Method for determining particle size distribution of aggregates and system therefor, and electronic equipment

Machine vision and Bayesian statistics are used to efficiently and accurately determine aggregate particle size distribution, addressing inefficiencies in traditional sieving methods and ensuring compliance with construction standards.

JP2024169406AActive Publication Date: 2024-12-05CHINA JILIANG UNIV
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
JP2024084356
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-05-24
Filing Date
2024-05-23
Publication Date
2024-12-05
Estimated Expiration
2044-05-23

AI Technical Summary

Technical Problem

Existing methods for determining aggregate particle size distribution in asphalt mixtures are inefficient due to clogging during sieving and require manual data analysis, failing to meet the need for quick and accurate determination of particle size distribution within ±5% error.

Method used

A method using machine vision detection and Bayesian statistics to determine aggregate particle size distribution by constructing aggregate particle subsets, calculating likelihood and posterior probabilities, and applying binomial and beta distributions to estimate particle size distribution accurately.

Benefits of technology

The method improves the efficiency and accuracy of determining aggregate particle size distribution, ensuring compliance with construction standards and reducing human error, enabling rapid and automated screening.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024169406000001_ABST
    Figure 2024169406000001_ABST
Patent Text Reader

Abstract

SOLUTION: A method for determining a particle size distribution of aggregates and a system therefor, and electronic equipment are disclosed, with the present disclosure relating to a road surface engineering technology field. The method includes: extracting a sample from a measurement object aggregate and constructing a set of aggregate particles, determining a likelihood function of the measurement object aggregate by using machine vision detection technology, and further determining a plurality of posterior probabilities by using a first Bayesian formula, then constructing a set of binomial distributions in accordance with particle quantities of different actual particle diameter intervals; determining a set of posterior distributions by using a second Bayesian formula, assuming a beta distribution as prior distribution, on the basis of the particle quantities of different actual particle size intervals and the set of binomial distributions, and determining the particle size distribution of the measurement object aggregate in accordance with the set of posterior distributions.EFFECT: The present invention determines a posterior probability by constructing a likelihood function and a prior probability and further determines a set of posterior distributions, enabling quickly determining the particle size distribution of aggregate in a coarse aggregate and further improve the efficiency of determining asphalt performance.SELECTED DRAWING: Figure 1
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to the field of road construction technology, and in particular to a method for determining the particle size distribution of aggregates, and a system and electronic device therefor. [Background technology]

[0002] Asphalt mixture is a multi-phase composite material composed of asphalt adhesive, aggregate particles and filler, etc., where the mass of coarse aggregate accounts for more than 90% of the total, and the performance of the asphalt mixture basically depends on the properties of the coarse aggregate. Through research, it is shown that whether the aggregate particle size distribution is in line with the specifications is the first major influencing factor that affects the performance of the asphalt mixture. In road construction standards, the aggregate particle size distribution is realized through the sieving method by sieving net, but during sieving, the situation where the aggregate clogs the sieve holes frequently occurs, and it is necessary to manually improve the passage, in this case, the detection efficiency is low, and the aggregate after sieving requires manual data analysis, which cannot meet the needs of the mixing station staff to quickly sieve the particle size of the aggregate. Usually, in actual construction, it is required that the absolute error of the aggregate particle size distribution at each interval does not exceed ±5%. Summary of the Invention [Problem to be solved by the invention]

[0003] The present invention aims to provide a method for determining the particle size distribution of aggregate, a system for the method, and an electronic device that can quickly determine the particle size distribution of aggregate in coarse aggregate and further improve the efficiency of determining the performance of asphalt. [Means for solving the problem]

