Optimization method and system for lightweight wallboard based on slag recycled aggregate

Abstract

The application relates to the technical field of building materials, in particular to a lightweight wallboard optimization method and system based on slag recycled aggregate, which comprises the following steps: acquiring basic raw material characteristic information of the slag recycled aggregate, performing proportioning optimization under multi-target constraints through a material collaborative optimization module, screening a target formula in combination with performance simulation, and optimizing a preparation process by using a dynamic process path planning module, so that a high-performance lightweight wallboard is finally prepared. The application can significantly improve the overall performance and production efficiency of the lightweight wallboard, and realizes efficient utilization of the slag recycled aggregate, and has environmental protection and economic benefits.

Description

Technical Field

[0001] This invention belongs to the field of building materials technology, specifically an optimization method and system for lightweight wall panels based on recycled aggregates from construction waste. Background Technology

[0002] With the increasing emphasis on environmental protection and resource recycling in the construction industry, lightweight wall panels based on recycled aggregates from construction waste are gradually becoming an important research direction in the field of building materials due to their excellent properties such as lightweight, thermal insulation, and sound insulation. However, existing lightweight wall panels still have shortcomings in terms of raw material optimization, preparation process, and performance control, which affect their comprehensive performance and widespread application. For example, patent CN106278377B discloses a self-insulating lightweight wall panel made of all-recycled lightweight aggregate concrete and its preparation method. By using all-recycled lightweight aggregate, ecological fiber, physical foaming agent, and other raw materials, a lightweight wall panel with a uniform porous structure is prepared, which has good waterproof, sound insulation, and thermal insulation properties. However, in this technical solution, the proportion of all-recycled lightweight aggregate is relatively wide (5-40%), which may lead to large fluctuations in material properties in actual production, making it difficult to achieve stable mechanical properties and thermal insulation effects. In addition, this solution does not fully consider the special properties of recycled aggregates from construction waste, such as irregular particle shape and high surface roughness, which may affect the fluidity of the slurry and the molding quality. On the other hand, patent CN114505325B discloses a method for preparing recycled aggregate for a new type of crack-resistant and heat-insulating wall panel. This method reduces the waste of new materials and improves the crack resistance and insulation effect of the wall panel by recycling and reusing the crack-resistant insulation layer within the waste wall panel. However, this technical solution mainly focuses on the preparation process of the recycled aggregate and lacks systematic research on optimizing the overall performance of the lightweight wall panel, especially the synergistic mechanism between recycled aggregate and other materials. Furthermore, this solution does not involve intelligent manufacturing processes for lightweight wall panels, which may limit its application in modern industrialized construction. These problems indicate that existing lightweight wall panels based on recycled aggregate still have shortcomings in raw material ratio optimization, molding process stability, comprehensive performance control, and intelligent manufacturing. Therefore, this invention aims to improve the overall performance and production efficiency of lightweight wall panels by precisely controlling the raw material ratio, optimizing the manufacturing process, enhancing the synergistic effect of material performance, and introducing an intelligent control system, thereby meeting the demands of modern construction for efficient, environmentally friendly, and intelligent building materials. Summary of the Invention

[0003] This invention provides a method and system for optimizing lightweight wall panels based on recycled aggregates from slag. Its main purpose is to improve the overall performance and production efficiency of lightweight wall panels through raw material ratio optimization, process improvement, and performance control.

[0004] To achieve the above objectives, the present invention provides an optimization method for lightweight wall panels based on recycled aggregates from construction waste, comprising:

[0005] Obtain basic raw materials based on recycled aggregates from slag, obtain particle distribution characteristics and surface morphology information of the basic raw materials, and obtain microstructure images of the basic raw materials;

[0006] Texture features are extracted from the microstructure image to obtain an aggregate texture feature set. The particle distribution features and surface morphology information are quantized and stitched together to obtain a particle morphology feature set. The aggregate texture feature set and the particle morphology feature set are merged to obtain a comprehensive feature set.

[0007] Using a pre-built material co-optimization module and material database, the basic raw materials are optimized under multi-objective constraints according to the comprehensive feature set to obtain a set of wall panel formulations. Combined with a pre-built set of historical experimental data, the performance of each formulation in the set of wall panel formulations is simulated within a preset period to obtain a formulation-performance mapping set.

[0008] Based on the formula-performance mapping set, the wall panel formula with the best overall performance is selected from the wall panel formula set to obtain the target wall panel formula;

[0009] Using a pre-built dynamic process path planning module, the target wall panel formulation is optimized based on process parameters and performance feedback to obtain the optimal process path.

[0010] According to the optimal process path, the target wall panel formulation is prepared to obtain the target lightweight wall panel.

