Building solid waste road performance index and multi-path utilization scene intelligent evaluation method

Abstract

The invention discloses a building solid waste road performance index and multi-path utilization scene intelligent evaluation method. The method comprises the steps that the mass content of red bricks, the mass content of needle-sheet-shaped particles and the particle gradation in building solid waste particles are obtained through a machine vision system; the pavement performance indexes, namely CBR value, water absorption rate and crushing value, of the building solid waste are obtained through tests; according to the influence rules of the mass content of the red bricks, the mass content of the needle-sheet-shaped particles and the grain composition in the building solid waste particles on the building solid waste road performance indexes, an estimation model of the building solid waste road performance indexes is established. According to the method, the mass content of the red bricks, the mass content of the needle-sheet-shaped particles and the grain composition in the building solid waste particles are obtained in real time on the basis of machine vision, the road performance indexes of the building solid waste are predicted through the prediction model of the road performance indexes of the building solid waste, and the detection efficiency and accuracy are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of road engineering and relates to an intelligent evaluation method for the road performance indexes of construction waste and multiple utilization scenarios. Background Technique

[0002] Since this century, China has experienced a wave of large-scale urbanization construction and transportation infrastructure construction. A large amount of construction solid waste (referred to as construction waste) has been left during the process of old city renovation, land expropriation and demolition, and road renovation. According to relevant data statistics, the annual production of construction waste in China currently reaches as high as 1.8 billion tons, and the resource utilization rate is less than 5%. Except for a small amount used for engineering backfilling and recycling, most of it is still simply stacked for treatment. Most construction waste is still treated by simple stacking or landfill, which not only occupies a large amount of land resources, but also may pose potential threats to the environment and human health.

[0003] To promote the transformation of China's transportation infrastructure construction towards resource conservation, environmental friendliness and green sustainable development, it is extremely crucial to improve the resource utilization rate of construction waste. Using it for road engineering construction is one of the important ways of resource utilization of construction waste. Among them, road engineering construction has a huge demand for building materials. In this process, the reuse of construction waste can not only effectively relieve the pressure of natural resource exploitation, reduce the dependence on traditional materials, but also significantly reduce the environmental pollution caused by construction waste, and achieve the goals of circular economy and sustainable development. However, due to the very complex composition of construction waste, how to quickly evaluate the quality of its road performance (such as CBR value, water absorption rate and crushing value) has become an important technical problem in the popularization and application of construction waste in subgrade engineering.

[0004] Machine vision, as an important branch of artificial intelligence, shows great potential and advantages in evaluating the road performance of construction waste. Through the application of technologies such as image recognition and processing, automatic detection and measurement, and intelligent analysis and prediction, the rapid and accurate evaluation of the road performance of construction waste can be realized. Machine vision technology uses an unmanned aerial vehicle to obtain the image signal of the object to be photographed and transmits it to a special image processing system, and the morphological information such as the brightness, color and texture of the object to be photographed can be obtained. Then, through various preset operations, the efficient detection of the characteristic attributes of the object to be photographed can be realized, and it has the characteristics of fast speed, large amount of information and multiple functions. Therefore, machine vision technology can be used to quickly predict the road performance of construction waste and provide guidance for the resource utilization of construction waste in highway construction. Summary of the Invention

[0005] To solve the above problems, the present invention provides a method for evaluating the road performance index of construction waste, which can obtain the mass content of red bricks, the mass content of needle-like and flaky particles, and the particle size distribution in construction waste particles in real time based on machine vision, and predict the road performance index of construction waste through a prediction model of the road performance index of construction waste, improving the detection efficiency and accuracy.

[0006] Another object of the present invention is to provide an intelligent evaluation method for multi-channel utilization scenarios of construction waste.

[0007] The technical solution adopted by the present invention is an intelligent evaluation method for the road performance index of construction waste, including the following steps:

[0008] S1. Obtain the mass content of red bricks, the mass content of needle-like and flaky particles, and the particle size distribution in construction waste particles through a machine vision system;

[0009] S2. Obtain the road performance indexes of the construction waste in S1 through tests, namely the CBR value, water absorption rate, and crushing value;

[0010] S3. Establish a prediction model for the road performance index of construction waste according to the influence laws of the mass content of red bricks, the mass content of needle-like and flaky particles, and the particle size distribution in construction waste particles on the road performance index of construction waste.

[0011] Further, the S1 includes the following steps:

[0012] S11. Obtain the image information of the original construction waste: Use a drone to collect construction waste images under simulated scattered and dense working conditions indoors; randomly adjust the plane position and three-dimensional posture of the construction solid waste under the two indoor working conditions, and arrange the construction waste in different stacking methods to simulate the complex situations that occur in the actual environment; then shoot from multiple regions, multiple angles, and multiple lighting conditions to construct an original data set of construction waste materials;

[0013] S12: Image processing: Convert the simulated images collected by the drone into digital images, preprocess the construction waste images to improve the image quality, reduce noise, and increase the image quantity and diversity;

[0014] S13. Perform image segmentation through a deep learning model, construct an FMD data set based on the image segmentation results, including material category samples of red brick particles, cement mortar particles, and gravel particles; divide the data set into a training set, a validation set, and a test set according to a ratio, and obtain the mass ratios of red brick particles, cement mortar particles, and gravel particles in the construction waste particles through a material composition recognition algorithm; obtain the mass content of needle-like and flaky particles and the particle size distribution in the construction waste particles through a particle geometric feature extraction algorithm.

[0015] Furthermore, the construction solid waste image preprocessing in S12 includes the following steps:

[0016] Improve image quality and reduce noise by graying, filtering, enhancing and edge detection;

[0017] Further data enhancement technology is used to perform optical transformation, geometric transformation, noise transformation and filtering transformation on the image to enhance the contrast of the picture, reduce the adverse effects of natural light on the image, improve the conditions for capturing and collecting images of construction solid waste, expand and increase the number and types of images in the original construction solid waste data set, and improve the database.

[0018] Furthermore, in S13, the material component identification algorithm is: the FMD material data set constructed after image segmentation is input into a support vector machine or a neural network model, and the model is trained and learned, and the extracted features are compared with the sample features in the FMD data set; for the support vector machine, by finding the optimal classification hyperplane, the input features are mapped to the corresponding material component categories, and the red brick particles, cement mortar particles and gravel particles in the construction solid waste particles and their mass proportions are output; for the neural network model, the data features are learned through complex connections and nonlinear transformations between a large number of neurons, and then the red brick particles, cement mortar particles and gravel particles in the construction solid waste particles and their mass proportions are output, thereby realizing the mapping from image features to material components and contents.

[0019] Furthermore, in S13, the particle geometric feature extraction algorithm is: based on the image segmentation result, the particle size of the solid waste particles is measured, the mass content of the needle-like particles and the particle roundness are calculated, and the number of particles in different particle size ranges is counted, so as to obtain the particle grading.

[0020] Furthermore, S2 is specifically: by conducting CBR tests, water absorption tests and crushing tests, the CBR value, water absorption rate and crushing value of construction solid waste materials under various red brick, cement mortar, crushed stone particle dosages, as well as non-uniformity coefficients and particle roundness conditions are determined.