[0004] In order to achieve the above object, the present invention provides the following solutions. The method for determining the particle size distribution of aggregate of the present invention comprises the steps of: Extracting a sample from the aggregate to be measured to construct a set of aggregate particles including a plurality of aggregate particle subsets, the actual particle sizes of all aggregate particles in the same aggregate particle subset all belong to the same actual particle size interval, the aggregate particle subsets correspond one-to-one to the actual particle size intervals, and the intersection of any two of the actual particle size intervals is an empty set; determining a likelihood function for the measured aggregate based on the aggregate particle set using machine vision detection technology, the likelihood function being used to describe a probability distribution relationship of aggregate particles in different actual particle size intervals and different ferret particle size intervals; Obtaining a prior probability of the aggregate to be measured, the prior probability being determined by the mass of a single aggregate particle in different actual particle size intervals; determining a plurality of posterior probabilities using a first Bayes' formula according to the likelihood function and the prior probability; and i |x i ) is the actual particle size interval c of each aggregate in the i-th ferret particle size interval xi i Denote the probability at converting the prior probabilities of all aggregate particles in the aggregate particle subset from sums of mass ratios to particle quantity ratios in the corresponding actual particle size intervals; constructing a set of binomial distributions according to particle quantity ratios of different actual particle size intervals, and the binomial distributions in the set of binomial distributions correspond one-to-one to the actual particle size intervals; Based on a set of particle quantities and binomial distributions of different actual particle size intervals, a set of posterior distributions is determined using a beta distribution as a prior distribution and a second Bayes formula, and the posterior distributions in the set of posterior distributions correspond one-to-one to the actual particle size intervals; and determining a particle size distribution of the aggregate to be measured in response to the set of posterior distributions.

[0005] Optionally, determining a likelihood function for a measured aggregate based on the set of aggregate particles using machine vision detection techniques includes: determining a ferret particle size of every particle in each of said aggregate collections utilizing machine vision detection techniques; constructing a plurality of ferret particle size intervals, and the intersection of any two of said ferret particle size intervals is an empty set; determining a plurality of likelihood function values ​​in response to a plurality of ferret particle sizes, i |c i ) is the i-th aggregate subset c i The ferret particle size of the aggregate in the i-th ferret particle size interval x i This indicates the probability that the

[0006] Optionally, determining a particle size distribution of the aggregate to be measured in response to the set of posterior distributions comprises: determining one of the posterior distributions in the set of posterior distributions as a current posterior distribution; Determining the expected value of the current posterior distribution as the quantity proportion of the actual particle size interval that the measured aggregate corresponds to.

[0007] Optionally, the first Bayes' formula is TIFF2024169406000002.tif24170, where N is the number of actual particle size intervals, and P(c i ) is the i-th aggregate subset c i is the prior probability of

[0008] The system for determining the particle size distribution of aggregates is an aggregate particle set construction module for extracting a sample from the aggregate to be measured and constructing an aggregate particle set; the aggregate particle set includes a plurality of aggregate particle subsets, the actual particle sizes of all aggregate particles in the same aggregate particle subset all belong to the same actual particle size interval, the aggregate particle subsets correspond one-to-one to the actual particle size intervals, and the common set of any two of the actual particle size intervals is an empty set; A likelihood function determination module for determining a likelihood function of the measured aggregate based on the aggregate particle set using machine vision detection technology, the likelihood function being used to describe the probability distribution relationship of the aggregate particles in different actual particle size intervals and different ferret particle size intervals; a prior probability determination module for obtaining a prior probability of the aggregate to be measured, the prior probability being determined by the mass of a single aggregate particle in different actual particle size intervals; a posterior probability determination module for determining a plurality of posterior probabilities using a first Bayes' formula according to the likelihood function and the prior probability; i |x i ) is the particle size interval x of the ith ferret i The actual particle size interval c of each aggregate in i Denote the probability at an actual particle size interval particle quantity determination module for converting the prior probabilities of all aggregate particles in the aggregate particle subset from a sum of mass ratios to a particle quantity ratio of the corresponding actual particle size interval; a binomial distribution set construction module for constructing a binomial distribution set according to particle quantity ratios of different actual particle size intervals, and the binomial distributions in the binomial distribution set correspond one-to-one to the actual particle size intervals; A posterior distribution set determination module for determining a posterior distribution set by using a second Bayes formula with a beta distribution as a prior distribution based on a set of particle quantities and binomial distributions of different actual particle size intervals, wherein the posterior distributions in the set of posterior distributions correspond one-to-one to the actual particle size intervals; The particle size distribution of the aggregate to be measured is determined according to the set of posterior distributions.

[0009] An electronic device comprising a memory and a processor, the memory being used to store a computer program, and the processor executing the computer program, such that the electronic device can perform the above-mentioned method for determining particle size distribution of aggregates of the present invention.

[0010] Optionally, said memory is a readable storage medium. Effect of the Invention

[0011] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects.

