Novel recycled lightweight aggregate concrete preparation method and system based on component optimization

Abstract

The invention relates to the technical field of artificial intelligence, and discloses a novel recycled lightweight aggregate concrete preparation method and system based on component optimization, and the method comprises the steps: carrying out the surface feature recognition of a microscopic image, obtaining an aggregate surface feature set, splicing pre-constructed physicochemical property features and the aggregate surface feature set, and obtaining an aggregate fusion feature set; according to the aggregate fusion feature set, performing component optimization proportioning operation based on multi-target constraint conditions on the recycled lightweight aggregate to obtain a concrete formula set, and screening a concrete formula with the highest performance score from the concrete formula set to obtain an expected concrete formula; performing process path simulation operation based on state, action and performance on the expected concrete formula to obtain an optimal process path of the expected concrete formula; and carrying out production operation on the expected concrete formula according to the optimal process path to obtain the expected recycled lightweight aggregate concrete. The material performance of the recycled lightweight aggregate concrete can be improved by improving the components.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a novel recycled lightweight aggregate concrete preparation method and system based on component optimization. Background Art

[0002] Concrete is a vital product that has driven the development of human society and is commonly used in building infrastructure. Concrete is composed of 60% to 70% aggregate and 30% to 40% gelling material (such as cement). The aggregate acts as a filler, supporting the gelling material to form a dense structure. It also distributes loads, reducing shrinkage and thermal stress.

[0003] Considering the green development concept of renewable resources and the demand for high-strength, low-density concrete in high-rise buildings, recycled lightweight aggregate has gradually evolved. By crushing, screening, and strengthening discarded lightweight concrete components and industrial waste, recycled lightweight aggregate with a density below 1200 kg / m³ can be obtained. However, this recycled lightweight aggregate easily forms residual microcracks between the recycled lightweight aggregate and the cement, reducing the concrete's strength. Furthermore, due to its production method, it absorbs water more easily, further reducing the material's strength. Summary of the Invention

[0004] The present invention provides a novel method for preparing recycled lightweight aggregate concrete based on component optimization, the main purpose of which is to improve the material properties of the recycled lightweight aggregate concrete by improving the components.

[0005] To achieve the above objectives, the present invention provides a novel method for preparing recycled lightweight aggregate concrete based on component optimization, comprising: Obtaining recycled lightweight aggregate based on waste ceramsite, obtaining physical parameters and chemical composition information of the recycled lightweight aggregate, and obtaining a microscopic image of the recycled lightweight aggregate; Performing surface feature recognition on the microscopic image to obtain an aggregate surface feature set, and performing splicing and quantification on the physical parameters and chemical composition information to obtain physicochemical property characteristics, and splicing the physicochemical property characteristics and the aggregate surface feature set to obtain an aggregate fusion feature set; Using a pre-built genetic algorithm and a material database, based on the aggregate fusion feature set, a component optimization ratio operation based on multi-objective constraints is performed on the recycled lightweight aggregate to obtain a concrete formula set, and based on a pre-built historical experimental data set, a performance prediction is performed on each concrete formula in the concrete formula set for a preset time length to obtain a formula-performance information set; According to the formula-performance information set, a concrete formula with the highest performance score is selected from the concrete formula set to obtain a desired concrete formula; Using a pre-built Markov decision process, a process path simulation operation based on state, action and performance is performed on the desired concrete formula to obtain an optimal process path for the desired concrete formula; According to the optimal process path, the desired concrete formula is produced to obtain the desired recycled lightweight aggregate concrete.

[0006] Optionally, obtaining recycled lightweight aggregate based on waste ceramsite includes: Obtaining waste ceramsite raw materials, crushing the waste ceramsite raw materials to obtain raw material powder, and screening the raw material powder to obtain initial lightweight aggregate; Using a pre-built impact crusher, the initial lightweight aggregate is shaped to obtain clean lightweight aggregate; The surface impurities of the clean lightweight aggregate are removed by using pre-constructed dilute sulfuric acid to obtain recycled lightweight aggregate.

[0007] Optionally, performing surface feature recognition on the microscopic image to obtain a set of aggregate surface features includes: grayscale the microscopic image to obtain a grayscale image, and perform a noise reduction operation on the grayscale image using a pre-built Gaussian filtering algorithm to obtain a noise-reduced image; Using a pre-trained lightweight aggregate feature extraction network, performing a convolution operation on the denoised image to obtain a set of convolution matrices; A dimensionality reduction operation based on pooling and flattening is performed on the convolution matrix set to obtain an aggregate surface feature set.

[0008] Optionally, before using the pre-trained lightweight aggregate feature extraction network, the method further includes: Obtain a concrete surface feature recognition model based on a pre-built concrete database and convolutional neural network; A collection of lightweight aggregate-based microstructure images is synthesized using a pre-built generative adversarial network. The concrete surface feature recognition model is fine-tuned and trained using the microstructure image set to obtain a trained lightweight aggregate feature extraction network.

[0009] Optionally, the use of a pre-built genetic algorithm and a material database, according to the aggregate fusion feature set, performs a component optimization ratio operation based on multi-objective constraints on the recycled lightweight aggregate to obtain a concrete formula set, including: Acquiring a material database, wherein the material database includes recycled lightweight aggregate information, cementitious material information, admixture information, and chemical additive information; A genetic algorithm is obtained, wherein the fitness function of the genetic algorithm is expressed as:

[0010] Where, represents the fitness score, Indicates the material strength, represents density, Indicates cost, represents the weight coefficient; Acquire multi-objective constraint conditions, wherein the multi-objective constraint conditions include: the material strength is greater than a preset standard strength value, the density is less than a preset light density value, and the cost is minimized; According to the material database and multi-objective constraints, the recycled lightweight aggregate is randomly configured to obtain an initialization formula set, and each initialization formula in the initialization formula set is quantized and encoded to obtain a formula chromosome set; Calculating the fitness score of each recipe chromosome in the recipe chromosome set according to the fitness function to obtain a fitness score set, and calculating the average score of the fitness score set; According to the fitness score set, the recipe chromosomes with fitness scores greater than the average score are screened to obtain a preferred recipe chromosome set, and the preferred recipe chromosome set is reproduced according to a preset crossover and mutation strategy to obtain a progeny recipe chromosome set; Using the preferred formula chromosome set and the offspring formula chromosome set, the formula chromosome set is updated to obtain an updated formula chromosome set, and the updated number of times is counted to obtain the number of updates; Determining whether the update number is greater than or equal to a preset reproduction threshold; When the number of updates is less than the reproduction threshold, returning to the above step of calculating the fitness score of each recipe chromosome in the recipe chromosome set according to the fitness function; When the number of updates is greater than or equal to the reproduction threshold, clustering the recipe chromosome set to obtain a recipe cluster, identifying the largest recipe cluster of the order of magnitude Top-N in the recipe clusters, and obtaining a desired recipe cluster; The cluster center of each expected formula cluster in the expected formula cluster is identified to obtain a concrete formula set.

[0011] Optionally, the method of performing performance prediction for a preset time period on each concrete formula in the concrete formula set based on the pre-constructed historical experimental data set to obtain a formula-performance information set includes: Constructing a simulation model of concrete performance changes based on the historical experimental data set; Extracting features of each concrete formula in the concrete formula set to obtain a concrete component ratio sequence set; Using the concrete performance change simulation model, a performance change simulation operation based on material strength and durability is performed on each concrete component ratio sequence in the concrete component ratio sequence set to obtain a performance change curve set; The performance change curve set is intercepted according to a preset time length to obtain a recipe-performance information set.