[0021] Furthermore, the estimation model of the road performance index of construction solid waste in S3 is:

[0022]

[0023] w=0.05(1+1.3θ)(1+Q e&f )

[0024]

[0025] Among them, δ a represents the crushing value; w represents the water absorption rate; CBR represents the California bearing ratio; θ represents the mass percentage of waste red bricks; Qe&f represents the mass percentage of flaky and needle-shaped particles; d 10 represents the particle size corresponding to 10% of the construction waste particles; d 60 represents the particle size corresponding to 60% of the construction waste particles; θ, Q e&f , d 10 and d 60 are obtained through the machine vision system;

[0026] Obtain θ, Q of the construction waste to be measured through the machine vision system e&f , d 10 and d 60 , input into the prediction model, and the predicted value of the road performance index of the construction waste can be obtained.

[0027] An intelligent evaluation method for multi-channel utilization scenarios of construction waste includes the following steps:

[0028] Step 1, determine the performance standards for using construction waste as road pavement structure and subgrade filler;

[0029] Step 2, use the water absorption rate, crushing value, mass content of flaky and needle-shaped particles, and CBR value obtained by the intelligent evaluation method for road performance indexes of construction waste described in claim 1 as evaluation indexes, denoted as x1, x2, x3,..., x n ; The corresponding weights of the evaluation indexes x1, x2, x3,..., x n are w1, w2, w3,..., w n , calculate the comprehensive evaluation coefficient S: S = x1w1 + x2w2 + x3w3 +... + x n w n ;

[0030] Determine the road utilization scenario of the construction waste according to the comparison table of the S value range and the road application scenario of the construction waste.

[0031] Furthermore, the acquisition method of the comparison table of the S value range and the road application scenario of the construction waste:

[0032] On the basis of the construction waste classification standard, further refine the classification of different grades of construction waste to determine the applicable standards for the construction of road surface courses, base courses, and sub-base courses;

[0033] Prepare a large number of construction waste samples with different component ratios and different physical property parameters, conduct tests on various performance indexes according to the specifications, and obtain the data of the water absorption rate, crushing value, mass content of flaky and needle-shaped particles, and CBR value of each sample; According to the calculation formula of the comprehensive evaluation coefficient S of the performance indexes S = x1w1 + x2w2 + x3w3 +... + x n w nPerform calculations and conduct statistical analysis on the large number of S values obtained from the calculations. Combine the actual requirements of different structural layers of road engineering for the performance of construction waste, divide the S values into different intervals, and thus obtain the corresponding relationship between the value range of medium S and the utilization scenarios of construction waste.

[0034] The beneficial effects of the present invention are as follows:

[0035] 1. The present invention proposes a method for identifying the physical properties parameters of construction waste based on machine vision, which helps to obtain in real time the physical properties parameters such as the material, color, content of each substance, and particle gradation of construction waste, and can significantly improve the detection efficiency and accuracy. Compared with the traditional receiving material detection method, the present invention can detect a large number of samples in real time, and the error can be controlled within a very small range, greatly solving the problems of cumbersome traditional manual detection and sorting processes and easy occurrence of judgment errors, saving a large amount of time cost and operation cost, and strongly supporting the resource utilization of construction waste. For example, in a large construction waste treatment site, drones can quickly traverse the whole site and collect thousands of images, while manual detection takes a lot of time to check one by one, and the efficiency gap is obvious.

[0036] 2. The present invention establishes a prediction model for the road use performance index of construction waste particles based on a large amount of test data. By carrying out indoor tests on engineering performance indexes, analyze the influence law of physical properties parameters such as the content of each substance component and particle gradation in construction waste particles on their road use performance indexes, and use MATLAB to perform fitting analysis and verification on the test data results. The fitting effect of this model is good and the accuracy is high, shortening the detection cycle from several days or even several weeks to several hours, greatly improving the project progress, and effectively solving the problems of long curing time, complex operation procedures, and large influence of human factors on data accuracy in engineering performance detection experiments.

[0037] 3. The present invention also formulates a classification method and standard for construction waste from different sources. Combining the physical properties parameter identification algorithm of construction waste and the prediction model for the road use performance index of construction waste particles, defines the comprehensive evaluation coefficient of engineering performance, and establishes an integrated prediction and evaluation system for the road use performance of construction waste particles relying on drones. This system can quickly and accurately propose suggestions on the applicable scenarios of construction waste by taking actual pictures of construction waste, greatly simplifying the cumbersome procedures of construction waste recycling, providing technical support for quickly judging the multi-channel application of construction waste in road engineering, significantly improving the utilization efficiency of construction waste resources, and realizing the high-value utilization of construction waste.

[0038] 4. Precise grading and classification methods and standards can effectively screen out construction waste of different qualities and clarify their applicable scenarios. For example, Grade I construction waste can replace natural aggregates for road surface structure construction, and Grade II construction waste can also be reasonably used for subgrade filling after evaluation, improving the resource utilization rate of construction waste, reducing the dependence on natural resources, and realizing the efficient recycling of resources. Reduce environmental pollution: Reduce the simple stacking and landfill of construction waste, and reduce the potential pollution risks to soil, water sources, and air. At the same time, reducing the extraction of natural aggregates helps protect the ecological environment, which is in line with the concept of green and sustainable development. It is estimated that by applying the technology of the present invention, a large amount of construction waste landfill can be reduced annually, and the ecological damage area caused by the extraction of natural aggregates can be reduced.

[0039] 5. Based on accurate performance evaluation and grading results, selecting suitable construction waste for corresponding parts of road engineering can ensure the stability and durability of the road structure. For example, using recycled construction waste with good performance in the road base can effectively improve the bearing capacity of the road surface, reduce the occurrence of road surface diseases, and extend the service life of the road. Save engineering costs: On the one hand, the resource utilization of construction waste reduces the procurement cost of traditional building materials; on the other hand, the efficient evaluation system reduces the detection time and labor costs, and at the same time avoids the project rework costs caused by using unqualified materials. Through comprehensive calculation, the total cost of road engineering construction can be significantly reduced, and the economic benefits can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0041] Figure 1 It is the overall working flow chart of the machine vision system in the embodiment of the present invention.

[0042] Figure 2 It is the overall flow chart of construction waste identification, estimation, and output in the embodiment of the present invention.

[0043] Figure 3a It is the original image of construction waste under scattered working conditions in the embodiment of the present invention.

[0044] Figure 3b is Figure 3a the grayscale image of

[0045] Figure 3c is Figure 3a the binary image of

[0046] Figure 3d is Figure 3a the segmented image of

[0047] Figure 4a This is the original image of construction waste under intensive working conditions in the embodiment of the present invention.

[0048] Figure 4b is Figure 4a the grayscale image of

[0049] Figure 4c is Figure 4a the binary image of

[0050] Figure 4d is Figure 4a the segmented image of

[0051] Figure 5 This is the curve relationship diagram of the crushing value, different red brick admixture ratios, and particle non-uniformity coefficient when the mass content of acicular and flaky particles is 10% in the embodiment of the present invention.

[0052] Figure 6 This is the curve relationship diagram of the CBR value, different red brick admixture ratios, and particle non-uniformity coefficient when the mass content of acicular and flaky particles is 10% in the embodiment of the present invention.

[0053] Figure 7 This is the curve relationship diagram of the water absorption rate, different red brick admixture ratios, and the mass content of acicular and flaky particles in the embodiment of the present invention. Specific Embodiments

[0054] Next, in combination with the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present invention.

[0055] In the embodiment of the present invention, a drone is used as a carrier, and physical property parameters such as color, material, content of various substances, and particle gradation in the image of construction waste recycled materials are quickly obtained based on the principle of machine vision. On the basis of a large number of tests, the influence laws of relevant physical property parameters on performance indicators such as CBR value, water absorption rate, and crushing value are analyzed, a prediction model for the performance indicators of construction waste particles is constructed to achieve rapid detection of performance indicators; subsequently, a classification standard for the performance of construction waste is formulated with reference to relevant design specifications, and suggestions for the applicable scenarios of construction waste are quickly and accurately proposed in combination with the predicted values of this model. This greatly simplifies the cumbersome procedures of the recycling and utilization of construction waste and provides technical support for the real-time judgment of the multi-channel application of construction waste in road engineering.