[0012] The present invention provides a method for determining the particle size distribution of aggregate, and a system and electronic device thereof, which extracts a sample from the aggregate to be measured, constructs a set of aggregate particles, and based on the set of aggregate particles, uses machine vision detection technology to determine a likelihood function of the aggregate to be measured, obtains a prior probability of the aggregate to be measured, and based on the likelihood function and the prior probability, uses the first Bayes formula to determine multiple posterior probabilities, and determines that the sum of the prior probabilities of all aggregate particles in the subset of aggregate particles is the particle quantity of the corresponding actual particle size interval, and based on the particle quantities of the different actual particle size intervals and the set of binomial distributions, uses the second Bayes formula with the beta distribution as the prior distribution to determine a set of posterior distributions, and based on the set of posterior distributions, determines the particle size distribution of the aggregate to be measured. The present invention determines the posterior probability by constructing the likelihood function and the prior probability, and further determines the set of posterior distributions, which can quickly determine the particle size distribution of aggregate in coarse aggregate, and further improves the efficiency of asphalt performance determination. [Brief description of the drawings]

[0013] In order to more clearly describe the embodiments of the present invention or the technical solutions in the prior art, the drawings necessary to be used in the embodiments are briefly described below. Obviously, the drawings in the following description are only some embodiments of the present invention, and those skilled in the art can further obtain other drawings based on these drawings without creative efforts.

[0014] [Figure 1] 1 is a flowchart showing a method for determining particle size distribution of aggregate in Example 1 of the present invention. [Diagram 2] FIG. 2 is a schematic diagram of aggregate image collection in the first embodiment of the present invention. [Diagram 3] FIG. 2 is a diagram showing the actual particle size intervals between different ferret particle sizes in Example 1 of the present invention. [Figure 4] 3 is a step flowchart for estimating particle size distribution of aggregate in the first embodiment of the present invention. [Diagram 5] FIG. 11 is a first posterior distribution diagram for estimating that aggregate with S9 specifications is in the actual particle size interval c1 in Example 1 of the present invention. [Figure 6] FIG. 2 is a second posterior distribution diagram estimating that aggregate with S9 specifications is in the actual particle size interval c1 in Example 1 of the present invention. [Figure 7] FIG. 3 is a third posterior distribution diagram estimating that aggregate with S9 specifications is in the actual particle size interval c1 in Example 1 of the present invention. [Figure 8] This is a fourth posterior distribution diagram estimating that aggregate with S9 specifications is in the actual particle size interval c1 in Example 1 of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] The technical solutions in the embodiments of the present invention will be described below clearly and completely with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts are all within the protection scope of the present invention.

[0016] The present invention aims to provide a method for determining the particle size distribution of aggregate, a system thereof, and an electronic device that can quickly determine the particle size distribution of aggregate in coarse aggregate and further improve the efficiency of determining the performance of asphalt.

[0017] In order to make the above objects, features and advantages of the present invention more apparent and comprehensible, the present invention will be described in more detail below with reference to the drawings and the description of the preferred embodiments of the present invention.

[0018] Example 1 The present invention uses the maximum Feret particle size of aggregate particles measured by machine vision as collected data, and there is an error between it and the actual particle size of aggregate particles, and the actual particle size of aggregates in different Feret particle size intervals still overlap (as shown in Figure 3). The present invention reduces the error caused by this by using Bayes' formula, and can correctly estimate the proportion of aggregates with different particle sizes (i.e., particle size distribution). In order to facilitate the understanding of aggregate particle size, i (0 <i<15,i∈N * ) is used to represent the actual particle size interval of the aggregate, and this interval range is determined by JTG F40-2004 "Technical Specifications for Road Asphalt Surface Construction", and x i (0 <i<15,i∈N * ) represents the Feret particle size interval of the aggregate.