[0011] Optionally, obtaining the basic raw materials based on recycled aggregate from waste soil includes:

[0012] Obtain recycled aggregate raw materials from slag and soil, perform layered crushing treatment on the recycled aggregate raw materials to obtain primary aggregate, and perform screening operation on the primary aggregate to obtain standard particle size aggregate.

[0013] The standard-size aggregate is surface-treated using a pre-constructed vibratory polishing device to obtain smooth aggregate;

[0014] The smooth aggregate is surface-activated using a pre-constructed alkaline solution to obtain the basic raw material.

[0015] Optionally, the step of extracting texture features from the microstructure image to obtain an aggregate texture feature set includes:

[0016] The microstructure image is subjected to edge enhancement processing to obtain an enhanced image, and a pre-constructed local contrast adjustment algorithm is used to perform detail optimization operation on the enhanced image to obtain an optimized image;

[0017] Using a pre-trained aggregate texture recognition network, feature extraction is performed on the optimized image to obtain a set of texture feature matrices;

[0018] The texture feature matrix set is subjected to dimensionality reduction and normalization to obtain the aggregate texture feature set.

[0019] Optionally, before utilizing the pre-trained aggregate texture recognition network, the method further includes:

[0020] Based on a pre-built database of building materials and a deep learning model, an aggregate texture recognition model is obtained.

[0021] A set of simulated images of microstructure based on aggregate surface properties is synthesized using a pre-built virtual generation module;

[0022] The aggregate texture recognition model is fine-tuned and trained using the microstructure simulation image set to obtain the trained aggregate texture recognition network.

[0023] Optionally, the step of utilizing a pre-built material co-optimization module and a material database to perform multi-objective ratio optimization operations on the basic raw materials based on the comprehensive feature set, thereby obtaining a wall panel formulation set, includes:

[0024] Obtain a materials database, wherein the materials database includes information on recycled aggregates from waste soil, binders, fiber-reinforced materials, and functional additives;

[0025] A material synergistic optimization module is obtained, wherein the core indicators of the material synergistic optimization module include mechanical strength, thermal conductivity and environmental protection cost;

[0026] Obtain multi-objective constraints, wherein the multi-objective constraints include: the mechanical strength is greater than a preset standard strength value, the thermal conductivity is less than a preset thermal insulation performance threshold, and the environmental protection cost is minimized.

[0027] Based on the material database and multi-objective constraints, the basic raw materials are randomly combined to obtain an initial formula set, and each initial formula in the initial formula set is coded and marked to obtain a formula code set;

[0028] The comprehensive score of each formula in the formula code set is calculated based on the core indicators to obtain a score set, and the baseline value of the score set is calculated.

[0029] Based on the scoring set, formula codes with comprehensive scores higher than the benchmark value are selected to obtain a preferred formula code set. Then, according to a preset iteration strategy, the preferred formula code set is optimized and reorganized to obtain an updated formula code set.

[0030] Determine whether the number of updates has reached a preset optimization threshold;

[0031] When the number of updates does not reach the optimization threshold, return to the steps described above for calculating the comprehensive score of each recipe in the recipe code set based on the core indicators;

[0032] When the number of updates reaches the optimization threshold, cluster analysis is performed on the formula code set to obtain formula clusters. The largest clusters in the top-N order of magnitude of the formula clusters are identified to obtain the target formula cluster.

[0033] Extract the center point of each target formula in the target formula cluster to obtain the wall panel formula set.

[0034] Optionally, the step of performing performance simulations on each formulation in the wall panel formulation set within a preset period based on a pre-constructed set of historical experimental data to obtain a formulation-performance mapping set includes:

[0035] Based on the aforementioned historical experimental data set, a simulation model for the performance evolution of the wall panel was constructed.

[0036] Feature analysis is performed on each formula in the wall panel formula set to obtain a set of formula composition sequences;

[0037] Using the wall panel performance evolution simulation model, a performance evolution simulation operation based on mechanical properties and durability is performed on each formula composition sequence in the formula composition sequence set to obtain a set of performance evolution curves;

[0038] According to a preset period, the set of performance evolution curves is truncated to obtain a formula-performance mapping set.

[0039] Optionally, the step of using a pre-built dynamic process path planning module to perform path optimization on the target wall panel formulation based on process parameters and performance feedback to obtain the optimal process path includes:

[0040] Using a pre-built dynamic process path planning module, a basic process state is constructed based on the preset sensor monitoring type. The basic process state is then extended according to the preset data processing rules to obtain an extended process state. The basic process state and the extended process state are then standardized to obtain a process state space.

[0041] Obtain the production process flow, and construct a process action space based on the production process flow, wherein the process action space includes stirring action, pressing and molding action and curing action;

[0042] The mechanical strength, thermal conductivity, and preset action efficiency are configured as core indicators, and an evaluation function based on the core indicators is obtained to obtain a preset optimization strategy.