[0056] Embodiment 1

[0057] An intelligent evaluation method for the road performance index of construction solid waste, as Figures 1-2 shown, includes the following steps:

[0058] S1. Obtain the content of red bricks, the content of needle-like and flaky particles, and the particle gradation in the construction solid waste particles through a machine vision system.

[0059] S11. Obtain the image information of the original construction solid waste materials;

[0060] In an indoor three-dimensional solid waste yard, use equipment such as drones to collect images of construction solid waste materials under sufficient light and appropriate angles. The purpose is to obtain clear and comprehensive images of construction solid waste for subsequent processing and analysis, and finally construct an original dataset of construction solid waste images. The traditional collection method is difficult to simulate the complex distribution of construction solid waste in the actual environment, and the collected image data is single, which is not conducive to the subsequent model learning comprehensive construction solid waste characteristics.

[0061] After obtaining a large amount of construction solid waste materials from a demolition site, process the construction solid waste through crushing and other processes using a jaw crusher, and sieve it with a sieve with a sieve hole of 50 mm to obtain the original construction solid waste particle raw materials; and use drones to collect a large number of images of common construction solid waste particles such as red brick particles, cement mortar particles, and gravel particles through digital image photography technology. To conform to the actual situation, two indoor three-dimensional solid waste yards in scattered conditions and dense conditions need to be constructed respectively. Among them, in the indoor three-dimensional solid waste yard in scattered conditions, ensure that the quantity of each solid waste is 1 - 3 pieces to simulate the scattered distribution of construction solid waste in the actual environment; in the indoor three-dimensional solid waste yard in dense conditions, ensure that the quantity of each solid waste is ≥6 pieces to simulate the scene of a large amount of construction solid waste accumulation and dense distribution in the actual environment.

[0062] Before shooting, randomly adjust the planar position and three-dimensional posture of construction solid waste under two indoor working conditions, which helps to increase the diversity and representativeness of the dataset, enabling the model to learn more varying situations; and use different stacking methods (such as laying flat, stacking, staggering, etc.) to arrange the construction solid waste to simulate various complex situations that may occur in the actual environment. Use a drone or other shooting equipment to take pictures at an appropriate height and angle, ensuring sufficient light to reduce the impact of shadows and reflections on the image quality; take multiple pictures of the construction solid waste, shooting from multiple regions, multiple angles, and multiple lighting conditions, and then construct a rich and diverse original dataset of construction solid waste images that is as close to the actual situation as possible, covering the image information of construction solid waste in different distribution states. Organize, classify, and annotate the collected images. According to actual needs, the waste in the images can be annotated (such as type, quantity, location, etc.) for subsequent processing and analysis, and an original dataset of construction solid waste materials is constructed. To ensure that the images in the construction solid waste image dataset can store sufficient and effective construction solid waste feature information, save the images at a resolution of 1920×1080.

[0063] S12, Image processing;

[0064] In the construction solid waste detection scenario, the construction solid waste images collected by the drone are initially in analog signals and cannot be directly processed and analyzed by a computer. After collecting the construction solid waste images, it is necessary to first convert the analog images into digital images, that is, image digitization. Preprocess the construction solid waste images. The preprocessing steps include grayscale conversion, filtering, enhancement, edge detection, etc. to improve the image quality and reduce the influence of factors such as noise, as Figures 3a-3d and Figures 4a-4d shown.

[0065] Grayscale conversion is the process of converting a color image into a grayscale image, that is, removing the color information of the image and only retaining the luminance information. In the construction solid waste image, the grayscale image can simplify the image information, reduce the calculation amount, and at the same time retain sufficient structural information for subsequent processing; Filtering is used to remove noise in the image or enhance certain features of the image. In the processing of construction solid waste images, filtering operations can help smooth the image, remove small particle noise, or sharpen the edges to make the image clearer. Commonly used filtering methods include mean filtering, median filtering, Gaussian filtering, etc.; Image enhancement aims to improve the visual effect of the image to make it more suitable for subsequent analysis and processing. In the construction solid waste image, enhancement may include contrast enhancement, brightness adjustment, etc. to highlight the useful information in the image, such as texture, shape, etc.; Edge detection is an important technology in image processing used to identify points with significant luminance changes in the image, and these points correspond to the boundaries of construction solid waste objects. In the construction solid waste image, edge detection helps to extract the shape features of the solid waste, providing an important basis for subsequent classification and recognition.

[0066] Further utilize data augmentation techniques to perform optical transformation, geometric transformation, noise transformation, filtering transformation, etc. on the images, so as to enhance the contrast of the pictures, reduce the adverse effects of natural light on the images, improve the shooting and acquisition conditions of construction waste images, expand and augment the number and types of images in the original construction waste dataset, and improve the database.

[0067] Optical transformation mainly refers to adjusting parameters such as the brightness, chromaticity, sharpness, and saturation of the images to enrich the images of construction waste under different lighting conditions and different sample states;

[0068] Geometric transformation is to transform the geometric features of the images to achieve the purpose of data augmentation. For example, perform image processing operations such as image rotation, mirror transposition, and scale scaling to enrich the spatial geometric features such as the angles and orientations of construction waste in the images;

[0069] Noise transformation is to perform noise reduction processing on the isolated pixels that are unnecessary and can cause visual interference in the images by adding Gaussian noise, which can increase the robustness of the model when classifying and recognizing noisy images;

[0070] Filtering transformation is to perform different forms of filtering transformation processing on the construction waste images to obtain filtered enhanced images with various different effects;

[0071] By using data augmentation techniques to increase the order of magnitude and diversity of training samples, the construction waste dataset can be made more in line with the actual scenarios, which can effectively enhance the classification robustness of the model for construction waste in different spatial positions and light source environments, thereby reducing the sensitivity of the model to certain special image attributes and improving its generalization ability. Preprocessing the construction waste images aims to improve the image quality, highlight key information, and reduce the complexity of subsequent image analysis through steps such as grayscale conversion, noise suppression, geometric correction, image enhancement, and image segmentation, so as to improve the processing efficiency and better meet various application scenarios; in practical applications, appropriate preprocessing methods should be selected according to specific requirements and scenarios to obtain better processing effects.

[0072] S13, obtain the physical property parameters of construction waste, that is, the content of flaky particles and particle gradation in the construction waste particles.

[0073] Based on the principle of machine vision, optimize the depth image through a depth map correction network to remove the noise in the original depth image. Use a priori network to perform convolution operations on the depth image and the visible light image, map the RGB / D image pair to a low-dimensional feature space, and extract the random feature S in the image e . Extract the salient features S of the optimized depth image and the visible light image through a saliency network dMultiple samplings are performed on the random features and combined with the significant features to obtain multiple image segmentation prediction results. Through the significant consensus network, the pixel values at each position are given according to each prediction result to obtain the final image segmentation result. Through data augmentation techniques and multi-network collaborative processing, the image data can be effectively expanded, the key features can be highlighted, and the model can stably and accurately identify construction waste in different scenarios.