[0012] Optionally, the method of using a pre-built Markov decision process to simulate a process path based on state, action, and performance of the desired concrete formula to obtain an optimal process path for the desired concrete formula includes: Using a pre-built Markov decision process, a base state is constructed according to a preset sensor data type, a state is derived from the base state according to preset feature engineering to obtain a derived state, and a normalization operation is performed on the base state and the derived state to obtain a state space; Obtaining a production process flow, and constructing an action space based on the production process flow, wherein the action space includes a stirring action, a molding action, and a curing action; Configuring the material strength, density, and preset action cost as core indicators, obtaining a reward function based on the core indicators, and obtaining a preset reward and punishment strategy; According to the reward function and the reward and punishment strategy, random path score identification is performed on the state space and the action space to obtain a path score identification result, and the path with the highest path score in the path score identification result is screened to obtain the optimal process path.

[0013] Optionally, after obtaining the desired recycled lightweight aggregate concrete, the method further comprises: Performing a performance test on the desired recycled lightweight aggregate concrete to obtain a performance test result; Determining whether the performance test result is greater than or equal to a preset pass threshold; When the performance test result is less than the qualified threshold, generating an updated microstructure image set according to the generative adversarial network, and optimizing the lightweight aggregate feature extraction network using the updated microstructure image set to obtain an optimized lightweight aggregate feature extraction network; The optimized lightweight aggregate feature extraction network is used to obtain an updated aggregate fusion feature set, and the updated aggregate fusion feature set is used to optimize the acquisition process of the optimal process path.

[0014] Optionally, after obtaining the desired recycled lightweight aggregate concrete, the method further comprises: Using pre-constructed Bacillus pasteurianus, spraying the desired recycled lightweight aggregate concrete; When cracks occur in the desired recycled lightweight aggregate concrete, the Bacillus pasteurianus is used to induce calcium carbonate precipitation in the cracks; The cracks are preliminarily repaired using the calcium carbonate precipitation.

[0015] To achieve the above objectives, the present invention further provides a novel recycled lightweight aggregate concrete preparation system based on component optimization, comprising: An aggregate observation module is used to obtain recycled lightweight aggregate based on waste ceramsite, obtain physical parameters and chemical composition information of the recycled lightweight aggregate, obtain a microscopic image of the recycled lightweight aggregate, perform surface feature recognition on the microscopic image to obtain an aggregate surface feature set, and perform splicing and quantification on the physical parameters and chemical composition information to obtain physicochemical property characteristics, and splice the physicochemical property characteristics with the aggregate surface feature set to obtain an aggregate fusion feature set; a formula component prediction module for performing a component optimization operation based on multi-objective constraints on the recycled lightweight aggregate using a pre-built genetic algorithm and a material database according to the aggregate fusion feature set to obtain a concrete formula set, and performing a performance prediction for each concrete formula in the concrete formula set for a preset time period based on a pre-built historical experimental data set to obtain a formula-performance information set, and screening the concrete formula with the highest performance score from the concrete formula set based on the formula-performance information set to obtain a desired concrete formula; A recipe process prediction module is used to use a pre-built Markov decision process to simulate the process path of the desired concrete recipe based on state, action and performance, and obtain the optimal process path of the desired concrete recipe; The concrete preparation module is used to perform production operations on the desired concrete formula according to the optimal process path to obtain the desired recycled lightweight aggregate concrete.

[0016] In order to solve the above problem, the present invention further provides an electronic device, comprising: a memory storing at least one instruction; The processor executes the instructions stored in the memory to implement the above-mentioned novel recycled lightweight aggregate concrete preparation method based on component optimization.

[0017] In order to solve the above problems, the present invention also provides a computer-readable storage medium, which stores at least one instruction, and the at least one instruction is executed by a processor in an electronic device to implement the above-mentioned new recycled lightweight aggregate concrete preparation method based on component optimization.

[0018] To address the problems described in the background art, the present invention first obtains physical parameters and chemical composition information of recycled lightweight aggregate, and obtains microscopic images of the recycled aggregate. This information can be used to preliminarily infer the physical and chemical properties of the recycled lightweight aggregate, thereby providing a data basis for subsequent concrete composition ratio determination. The present invention uses a genetic algorithm to randomly configure each material in the material database, and then uses concrete simulation to determine the effectiveness of each material ratio, thereby screening the desired concrete formula. In addition, after finding the desired concrete formula, the present invention also configures the process path of the desired concrete formula to find the process path that best maximizes the performance of the desired concrete formula, obtaining the optimal process path. The process path selection adopts a Markov decision process to improve the accuracy and efficiency of decision selection. Finally, a new type of recycled lightweight aggregate concrete can be constructed using the optimal process path and the desired concrete formula. Therefore, the present invention can improve the material properties of recycled lightweight aggregate concrete by improving the composition. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A schematic flow chart of a novel method for preparing recycled lightweight aggregate concrete based on component optimization provided by one embodiment of the present invention; Figure 2 This is a functional module diagram of a novel recycled lightweight aggregate concrete preparation system based on component optimization provided by one embodiment of the present invention; Figure 3 A schematic structural diagram of an electronic device for implementing the novel recycled lightweight aggregate concrete preparation method based on component optimization provided by one embodiment of the present invention.

[0020] Description of reference numerals: 1. Electronic device; 10. Processor; 11. Memory; 12. Bus.

[0021] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0022] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0023] The embodiment of the present application provides a novel method for preparing recycled lightweight aggregate concrete based on component optimization. The execution subject of the novel method for preparing recycled lightweight aggregate concrete based on component optimization includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the method provided by the embodiment of the present application. In other words, the novel method for preparing recycled lightweight aggregate concrete based on component optimization can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc.

[0024] Reference Figure 1 FIG. 1 is a flow chart of a novel method for preparing recycled lightweight aggregate concrete based on component optimization according to an embodiment of the present invention. In this embodiment, the novel method for preparing recycled lightweight aggregate concrete based on component optimization includes: S1. Obtain recycled lightweight aggregate based on waste ceramsite, obtain physical parameters and chemical composition information of the recycled lightweight aggregate, and obtain a microscopic image of the recycled lightweight aggregate.

[0025] Among them, the waste expanded clay refers to expanded clay waste or waste expanded clay products generated during use, production or demolition, which need to be recycled and processed to achieve resource reuse.

[0026] The recycled lightweight aggregate refers to aggregate produced from abandoned buildings, and the aggregate density is less than 1200 kg / m³.

[0027] The physical parameters include density and particle size.

[0028] The chemical composition information refers to the chemical components and proportions in the recycled lightweight aggregate.

[0029] The microscopic image refers to an image of the recycled lightweight aggregate obtained through high-magnification microscopy technology.

[0030] In detail, in an embodiment of the present invention, the step of obtaining recycled lightweight aggregate based on waste ceramsite includes: Obtaining waste ceramsite raw materials, crushing the waste ceramsite raw materials to obtain raw material powder, and screening the raw material powder to obtain initial lightweight aggregate; Using a pre-built impact crusher, the initial lightweight aggregate is shaped to obtain clean lightweight aggregate; The surface impurities of the clean lightweight aggregate are removed by using pre-constructed dilute sulfuric acid to obtain recycled lightweight aggregate.