[0043] Based on the evaluation function and optimization strategy, path scoring and identification are performed on the process state space and process action space to obtain path scoring results. The path with the highest score in the path scoring results is then selected to obtain the optimal process path.

[0044] Optionally, after obtaining the target lightweight wall panel, the method further includes:

[0045] The target lightweight wall panel was subjected to performance testing, and the performance test results were obtained.

[0046] Determine whether the performance test results meet the preset pass / fail criteria;

[0047] When the performance test results do not meet the qualification criteria, an updated set of microstructure simulation images is generated according to the virtual generation module. The aggregate texture recognition network is then optimized using the updated set of microstructure simulation images to obtain an optimized aggregate texture recognition network.

[0048] The optimized aggregate texture recognition network is used to obtain an updated comprehensive feature set, and the process of obtaining the optimal process path is optimized using the updated comprehensive feature set.

[0049] Optionally, after obtaining the target lightweight wall panel, the method further includes:

[0050] The target lightweight wall panel is surface-sprayed using a pre-constructed microbial mineralization technology.

[0051] When cracks appear in the target lightweight wall panel, the microbial mineralization technology is used to induce the formation of calcium carbonate deposits within the cracks;

[0052] The crack was initially repaired using the calcium carbonate deposition.

[0053] To achieve the above objectives, the present invention also provides a lightweight wall panel optimization system based on recycled aggregate from slag, comprising:

[0054] The raw material observation module is used to acquire basic raw materials based on recycled aggregates from slag, acquire particle distribution characteristics and surface morphology information of the basic raw materials, acquire microstructure images of the basic raw materials, extract texture features from the microstructure images to obtain aggregate texture feature set, quantize and stitch the particle distribution characteristics and surface morphology information to obtain particle morphology feature set, and merge the aggregate texture feature set and particle morphology feature set to obtain comprehensive feature set.

[0055] The formulation optimization module is used to utilize a pre-built material co-optimization module and material database to perform multi-objective constraint optimization operations on the basic raw materials according to the comprehensive feature set, thereby obtaining a set of wall panel formulations. Combined with a pre-built set of historical experimental data, the module performs performance simulations on each formulation in the set of wall panel formulations within a preset period to obtain a formulation-performance mapping set. Based on the formulation-performance mapping set, the module selects the wall panel formulation with the best comprehensive performance from the set of wall panel formulations to obtain the target wall panel formulation.

[0056] The process path optimization module is used to perform path optimization operations on the target wall panel formula based on process parameters and performance feedback using a pre-built dynamic process path planning module to obtain the optimal process path.

[0057] The wall panel preparation module is used to prepare the target wall panel formula according to the optimal process path to obtain the target lightweight wall panel.

[0058] To address the above problems, the present invention also provides an electronic device, the electronic device comprising:

[0059] Memory, storing at least one instruction;

[0060] The processor executes the instructions stored in the memory to implement the above-described method for optimizing lightweight wall panels based on recycled aggregates.

[0061] To address the aforementioned problems, the present invention also provides a computer-readable storage medium storing at least one instruction, which is executed by a processor in an electronic device to implement the aforementioned method for optimizing lightweight wall panels based on recycled aggregates.

[0062] To address the problems described in the background section, this invention first acquires the particle distribution characteristics and surface morphology information of recycled aggregates from slag, and obtains their microstructure images. This information provides data support for subsequent proportioning optimization. The invention then uses a material co-optimization module to randomly proportion raw materials under multi-objective constraints, and combines this with performance simulation to evaluate the effectiveness of each formulation, thereby selecting the target wall panel formulation. Furthermore, after finding the target wall panel formulation, the invention optimizes its process path using a dynamic process path planning module to ensure maximum formulation performance. Finally, through the optimal process path and the target wall panel formulation, a high-performance target lightweight wall panel can be prepared. Therefore, this invention significantly improves the overall performance and production efficiency of lightweight wall panels through raw material optimization, process improvement, and performance control. Attached Figure Description

[0063] Figure 1This is a flowchart illustrating an embodiment of the present invention for an optimized method of lightweight wall panels based on recycled aggregates from slag.

[0064] Figure 2 This is a functional module diagram of a lightweight wall panel optimization system based on recycled aggregate provided in an embodiment of the present invention;

[0065] Figure 3 This is a schematic diagram of the structure of an electronic device for implementing the optimized method for lightweight wall panels based on recycled aggregates, according to an embodiment of the present invention. Detailed Implementation

[0066] This invention provides an optimized method and system for lightweight wall panels based on recycled aggregates from slag. Its core objective is to significantly improve the overall performance and production efficiency of lightweight wall panels through raw material ratio optimization, process improvement, and performance control. The following is in conjunction with the appendix... Figure 1 To be continued Figure 3 The specific embodiments of the present invention will be described in detail below.