[0074] Based on the above image segmentation results, an FMD material dataset is constructed, including material category samples of red brick particles, cement mortar particles, and gravel particles. There are 1000 samples for each type of material. In the experiment, the dataset is divided into a training set, a validation set, and a test set according to the ratio of 7:1:2. Each group of experiments is repeated 3 times, and the average value of the 3 experimental results is taken as the final experimental result. Hyperparameters such as the learning rate, momentum, and weight decay in the training process are set to 0.01, 0.9, and 0.001 respectively, and each training cycle has 20 epochs. Finally, a method based on the principle of machine vision is formed, which can accurately and quickly identify the physical property parameters such as the material composition, color, particle shape, and particle gradation of construction waste particles through image acquisition, image segmentation, identification of substance composition components, and extraction of particle geometric features.

[0075] Substance composition recognition algorithm: Based on the image segmentation results, a support vector machine or a neural network model is used for feature extraction. The FMD material dataset constructed after image segmentation is input into the support vector machine or the neural network model. After training and learning, the model will compare the extracted features with the sample features in the FMD dataset. For the support vector machine, by finding the optimal classification hyperplane, the input features are mapped to the corresponding substance composition categories, and the mass ratios of red brick particles, cement mortar particles, and gravel particles in the construction waste particles are output; for the neural network model, through the complex connections and non-linear transformations between a large number of neurons, the data features are learned, and then the substance composition and its content information are output to realize the mapping from image features to substance composition and content.

[0076] Particle geometric feature extraction algorithm: Based on the image segmentation results, the pixels of the segmented construction waste particle images are analyzed. Using the mathematical morphology method, first measure the particle size of the construction waste particles. Specifically, according to the conversion relationship between image pixels and actual dimensions, combined with the image resolution, calculate the equivalent diameter of the particles (such as using area conversion for circular particles and the Feret diameter for irregular particles). Then calculate the mass content of needle-like and flaky particles. By calculating shape parameters such as the rectangularity (the ratio of the particle area to the area of the minimum circumscribed rectangle) and the elongation (the ratio of the major axis length to the minor axis length of the particle) of the particles, set appropriate thresholds, and count the proportion of the number of particles that meet the needle-like and flaky characteristics in the total number of particles to obtain the mass content of needle-like and flaky particles. At the same time, calculate the roundness of the particles, using a formula similar to the one for calculating roundness (such as C = 4πA / P 2), where A is the particle area, P is the particle perimeter, the area and perimeter are obtained through morphological operations, and C represents the particle roundness). Finally, count the number of particles in different particle size ranges, divide them according to a certain particle size interval, such as (0 - 2.36 mm, 2.36 - 4.75 mm), etc., to obtain the particle gradation information and achieve the output from image pixel information to particle geometric feature parameters.

[0077] S2. Obtain the road - using performance index parameters of construction waste, that is, obtain the road - using performance indexes of the construction waste in S1 through tests, namely the CBR value, water absorption rate, and crushing value.

[0078] Through indoor tests such as CBR tests, water absorption tests, and crushing tests, measure the CBR value, water absorption rate, and crushing value of construction waste materials under various conditions of red brick, cement mortar, crushed stone particle content, coefficient of uniformity, and particle roundness.

[0079] In the embodiment, the construction waste includes red bricks, cement mortar, and crushed stones. The mass ratio of the particle content of red bricks, crushed stones, and cement mortar is 1:2:2, that is, the mass ratio of the red brick content is 20%. The construction waste particles should meet the road construction standards to ensure the stability and durability of the road.

[0080] In the embodiment, the test schemes and main steps of the CBR value, water absorption rate, and crushing value are as follows:

[0081] S2 - 1. CBR test: Conduct the CBR test according to the "Code for Highway Geotechnical Tests" (JTG3430—2020).

[0082] S2 - 1 - 1. Take the air - dried construction waste test sample (dried in an oven at 65°C) for crushing treatment. After crushing, sieve the particles smaller than 40 mm with a sieve for testing. Prepare the specimen according to the optimal moisture content of the construction waste, mix the test sample thoroughly with water, and then put it into a plastic bag and soak it for about 2 h.

[0083] S2 - 1 - 2. Use the static pressure forming method to form the required specimen amount at one time according to the compaction degrees of the upper roadbed, lower roadbed, upper embankment, and lower embankment, and soak the specimen in water for 4 days.

[0084] S2 - 1 - 3. Then place the specimen on the lifting platform of the 14 / 21 penetrometer, make the top of the specimen contact the penetrometer, place 8 load plates on the top surface of the specimen, install and record the initial reading of the dial gauge. Apply a load to make the penetrometer rod press into the specimen at a speed of 1 to 1.25 mm / min, record the readings, and ensure that there are more than 5 readings when the penetration amount is 2.5 mm, and calculate the CBR value of the construction waste.

[0085] The requirements for the compaction degrees of the upper roadbed, lower roadbed, upper embankment, and lower embankment are shown in Table 1 respectively:

[0086] Table 1 Different Compaction Requirements for Subgrade

[0087]

[0088] S2-2, Water Absorption Test: Take 4L of construction waste samples, sieve them through a sieve with a sieve hole of 2.36mm, and take the residue for drying and standby. Mix the samples evenly, divide them into three equal parts, weigh them respectively, and then put them into a container filled with water. If there are particles floating on the water surface, press them into the water. After soaking the samples for 24h, make the samples into saturated surface dry, and then weigh them. Calculate the water absorption rate, and take the arithmetic mean of the three measured values as the test result.

[0089] S2-3, Crushing Test: Take samples according to the regulations, air-dry or oven-dry them, and then sieve out the particles larger than 19.0mm and smaller than 9.50mm. Divide them into 3 equal parts for standby, each part is about 3000g. Take one part of the sample and put the sample into a circular mold (placed on the chassis) in two layers. After each layer of the sample is filled, place a pad rod under the chassis. Hold the cylinder and alternately strike the ground 25 times on the left and right. After compaction of the two layers, level the surface of the sample in the mold and cover it with a pressure head. When the circular mold cannot hold 3000g of the sample, fill it to 10mm from the upper mouth of the circular mold. Place the circular mold with the sample on the pressure testing machine, start the pressure testing machine, apply a uniform load at a speed of 1kN / s until 200kN and keep the load stable for 5s, and then unload the load. Remove the pressure head, pour out the sample, and weigh its mass; sieve out the crushed fine particles with a sieve with a pore size of 2.36mm, weigh the mass of the sample remaining on the sieve, and calculate the crushing value.

[0090] The CBR test, water absorption test and crushing test are respectively used to reflect the ability of the filler to resist local deformation, the pore structure and permeability of the concrete surface, and the ability of the construction waste material to resist crushing under gradually increasing loads in road construction, and jointly provide important technical support for the quality and performance evaluation of road engineering.

[0091] S3. According to the influence laws of the red brick content, flaky particle content and particle gradation in the construction waste particles on the road use performance indexes of the construction waste respectively, establish a prediction model for the road use performance indexes of the construction waste.

[0092] After obtaining the physical property parameters and performance index parameters of construction waste, analyze the influence laws of the content of each substance component and particle gradation in construction waste particles and other physical property parameters on its performance indexes such as CBR value, water absorption rate, and crushing value. Use MATLAB to perform fitting analysis on its data results, establish a prediction model for the road performance indexes of construction waste and verify it; thus establish the connection between the physical property parameters of construction waste and the road performance indexes, and can quickly predict the road performance of construction waste, solving problems such as long curing time, complex operation procedures, and large influence of human factors on data accuracy in traditional engineering performance detection experiments.