[0019] As shown in FIG. 1 and FIG. 4, the present embodiment provides a method for determining particle size distribution of aggregate, which includes the following steps: Step 101: Extract a sample from the aggregate to be measured to construct a set of aggregate particles, the set of aggregate particles includes multiple aggregate particle subsets, the actual particle sizes of all aggregate particles in the same aggregate particle subset all belong to the same actual particle size interval, the aggregate particle subsets correspond one-to-one to the actual particle size intervals, and the common set of any two actual particle size intervals is an empty set. Step 102: Based on the set of aggregate particles, a likelihood function of the measured aggregate is determined using machine vision detection technology, and the likelihood function is used to describe the probability distribution relationship of aggregate particles in different actual particle size intervals and different ferret particle size intervals. Step 102 includes: Step 1021: Using machine vision detection technology, the ferret particle size of every particle in each aggregate set was determined. Step 1022: A plurality of ferret particle size intervals were constructed, and the intersections of any two ferret particle size intervals were all empty sets. Step 1023: Determine a plurality of likelihood function values ​​according to a plurality of ferret particle sizes, and estimate the likelihood function value P(x i |ci ) is the i-th aggregate subset c i The ferret particle size of the aggregate in the i-th ferret particle size interval x i The probability of belonging to

[0020] Regarding how to obtain the likelihood function in Bayes' formula, first, we use a sieve net to find different c i Then, 1000 pieces of each experimental aggregate are obtained, and a set D is obtained using machine vision measurement. The set D is i All experimental aggregates belonging to the group are shown, where x i The number of particles in the interval is n i It was.

[0021] Where: TIFF2024169406000003.tif16170. This gives the likelihood function Listed TIFF2024169406000004.tif16170.

[0022] By taking ln from the above equation, we obtain the following equation. TIFF2024169406000005.tif16170

[0023] Where: Since TIFF2024169406000006.tif20170 is a known condition, the Lagrange undetermined multiplier method is used for the above equation. TIFF2024169406000007.tif18170 was used.

[0024] Find the partial derivative with respect to θ, set the partial derivative function equal to 0, I got TIFF2024169406000008.tif17170.

[0025] As a result, TIFF2024169406000009.tif23170 can be derived.

[0026] therefore I get TIFF2024169406000010.tif15170.

[0027] The solution to this problem is the degree function value θ i It is possible to obtain aggregates of c i If the Feret particle size measured by machine vision belongs to x i The probability of belonging to is P(x i |c i ) was shown.

[0028] Step 103: Obtain the prior probability of the aggregate to be measured, and the prior probability is determined by the mass of a single aggregate particle in different actual particle size intervals. Since the aggregate specification is judged by the mixing station staff to be in compliance with the standard (JTG F40-2004 "Technical Specifications for Road Asphalt Surface Construction") by the sieve mesh sieving method, the present invention takes the aggregate specification in JTG F40-2004 as the prior probability. In actual construction, the sieve mesh sieving method is used to estimate the aggregate particle size distribution, and the mass of the aggregate passing through each level of the sieve mesh is judged according to the percentage of the total mass. In the present invention, the aggregate particle size distribution is judged according to the percentage of the aggregate quantity in different particle size intervals that occupies the total quantity, so in order to make the calculation of the present invention convenient, it is necessary to estimate the relationship between the aggregate material mass and quantity.

[0029] Varies depending on the screening method using sieve meshc i Obtain 1000 particles for each of the experimental aggregates, and then i Weigh the aggregates and calculate the different c i The bone mass of the bone is obtained, and the different ci in accordance with the JTG F40-2004 standard with the quantity as a unit are calculated using the prior probability P(c i ) was obtained.

[0030] Step 104: Determine a plurality of posterior probabilities using the first Bayes' formula according to the likelihood function and the prior probability, and calculate the posterior probability P(c i |x i ) is the particle size interval x of the ith ferreti The probability of each aggregate in the i-th actual particle size interval ci is shown.

[0031] Here, the first Bayes formula is TIFF2024169406000011.tif26170, where N is the number of actual particle size intervals, and P(c i ) is the i-th aggregate subset c i was the prior probability.

[0032] Step 105: The prior probabilities of all aggregate particles in the subset of aggregate particles were converted from sums of mass ratios to particle quantity ratios of the corresponding actual particle size intervals. Step 106: A set of binomial distributions is constructed according to the particle quantity ratios of different actual particle size intervals, and the binomial distributions in the set of binomial distributions correspond one-to-one to the above actual particle size intervals. The quality inspector at the mixing station obtained the test sample (the number of particles was about 3000, and the quantity was recorded as M) by sampling. The test sample was visually measured, and then the Feret particle size data set under visual measurement was obtained. Then, the Bayes' rule was applied to calculate the test sample, and the posterior probability P(c i |x i ) was obtained. P(c i |x i ) are different c i Similarly, the probability p value of each individual aggregate in i It can be understood as the distribution value of a single stone in each c i By accumulating the p-values ​​of the measured samples in the interval, the measured samples c with different aggregate specifications i We obtain the possible particle number of each c i Quantity m i The probability in this interval is considered to be binomial distribution B(M,θ), so that the set of binomial distributions {B1(M,θ1),B2(M,θ2),B3(M,θ3)…,B 15 (M,θ 15 )} was obtained.