[0067] In the specific implementation process, the first step is to obtain the basic raw materials based on recycled aggregates from waste soil; this is the starting point of the entire optimization method. For example... Figure 1 As shown, this process includes several steps: First, the recycled aggregate raw material is obtained and subjected to layered crushing to obtain primary aggregate. Layered crushing uses multi-stage crushing equipment, typically divided into coarse and fine crushing stages. The coarse crushing stage mainly uses a jaw crusher, while the fine crushing stage uses a hammer crusher or cone crusher to ensure uniform aggregate particle size distribution. Second, the primary aggregate is screened to obtain standard-sized aggregate. Screening is performed using vibrating screens with screen apertures ranging from 5mm to 10mm, depending on actual needs, to meet the requirements for subsequent lightweight wall panel production. Third, the standard-sized aggregate is surface-treated using a pre-constructed vibratory polishing device to obtain smooth aggregate. The vibratory polishing device uses high-frequency vibration to ensure full contact between the aggregate particles and the polishing medium, thereby removing rough surfaces and improving the surface quality of the aggregate. Fourth, the smooth aggregate is surface-activated using a pre-constructed alkaline solution to obtain the base raw material. Alkaline solutions are typically prepared from sodium hydroxide or potassium hydroxide, with a concentration ranging from 5% to 10%, and a treatment time of 1 to 2 hours. The purpose is to enhance the chemical activity of the aggregate surface, thereby improving its bonding ability with the binder.

[0068] After obtaining the basic raw materials, the next step is to extract texture features from the microstructure images of the basic raw materials to obtain a set of aggregate texture features. This process is as follows: Figure 1 As shown, the process includes the following steps: First, edge enhancement processing is performed on the acquired microstructure image to highlight the texture details of the aggregate surface. Edge enhancement processing uses the Sobel operator algorithm, G...x and G y Let these represent the gradient values ​​in the horizontal and vertical directions, respectively, expressed by the formula:

[0069]

[0070] The gradient value G of the enhanced image is calculated. Then, a pre-constructed local contrast adjustment algorithm is used to perform detail optimization on the enhanced image, resulting in an optimized image. The core idea of ​​the local contrast adjustment algorithm is to further improve image sharpness by dynamically adjusting the brightness and contrast of local regions of the image. Specifically, this algorithm uses formula I... out (x,y)=α×I in (x,y)+β is implemented, where I in (x,y) represents the pixel values ​​of the input image, I out (x, y) represents the pixel values ​​of the output image, and α and β are the gain coefficient and offset, respectively, with values ​​ranging from 1.2 to 1.5 and -20 to 20. Next, a pre-trained aggregate texture recognition network is used to extract features from the optimized image, resulting in a texture feature matrix set. The aggregate texture recognition network is based on a convolutional neural network (CNN) architecture. Its input layer receives the optimized image, and after multiple convolution and pooling operations, it outputs a texture feature matrix. To improve the accuracy of feature extraction, the aggregate texture recognition network needs to be fine-tuned before use. The fine-tuning process includes obtaining an aggregate texture recognition model from a building materials database and using a virtual generation module to synthesize a set of simulated images of the microstructure based on the surface properties of aggregates. These simulated images are then used for model training and optimization. Finally, the texture feature matrix set is dimensionality reduced and normalized to obtain the aggregate texture feature set. Dimensionality reduction uses Principal Component Analysis (PCA) algorithm, with the core formula Y = XW, where X is the original feature matrix, W is the projection matrix, and Y is the dimensionality-reduced feature matrix. Normalization is achieved using the formula Z = (X - μ) / σ, where μ and σ are the mean and standard deviation of the feature matrix, respectively.

[0071] After extracting the aggregate texture feature set, it is necessary to quantify and stitch together the particle distribution features and surface morphology information to obtain the particle morphology feature set. The quantification of particle distribution features mainly relies on the statistical analysis of particle size distribution curves, and its calculation formula is D. n =∑d i / n, where D n d represents the average particle size. iLet represent the particle size value of the i-th particle, and n represent the total number of particles. Surface morphology information is quantified by acquiring the height distribution data of the aggregate surface using a 3D scanner, and then smoothing it using a Gaussian filter to obtain the surface roughness parameter Ra. The particle morphology feature set is obtained by concatenating the two quantification results to form a unified feature vector. Subsequently, the aggregate texture feature set and the particle morphology feature set are merged to obtain a comprehensive feature set, providing data support for subsequent formulation optimization.