[0093] The prediction model of the road performance index of construction waste is as follows:

[0094]

[0095] w = 0.05(1 + 1.3θ)(1 + Q e&f ) (2)

[0096]

[0097] The meanings of the symbols in the above formula are as follows:

[0098] δ a ——Crushing value / %;

[0099] w——Water absorption rate / %;

[0100] CBR——California Bearing Ratio / %;

[0101] θ——Content of waste red bricks / %, obtained from the image of construction waste particles;

[0102] Q e&f ——Content of needle-like and flaky particles / % (referring to the total content of needle-like particles and flaky particles), obtained from the image of construction waste particles;

[0103] d 10 ——Particle size corresponding to 10% (mass percentage) of construction waste particles, obtained from the image of construction waste particles;

[0104] d 60 ——Particle size corresponding to 60% (mass percentage) of construction waste particles, obtained from the image of construction waste particles.

[0105] In this embodiment, the R of the correlation between the crushing value and the proportion of red brick admixture (θ) and the mass content of needle-like and flaky particles (Q e&f ) is 0.96, and the R of the correlation between the CBR value and the proportion of red brick admixture (θ) and the mass content of needle-like and flaky particles (Q 2 ) is e&f ) is 2is 0.98, and the R value of the correlation between the water absorption rate and the proportion of red brick content (θ) and the content of flaky and needle-shaped particles (Q e&f ) 2 is 0.94. The fitting effect is good. Therefore, the crushing value, CBR value, and water absorption rate obtained from the prediction model of the performance indicators of construction waste particles all have strong representativeness and meet the engineering requirements.

[0106] The machine vision system obtains the red brick content, the content of flaky and needle-shaped particles, and the particle gradation (θ, Q e&f , d 10 and d 60 ) in construction waste particles through non-contact image acquisition, combined with the substance composition recognition algorithm and the particle geometric feature extraction algorithm. By inputting these into the prediction model, the predicted values of the road performance indicators of construction waste can be obtained, realizing efficient and high-speed detection, far exceeding the manual speed, being able to work continuously and being free from human interference, significantly improving the evaluation efficiency.

[0107] Example 2

[0108] An intelligent evaluation method for multiple utilization scenarios of construction waste includes the following steps:

[0109] Step 1, determine the standards for using construction waste in different parts (pavement structure and subgrade filler) of road engineering;

[0110] First, based on the coarse aggregate grading method, combined with road-related design specifications such as the "Technical Specification for Construction of Highway Pavement Bases" (JTJ 034 - 2000) and the "Code for Design of Highway Subgrades" (JTG D30 - 2015), and based on the test results of the road performance of construction waste particles, determine the water absorption rate, crushing value, and the content of flaky and needle-shaped particles as the grading control indicators. The grading standards are shown in Table 2:

[0111] Table 2 Construction waste grading standards

[0112] Construction waste grade Water absorption rate / % Crushing value / % Content of needle-like and flaky particles / % Grade I ≤5.0 ≤40.0 ≤25.0 Grade II >5.0 >40.0 >25.0

[0113] Among them, the quality of Class I construction waste is good and can be used as recycled aggregate to replace natural aggregate for building pavement structures; the quality of Class II construction waste is relatively poor, and it can be considered as subgrade filler according to relevant road performance indicators. Then, based on the construction waste grading standards and according to road-related design specifications, Class I and Class II construction waste are further refined and classified.

[0114] For Class I construction waste, in this embodiment of the present invention, according to the "Technical Rules for Construction of Highway Pavement Bases" (JTG / TF20 - 2015), it is classified using performance indicators such as the crushing value and particle size, and is respectively applied to the construction of road surface courses, bases, and sub-bases. The specific classification standards are shown in Table 3:

[0115] Table 3 Classification Criteria for Roadbed Fillers

[0116]

[0117]

[0118] For Class II construction solid waste, in the embodiments of the present invention, according to the "Code for Design of Highway Subgrade" (JTG D30-2015), its grading is carried out by using road performance indicators such as CBR and particle size, and it is respectively applied to the filling of the upper subgrade, lower subgrade, upper embankment, and lower embankment. The specific classification criteria are shown in Table 4:

[0119] Table 4 Classification Criteria for Subgrade Fillers

[0120]

[0121] Table 2-4 is a known classification criterion in the art.

[0122] Step 2: Compare the predicted road performance indicators of the construction solid waste in Example 1 with the standards in Table 2-4 to determine the grade of the construction solid waste and its applicability in different structural layers of the road project, so as to provide guidance for the rational application of the construction solid waste in the road project.

[0123] Step 2.1: Determine the evaluation indicators: Select the indicators that have a significant impact on the road performance of the construction solid waste. In the embodiments of the present invention, the water absorption rate, crushing value, mass content of needle-like and flaky particles, and CBR value are selected. These indicators reflect the performance characteristics of the construction solid waste from different aspects and are crucial for the evaluation of its applicability in the road project. For example, the water absorption rate reflects the water absorption of the construction solid waste, and too high a water absorption rate may affect the stability of the road structure; the crushing value reflects the ability of the material to resist crushing and is related to the bearing performance of the road.

[0124] Step 2.2: Determine the indicator weights: Determine the weights according to the importance of each indicator in different application scenarios of the road project. The analytic hierarchy process is used to determine. For example, in the pavement project, the crushing value has a greater impact on the strength and stability of the material, so a higher weight is given; in the subgrade project, the CBR value is more critical for ensuring the bearing capacity of the subgrade, and the corresponding weight can be set higher. By reasonably setting the weights, the relative importance of different indicators in the comprehensive evaluation is highlighted, making the evaluation result more in line with the actual project requirements.

[0125] In the evaluation of the applicable scenarios of construction solid waste, the analytic hierarchy process (AHP) is used to set the weights. For pavement engineering: Based on a large amount of test data (for each sample, at least 3 CBR tests, water absorption tests, and crushing tests were carried out according to standard test methods, and rich performance data were obtained) and engineering experience, it is considered that the degree of influence on pavement engineering is: crushing value > water absorption rate > content of needle-like and flaky particles > CBR value. The judgment matrix constructed is shown in Table 5. Taking the 1-9 scale method as an example, 1 means that two factors are equally important when compared; 9 means that one factor is extremely important compared to the other factor, and the intermediate values represent different degrees of importance differences.

[0126] Table 5 Judgment Matrix for Pavement Engineering

[0127] Crushing value Water absorption rate Content of needle-like and flaky particles CBR Crushing value 1 3 5 7 Water absorption rate 1 / 3 1 3 5 Content of needle-like and flaky particles 1 / 5 1 / 3 1 3 CBR value 1 / 7 1 / 5 1 / 3 1

[0128] The reason for taking these values (1, 3, 5, 7) is based on (the degree of influence on pavement engineering: crushing value > water absorption rate > content of needle-like and flaky particles > CBR value). In Table 5, "1 / 3" means that the crushing value is more important than the water absorption rate, that is, the importance of the water absorption rate is 1 / 3 of the crushing value; "1 / 5" means that the crushing value is more important than the content of needle-like and flaky particles, and the importance of the content of needle-like and flaky particles is 1 / 5 of the crushing value; "1 / 7" means that the crushing value is more important than the CBR value, and the importance of the CBR value is 1 / 7 of the crushing value. "1 / 3" and "3" represent opposite relative importance.

[0129] By calculating the maximum eigenvalue and eigenvector of the judgment matrix, and normalizing the eigenvector, the weights of each index are obtained. After calculation, the weight of the crushing value is 0.52, the weight of the water absorption rate is 0.27, the weight of the content of needle-like and flaky particles is 0.13, and the weight of the CBR value is 0.08.

[0130] For subgrade engineering: Based on a large amount of test data and engineering experience, it is considered that the CBR value is more critical for ensuring the bearing capacity of the subgrade. The judgment matrix constructed is shown in Table 6.