[0033] Step 107: Based on the particle quantities of different actual particle size intervals and the set of binomial distributions, a set of posterior distributions is determined using the second Bayes' formula with the beta distribution as the prior distribution, and the posterior distributions in the set of posterior distributions correspond one-to-one to the actual particle size intervals. Step 108: The particle size distribution of the measured aggregate is determined according to the set of posterior distributions.

[0034] Step 108 is Step 1081 of determining one of the posterior distributions in the set of previous posterior distributions as a current posterior distribution; and step 1082 of determining the expected value of the current posterior distribution as the quantity ratio in the actual particle size interval corresponding to the measured aggregate. Here, in actual construction, estimating the particle size distribution of aggregate is judged according to the proportion of bone material mass in different intervals, and the relationship ratio between bone material mass and quantity is obtained through experiments, and the mass ratio of aggregate in different intervals can be obtained by calculation, and the accuracy of the present invention is judged by comparing with the results obtained by the manual sieving method using a sieve net.

[0035] To estimate the particle size distribution of aggregates, i The purpose of this study is to estimate the quantity ratio of the measured sample aggregate in the sample. This is called the estimation target parameter θ i According to step 3, each c i Obtain a set of binomial distributions of the measured samples in the interval and calculate the beta distribution f(θ i ) as the prior distribution, and apply Bayesian to obtain the posterior distribution f(θ i |m i ) was estimated, where f(θ i |m i ) ∝f(θ i )p(m i |θ i ), the probability density function of the prior distribution is TIFF2024169406000012.tif24170. The probability density function of the binomial distribution is The file is TIFF2024169406000013.tif24170.

[0036] therefore, The formula I got was TIFF2024169406000014.tif18170.

[0037] Since the prior distribution has already been determined, the parameters α and β are known, as are the total number of samples to be measured M and c i The quantity m belonging to i is known, so TIFF2024169406000015.tif17170 and TIFF2024169406000016.tif19170 is the parameter θ to be estimated i It was confirmed that this is not related to the above equation, and the above equation was combined and normalized to obtain the following equation. TIFF2024169406000017.tif17170

[0038] As can be judged from the above formula, the obtained posterior distribution is a beta distribution, whose parameter is α'=α+m i , β'=β+Mm i where the expectation of the beta distribution is expressed as α' / (α'+β'), and the expected value obtained is the parameter θ i is an estimate of, and for all c i By estimating the quantitative proportion of the measured samples, it was possible to estimate the particle size distribution of aggregates based on Bayesian statistics.

[0039] In the description of the present invention, as shown in Fig. 2, the maximum Feret particle size of the aggregate was obtained by using a vibration material feeder to uniformly drop the experimental aggregate, and using the multi-position images of each aggregate obtained by image collection to calibrate and convert the pixel size of the aggregate image. The present invention uses the maximum Feret particle size of the aggregate particles obtained by machine vision measurement as the collected data, and there is an error between it and the actual particle size of the aggregate particles. As shown in Fig. 3, aggregates with different Feret particle size intervals still overlap, so there is an error in the equivalent actual particle size interval to the maximum Feret particle size interval, and the error caused by this is reduced by Bayes' formula. Here, the aggregate particle size obtained by sieving using a sieve net is authoritative, and the present invention takes the aggregate particle size obtained by this method as the actual particle size.

[0040] As shown in Figure 4, the present invention takes the standard S9 specification coarse aggregate as an example, and the actual particle size intervals of the aggregate of this specification are c1 (19-26.5) mm, c2 (16-19) mm, c3 (13.2-16) mm, c4 (9.5-13.2) mm and c5 (4.75-9.5) mm.