[0072] After obtaining the comprehensive feature set, the next step is to utilize the pre-built material collaborative optimization module and material database to perform multi-objective ratio optimization operations on the basic raw materials to obtain the wall panel formulation set. This process is as follows: Figure 2 As shown, the main steps include: First, obtaining relevant material information from the material database, including information on recycled aggregates, binders, fiber-reinforced materials, and functional additives. The content of recycled aggregates typically ranges from 40% to 60%, binders from 20% to 30%, fiber-reinforced materials from 5% to 10%, and functional additives from 1% to 5%. Second, obtaining the material synergistic optimization module, whose core indicators include mechanical strength, thermal conductivity, and environmental cost. The target value for mechanical strength is typically set to be greater than 5 MPa, the target value for thermal conductivity is set to be less than 0.2 W / (m·K), and the target value for environmental cost is set to be minimized. Next, based on the material database and multi-objective constraints, the basic raw materials are randomly combined to obtain an initial formula set. The initial formula set was generated using a genetic algorithm, with the core formula being F(x) = w1×f1(x) + w2×f2(x) + w3×f3(x), where F(x) represents the comprehensive scoring function, f1(x), f2(x), and f3(x) represent the scoring functions for mechanical strength, thermal conductivity, and environmental cost, respectively, and w1, w2, and w3 are the weight coefficients for each indicator, with values ​​ranging from 0.4 to 0.6, 0.2 to 0.4, and 0.1 to 0.3, respectively. Subsequently, each initial formula in the initial formula set was coded to obtain a formula code set. The coding used a binary encoding method, with each formula's code length being 16 bits. The first 8 bits represent the aggregate content, the middle 4 bits represent the binder content, and the last 4 bits represent the fiber reinforcement content. Next, the comprehensive score of each formula in the formula code set was calculated based on the core indicators to obtain a score set, and a baseline value for the score set was calculated. The formula for calculating the baseline value is: B = ∑F i / n, where B represents the baseline value, F iLet represent the overall score of the i-th formula, and n represent the total number of formulas. Then, formula codes with overall scores higher than the benchmark value are selected from the score set to obtain a preferred formula code set. This preferred formula code set is then optimized and reorganized according to a preset iterative strategy to obtain an updated formula code set. The iterative strategy uses the simulated annealing algorithm, whose core formula is:

[0073]

[0074] Where P represents the probability of accepting the new solution, F new and F old , representing the combined scores of the new and old solutions respectively, and T representing the current temperature. During the iteration process, if the number of updates does not reach the preset optimization threshold, the process returns to the step of calculating the combined score; if the number of updates reaches the optimization threshold, cluster analysis is performed on the formula coding set to obtain formula clusters, and the top-N largest clusters are identified to obtain the target formula clusters. Finally, the center points of each target formula in the target formula clusters are extracted to obtain the wall panel formula set.

[0075] After obtaining the wall panel formulation set, the next step is to combine it with a pre-constructed set of historical experimental data to perform performance simulations on each formulation within a preset period, in order to obtain a formulation-performance mapping set. This process is as follows: Figure 1 As shown, the main steps include the following: First, a simulation model of wall panel performance evolution is constructed based on historical experimental data. This model is based on the finite element analysis method, with the core formula being [K]{u}={F}, where [K] represents the stiffness matrix, {u} represents the displacement vector, and {F} represents the external force vector. Second, feature analysis is performed on each formulation in the wall panel formulation set to obtain a set of formulation composition sequences. The set of formulation composition sequences is generated using a unique thermal coding method, with each formulation having a 20-bit code length. The first 10 bits represent the aggregate content, the middle 5 bits represent the binder content, and the last 5 bits represent the fiber reinforcement content. Next, the wall panel performance evolution simulation model is used to perform performance evolution simulation operations based on mechanical properties and durability on each formulation composition sequence in the set, resulting in a set of performance evolution curves. The set of performance evolution curves is generated using the time step method, with the core formula being u(t+Δt)=u(t)+v(t)Δt+0.5a(t)Δt. 2 Here, u(t), v(t), and a(t) represent the time series of displacement, velocity, and acceleration, respectively, and Δt represents the time step. Finally, the set of performance evolution curves is truncated according to a preset period to obtain the formula-performance mapping set. The preset period is usually set to 28 days to simulate the performance changes of the wall panel in the actual use environment.

[0076] After obtaining the formula-performance mapping set, the next step is to select the wall panel formula with the best overall performance from the wall panel formula set to obtain the target wall panel formula. This process is as follows: Figure 1 As shown, the main steps include the following: First, based on the performance data in the formula-performance mapping set, calculate the comprehensive performance score for each formula. The formula for calculating the comprehensive performance score is S = w1 × s1 + w2 × s2 + w3 × s3, where S represents the comprehensive performance score, s1, s2, and s3 represent the scores for mechanical strength, thermal conductivity, and durability, respectively, and w1, w2, and w3 are the weighting coefficients for each indicator, with values ​​ranging from 0.4 to 0.6, 0.2 to 0.4, and 0.1 to 0.3, respectively. Second, rank the formulas in the wall panel formula set according to the comprehensive performance score, and select the formula with the highest score as the target wall panel formula. The determination of the target wall panel formula considers not only mechanical properties and thermal insulation performance, but also environmental costs and the feasibility of the production process.