[0131] Table 6 Judgment Matrix for Subgrade Engineering

[0132] Crushing value Water absorption rate Content of needle-like and flaky particles CBR Crushing value 1 1 / 3 1 / 5 1 / 7 Water absorption rate 3 1 1 / 3 1 / 5 Content of needle-like and flaky particles 5 3 1 1 / 3 CBR value 7 5 3 1

[0133] After calculation, the weight of the crushing value is 0.08, the weight of the water absorption rate is 0.13, the weight of the content of needle-like and flaky particles is 0.27, and the weight of the CBR value is 0.52.

[0134] Step 2.3, calculate the comprehensive evaluation coefficient: The evaluation indicators are x1, x2, x3,..., x n , and the corresponding weights are w1, w2, w3,..., w n, the calculation formula for the comprehensive evaluation coefficient S of performance indicators is S = x1w1 + x2w2 + x3w3... + x n w n .

[0135] Table 7 Comparison of the value range of S and the application scenarios of construction waste

[0136]

[0137] Note: The value range of S in Table 7 is calculated based on the performance requirements of construction waste in different road engineering structural layers and combined with the weights of various performance indicators. In actual applications, it can be further optimized and adjusted according to more test data and engineering experience. The specific process is as follows:

[0138] (1) Based on the "Technical Rules for Construction of Highway Pavement Bases", on the basis of the classification standard of construction waste in Table 2, further refine the classification of different grades of construction waste to determine the applicable standards for the construction of road surface courses, bases, and sub-bases. For example, for the surface course, the particle size for expressways, first-class highways, and highways of grade two and below should be ≤ 37.5 mm, the crushing value for expressways and first-class highways should be ≤ 17.0%, and the crushing value for highways of grade two and below should be ≤ 19.0%.

[0139] For grade II construction waste, according to the "Code for Design of Highway Subgrades", use indicators such as CBR and particle size to clarify its standards for filling each part of the subgrade. For example, for the upper roadbed of the subgrade, the particle size for expressways and first-class highways should be ≤ 100 mm and CBR ≥ 8%, for second-class highways, CBR ≥ 6%, and for third- and fourth-class highways, CBR ≥ 5%.

[0140] (2) Prepare a large number of construction waste samples with different component ratios and different physical property parameters, conduct tests on various performance indicators according to the specifications, and obtain the water absorption rate, crushing value, mass content of needle-like and flaky particles, and CBR value data for each sample.

[0141] According to the calculation formula for the comprehensive evaluation coefficient S of performance indicators S = x1w1 + x2w2 + x3w3... + x n w n Perform the calculation.

[0142] Conduct statistical analysis on a large number of calculated S values (1000 construction waste samples with different component ratios and different physical property parameters); combined with the actual requirements of different structural layers of road engineering (pavement surface course, base, sub-base, upper roadbed, lower roadbed, upper embankment, lower embankment) for the performance of construction waste, divide the S values into different intervals. Thus, the corresponding relationship between the value range of S in Table 7 and the application scenarios of construction waste is obtained.

[0143] Pavement surface course: For the pavement surface course of expressways and first-class highways, due to its extremely high requirements for strength and stability, after calculating the S values of these 1000 samples and sorting them from high to low, it is found that the top 200 samples (accounting for 20%, this ratio is an assumption and needs to be determined based on a large amount of engineering experience and tests in practice) perform excellently in long-term simulated traffic load tests, water damage resistance tests, etc., and can meet the performance requirements of the pavement surface course of expressways and first-class highways. Among these 200 samples, the lowest S value is 85, so it is set that S ≥ 85 for the pavement surface course of expressways and first-class highways. The performance requirements for the pavement surface course of second-class highways and below are relatively lower. Assuming that after sorting, it is found that the 201st - 400th samples (accounting for 20%) can meet its performance requirements, among which the lowest S value is 75 and the highest is 84, so it is set that 75 ≤ S < 85 for the pavement surface course of second-class highways and below.

[0144] Pavement base course: The pavement base course of expressways and first-class highways also has relatively high requirements for strength and stability, but slightly lower than the surface course. Assuming that after sorting, the 401st - 600th samples (accounting for 20%) can meet its performance requirements, among which the lowest S value is 65 and the highest is 84, so it is set that 65 ≤ S < 85 for the pavement base course of expressways and first-class highways. The performance requirements for the pavement base course of second-class highways and below are even lower. Assuming that the 601st - 800th samples (accounting for 20%) can meet its performance requirements, the lowest S value is 55 and the highest is 74, so it is set that 55 ≤ S < 75 for the pavement base course of second-class highways and below.

[0145] Subbase course: For the subbase course of expressways and first-class highways, assuming that after sorting, the 801st - 900th samples (accounting for 10%) can meet its performance requirements, the lowest S value is 50 and the highest is 64, so it is set that 50 ≤ S < 65 for the subbase course of expressways and first-class highways. For the subbase course of second-class highways and below, assuming that the 901st - 950th samples (accounting for 5%) can meet its performance requirements, the lowest S value is 40 and the highest is 54, so it is set that 40 ≤ S < 55 for the subbase course of second-class highways and below.

[0146] Subgrade part: Taking the upper roadbed of the subgrade as an example, high-speed and first-class highways have high requirements for its bearing capacity. Assuming that after sorting, it is found that the first 150 samples (accounting for 15%) can meet its performance requirements, and the lowest S value is 70. Therefore, it is set that S ≥ 70 for the upper roadbed of high-speed and first-class highways. The requirements for second-class highways are slightly lower. Assuming that the 151st - 300th samples (accounting for 15%) can meet its performance requirements, the lowest S value is 60 and the highest is 69. Therefore, it is set that 60 ≤ S < 70 for the upper roadbed of second-class highways. The requirements for third- and fourth-class highways are even lower. Assuming that the 301st - 450th samples (accounting for 15%) can meet its performance requirements, the lowest S value is 50 and the highest is 59. Therefore, it is set that 50 ≤ S < 60 for the upper roadbed of third- and fourth-class highways. For other subgrade structural layers (lower roadbed, upper embankment, lower embankment), and so on. According to the performance requirements of different grades of highways, combined with the distribution of sample S values, the corresponding range is determined.

[0147] In actual engineering applications, engineers optimize and adjust the initially determined S value range based on long-term accumulated engineering experience. For example, considering the influence of factors such as climate conditions, traffic flow, and road service life in different regions on the performance of construction waste, the S value range is slightly adjusted to make it more in line with the actual engineering situation.

[0148] Taking Example 1 as an example: When the content of flaky and needle-shaped particles is 20% and the coefficient of uniformity is 5, the crushing value is 21.8%, the water absorption rate is 7.3%, and the CBR value is 91.1%, that is, the values of x1, x2, x3... x n At this time, in the evaluation of the pavement engineering scenario, calculate the comprehensive evaluation coefficient S of the performance index according to the pavement engineering weights set in Table 5 (crushing value weight: 0.52, water absorption rate weight: 0.27, content of flaky and needle-shaped particles weight: 0.13, CBR weight: 0.08):

[0149] S = 21.8×0.52 + 7.3×0.27 + 20×0.13 + 91.1×0.08 = 23.195 (does not meet the pavement engineering requirements)

[0150] In the evaluation of the subgrade engineering scenario, calculate the comprehensive evaluation coefficient S of the performance index according to the subgrade engineering weights set in Table 6 (crushing value weight: 0.08, water absorption rate weight: 0.13, content of flaky and needle-shaped particles weight: 0.27, CBR weight: 0.52):

[0151] S = 21.8×0.08 + 7.3×0.13 + 20×0.27 + 91.1×0.52 = 55.465 (suitable for the upper roadbed of third- and fourth-class highways or other subgrade structural parts that meet its performance indicators).