[0041] (1) In order to make it easier to estimate the actual particle size intervals of the aggregates, the aggregates were sieved uniformly at 1 mm intervals, and 1,000 experimental aggregates were obtained for each interval. By solving the likelihood function, it was found that the aggregates were c i If i∈{1,2,3,4,5}, the Feret particle size measured by machine vision is x i The probability of belonging to (i∈{1,2,3,4,5}) was obtained and is shown in Table 1.

[0042] [Table 1]

[0043] (2) Prior probability P(c i In order to obtain the c value, the present invention estimates the relationship between bone material content and quantity through experiments. iObtain 1000 particles for each of the experimental aggregates, and then i Weigh the aggregates and calculate the corresponding different c i The present invention repeats the above steps three times, selects the average value for the obtained results, and then performs calculations to obtain the aggregate quantity ratio that complies with the aggregate specifications in JTG F40-2004, as shown in Table 2.

[0044] [Table 2]

[0045] (3) The quality inspector at the mixing station obtained the measured sample (the number of particles was about 3000, and the quantity was recorded as M) by sample sampling, and the measured sample was visually measured, and then a collection of Feret particle size data under visual measurement was obtained. Then, the measured sample was calculated by applying Bayes' rule. According to Bayes' rule, the posterior probability can be determined as the product of the prior probability and the likelihood divided by the sum of the product of the prior probability and the likelihood. TIFF2024169406000020.tif30170

[0046] Here, N is the number of actual particle size intervals of the aggregate, where N=5.

[0047] [Table 3]

[0048] As shown in Table 3, by applying Bayes' rule, the posterior probability P(c i |x i ) and obtain different c i Similarly, the probability value of each individual aggregate in c i It can be understood as the distribution value of a single stone in each c i By accumulating the probability values ​​of the measured samples in the interval, c iThe possible particle numbers of the measured sample were obtained. For the particle numbers of the measured sample, when M = 9, 90, 900, and 3000, the estimated and calculated c i The possible numbers of aggregate particles in were obtained and are shown in Table 4.

[0049] [Table 4]

[0050] (4) Estimation is performed to obtain the posterior distribution, and the expected value obtained is the estimated quantity proportion of aggregate particle size belonging to c1. As shown in Figures 5 to 8, the abscissa indicates the likelihood estimation, and the ordinate indicates the posterior distribution. It can be seen that as the aggregate particle quantity M increases, the obtained posterior distribution becomes more and more concentrated, and the certainty of the estimation target parameter θ1 increases. At the same time, the aggregate quantity proportion estimated to belong to c1 gradually approaches the actual quantity proportion. Here, the parameters α and β of the prior beta distribution are selected as 1, and the possible particle quantity of the measured sample belonging to the c1 interval is calculated to be 29.926. The formula obtained in step 4 Substitute TIFF2024169406000023.tif28170, Similarly, the other intervals were calculated and obtained as shown in Table 5, and the estimation of the aggregate particle size distribution was realized by calculating and obtaining the mass proportion of each interval using the relationship ratio between the bone material content and quantity obtained in the experiment. Here, the absolute error was basically kept within ±5%.

[0051] [Table 5]

[0052] From the above analysis, it can be seen that when the number of aggregates is small, the influence from prior distribution to posterior distribution is the largest, but as the number of aggregates increases, the influence from prior to posterior distribution gradually decreases, so it can be reasonably assumed that after the number of aggregates increases to a certain extent, the influence from different prior to posterior distribution is negligible. The present invention uses Bayesian statistics to estimate the particle size distribution of aggregates, which can effectively improve the screening efficiency. Compared with the traditional manual screening of aggregates by humans, this method, combined with machine vision measurement technology, realizes rapid screening, and at the same time, the entire detection process is not affected by human factors, effectively ensuring the quality of aggregate screening.

[0053] Based on the rapid development in the field of machine vision and the improved accuracy of image collection information, the present invention uses the Bayesian statistical estimation method to estimate the particle size distribution of aggregates, which has a higher automation level than the conventional methods, and can meet the requirement of quickly detecting whether the conveyed aggregate complies with the JTG F40-2004 standard at the mixing station. In addition, when the number of aggregates is large enough, the present invention can estimate the particle size distribution of unacceptable aggregates, so that it can be applied to aggregates of all specifications, and the accuracy of the particle size distribution of aggregates estimated by this method is high, which also provides a new method and perspective for future aggregate particle size screening.