[0031] The waste ceramsite raw materials refer to abandoned buildings such as light floor slabs. The crushing refers to crushing large pieces of waste ceramsite raw materials into pieces less than 1 cm by a crusher.3 The raw material powder is the result of crushing the waste ceramsite raw material.

[0032] The screening refers to the process of removing metal and wood impurities from the raw material powder. The initial lightweight aggregate is the aggregate after the metal and wood impurities are removed from the raw material powder.

[0033] The impact crusher is a machine used to break up surface deposits. Shaping is the process of removing cement from the surface of recycled lightweight aggregate. Clean lightweight aggregate refers to aggregate with no other materials attached to its surface.

[0034] The concentration of the dilute sulfuric acid is configured to be 3% to 5%. The surface impurity removal refers to the process of corroding some impurities in the cavity of the clean lightweight aggregate.

[0035] Specifically, in an embodiment of the present invention, waste lightweight floor slabs and other building materials are first recycled and then shredded using a jaw crusher to obtain a raw powder. Metal and wood particles are then removed from the raw powder to obtain initial lightweight aggregate. The present invention utilizes a vertical impact crusher to shape the initial lightweight aggregate and remove loose cement mortar from the aggregate surface, thereby obtaining clean lightweight aggregate. Finally, to enhance the chemical bonding between the primary pores of the aggregate and the cementitious material, the present invention soaks the clean lightweight aggregate in 3% dilute sulfuric acid, thereby obtaining the recycled lightweight aggregate required by the present invention.

[0036] The recycled lightweight aggregate obtained according to the above process has the characteristics of pure quality and clear structure, and can improve the accuracy of subsequent concrete batching.

[0037] Furthermore, in the embodiments of the present invention, physical parameters such as density and particle size are obtained through weighing and measurement. The chemical composition of the recycled lightweight aggregate is then determined through X-ray fluorescence spectroscopy. Finally, microscopic images are captured using a high-power microscope.

[0038] S2. Perform surface feature recognition on the microscopic image to obtain an aggregate surface feature set, and perform splicing and quantification on the physical parameters and chemical composition information to obtain physicochemical property characteristics, and splice the physicochemical property characteristics and the aggregate surface feature set to obtain an aggregate fusion feature set.

[0039] The surface feature recognition operation refers to the process of extracting features from a microscopic image through a neural network. The aggregate surface feature set is the feature extraction result of recycled lightweight aggregate.

[0040] The splicing and quantization operation refers to the process of splicing the physical parameters and chemical composition information and then uniformly quantizing and encoding them. The physicochemical property characteristics are the splicing and quantization results of the physical parameters and chemical composition information.

[0041] The process of splicing the physical and chemical property characteristics and the aggregate surface feature set is the same as the above splicing process, in which two encoding vectors are spliced. The aggregate fusion feature set represents the splicing result of the physical and chemical property characteristics and the aggregate surface feature set.

[0042] In detail, in an embodiment of the present invention, performing surface feature recognition on the microscopic image to obtain a set of aggregate surface features includes: grayscale the microscopic image to obtain a grayscale image, and perform a noise reduction operation on the grayscale image using a pre-built Gaussian filtering algorithm to obtain a noise-reduced image; Using a pre-trained lightweight aggregate feature extraction network, performing a convolution operation on the denoised image to obtain a set of convolution matrices; A dimensionality reduction operation based on pooling and flattening is performed on the convolution matrix set to obtain an aggregate surface feature set.

[0043] The grayscale processing refers to the process of converting a color image into a black and white image. The grayscale image is the grayscale result of the microscopic image.

[0044] The Gaussian filtering algorithm is an algorithm that performs weighted averaging on an image using the weight distribution of a Gaussian function. Noise reduction is an operation that removes noise points from a grayscale image. The denoised image is the denoised result of the grayscale image.

[0045] The lightweight aggregate feature extraction network is a CNN-based neural network model used to identify various features on the lightweight aggregate surface. The convolution operation is the process of performing a traversal calculation on the denoised image using a convolution kernel for feature extraction. The convolution matrix set represents the result of the convolution kernel traversal of the denoised image.

[0046] The pooling and flattening are common dimensionality reduction structures in CNN, which are used to reduce the order of magnitude of data while maintaining valid information. The aggregate surface feature set represents the dimensionality reduction result of the convolution matrix set.

[0047] Specifically, in the embodiment of the present invention, the RGB color channel values in the microscopic image are converted into grayscale values using a grayscale formula, thereby obtaining a grayscale image. The grayscale image is then input into a Gaussian filter to obtain a noise-reduced image.

[0048] Then, the present invention uses a pre-trained lightweight aggregate feature extraction network to perform a feature extraction process to obtain a convolution matrix set, and performs dimensionality reduction through average pooling operations and flattening operations to obtain an aggregate surface feature set.

[0049] Furthermore, in an embodiment of the present invention, a pre-built one-hot encoding model is used to splice and quantize the physical parameters and chemical composition information to obtain physicochemical property characteristics, and then the physicochemical property characteristics and the aggregate surface feature set are spliced to obtain an aggregate fusion feature set.

[0050] In detail, in an embodiment of the present invention, before using the pre-trained lightweight aggregate feature extraction network, the method further includes: Obtain a concrete surface feature recognition model based on a pre-built concrete database and convolutional neural network; A collection of lightweight aggregate-based microstructure images is synthesized using a pre-built generative adversarial network. The concrete surface feature recognition model is fine-tuned and trained using the microstructure image set to obtain a trained lightweight aggregate feature extraction network.

[0051] The concrete database is a database storing information about the surface structure of concrete. The convolutional neural network is a CNN network. The concrete surface feature recognition model is the result of training the CNN network based on the concrete database and is used to identify the surface structure information of concrete.

[0052] The generative adversarial network (GAN) is a network that can maximize the generation of realistic data. For example, it can synthesize images of aggregate surface structures that are as realistic as possible. It comprises a generator and a discriminator: the generator generates images, and the discriminator determines whether the images are realistic. Ultimately, the discriminator determines that all images generated by the generator are realistic, thereby expanding the order of magnitude of concrete data. The microstructural image set represents the set of images generated by the GAN for lightweight aggregate.

[0053] The training process of the fine-tuning training is the same as the normal training process, except that fewer data samples are used, which is more suitable for the application scenario of lightweight aggregate recognition. The lightweight aggregate feature extraction network represents the fine-tuning result of the concrete surface feature recognition model.

[0054] S3. Using a pre-built genetic algorithm and a material database, and based on the aggregate fusion feature set, performing a component optimization ratio operation on the recycled lightweight aggregate based on multi-objective constraints to obtain a concrete formula set. Furthermore, based on a pre-built historical experimental data set, performing a performance prediction for each concrete formula in the concrete formula set for a preset time length to obtain a formula-performance information set.

[0055] The genetic algorithm is an intelligent optimization algorithm that simulates the biological evolution process. It searches for the optimal solution in complex problems through "natural selection" and "genetic mechanism".