[0152] The introduction of grading and classification can accurately identify the applicable scenarios of construction waste, thus significantly reducing the demand for natural resource extraction and lowering production costs. At the same time, reasonable classification also helps to reduce environmental pollution and relieve the pressure of construction waste treatment. Secondly, clear classification criteria contribute to quickly distinguishing construction waste of different qualities and improving the treatment efficiency. This efficient classification and treatment method helps to achieve the goals of reducing, recycling, and harmless treatment of construction waste, laying a solid foundation for the green development of the construction industry. Moreover, the grading and classification of construction waste play a positive role in promoting the development of the circular economy. Through the recycling and reuse of resources, the maximization of economic value can be achieved, bringing significant economic benefits to the construction industry and even the whole society.

[0153] The embodiments of the present invention can accurately extract key physical property characteristics, combine with the road use performance of construction waste, and compare with the grading and classification standards to improve the accuracy and reliability of the evaluation results. The system can flexibly adapt to changes by updating algorithms and models to meet the ever-changing requirements of the road use performance evaluation of construction waste. The machine vision-based evaluation method for construction waste has the advantages of non-contact, high efficiency, high precision, flexibility, automation, and intelligence, providing strong support for the rational utilization of construction waste and environmental protection.

[0154] Construction waste usually has certain processability, such as red bricks, cement mortar, and crushed stones, which can be recycled through processes such as crushing and screening. If construction waste is used as road construction materials, it needs to meet a series of performance index requirements, including the physical property parameters and road use performance indicators of construction waste. By comparing the grading and classification standards with the performance requirements and using optimization algorithms, the proportions of various components (red bricks, cement mortar, crushed stones, etc.) in construction waste are adjusted on the premise of meeting the quality and performance of road engineering, so as to calculate a reasonable mix design. For example, in the construction of the road base, according to the requirements of the base for crushing value, particle size, etc., combined with the physical property parameters and estimated performance indicators of construction waste, the optimal dosage of each component is determined to achieve the efficient utilization of construction waste and the sustainable development of the environment.

[0155] The embodiments of the present invention use drones as carriers, quickly obtain physical property parameters such as color, material, content of each substance, and particle gradation in the images of construction waste recycled materials based on the principle of machine vision. On the basis of a large number of experiments, the influence laws of relevant physical property parameters on performance indicators such as CBR value, water absorption rate, and crushing value are analyzed, and a prediction model for the road use performance of construction waste particles is constructed to achieve rapid detection of performance indicators. Subsequently, referring to relevant design specifications, a quality grading and classification standard for construction waste is formulated, and combined with the predicted values of this model, suggestions for the applicable scenarios of construction waste are put forward quickly and accurately. This greatly simplifies the cumbersome processes of the recycling and utilization of construction waste and provides technical support for quickly judging the multi-channel applications of construction waste in road engineering.

[0156] Example 3

[0157] The mass ratio of the red brick content is 40%, the mass ratio of the crushed stone content is 30%, and the mass ratio of the cement mortar content is 30%; the remaining steps are the same as those in Example 1.

[0158] Example 4

[0159] The mass ratio of the red brick content is 60%, the mass ratio of the crushed stone content is 20%, and the mass ratio of the cement mortar content is 20%; the remaining steps are the same as those in Example 1.

[0160] Example 5

[0161] The mass ratio of the red brick content is 80%, the mass ratio of the crushed stone content is 10%, and the mass ratio of the cement mortar content is 10%; the remaining steps are the same as those in Example 1.

[0162] Example 6

[0163] The mass ratio of the red brick content is 100%, the mass ratio of the crushed stone content is 0%, and the mass ratio of the cement mortar content is 0%; the remaining steps are the same as those in Example 1.

[0164] Example 7

[0165] The mass ratio of the red brick content is 0%, the mass ratio of the crushed stone content is 50%, and the mass ratio of the cement mortar content is 50%; the remaining steps are the same as those in Example 1.

[0166] Examples 3 - 7 are used to explore the effects of different red brick, crushed stone, and cement mortar contents on the road - use performance indexes (crushing value, water absorption rate, CBR value) of construction waste, so as to more comprehensively analyze the relationship between the proportion of each component of construction waste and its road - use performance, and provide a basis for determining the reasonable mix ratio of construction waste.

[0167] The road - use performance data of the construction waste particles in Examples 1 and 3 - 7 are shown in Tables 8 - 10, which confirm that the construction waste has good road - use performance. The obtained road - use performance data of the construction waste can be used to provide data support. By combining the test results with the developed classification method and standard for construction waste, the performance index prediction model of construction waste particles is combined with the classification method to develop a detection and evaluation system suitable for construction waste, so that the system can conveniently judge in real - time according to the prediction results and recommend its most suitable application route.

[0168] Table 8 Road - use performance data of construction waste particles (crushing value / %)

[0169]

[0170] Table 9 Road - use performance data of construction waste particles (water absorption rate / %)

[0171]

[0172] Table 10 Road performance data of construction waste particles (CBR / %)

[0173]

[0174] The CBR value and crushing value gradually decrease with the increase of the red brick content; the water absorption rate gradually increases with the increase of the red brick content; the effects of Examples 3-6 are worse than those of Example 1, and Example 7 is higher than Example 1.

[0175] Figure 5 It is a curve relationship diagram of the crushing value, different red brick content ratios, and particle unevenness coefficients when the mass content of flaky particles in the particles is 10% in the embodiment of the present invention. Corresponding to the relevant content in the embodiment of the present invention regarding the study of the relationship between the crushing value and different construction waste components (red brick content ratio) and particle characteristics (particle unevenness coefficient) when the mass content of flaky particles in the particles is 10%, through Figure 5 , it can be intuitively seen how the changes in the red brick content ratio and particle unevenness coefficient affect the crushing value under the condition that the mass content of flaky particles in the particles is fixed, thereby providing a basis for evaluating the crushing resistance ability of construction waste, helping to judge the stability of construction waste when bearing loads in road engineering, and providing a reference for the application of construction waste in different parts of the road structure.

[0176] Figure 6 It is a curve relationship diagram of the CBR value, different red brick content ratios, and particle unevenness coefficients when the mass content of flaky particles in the particles is 10% in the embodiment of the present invention. Corresponding to the research content in the embodiment of the present invention on the relationship between the CBR value and different construction waste components (red brick content ratio) and particle characteristics (particle unevenness coefficient) when the mass content of flaky particles in the particles is 10%, Figure 6 Corresponding to the research in the embodiment of the present invention on analyzing the correlation between the CBR value, red brick content ratio, and particle unevenness coefficient. By observing the curve changes, the change trend of the CBR value of construction waste under different factor combinations can be understood, providing data support for judging whether the construction waste can meet the requirements of different road structure layers for the ability to resist local deformation, and thus determining its applicability in road engineering.

[0177] Figure 7It is a curve relationship diagram of water absorption rate, different red brick admixture ratios, and the mass content of flaky and needle-shaped particles in the embodiments of the present invention. Corresponding to the research part in the embodiments of the present invention on the relationship between water absorption rate and different building waste composition components (red brick admixture ratios) and particle characteristics (particle non-uniformity coefficient), by analyzing the curve, the influence law of different factors on the water absorption rate of building waste can be clarified, which helps to evaluate problems such as strength reduction that may be caused by water absorption during the road use process of building waste, and provides a basis for reasonably selecting the application scenarios of building waste in road engineering.