[0077] After obtaining the target wall panel formulation, the next step is to utilize a pre-built dynamic process path planning module to perform path optimization based on process parameters and performance feedback to obtain the optimal process path. This process is as follows: Figure 2As shown, the main steps include: First, a basic process state is constructed using the dynamic process path planning module based on preset sensor monitoring types. Sensor monitoring types include temperature sensors, humidity sensors, and pressure sensors, whose monitoring data are used to reflect the real-time status of the production process. Second, the basic process state is extended according to preset data processing rules to obtain an extended process state. The state extension process uses a Kalman filter algorithm, with the core formula being x(k|k)=x(k|k-1)+K(k)(z(k)-Hx(k|k-1)), where x(k|k) represents the current state estimate, x(k|k-1) represents the previous state prediction, K(k) represents the Kalman gain, z(k) represents the observed value, and H represents the observation matrix. Subsequently, the basic and extended process states are standardized to obtain the process state space. Standardization is achieved using the formula Z=(X-μ) / σ, where μ and σ are the mean and standard deviation of the state data, respectively. Next, the production process flow is obtained, and a process action space is constructed based on the production process flow. The process motion space includes stirring, pressing, and curing actions, with stirring time, pressing pressure, and curing temperature as the core parameters. Subsequently, mechanical strength, thermal conductivity, and preset action efficiency are configured as core indicators, and an evaluation function based on these indicators is obtained. The calculation formula for the evaluation function is E = w1×e1 + w2×e2 + w3×e3, where E represents the evaluation score, e1, e2, and e3 represent the scores for mechanical strength, thermal conductivity, and action efficiency, respectively, and w1, w2, and w3 are the weight coefficients for each indicator, with values ​​ranging from 0.4 to 0.6, 0.2 to 0.4, and 0.1 to 0.3, respectively. Finally, based on the evaluation function and optimization strategy, path scoring is performed on the process state space and process motion space to obtain path scoring results, and the path with the highest score is selected as the optimal process path. The optimization strategy uses the ant colony algorithm, with the core formula being τ. ij (t+1)=(1-ρ)τ ij (t)+Δτ ij , where τ ij The pheromone concentration at path (i,j) is represented by ρ, and the pheromone evaporation coefficient is represented by Δτ. ij The pheromone increment represents the path (i,j).

[0078] After obtaining the optimal process route, the next step is to prepare the target wall panel formulation according to the optimal process route to obtain the target lightweight wall panel. This process is as follows: Figure 1As shown, the main steps include: First, mixing is performed according to the mixing parameters in the optimal process path. A twin-shaft mixer is used, with a mixing time of 3 to 5 minutes and a mixing speed of 60 to 80 r / min to ensure thorough mixing of the aggregate and binder. Second, pressing is performed according to the pressing parameters in the optimal process path. A hydraulic molding machine is used, with a pressing pressure of 10 to 15 MPa and a holding time of 10 to 15 seconds to ensure the density and strength of the wall panel. Finally, curing is performed according to the curing parameters in the optimal process path. A steam curing chamber is used, with a curing temperature of 60°C to 80°C and a curing time of 24 to 48 hours to accelerate the hardening process of the wall panel.

[0079] After preparing the target lightweight wall panel, performance testing is required to verify whether it meets the preset qualification standards. Performance testing mainly includes mechanical property testing, thermal conductivity testing, and durability testing. Mechanical property testing uses a universal testing machine, and the test result must be greater than 5 MPa; thermal conductivity testing uses a heat flow meter method, and the test result must be less than 0.2 W / (m·K); durability testing uses a freeze-thaw cycle test, and the test result must meet the requirement of no obvious cracks or spalling. If the performance test results do not meet the qualification standards, an updated set of simulated microstructure images needs to be generated based on the virtual generation module. This updated set of simulated microstructure images is then used to optimize the aggregate texture recognition network, resulting in an optimized aggregate texture recognition network. Subsequently, the optimized aggregate texture recognition network is used to obtain an updated comprehensive feature set, and this updated comprehensive feature set is used to optimize the process for obtaining the optimal process path.

[0080] Furthermore, to further enhance the performance of the target lightweight wall panel, a pre-constructed microbial mineralization technology can be used for surface spraying. This technology induces calcium carbonate deposition on the wall panel surface by spraying a microbial solution containing carbonic anhydrase, thereby improving its crack resistance and durability. When cracks appear in the target lightweight wall panel, microbial mineralization technology can be used to induce calcium carbonate deposition within the cracks, and this deposition can then be used for initial repair.