[0056] The material database includes various common cementitious materials, admixtures, and chemical additives. Common cementitious materials include cement grade (P·O 42.5 / SAC 42.5), mineral powder activity index (S95 / S105), and silica fume specific surface area (15-25 m² / g). Common admixtures include fly ash loss on ignition (≤5%), microbead particle size (D50 = 5μm), and nano-SiO2 purity (≥99%). Common chemical additives include water-reducing agent water-reducing rate (20%-35%) and retarder setting time adjustment range (1-4h).

[0057] The multi-objective constraints refer to the constraints on material strength, density and cost. The multi-objective constraints can ensure that the concrete formula constructed in the subsequent process meets the performance standards of high-altitude buildings.

[0058] The component optimization ratio operation refers to selecting and allocating the proportions of recycled lightweight aggregate information, cementitious material information, admixture information, and chemical additive information, thereby randomly generating various concrete formulas, and then selecting some formulas whose performance meets the requirements from the randomly generated concrete formulas.

[0059] The concrete formula set represents the final formula screening result of the component optimization ratio operation.

[0060] The historical experimental data set is a database storing performance parameters of concrete of various formulations over the next 30 days, wherein the performance parameters mainly include strength and durability.

[0061] The time length is configured as 28 days.

[0062] The performance prediction refers to the operation of predicting the strength and durability of concrete corresponding to each recipe in the concrete recipe set after 28 days. The recipe-performance information set represents a set of corresponding relationships between each recipe and performance.

[0063] In detail, in an embodiment of the present invention, the pre-built genetic algorithm and material database are used to perform a component optimization ratio operation based on multi-objective constraints on the recycled lightweight aggregate according to the aggregate fusion feature set to obtain a concrete formula set, including: Acquiring a material database, wherein the material database includes recycled lightweight aggregate information, cementitious material information, admixture information, and chemical additive information; A genetic algorithm is obtained, wherein the fitness function of the genetic algorithm is expressed as:

[0064] Where, represents the fitness score, Indicates the material strength, represents density, Indicates cost, represents the weight coefficient; Acquire multi-objective constraint conditions, wherein the multi-objective constraint conditions include: the material strength is greater than a preset standard strength value, the density is less than a preset light density value, and the cost is minimized; According to the material database and multi-objective constraints, the recycled lightweight aggregate is randomly configured to obtain an initialization formula set, and each initialization formula in the initialization formula set is quantized and encoded to obtain a formula chromosome set; Calculating the fitness score of each recipe chromosome in the recipe chromosome set according to the fitness function to obtain a fitness score set, and calculating the average score of the fitness score set; According to the fitness score set, the recipe chromosomes with fitness scores greater than the average score are screened to obtain a preferred recipe chromosome set, and the preferred recipe chromosome set is reproduced according to a preset crossover and mutation strategy to obtain a progeny recipe chromosome set; Using the preferred formula chromosome set and the offspring formula chromosome set, the formula chromosome set is updated to obtain an updated formula chromosome set, and the updated number of times is counted to obtain the number of updates; Determining whether the update number is greater than or equal to a preset reproduction threshold; When the number of updates is less than the reproduction threshold, returning to the above step of calculating the fitness score of each recipe chromosome in the recipe chromosome set according to the fitness function; When the number of updates is greater than or equal to the reproduction threshold, clustering the recipe chromosome set to obtain a recipe cluster, identifying the largest recipe cluster of the order of magnitude Top-N in the recipe clusters, and obtaining a desired recipe cluster; The cluster center of each expected formula cluster in the expected formula cluster is identified to obtain a concrete formula set.

[0065] The fitness function is a method for calculating the fitness value of concrete corresponding to each recipe. The fitness value is a comprehensive score of the concrete in terms of hardness, density and cost.

[0066] The standard strength value is C30, and the light density value is 1800 kg / m 3 .

[0067] The random configuration operation refers to randomly extracting a cementitious material from the cementitious material information, randomly extracting an admixture from the admixture information, and randomly extracting a chemical additive from the chemical additive information to form a contributing component, and then randomly configuring the content ratio of recycled lightweight aggregate, cementitious material, admixture and chemical additive.

[0068] The initialized formula set represents a set of randomly constructed concrete formulas.

[0069] The quantization encoding refers to the process of representing the content of each recipe by fixed characters. The recipe chromosome set refers to the quantization result of the initialization recipe set.

[0070] The fitness score set is the set of fitness scores corresponding to each recipe chromosome, and the average score is the result of summing up the fitness scores of each fitness score in the fitness score set and then dividing the sum equally.

[0071] In the crossover and mutation strategies, crossover simulates biological reproduction, exchanging parts of the genes of two parents to generate new individuals. Mutation randomly changes certain bits of a gene with a low probability to increase population diversity.

[0072] For example, crossover simulates a binary crossover (SBX, probability = 0.8) for mixed measurements of continuous variables, and variation simulates a polynomial variation (probability = 0.1) for aggregate ratio ±5% and water-binder ratio ±0.02.

[0073] The reproduction refers to the process of generating offspring chromosomes from a parent chromosome. The offspring formula chromosome set refers to the set of generated offspring chromosomes.

[0074] The updating refers to a direct replacement operation process.

[0075] The update times is a value recorded along with the number of changes in the recipe chromosome set.

[0076] The reproduction threshold is configured as 100 times.

[0077] The clustering operation refers to the operation of grouping chromosomes with similar components and proportions into one cluster. The recipe cluster represents the set of clusters identified by the clustering operation.

[0078] The order of magnitude Top-N refers to screening N clusters with the largest number of elements in the cluster. The desired recipe clustering cluster is a set of N clusters screened.

[0079] The operation of identifying cluster centers can be processed by a cosine similarity algorithm. The concrete formula set represents a set of cluster centers of N clusters.

[0080] Specifically, in this embodiment of the present invention, information about recycled lightweight aggregate, cementitious materials, admixtures, and chemical additives is first collected to construct a materials database. Then, a fitness function is constructed based on the concrete standards (strength, density, and cost) required for high-rise buildings.

[0081] In the embodiment of the present invention, the random configuration process is first constrained according to multi-objective constraints, thereby reducing the order of magnitude of the initialization recipe set, and then encoding is performed through an encoding tool to obtain a recipe chromosome set.

[0082] Specifically, in this embodiment of the present invention, the recipe chromosome set is first screened to obtain the optimal recipe chromosome set. Then, through mutation and crossover, the offspring recipe chromosome set is obtained. For example, if parent recipe A is [60% aggregate, 0.35 water-binder ratio] and recipe B is [55% aggregate, 0.40 water-binder ratio], the offspring recipe chromosome after crossover might be [55% aggregate, 0.35 water-binder ratio]. Furthermore, the water-binder ratio is randomly adjusted from 0.35 to 0.34 or 0.36 to prevent the algorithm from falling into a local optimum.

[0083] In the embodiment of the present invention, the above process of generating offspring recipe chromosomes from parent recipe chromosomes is repeated 100 times to obtain the desired recipe cluster.

[0084] Specifically, in this embodiment of the present invention, a clustering operation can be used to group the chromosome types (recipe types) within the desired recipe clusters to obtain recipe clusters. The three largest clusters (N = 3) are then selected to obtain the desired recipe clusters. The cluster center of each desired recipe cluster represents the recipe type with the highest weight among all recipe types. Therefore, by calculating the cluster centers, a set of concrete recipes can be obtained.