[0054] Example 2 To carry out a method corresponding to the above Example 1 so as to achieve corresponding functions and technical effects, a system for determining particle size distribution of aggregate is provided below. The system for determining particle size distribution of aggregate in the second embodiment includes an aggregate particle set construction module for extracting a sample from the aggregate to be measured and constructing an aggregate particle set including a plurality of aggregate particle subsets; the actual particle sizes of all aggregate particles in the same aggregate particle subset all belong to the same actual particle size interval, the aggregate particle subsets correspond one-to-one with the actual particle size intervals, and the common sets of any two actual particle size intervals are all empty sets; A likelihood function determination module for determining a likelihood function of the measured aggregate based on the aggregate particle set using machine vision detection technology, the likelihood function being used to describe the probability distribution relationship of the aggregate particles in different actual particle size intervals and different ferret particle size intervals; a prior probability determination module for obtaining a prior probability of the aggregate to be measured, the prior probability being determined by the mass of a single aggregate particle in different actual particle size intervals; a posterior probability determination module for determining a plurality of posterior probabilities using the first Bayes formula according to the likelihood function and the prior probability; i |x i ) is the particle size interval x of the ith ferret i The i-th actual particle size interval c i Denote the probability of each aggregate in an actual particle size interval particle count determination module for determining that the sum of prior probabilities of all aggregate particles in the subset of aggregate particles is a particle count for the corresponding actual particle size interval; A binomial distribution set construction module for constructing a binomial distribution set according to the particle quantity of different actual particle size intervals, and the binomial distribution in the binomial distribution set corresponds one-to-one with the actual particle size intervals; and a posterior distribution set determination module for determining a set of posterior distributions based on a set of particle quantities and binomial distributions of different actual particle size intervals, using a beta distribution as a prior distribution and the second Bayes' formula, wherein the posterior distributions in the set of posterior distributions correspond one-to-one to the actual particle size intervals.

[0055] The particle size distribution of the measured aggregate was determined according to the set of posterior distributions.

[0056] Example 3 This embodiment provides an electronic device, comprising a memory and a processor, the memory being used to store a computer program, and the processor being configured to execute the computer program, such that the electronic device performs the method for determining particle size distribution of aggregates described in Example 1.

[0057] Here, the memory is a readable storage medium.

[0058] Each embodiment in this specification is described step by step, and each embodiment mainly describes the difference between the other embodiments, and the same or similar parts between the embodiments can be referred to each other. The system disclosed in the embodiment corresponds to the method disclosed in the embodiment, so the explanation is relatively simple, and the relevant information can be referred to the explanation of the method part.

[0059] In this specification, the principle and embodiment of the present invention are described by applying specific examples, and the description of the above examples is only useful for understanding the method of the present invention and its core idea, and at the same time, those skilled in the art can change the form for implementing the invention and the scope of application according to the idea of ​​the present invention. Therefore, the contents of this specification should not be understood as limitations on the present invention.

Claims

1. 1. A method for determining particle size distribution of aggregate, comprising the steps of: Extracting a sample from the aggregate to be measured to construct a set of aggregate particles including a plurality of aggregate particle subsets, the actual particle sizes of all aggregate particles in the same aggregate particle subset all belong to the same actual particle size interval, the aggregate particle subsets correspond one-to-one to the actual particle size intervals, and the intersection of any two of the actual particle size intervals is an empty set; determining a likelihood function for the measured aggregate based on the aggregate particle set using machine vision detection technology, the likelihood function being used to describe a probability distribution relationship of aggregate particles in different actual particle size intervals and different ferret particle size intervals; Obtaining a prior probability of the aggregate to be measured, the prior probability being determined by the mass of a single aggregate particle in different actual particle size intervals; determining a plurality of posterior probabilities using a first Bayes' rule according to the likelihood function and the prior probability; and i |x i ) is the particle size interval x of the i-th ferret i The i-th actual particle size interval c of each aggregate in i Denote the probability at converting the prior probabilities of all aggregate particles in the aggregate particle subset from sums of mass ratios to particle quantity ratios in the corresponding actual particle size intervals; constructing a set of binomial distributions according to particle quantity ratios of different actual particle size intervals, and the binomial distributions in the set of binomial distributions correspond one-to-one to the actual particle size intervals; Based on a set of particle quantities and binomial distributions of different actual particle size intervals, a set of posterior distributions is determined using a beta distribution as a prior distribution and a second Bayes' formula, and the posterior distributions in the set of posterior distributions correspond one-to-one to the actual particle size intervals; and determining the particle size distribution of the aggregate to be measured in accordance with the set of posterior distributions.