[0178] Figures 5-7 It is related to the analysis in the embodiments of the present invention of the influence law of physical property parameters such as the content of each substance component and particle gradation in building waste particles on road use performance indicators (crushing value, CBR value, water absorption rate, etc.). Among them, only one curve relationship diagram is listed under different road use performance indicators (crushing value, CBR value, water absorption rate, etc.). Figures 5-7 R in 2 Only corresponds to the research content in the figure, such as Figures 5-7 R in 2 Correspond to when the mass content of flaky and needle-shaped particles is 10% (R 2 =0.98); the crushing value, different red brick admixture ratios, and particle non-uniformity coefficient, when the mass content of flaky and needle-shaped particles is 10%, the CBR value, different red brick admixture ratios, and particle non-uniformity coefficient (R 2 =0.95); the water absorption rate, different red brick admixture ratios, and the mass content of flaky and needle-shaped particles (R 2 =0.98).

[0179] The above is only the preferred embodiment of the present invention and is not used to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention are included in the protection scope of the present invention.

Claims

1. An intelligent evaluation method for the road performance index of construction solid waste, characterized in that, It includes the following steps: S1. Obtain the mass content of red bricks, the mass content of needle-like and flaky particles, and the particle gradation in construction waste particles through a machine vision system; S2. Obtain the road performance indexes of the construction waste in S1 through tests, namely the CBR value, water absorption rate, and crushing value; S3. Establish a prediction model for the road performance indexes of construction waste according to the influence laws of the mass content of red bricks, the mass content of needle-like and flaky particles, and the particle gradation in construction waste particles on the road performance indexes of construction waste.

2. The intelligent evaluation method for the road performance index of construction solid waste according to claim 1, characterized in that, The S1 includes the following steps: S11. Obtain the image information of the original construction waste: Use a drone to collect construction waste images under the simulated scattered and dense working conditions indoors; randomly adjust the planar position and three-dimensional posture of the construction solid waste under the two indoor working conditions, and arrange the construction waste using different stacking methods to simulate the complex situations that occur in the actual environment; then shoot from multiple regions, multiple angles, and multiple lighting conditions to construct the original data set of construction waste materials; S12: Image processing: Convert the simulated images collected by the drone into digital images, and preprocess the construction waste images to improve the image quality, reduce noise, and increase the image quantity and diversity; S13. Perform image segmentation through a deep learning model, and construct an FMD data set based on the image segmentation results, including material category samples of red brick particles, cement mortar particles, and gravel particles; divide the data set into a training set, a validation set, and a test set according to a certain proportion, and obtain the mass ratios of red brick particles, cement mortar particles, and gravel particles in the construction waste particles through a material composition recognition algorithm; obtain the mass content of needle-like and flaky particles and the particle gradation in the construction waste particles through a particle geometric feature extraction algorithm.

3. The intelligent evaluation method for road performance indicators of construction solid waste according to claim 2, characterized in that, The preprocessing of the construction waste images in the S12 includes the following steps: Successively perform grayscale conversion, filtering, enhancement, and edge detection to improve the image quality and reduce noise; Further perform optical transformation, geometric transformation, noise transformation, and filtering transformation on the images through data enhancement technology to enhance the contrast of the pictures, reduce the adverse effects of natural light on the images, improve the shooting and collection conditions of the construction waste images, expand and augment the number and types of images in the original data set of construction waste, and improve the database.

4. The intelligent evaluation method for the road performance indicators of construction solid waste according to claim 2, wherein, In the S13, the material composition recognition algorithm is as follows: Input the FMD material data set constructed after image segmentation into a support vector machine or a neural network model. The model undergoes training and learning, and compares the extracted features with the sample features in the FMD data set; for the support vector machine, by finding the optimal classification hyperplane, map the input features to the corresponding material composition categories, and output the red brick particles, cement mortar particles, and gravel particles in the construction waste particles and the mass ratios they occupy; for the neural network model, through the complex connections and nonlinear transformations between a large number of neurons, learn the data features, and then output the red brick particles, cement mortar particles, and gravel particles in the construction waste particles and the mass ratios they occupy, realizing the mapping from image features to material composition and content.

5. The intelligent evaluation method for the road performance index of construction solid waste according to claim 2, characterized in that, In S13, the particle geometric feature extraction algorithm is: based on the image segmentation result, the particle size of the solid waste particles is measured, the mass content of the needle-like particles and the particle roundness are calculated, and the number of particles in different particle size ranges is counted, so as to obtain the particle grading.

6. The intelligent evaluation method for the road performance index of construction waste according to claim 1, characterized in that Specifically, S2 is as follows: by carrying out CBR test, water absorption test and crushing test, the CBR value, water absorption rate and crushing value of construction solid waste materials under various red brick, cement mortar, crushed stone particle dosage, non-uniformity coefficient and particle roundness conditions are determined.

7. The intelligent evaluation method for the road performance index of construction solid waste according to claim 1, characterized in that The estimation model of the road performance index of construction solid waste in S3 is: w = 0.05(1 + 1.3θ)(1 + Q e&f ) Among them, δ a represents the crushing value; w represents the water absorption rate; CBR represents the California Bearing Ratio; θ represents the mass percentage of waste red bricks; Q e&f represents the mass percentage of flaky particles; d 10 represents the particle size corresponding to 10% of the construction waste particles; d 60 represents the particle size corresponding to 60% of the construction waste particles; θ, Q e&f , d 10 and d 60 are obtained through the machine vision system; Obtain θ, Q of the construction solid waste to be measured through a machine vision system e&f , d 10 and d 60 , input into the prediction model, and the predicted values of the road performance indicators of the construction solid waste can be obtained.

8. An intelligent evaluation method for multi-channel utilization scenarios of construction waste, characterized in that, The following steps are involved: Step 1: Determine the performance standards for construction solid waste used as road pavement structure and roadbed filler; Step 2: Take the water absorption rate, crushing value, mass content of needle-like and flaky particles, and CBR value obtained by the intelligent evaluation method for the road performance indicators of construction solid waste described in Claim 1 as evaluation indicators, denoted as x1, x2, x3,..., x n ; The corresponding weights of the evaluation indicators x1, x2, x3,..., x n are w1, w2, w3,..., w n , and calculate the comprehensive evaluation coefficient S: S = x1w1 + x2w2 + x3w3 +... + x n w n ; Determine the road utilization scenario of construction solid waste based on the S value range and the comparison table of construction solid waste road application scenarios.

9. The intelligent evaluation method for multi-path utilization scenarios of construction solid waste according to claim 8, wherein, Method for obtaining the comparison table of the S value range and construction solid waste road application scenarios: On the basis of the classification standards for construction solid waste, further refine the classification of different levels of construction solid waste and determine the applicable standards for construction solid waste when constructing road surface, base and subbase layers; Prepare a large number of construction waste samples with different component ratios and different physical property parameters, conduct various performance index tests according to the specifications, and obtain the water absorption rate, crushing value, mass content of flaky and needle-shaped particles, and CBR value data of each sample; according to the calculation formula of the comprehensive evaluation coefficient S of performance indexes S = x1w1 + x2w2 + x3w3... + x n w n Perform calculations, and conduct statistical analysis on the large number of S values obtained from the calculations; combined with the actual requirements of different structural layers of road engineering for the performance of construction waste, divide the S values into different intervals, so as to obtain the corresponding relationship between the value range of medium S values and the utilization scenarios of construction waste.

Images