[0085] Specifically, in an embodiment of the present invention, the performance prediction of each concrete recipe in the concrete recipe set for a preset time period is performed based on the pre-constructed historical experimental data set to obtain a recipe-performance information set, including: Constructing a simulation model of concrete performance changes based on the historical experimental data set; Extracting features of each concrete formula in the concrete formula set to obtain a concrete component ratio sequence set; Using the concrete performance change simulation model, a performance change simulation operation based on material strength and durability is performed on each concrete component ratio sequence in the concrete component ratio sequence set to obtain a performance change curve set; The performance change curve set is intercepted according to a preset time length to obtain a recipe-performance information set.

[0086] The concrete performance change simulation model is a regression network, which is formed by fitting the performance change curves of each concrete at each time point.

[0087] The feature extraction refers to the convolution and pooling process of the neural network. The concrete component ratio sequence set refers to the set of feature data of each concrete formula in the concrete formula set.

[0088] The performance change simulation operation based on material strength and durability refers to the forward network calculation process of the concrete performance change simulation model, and outputs relevant data on material strength and durability. The performance change curve set represents the set of material strength curves and durability curves corresponding to each concrete component ratio sequence.

[0089] The interception refers to the process of saving the content data of the 28th day in the performance change curve set within 0 to 30 days.

[0090] Specifically, in this embodiment of the present invention, a regression network is first trained based on a set of historical experimental data to generate a simulation model for concrete performance changes. Then, through the forward network calculation process of the concrete performance change simulation model, a set of concrete component ratio sequences and a set of performance change curves are obtained. Finally, by intercepting the data at the 28-day time point, a set of formula-performance information corresponding to each concrete formula is obtained.

[0091] S4. Filtering the concrete formula with the highest performance score from the concrete formula set according to the formula-performance information set to obtain a desired concrete formula.

[0092] Specifically, in the embodiment of the present invention, the process of screening the concrete with the highest performance score is to weight the strength score and durability score of each concrete formula, then sort them by size, and select the concrete formula with the largest score as the desired concrete formula.

[0093] S5. Using a pre-built Markov decision process, simulate the process path of the desired concrete formula based on the state, action and performance to obtain the optimal process path of the desired concrete formula.

[0094] Among them, the Markov decision process refers to a detailed implementation plan for optimizing the preparation process of recycled lightweight aggregate concrete. By defining the sensor data and derived features collected in real time during the production process and designing a multi-objective reward function, balancing strength, density and cost can be achieved.

[0095] The process path simulation operation refers to the process of performing process simulation according to the assigned state and action. The optimal process path refers to the process path with the highest performance among the various simulated process paths.

[0096] In detail, in an embodiment of the present invention, the process path simulation operation based on the state, action and performance of the desired concrete formula is performed using a pre-built Markov decision process to obtain the optimal process path of the desired concrete formula, including: Using a pre-built Markov decision process, a base state is constructed according to a preset sensor data type, a state is derived from the base state according to preset feature engineering to obtain a derived state, and a normalization operation is performed on the base state and the derived state to obtain a state space; Obtaining a production process flow, and constructing an action space based on the production process flow, wherein the action space includes a stirring action, a molding action, and a curing action; Configuring the material strength, density, and preset action cost as core indicators, obtaining a reward function based on the core indicators, and obtaining a preset reward and punishment strategy; According to the reward function and the reward and punishment strategy, random path score identification is performed on the state space and the action space to obtain a path score identification result, and the path with the highest path score in the path score identification result is screened to obtain the optimal process path.

[0097] The sensor data types include temperature, humidity and vibration intensity. The basic state refers to three sensor types.

[0098] The feature engineering refers to an algorithm for mining potential features in data features, such as calculating the rate of change, mean, etc. The derived state refers to the type set of potential features between data.

[0099] The normalization operation refers to the process of mapping a numerical value to a value between 0 and 1. The state space refers to the set of normalization results of each data in the basic state and the derived state.

[0100] The production process includes stirring, forming and curing in sequence.

[0101] Among them, constructing the action space means taking the three types of mixing, forming and curing as spatial dimensions.

[0102] Among them, the core indicators are: compressive strength (MPa, target ≥C30); apparent density (kg / m³, target ≤1900); action cost (yuan / m³, positively correlated with process time and temperature), which need to be distinguished from the cost of ingredients.

[0103] The reward and punishment strategy is: if the final intensity meets the standard and the density is qualified, then +50; if any indicator does not meet the standard, then -30.

[0104] The reward function is expressed as:

[0105] Where, (Adjustable), Indicates compressive strength, represents density, Indicates cost, , .

[0106] The random path score identification refers to the process of calculating rewards for each random path to obtain scores.

[0107] Specifically, in the embodiment of the present invention, the sensor data and derived features collected in real time during the production process can be defined through the state space as follows: The basic sensor data include: temperature (mixer internal temperature, curing room temperature, range: 10°C ~ 80°C); humidity (ambient humidity, aggregate moisture content, range: 30% ~ 95% RH); vibration frequency (molding vibration table frequency, range: 0 ~ 100 Hz).

[0108] The derived state characteristics include: temperature change rate (ΔT / Δt); deviation of aggregate moisture content from the target value (ΔW); and vibration energy density (square of vibration frequency × amplitude).

[0109] Among them, the state representation method is: continuous variable normalization: mapping the original data to the [0, 1] interval; discretization classification (optional): for example, dividing the temperature into three levels: low temperature (<20℃), normal temperature (20-40℃), and high temperature (>40℃).

[0110] Furthermore, in an embodiment of the present invention, the action space can be used to define adjustable process parameters as actions of the agent, for example: The stirring stage: stirring speed (low / medium / high, corresponding to 100 / 200 / 300 rpm); stirring time adjustment (±10 seconds).

[0111] The molding stage includes: vibration time (10-60 seconds, step length 5 seconds); vibration mode (continuous vibration / intermittent vibration).

[0112] The curing stage: steam curing temperature setting (40℃ / 60℃ / 80℃); curing humidity setting (50% / 70% / 90% RH).

[0113] Specifically, in an embodiment of the present invention, after the state space and the action space are constructed, the various values in the state space and the action space are randomly configured to obtain various random paths, and then the scores of the various random paths are identified according to the reward function and the reward and punishment measurement to obtain the path score identification results, and then the path with the highest score is screened to obtain the optimal process path.

[0114] S6. According to the optimal process path, the desired concrete formula is produced to obtain the desired recycled lightweight aggregate concrete.

[0115] Specifically, in an embodiment of the present invention, the desired concrete formula is produced according to the optimal process path: cement (PO 42.5), silica fume (10%), and mineral powder (S95 grade, 15%) are dry-mixed for 60 seconds. Recycled lightweight aggregate, pre-wetting water (70% of the total water content), and a water reducer (polycarboxylic acid type, 1.0%) are then mixed for 90 seconds. The binder and the remaining water (30% of the total water content) are then added and mixed for 180 seconds until the slump reaches 180±20mm. The final vibration time is 30 seconds at a pressure of 10 MPa to ensure density (porosity <25%). Steam curing is then performed at 50°C for 12 hours (heating rate 15°C / h), with a humidity ≥90% and a demolding strength ≥15 MPa. Curing is continued for 28 days, with daily water spraying for moisture retention, to produce the desired recycled lightweight aggregate concrete.