[0081] In summary, this invention, through a series of detailed implementation steps, realizes the specific implementation of an optimized method and system for lightweight wall panels based on recycled aggregates from waste soil. From the acquisition of basic raw materials to the final preparation of the lightweight wall panels, each step has undergone scientific design and strict control, ensuring a significant improvement in the overall performance and production efficiency of the wall panels.

Claims

1. An optimization method for lightweight wall panels based on recycled aggregate from construction waste, characterized in that, The method includes: Obtain basic raw materials based on recycled aggregates from slag, obtain particle distribution characteristics and surface morphology information of the basic raw materials, and obtain microstructure images of the basic raw materials; Texture features are extracted from the microstructure image to obtain an aggregate texture feature set. The particle distribution features and surface morphology information are quantized and stitched together to obtain a particle morphology feature set. The aggregate texture feature set and the particle morphology feature set are merged to obtain a comprehensive feature set. Using a pre-built material co-optimization module and material database, the basic raw materials are optimized under multi-objective constraints according to the comprehensive feature set to obtain a set of wall panel formulations. Combined with a pre-built set of historical experimental data, the performance of each formulation in the set of wall panel formulations is simulated within a preset period to obtain a formulation-performance mapping set. The method utilizes a pre-built material collaborative optimization module and a material database to perform multi-objective ratio optimization operations on the basic raw materials based on the comprehensive feature set, thereby obtaining a wall panel formulation set, including: Obtain a materials database, wherein the materials database includes information on recycled aggregates from waste soil, binders, fiber-reinforced materials, and functional additives; A material synergistic optimization module is obtained, wherein the core indicators of the material synergistic optimization module include mechanical strength, thermal conductivity and environmental protection cost; Obtain multi-objective constraints, wherein the multi-objective constraints include: the mechanical strength is greater than a preset standard strength value, the thermal conductivity is less than a preset thermal insulation performance threshold, and the environmental protection cost is minimized. Based on the material database and multi-objective constraints, the basic raw materials are randomly combined to obtain an initial formula set, and each initial formula in the initial formula set is coded and marked to obtain a formula code set; The comprehensive score of each formula in the formula code set is calculated based on the core indicators to obtain a score set, and the baseline value of the score set is calculated. Based on the score set, formula codes with comprehensive scores higher than the benchmark value are selected to obtain a preferred formula code set. The preferred formula code set is then optimized and reorganized according to a preset iteration strategy to obtain an updated formula code set. The number of times the optimized and reorganized set is updated is recorded. Determine whether the number of updates has reached a preset optimization threshold; When the number of updates does not reach the optimization threshold, return to the steps described above for calculating the comprehensive score of each recipe in the recipe code set based on the core indicators; When the number of updates reaches the optimization threshold, cluster analysis is performed on the formula code set to obtain formula clusters. The largest clusters in the top-N order of magnitude of the formula clusters are identified to obtain the target formula cluster. Extract the center point of each target formula in the target formula cluster to obtain the wall panel formula set; Based on the formula-performance mapping set, the wall panel formula with the best overall performance is selected from the wall panel formula set to obtain the target wall panel formula; Using a pre-built dynamic process path planning module, the target wall panel formulation is optimized based on process parameters and performance feedback to obtain the optimal process path. According to the optimal process path, the target wall panel formulation is prepared to obtain the target lightweight wall panel.

2. The method for optimizing lightweight wall panels based on recycled aggregates as described in claim 1, characterized in that, The acquisition of basic raw materials based on recycled aggregate from waste soil includes: Obtain recycled aggregate raw materials from slag and soil, perform layered crushing treatment on the recycled aggregate raw materials to obtain primary aggregate, and perform screening operation on the primary aggregate to obtain standard particle size aggregate. The standard-size aggregate is surface-treated using a pre-constructed vibratory polishing device to obtain smooth aggregate; The smooth aggregate is surface-activated using a pre-constructed alkaline solution to obtain the basic raw material.

3. The method for optimizing lightweight wall panels based on recycled aggregates as described in claim 1, characterized in that, The step of extracting texture features from the microstructure image to obtain an aggregate texture feature set includes: The microstructure image is subjected to edge enhancement processing to obtain an enhanced image, and a pre-constructed local contrast adjustment algorithm is used to perform detail optimization operation on the enhanced image to obtain an optimized image; Using a pre-trained aggregate texture recognition network, feature extraction is performed on the optimized image to obtain a set of texture feature matrices; The texture feature matrix set is subjected to dimensionality reduction and normalization to obtain the aggregate texture feature set.

