hot in summer, so the high temperature performance of asphalt mixture is the performance index that is focused on. Five experts were invited to conduct a questionnaire survey on the best asphalt content and various performance indexes using the expert survey method, and the results are shown in Table 8:

[0134] Table 8 Expert questionnaire survey results

[0135]

[0136] The best asphalt content and various road performance indexes have a total of 7, so they can be converted into a 5x7 matrix to obtain the index importance sequence matrix A:

[0137]

[0138] 402. Calculate the membership degree

[0139] According to formula (9), the membership degrees of the constant quantity index and the road performance index are calculated, where j* = 7 and m = 9, and the following membership degree matrix C can be obtained.

[0140]

[0141] 403. "Blindness analysis" and calculation of overall understanding degree

[0142] According to formula (10), the average understanding degrees of the reorganized indexes are calculated as follows:

[0143] c j* = [0.874 0.906 1.000 0.605 0.699 0.682 0.333]

[0144] According to formula (11), the "blindness" B caused by uncertainty is obtained j* ,

[0145]

[0146] According to formula (12), the overall understanding degrees x of the two groups of indexes by the five experts are calculated as follows: j*

[0147] x j* = [0.890 0.913 1.000 0.577 0.703 0.703 0.333]

[0148] 404. Normalization processing.

[0149] After normalization processing, the final constant and road performance index weight matrix Q is obtained as follows:

[0150] ​Q=[0.174 0.178 0.195 0.113 0.137 0.137 0.065]

[0151] Visualizing indicator weights, such as Figure 3 :

[0152] Depend on Figure 3 It is known that in high-temperature regions, experts believe that dynamic stability is the most important indicator, followed by Marshall stability, and these weights can be used to evaluate subsequent routine and road performance.

[0153] Step 5: Calculate the comprehensive performance evaluation value of each candidate target mix based on the evaluation value and weight of each indicator, and select the optimal target mix ratio.

[0154] Combining the evaluation matrix M of each performance index of the candidate target gradation with the weight matrix Q of each performance index, the comprehensive evaluation matrix Z of the candidate target gradation of asphalt mixture is obtained:

[0155] Z = [0.545 0.742 0.671]

[0156] Visualize the comprehensive evaluation value of each candidate target gradation, such as Figure 4 :

[0157] Depend on Figure 4 It can be seen that the final optimal evaluation result is: gradation 2 > gradation 3 > gradation 1. Gradation 2 is the optimal target gradation, with the highest evaluation values ​​for both dynamic stability and Marshall stability, consistent with the weighting ratio assigned by the experts. The target mix proportion corresponding to gradation 2 is:

[0158] 15-25mm: 10-15mm: 5-10mm: 0-5mm: mineral powder = 20: 22: 15: 35: 8

[0159] The corresponding optimal asphalt content is 4.7%. Using this as the final target mix design result, the target mix gradation optimization is complete.

[0160] This invention is not limited to the embodiments described above. The above description of specific embodiments is intended to illustrate and explain the technical solutions of this invention. The specific embodiments described above are merely illustrative and not restrictive. Without departing from the spirit and scope of the claims, those skilled in the art can make many specific modifications based on the teachings of this invention, and these modifications all fall within the scope of protection of this invention.