[0116] In detail, in an embodiment of the present invention, after obtaining the desired recycled lightweight aggregate concrete, the method further includes: Performing a performance test on the desired recycled lightweight aggregate concrete to obtain a performance test result; Determining whether the performance test result is greater than or equal to a preset pass threshold; When the performance test result is less than the qualified threshold, generating an updated microstructure image set according to the generative adversarial network, and optimizing the lightweight aggregate feature extraction network using the updated microstructure image set to obtain an optimized lightweight aggregate feature extraction network; The optimized lightweight aggregate feature extraction network is used to obtain an updated aggregate fusion feature set, and the updated aggregate fusion feature set is used to optimize the acquisition process of the optimal process path.

[0117] The performance test refers to the process of testing the performance of the sample of the desired recycled lightweight aggregate concrete by means of impact, vibration, friction, etc. The performance test result represents the experimental test result of the desired recycled lightweight aggregate concrete.

[0118] Among them, the qualified threshold is strength ≥30MPa, density ≤1900kg / m³, and cost less than 600 yuan per cubic meter.

[0119] Specifically, in the embodiment of the present invention, after obtaining the desired recycled lightweight aggregate concrete, it should be verified whether the desired recycled lightweight aggregate concrete is qualified.

[0120] Through actual performance testing, performance test results are obtained, and then the performance test results are compared with the qualified threshold. Among them, as long as one indicator in the performance test result fails, it is determined that the performance test result is less than the qualified threshold, and the solution process of this application needs to be re-optimized.

[0121] The recognition of lightweight aggregate surface features forms the data foundation for this approach. Therefore, it is necessary to regenerate samples using a generative adversarial network to obtain an updated set of microstructure images, which can then be used to re-implement this approach and optimize subsequent processes.

[0122] In detail, in an embodiment of the present invention, after obtaining the desired recycled lightweight aggregate concrete, the method further includes: Using pre-constructed Bacillus pasteurianus, spraying the desired recycled lightweight aggregate concrete; When cracks occur in the desired recycled lightweight aggregate concrete, the Bacillus pasteurianus is used to induce calcium carbonate precipitation in the cracks; The cracks are preliminarily repaired using the calcium carbonate precipitation.

[0123] The Bacillus pasteurianus is a Gram-positive bacterium with special biomineralization ability and can generate calcium carbonate precipitation.

[0124] The spraying operation refers to adding microorganisms in a vaporized and dispersed liquid form during the construction of the desired recycled lightweight aggregate concrete.

[0125] Specifically, in the embodiment of the present invention, a biological characteristic method is used to further protect the desired recycled lightweight aggregate concrete, and Bacillus pasteurianus releases calcium carbonate at the cracks when it comes into contact with water to repair microcracks.

[0126] To address the problems described in the background art, the present invention first obtains physical parameters and chemical composition information of recycled lightweight aggregate, and obtains microscopic images of the recycled aggregate. This information can be used to preliminarily infer the physical and chemical properties of the recycled lightweight aggregate, thereby providing a data basis for subsequent concrete composition ratio determination. The present invention uses a genetic algorithm to randomly configure each material in the material database, and then uses concrete simulation to determine the effectiveness of each material ratio, thereby screening the desired concrete formula. In addition, after finding the desired concrete formula, the present invention also configures the process path of the desired concrete formula to find the process path that best maximizes the performance of the desired concrete formula, obtaining the optimal process path. The process path selection adopts a Markov decision process to improve the accuracy and efficiency of decision selection. Finally, a new type of recycled lightweight aggregate concrete can be constructed using the optimal process path and the desired concrete formula. Therefore, the present invention can improve the material properties of recycled lightweight aggregate concrete by improving the composition.

[0127] like Figure 2 , which is a functional module diagram of a novel recycled lightweight aggregate concrete preparation system based on component optimization provided by one embodiment of the present invention.

[0128] The novel recycled lightweight aggregate concrete production system 100 based on component optimization described in the present invention can be installed in an electronic device. Depending on the functionality implemented, the novel recycled lightweight aggregate concrete production system 100 based on component optimization can include an aggregate observation module 101, a recipe component prediction module 102, a recipe process prediction module 103, and a concrete preparation module 104. A module, also referred to as a unit, is a series of computer program segments that can be executed by an electronic device processor and perform a fixed function. These are stored in the electronic device's memory.

[0129] The aggregate observation module 101 is used to obtain recycled lightweight aggregate based on waste ceramsite, obtain physical parameters and chemical composition information of the recycled lightweight aggregate, obtain a microscopic image of the recycled lightweight aggregate, and perform surface feature recognition on the microscopic image to obtain an aggregate surface feature set, and perform splicing and quantification on the physical parameters and chemical composition information to obtain physicochemical property characteristics, and splice the physicochemical property characteristics and the aggregate surface feature set to obtain an aggregate fusion feature set; The formula component prediction module 102 is configured to utilize a pre-built genetic algorithm and a material database to perform a component optimization operation based on multi-objective constraints on the recycled lightweight aggregate according to the aggregate fusion feature set to obtain a concrete formula set, and to perform a performance prediction for each concrete formula in the concrete formula set for a preset time period based on a pre-built historical experimental data set to obtain a formula-performance information set, and to select a concrete formula with the highest performance score from the concrete formula set based on the formula-performance information set to obtain a desired concrete formula; The recipe process prediction module 103 is used to use a pre-built Markov decision process to simulate the process path of the desired concrete recipe based on state, action and performance, and obtain the optimal process path of the desired concrete recipe; The concrete preparation module 104 is configured to perform production operations on the desired concrete formula according to the optimal process path to obtain desired recycled lightweight aggregate concrete.

[0130] In detail, each module in the novel recycled lightweight aggregate concrete preparation system 100 based on component optimization in the embodiment of the present invention adopts the same method as above when in use. Figure 1 The same technical means are used as the preparation method of the new recycled lightweight aggregate concrete based on component optimization described in , and can produce the same technical effects, so they will not be repeated here.

[0131] like Figure 3 , which is a structural diagram of an electronic device for implementing a novel recycled lightweight aggregate concrete preparation method based on component optimization provided by an embodiment of the present invention.

[0132] The electronic device 1 may include a processor 10, a memory 11 and a bus 12, and may also include a computer program stored in the memory 11 and executable on the processor 10, such as a program for preparing a novel recycled lightweight aggregate concrete based on component optimization.

[0133] The memory 11 includes at least one type of readable storage medium, including flash memory, a removable hard drive, a multimedia card, a card-type memory (e.g., SD or DX memory), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the memory 11 may be an internal storage unit of the electronic device 1, such as a removable hard drive of the electronic device 1. In other embodiments, the memory 11 may also be an external storage device of the electronic device 1, such as a plug-in removable hard drive, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card, etc. Furthermore, the memory 11 includes both internal storage units and external storage devices of the electronic device 1. The memory 11 can be used not only to store application software installed in the electronic device 1 and various data, such as the code of a program for a novel recycled lightweight aggregate concrete preparation method based on component optimization, but also to temporarily store data that has been output or is about to be output.