4. The method for optimizing lightweight wall panels based on recycled aggregates as described in claim 3, characterized in that, Prior to utilizing the pre-trained aggregate texture recognition network, the method further includes: Based on a pre-built database of building materials and a deep learning model, an aggregate texture recognition model is obtained. A set of simulated images of microstructure based on aggregate surface properties is synthesized using a pre-built virtual generation module; The aggregate texture recognition model is fine-tuned and trained using the microstructure simulation image set to obtain the trained aggregate texture recognition network.

5. The method for optimizing lightweight wall panels based on recycled aggregates as described in claim 1, characterized in that, Based on the pre-constructed set of historical experimental data, performance simulations are performed on each formulation in the wall panel formulation set within a preset period to obtain a formulation-performance mapping set, including: Based on the aforementioned historical experimental data set, a simulation model for the performance evolution of the wall panel was constructed. Feature analysis is performed on each formula in the wall panel formula set to obtain a set of formula composition sequences; Using the wall panel performance evolution simulation model, a performance evolution simulation operation based on mechanical properties and durability is performed on each formula composition sequence in the formula composition sequence set to obtain a set of performance evolution curves; According to a preset period, the set of performance evolution curves is truncated to obtain a formula-performance mapping set.

6. The method for optimizing lightweight wall panels based on recycled aggregates as described in claim 1, characterized in that, The method of using a pre-built dynamic process path planning module to perform path optimization on the target wall panel formulation based on process parameters and performance feedback to obtain the optimal process path includes: Using a pre-built dynamic process path planning module, a basic process state is constructed based on the preset sensor monitoring type. The basic process state is then extended according to the preset data processing rules to obtain an extended process state. The basic process state and the extended process state are then standardized to obtain a process state space. Obtain the production process flow, and construct a process action space based on the production process flow, wherein the process action space includes stirring action, pressing and molding action and curing action; The mechanical strength, thermal conductivity, and preset action efficiency are configured as core indicators, and an evaluation function based on the core indicators is obtained to obtain a preset optimization strategy. Based on the evaluation function and optimization strategy, path scoring and identification are performed on the process state space and process action space to obtain path scoring results. The path with the highest score in the path scoring results is then selected to obtain the optimal process path.

7. The method for optimizing lightweight wall panels based on recycled aggregates as described in claim 4, characterized in that, After obtaining the target lightweight wall panel, the method further includes: The target lightweight wall panel was subjected to performance testing, and the performance test results were obtained. Determine whether the performance test results meet the preset pass / fail criteria; When the performance test results do not meet the qualification criteria, an updated set of microstructure simulation images is generated according to the virtual generation module. The aggregate texture recognition network is then optimized using the updated set of microstructure simulation images to obtain an optimized aggregate texture recognition network. The optimized aggregate texture recognition network is used to obtain an updated comprehensive feature set, and the process of obtaining the optimal process path is optimized using the updated comprehensive feature set.

8. The method for optimizing lightweight wall panels based on recycled aggregates as described in claim 1, characterized in that, After obtaining the target lightweight wall panel, the method further includes: The target lightweight wall panel is surface-sprayed using a pre-constructed microbial mineralization technology. When cracks appear in the target lightweight wall panel, the microbial mineralization technology is used to induce the formation of calcium carbonate deposits within the cracks; The crack was initially repaired using the calcium carbonate deposition.

9. A lightweight wall panel optimization system based on recycled aggregate from slag, used to implement the method as described in any one of claims 1-8, characterized in that, The system includes: The raw material observation module is used to acquire basic raw materials based on recycled aggregates from slag, acquire particle distribution characteristics and surface morphology information of the basic raw materials, acquire microstructure images of the basic raw materials, extract texture features from the microstructure images to obtain aggregate texture feature set, quantize and stitch the particle distribution characteristics and surface morphology information to obtain particle morphology feature set, and merge the aggregate texture feature set and particle morphology feature set to obtain comprehensive feature set. The formulation optimization module is used to utilize a pre-built material co-optimization module and material database to perform multi-objective constraint optimization operations on the basic raw materials according to the comprehensive feature set, thereby obtaining a set of wall panel formulations. Combined with a pre-built set of historical experimental data, the module performs performance simulations on each formulation in the set of wall panel formulations within a preset period to obtain a formulation-performance mapping set. Based on the formulation-performance mapping set, the module selects the wall panel formulation with the best comprehensive performance from the set of wall panel formulations to obtain the target wall panel formulation. The process path optimization module is used to perform path optimization operations on the target wall panel formula based on process parameters and performance feedback using a pre-built dynamic process path planning module to obtain the optimal process path. The wall panel preparation module is used to prepare the target wall panel formula according to the optimal process path to obtain the target lightweight wall panel.

Images