Claims

1. A gradation optimization method based on asphalt mixture performance index evaluation, characterized in that, Comprise: S1. Combined with the screening results of mineral aggregate, set the target mix ratio, according to the demand to try out several groups of candidate target gradation to meet the specification requirements of mineral aggregate gradation range, and determine the best asphalt content of each group of candidate target gradation curve according to OAC method; S2. The candidate target gradation of trial is molded according to the specification, and the road performance test required by the specification is carried out, 5 parallel test pieces are made for each road performance test, and the road performance test results are collected; the road performance test includes Marshall test, immersion Marshall test, track test, small beam bending test, freeze-thaw splitting test and water permeability test; S3. The performance index evaluation value of the candidate target gradation is determined by using entropy method and grey correlation analysis method; specifically including: S301. For each performance index, the average value and the coefficient of variation of the test results of each group of candidate target gradation parallel test pieces are calculated respectively; the coefficient of variation The calculation formula is: （1）； wherein is the standard deviation of the experimental results of the performance index of the parallel test pieces, and μ is the average value of the experimental results of the performance index of the parallel test pieces. S302. The average value and the coefficient of variation of the performance index of the parallel test pieces are normalized, and the actual value is converted to the range of [0, 1], and the normalization formula is （2）； is the average value, or the normalized value of the coefficient of variation value; is the average value, or the coefficient of variation of the performance index of each group of candidate target gradation; is the maximum value of the average value, or the coefficient of variation of the performance index; is the minimum value of the average value, or the coefficient of variation of the performance index; j = 1, 2, j = 1 represents the average value, and j = 2 represents the coefficient of variation; The correlation analysis between reference and comparative data is conducted based on the analysis of individual indicators; the degree of correlation of a single indicator is the correlation coefficient; with the optimal value... For reference purposes, For the data being compared, for a certain performance index, j is the average or coefficient of variation, and the correlation coefficient between the average or coefficient of variation of the k-th candidate target gradation and the optimal value is... The calculation formula is: （3）； is the optimal value of the average value or the coefficient of variation, is the normalized value of the average value or the coefficient of variation of each group of candidate target gradations, is the resolution coefficient, the value range is 0<ρ<1, and 0.5 is taken, and k is the number of candidate target gradations; Obtaining the correlation coefficient between each index , forming the evaluation matrix ; S303. The weight of the average value and the coefficient of variation is determined by using the entropy method, and the index value of the average value and the coefficient of variation of each group of candidate target gradation is calculated The calculation formula is as follows: （4）； post-computing entropy values , representing individual index decision information: （5）； wherein q is the number of candidate target gradations; The utility value of each index is calculated , i.e. the deviation degree of the index, the greater the deviation degree, the greater the utility value , the greater the value, the greater the weight; （6）； Entropy values for each index The weight factor is calculated to obtain the weight; the weight factor of the test average value or the coefficient of variation is expressed as: （7）； from which an average and coefficient of variation weight matrix is formed ; n i representing the utility value ; S304. Determine the evaluation value of each performance index; according to the correlation coefficient and weight of the average value and the coefficient of variation of each performance index obtained above, the evaluation matrix of each performance index of the candidate target gradation is calculated: （8）； F is the evaluation matrix; S4. Combined with the local climate, the weight of the conventional and road performance index is determined by using expert investigation method; specifically including: S401. Collect expert opinions and establish index importance sequence matrix; Referring to Delphi expert investigation method, questionnaire survey is conducted to experts, and their opinions are sorted and scored respectively for secondary indexes under common quantitative indexes and recruitment performance indexes; scoring rules are as follows: if an expert thinks that a secondary index is "the first consideration", 1 score is given; if he thinks that it is "the second choice", 2 score is given, and so on, until each index is scored; finally, n×j matrix composed of n experts and j indexes is obtained, which is the index importance sequence matrix. * * ;​​ S402. Calculate the membership degree; the membership function for converting the expert score value is: （9）； wherein, n is the number of experts, and l is the score value of each indicator given by the experts, l = {1, 2, 3, …, j * }, j * is the total number of indicators, ; is a variable, the importance of a certain indicator The closer to 1, the more important it indicates; The membership degrees of the constant quantity index and the road performance index are calculated according to the formula, respectively, and the membership degrees of the constant quantity index and the road performance index are set as is the membership degree of the constant quantity index and the road performance index, respectively, to obtain a membership degree matrix. ； S403. Blindness analysis and calculation of overall understanding degree; Define an average level of awareness of the expert for each indicator ; （10）； Defining a blind spot due to uncertainty created by an expert's cognition , （11）； Finally, the overall degree of awareness of the two groups of indicators by n-bit experts is calculated , （12）； S404. Normalization processing; Overall awareness After normalization, the normalization formula is as follows: ， （13）； Finally, the weight matrix of each performance index is obtained: ； S5. According to the evaluation value and weight of each index, the comprehensive performance evaluation value of each candidate target gradation is calculated, and the optimal target mix ratio is selected.

2. The gradation optimization method based on the performance index evaluation of asphalt mixture according to claim 1, characterized in that, Step S5 is based on the evaluation matrix of each performance index of the candidate target gradation and the weight matrix of each performance index The comprehensive evaluation matrix of the candidate target gradation of the asphalt mixture is obtained by the following formula: （14）； Thus the comprehensive evaluation matrix Z is obtained, and the corresponding mix ratio of the optimal gradation is the target mix ratio.

Images