[0134] In some embodiments, the processor 10 may be comprised of an integrated circuit, such as a single packaged integrated circuit or a combination of multiple packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips. The processor 10 is the control core (control unit) of the electronic device, connecting the various components of the electronic device using various interfaces and circuits. It executes programs or modules stored in the memory 11 (e.g., a program for a novel recycled lightweight aggregate concrete preparation method based on component optimization) and accesses data stored in the memory 11 to perform various functions and process data.

[0135] The bus 12 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus. The bus 12 may be divided into an address bus, a data bus, a control bus, etc. The bus 12 is configured to enable communication between the memory 11 and at least one processor 10, etc.

[0136] Figure 3 Only the electronic device with components is shown, and it can be understood by those skilled in the art that Figure 3The structure shown does not constitute a limitation on the electronic device 1 , and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0137] For example, although not shown, the electronic device 1 may further include a power supply (e.g., a battery) to power various components. Preferably, the power supply may be logically connected to the at least one processor 10 via a power management device, thereby enabling functions such as charge management, discharge management, and power consumption management via the power management device. The power supply may further include any components such as one or more DC or AC power supplies, a recharging device, a power failure detection circuit, a power converter or inverter, and a power status indicator. The electronic device 1 may also include various sensors, Bluetooth modules, Wi-Fi modules, etc., which are not further described here.

[0138] Furthermore, the electronic device 1 may also include a network interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the electronic device 1 and other electronic devices.

[0139] Optionally, the electronic device 1 may further include a user interface, which may be a display or an input unit (such as a keyboard). Optionally, the user interface may also be a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen. The display may also be appropriately referred to as a display screen or a display unit, and is used to display information processed by the electronic device 1 and to display a visual user interface.

[0140] The program of the novel recycled lightweight aggregate concrete preparation method based on component optimization stored in the memory 11 of the electronic device 1 is a combination of multiple instructions. When running in the processor 10, it can achieve the following: Obtaining recycled lightweight aggregate based on waste ceramsite, obtaining physical parameters and chemical composition information of the recycled lightweight aggregate, and obtaining a microscopic image of the recycled lightweight aggregate; Performing surface feature recognition on the microscopic image to obtain an aggregate surface feature set, and performing splicing and quantification on the physical parameters and chemical composition information to obtain physicochemical property characteristics, and splicing the physicochemical property characteristics and the aggregate surface feature set to obtain an aggregate fusion feature set; Using a pre-built genetic algorithm and a material database, based on the aggregate fusion feature set, a component optimization ratio operation based on multi-objective constraints is performed on the recycled lightweight aggregate to obtain a concrete formula set, and based on a pre-built historical experimental data set, a performance prediction is performed on each concrete formula in the concrete formula set for a preset time length to obtain a formula-performance information set; According to the formula-performance information set, a concrete formula with the highest performance score is selected from the concrete formula set to obtain a desired concrete formula; Using a pre-built Markov decision process, a process path simulation operation based on state, action and performance is performed on the desired concrete formula to obtain an optimal process path for the desired concrete formula; According to the optimal process path, the desired concrete formula is produced to obtain the desired recycled lightweight aggregate concrete.

[0141] Specifically, the specific implementation method of the processor 10 for the above instructions can refer to Figures 1 to 3 The description of the relevant steps in the corresponding embodiments will not be repeated here.

[0142] Furthermore, if the modules / units integrated into the electronic device 1 are implemented as software functional units and sold or used as independent products, they may be stored in a computer-readable storage medium. The computer-readable storage medium may be volatile or non-volatile. For example, the computer-readable medium may include any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard drive, a magnetic disk, an optical disk, a computer memory, or a read-only memory (ROM).

[0143] The present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program. When the computer program is executed by a processor of an electronic device, the computer program can implement: Obtaining recycled lightweight aggregate based on waste ceramsite, obtaining physical parameters and chemical composition information of the recycled lightweight aggregate, and obtaining a microscopic image of the recycled lightweight aggregate; Performing surface feature recognition on the microscopic image to obtain an aggregate surface feature set, and performing splicing and quantification on the physical parameters and chemical composition information to obtain physicochemical property characteristics, and splicing the physicochemical property characteristics and the aggregate surface feature set to obtain an aggregate fusion feature set; Using a pre-built genetic algorithm and a material database, based on the aggregate fusion feature set, a component optimization ratio operation based on multi-objective constraints is performed on the recycled lightweight aggregate to obtain a concrete formula set, and based on a pre-built historical experimental data set, a performance prediction is performed on each concrete formula in the concrete formula set for a preset time length to obtain a formula-performance information set; According to the formula-performance information set, a concrete formula with the highest performance score is selected from the concrete formula set to obtain a desired concrete formula; Using a pre-built Markov decision process, a process path simulation operation based on state, action and performance is performed on the desired concrete formula to obtain an optimal process path for the desired concrete formula; According to the optimal process path, the desired concrete formula is produced to obtain the desired recycled lightweight aggregate concrete.

[0144] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, systems and methods can be implemented in other ways. For example, the system embodiments described above are only exemplary, and actual implementations may have other division methods.

[0145] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed across multiple network elements. Some or all of the modules may be selected to achieve the purpose of the solution of this embodiment according to actual needs.

[0146] In addition, the functional modules in various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or hardware plus software functional modules.

[0147] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0148] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A novel method for preparing recycled lightweight aggregate concrete based on component optimization, characterized in that: The method comprises: Obtaining recycled lightweight aggregate based on waste ceramsite, obtaining physical parameters and chemical composition information of the recycled lightweight aggregate, and obtaining a microscopic image of the recycled lightweight aggregate; Performing surface feature recognition on the microscopic image to obtain an aggregate surface feature set, and performing splicing and quantification on the physical parameters and chemical composition information to obtain physicochemical property characteristics, and splicing the physicochemical property characteristics and the aggregate surface feature set to obtain an aggregate fusion feature set; Using a pre-built genetic algorithm and a material database, based on the aggregate fusion feature set, a component optimization ratio operation based on multi-objective constraints is performed on the recycled lightweight aggregate to obtain a concrete formula set, and based on a pre-built historical experimental data set, a performance prediction is performed on each concrete formula in the concrete formula set for a preset time length to obtain a formula-performance information set; According to the formula-performance information set, a concrete formula with the highest performance score is selected from the concrete formula set to obtain a desired concrete formula; Using a pre-built Markov decision process, a process path simulation operation based on state, action and performance is performed on the desired concrete formula to obtain an optimal process path for the desired concrete formula; According to the optimal process path, the desired concrete formula is produced to obtain the desired recycled lightweight aggregate concrete.

2. The method for preparing a novel recycled lightweight aggregate concrete based on component optimization according to claim 1, characterized in that: The method of obtaining recycled lightweight aggregate based on waste ceramsite comprises: Obtaining waste ceramsite raw materials, crushing the waste ceramsite raw materials to obtain raw material powder, and screening the raw material powder to obtain initial lightweight aggregate; Using a pre-built impact crusher, the initial lightweight aggregate is shaped to obtain clean lightweight aggregate; The surface impurities of the clean lightweight aggregate are removed by using pre-constructed dilute sulfuric acid to obtain recycled lightweight aggregate.