2. determining a likelihood function for the measured aggregate based on the set of aggregate particles and utilizing machine vision detection techniques; determining a ferret particle size of every particle in each of said aggregate particle populations utilizing machine vision detection techniques; constructing a plurality of ferret particle size intervals, and the intersection of any two of said ferret particle size intervals is an empty set; determining a plurality of likelihood function values ​​P(x i |c i ) is the i-th aggregate subset c i The ferret particle size of the aggregate in the i-th ferret particle size interval x i The method for determining the particle size distribution of aggregates according to claim 1, characterized in that the method indicates the probability that the particle size distribution of aggregates belongs to the above range.

3. Determining a particle size distribution of the aggregate to be measured according to the set of posterior distributions includes: determining one of the posterior distributions in the set of posterior distributions as a current posterior distribution; 2. A method for determining the particle size distribution of aggregates as described in claim 1, further comprising: determining an expected value of the current posterior distribution as a quantity proportion in the actual particle size interval corresponding to the measured aggregates.

4. The first Bayes formula is and where N is the number of actual particle size intervals, and P(c i ) is the i-th aggregate subset c i 3. The method for determining particle size distribution of aggregate according to claim 2, characterized in that the prior probability is .

5. 1. A system for determining particle size distribution of aggregates, comprising: an aggregate particle set construction module for extracting a sample from the aggregate to be measured and constructing a set of aggregate particles, the set of aggregate particles including a plurality of aggregate particle subsets, where the actual particle sizes of all aggregate particles in the same aggregate particle subset all belong to the same actual particle size interval, the aggregate particle subsets correspond one-to-one to the actual particle size intervals, and the intersection of any two of the actual particle size intervals is an empty set; A likelihood function determination module for determining a likelihood function of the measured aggregate based on the aggregate particle set using a machine vision detection technique, the likelihood function being used to describe the probability distribution relationship of the aggregate particles in different actual particle size intervals and different ferret particle size intervals; a prior probability determination module for obtaining a prior probability of the aggregate to be measured, the prior probability being determined by the mass of a single aggregate particle in different actual particle size intervals; a posterior probability determination module for determining a plurality of posterior probabilities using a first Bayes' formula according to the likelihood function and the prior probability; i |x i ) is the particle size interval x of the i-th ferret i The i-th actual particle size interval c of each aggregate in i Denote the probability at a particle quantity determination module for determining the particle quantity of an actual particle size interval for converting the prior probability of all aggregate particles in the aggregate particle subset from a sum of mass ratios to a particle quantity ratio of a corresponding actual particle size interval; a binomial distribution set construction module for constructing a binomial distribution set according to particle quantity ratios of different actual particle size intervals, and the binomial distributions in the binomial distribution set correspond one-to-one to the actual particle size intervals; A posterior distribution set determination module for determining a posterior distribution set by using a second Bayes' formula with a beta distribution as a prior distribution based on a set of particle quantities and binomial distributions of different actual particle size intervals, wherein the posterior distributions in the set of posterior distributions correspond one-to-one to the actual particle size intervals; A system for determining the particle size distribution of aggregates, characterized in that the particle size distribution of the aggregates to be measured is determined according to the set of posterior distributions.

6. 5. An electronic device comprising a memory and a processor, the memory being used to store a computer program, and the processor being adapted to execute the computer program so that the electronic device can perform the method for determining the particle size distribution of aggregates according to any one of claims 1 to 4.

7. 7. The electronic device according to claim 6, wherein the memory is a readable storage medium.

Citation Information

Patent Citations

  • Quick test method for aggregate grade of asphalt concrete pavement

    CN102012356A

  • Coarse-grained soil gradation determination method based on image technology

    CN111504862A

  • Method and system for measuring freshly mixed concrete aggregate particle size

    JP2008268051A

  • System and method for measuring particle size distribution of granular material

    JP2010066127A

  • Asphalt mixture component information estimation system and component information estimation method

    JP2021116649A