3. The method for preparing a novel recycled lightweight aggregate concrete based on component optimization according to claim 2, characterized in that: The performing surface feature recognition on the microscopic image to obtain a set of aggregate surface features includes: grayscale the microscopic image to obtain a grayscale image, and perform a noise reduction operation on the grayscale image using a pre-built Gaussian filtering algorithm to obtain a noise-reduced image; Using a pre-trained lightweight aggregate feature extraction network, performing a convolution operation on the denoised image to obtain a set of convolution matrices; A dimensionality reduction operation based on pooling and flattening is performed on the convolution matrix set to obtain an aggregate surface feature set.

4. The method for preparing a novel recycled lightweight aggregate concrete based on component optimization according to claim 3, characterized in that: Before using the pre-trained lightweight aggregate feature extraction network, the method further includes: Obtain a concrete surface feature recognition model based on a pre-built concrete database and convolutional neural network; A collection of lightweight aggregate-based microstructure images is synthesized using a pre-built generative adversarial network. The concrete surface feature recognition model is fine-tuned and trained using the microstructure image set to obtain a trained lightweight aggregate feature extraction network.

5. The method for preparing a novel recycled lightweight aggregate concrete based on component optimization according to claim 4, characterized in that: The method utilizes a pre-built genetic algorithm and a material database to perform a component optimization ratio operation based on multi-objective constraints on the recycled lightweight aggregate according to the aggregate fusion feature set, thereby obtaining a concrete formula set, including: Acquiring a material database, wherein the material database includes recycled lightweight aggregate information, cementitious material information, admixture information, and chemical additive information; A genetic algorithm is obtained, wherein the fitness function of the genetic algorithm is expressed as: Where, represents the fitness score, Indicates the material strength, represents density, Indicates cost, represents the weight coefficient; Acquire multi-objective constraint conditions, wherein the multi-objective constraint conditions include: the material strength is greater than a preset standard strength value, the density is less than a preset light density value, and the cost is minimized; According to the material database and multi-objective constraints, the recycled lightweight aggregate is randomly configured to obtain an initialization formula set, and each initialization formula in the initialization formula set is quantized and encoded to obtain a formula chromosome set; Calculating the fitness score of each recipe chromosome in the recipe chromosome set according to the fitness function to obtain a fitness score set, and calculating the average score of the fitness score set; According to the fitness score set, the recipe chromosomes with fitness scores greater than the average score are screened to obtain a preferred recipe chromosome set, and the preferred recipe chromosome set is reproduced according to a preset crossover and mutation strategy to obtain a progeny recipe chromosome set; Using the preferred formula chromosome set and the offspring formula chromosome set, the formula chromosome set is updated to obtain an updated formula chromosome set, and the updated number of times is counted to obtain the number of updates; Determining whether the update number is greater than or equal to a preset reproduction threshold; When the number of updates is less than the reproduction threshold, returning to the above step of calculating the fitness score of each recipe chromosome in the recipe chromosome set according to the fitness function; When the number of updates is greater than or equal to the reproduction threshold, clustering the recipe chromosome set to obtain a recipe cluster, identifying the largest recipe cluster of the order of magnitude Top-N in the recipe clusters, and obtaining a desired recipe cluster; The cluster center of each expected formula cluster in the expected formula cluster is identified to obtain a concrete formula set.

6. The method for preparing a novel recycled lightweight aggregate concrete based on component optimization according to claim 5, characterized in that: The method of performing performance prediction for each concrete formula in the concrete formula set for a preset time period based on the pre-constructed historical experimental data set to obtain a formula-performance information set includes: Constructing a simulation model of concrete performance changes based on the historical experimental data set; Extracting features of each concrete formula in the concrete formula set to obtain a concrete component ratio sequence set; Using the concrete performance change simulation model, a performance change simulation operation based on material strength and durability is performed on each concrete component ratio sequence in the concrete component ratio sequence set to obtain a performance change curve set; The performance change curve set is intercepted according to a preset time length to obtain a recipe-performance information set.

7. The method for preparing a novel recycled lightweight aggregate concrete based on component optimization according to claim 6, characterized in that: The method utilizes a pre-built Markov decision process to simulate a process path based on state, action, and performance of the desired concrete formula to obtain an optimal process path for the desired concrete formula, including: Using a pre-built Markov decision process, a base state is constructed according to a preset sensor data type, a state is derived from the base state according to preset feature engineering to obtain a derived state, and a normalization operation is performed on the base state and the derived state to obtain a state space; Obtaining a production process flow, and constructing an action space based on the production process flow, wherein the action space includes a stirring action, a molding action, and a curing action; Configuring the material strength, density, and preset action cost as core indicators, obtaining a reward function based on the core indicators, and obtaining a preset reward and punishment strategy; According to the reward function and the reward and punishment strategy, random path score identification is performed on the state space and the action space to obtain a path score identification result, and the path with the highest path score in the path score identification result is screened to obtain the optimal process path.

8. The method for preparing a novel recycled lightweight aggregate concrete based on component optimization according to claim 7, characterized in that: After obtaining the desired recycled lightweight aggregate concrete, the method further comprises: Performing a performance test on the desired recycled lightweight aggregate concrete to obtain a performance test result; Determining whether the performance test result is greater than or equal to a preset pass threshold; When the performance test result is less than the qualified threshold, generating an updated microstructure image set according to the generative adversarial network, and optimizing the lightweight aggregate feature extraction network using the updated microstructure image set to obtain an optimized lightweight aggregate feature extraction network; The optimized lightweight aggregate feature extraction network is used to obtain an updated aggregate fusion feature set, and the updated aggregate fusion feature set is used to optimize the acquisition process of the optimal process path.

9. The method for preparing a novel recycled lightweight aggregate concrete based on component optimization according to claim 8, characterized in that: After obtaining the desired recycled lightweight aggregate concrete, the method further comprises: Using pre-constructed Bacillus pasteurianus, spraying the desired recycled lightweight aggregate concrete; When cracks occur in the desired recycled lightweight aggregate concrete, the Bacillus pasteurianus is used to induce calcium carbonate precipitation in the cracks; The cracks are preliminarily repaired using the calcium carbonate precipitation.

10. A new type of recycled lightweight aggregate concrete preparation system based on component optimization, characterized in that: The system comprises: An aggregate observation module is used to obtain recycled lightweight aggregate based on waste ceramsite, obtain physical parameters and chemical composition information of the recycled lightweight aggregate, obtain a microscopic image of the recycled lightweight aggregate, perform surface feature recognition on the microscopic image to obtain an aggregate surface feature set, and perform splicing and quantification on the physical parameters and chemical composition information to obtain physicochemical property characteristics, and splice the physicochemical property characteristics with the aggregate surface feature set to obtain an aggregate fusion feature set; a formula component prediction module for performing a component optimization operation based on multi-objective constraints on the recycled lightweight aggregate using a pre-built genetic algorithm and a material database according to the aggregate fusion feature set to obtain a concrete formula set, and performing a performance prediction for each concrete formula in the concrete formula set for a preset time period based on a pre-built historical experimental data set to obtain a formula-performance information set, and screening the concrete formula with the highest performance score from the concrete formula set based on the formula-performance information set to obtain a desired concrete formula; A recipe process prediction module is used to use a pre-built Markov decision process to simulate the process path of the desired concrete recipe based on state, action and performance, and obtain the optimal process path of the desired concrete recipe; The concrete preparation module is used to perform production operations on the desired concrete formula according to the optimal process path to obtain the desired recycled lightweight aggregate concrete.

Images