Building material detection and acceptance method and system

Abstract

The invention provides a building material detection and acceptance method and system, and the method comprises the steps: receiving related data of a to-be-detected building material; based on the data and a preset multi-dimensional acceptance standard, a virtual environment is constructed, parameters of the virtual environment are optimized, and it is ensured that simulation conditions are close to an actual use scene; in the optimized virtual environment, stress distribution of the material is calculated through finite element analysis, the degradation rate is predicted in combination with a machine learning algorithm, and a degradation rate predicted value is obtained; according to the predicted value, relevant data are adjusted through a multi-scale modeling technology and a random event triggering mechanism, and a response data set is determined; a support vector machine classifier is used for evaluating whether the response data set meets the acceptance standard or not, and an improvement suggestion report is generated for the non-conformity condition; and comprehensively improving the suggestion report, the response data set and the evaluation result, and generating a non-tampering detection report by means of a block chain technology. According to the invention, efficient and accurate detection and quality control of the building material are realized.

Description

Technical Field

[0001] The embodiments of the present application relate to the technical field of building material inspection and acceptance, and in particular, to a method and system for inspecting and accepting building materials. Background Art

[0002] In the modern construction industry, the quality and performance of building materials are crucial for the long-term stability and safety of buildings. With the continuous improvement of building standards, higher requirements are put forward for the inspection and acceptance of building materials. Specific application scenarios include, but are not limited to, the research and development of new building materials, the quality control of existing materials, and the acceptance of building engineering projects. In these scenarios, a technical means capable of comprehensively evaluating the physical durability, chemical stability, and ecological impact of building materials is required to ensure the reliability and environmental friendliness of materials under various actual use conditions. In addition, the ability to predict the deterioration rate of materials is also needed to take preventive measures in advance and extend the service life of materials.

[0003] Currently, the inspection and acceptance of building materials mainly rely on laboratory tests and on-site sampling inspections. Traditional methods usually include the following steps: First, collect relevant data of the building materials to be inspected, such as identification information and expected performance parameters; second, conduct a series of physical and chemical experiments according to industry standards to evaluate the basic properties of the materials; finally, judge whether the materials meet the acceptance standards according to the experimental results. Some advanced detection systems have begun to introduce computer simulation technology to preliminarily simulate the performance of materials in a specific environment and calculate the stress distribution using finite element analysis.

[0004] The existing solutions have many limitations in practical applications and are difficult to meet the current requirements for efficient and accurate detection of building materials. First, traditional experimental methods can only reflect the static performance of materials under specific conditions and cannot simulate complex and variable actual use environments, especially the influence of long-term dynamic changes, which may lead to deviations between the detection results and the actual situation. Second, laboratory tests are time-consuming, costly, and cannot achieve large-scale rapid screening, restricting their wide application in industrial production. In addition, the existing methods have insufficient ability to predict the deterioration rate of materials. Even if there are relevant studies, they are mostly based on historical data statistics and do not fully utilize advanced technologies such as machine learning to provide accurate predictions. Finally, the data management and storage methods of inspection reports are relatively traditional, prone to tampering or loss, and lack transparency and credibility. Summary of the Invention

[0005] The embodiments of the present application provide a method and system for inspecting and accepting building materials to solve the problems of inefficient and error-prone detection and quality control of building materials in the prior art.

[0006] In a first aspect, an embodiment of the present application provides a method for detecting and accepting building materials, including: receiving relevant data of the building materials to be inspected, where the relevant data includes identification information and expected performance parameters;

[0007] Based on the relevant data, combined with a preset multi-dimensional acceptance standard, a virtual environment is constructed, and the parameters of the virtual environment are adjusted to obtain an optimized virtual environment, where the acceptance standard includes physical durability, chemical stability, and ecological impact;

[0008] Based on the optimized virtual environment, the stress distribution of the building materials to be inspected is calculated using finite element analysis, and the deterioration rate of the building materials to be inspected is predicted by combining a machine learning algorithm to obtain a predicted deterioration rate value. Based on the predicted deterioration rate value, combined with multi-scale modeling technology and a random event triggering mechanism, the relevant data is adjusted to determine a response data set;

[0009] Based on the response data set, a support vector machine classifier is used to determine whether the relevant data meets the multi-dimensional acceptance standard. If it is shown that the relevant data does not meet the multi-dimensional acceptance standard, an improvement opinion report is generated;

[0010] Based on the improvement opinion report, the response data set, and the evaluation result, and using blockchain technology, a detection report is generated.

[0011] Optionally, the step of calculating the stress distribution of the building materials to be inspected using finite element analysis based on the optimized virtual environment, predicting the deterioration rate of the building materials to be inspected by combining a machine learning algorithm to obtain a predicted deterioration rate value, and adjusting the relevant data based on the predicted deterioration rate value by combining multi-scale modeling technology and a random event triggering mechanism to determine a response data set includes:

[0012] Using high-precision finite element analysis combined with preset load conditions to calculate the stress distribution of the relevant data to obtain a preliminary stress distribution map, identifying key areas with stress concentration from the preliminary stress distribution map, and using adaptive mesh refinement technology to adjust the key areas to generate a target stress distribution map;

[0013] Based on the target stress distribution map combined with a topology optimization algorithm, the optimal material distribution pattern is optimized from the structure of the building materials to be inspected. Based on the optimal material distribution pattern and the historical performance data of the building materials to be inspected, a deep neural network model is used to predict the deterioration rate of the building materials to be inspected, and combined with complex behaviors with nonlinearity and time dependence, a predicted deterioration rate value is obtained;

[0014] Using principal component analysis and cluster analysis, dimensionality reduction and classification are performed on the predicted deterioration rate values to identify key influencing factors and similar performance groups. Based on the key influencing factors and the similar performance groups, a multi-scale modeling technique is combined to evaluate the structural changes and performance impacts of the building materials to be inspected, and a comprehensive evaluation report is generated;

[0015] According to the random event trigger mechanism, preset external factors are simulated in the virtual environment to evaluate the performance of the building materials to be inspected, and a Monte Carlo simulation is used to generate a combination of external events to evaluate the response data of the building materials to be inspected. Combined with survival analysis, the life distribution of the building materials to be inspected is evaluated to generate an evaluation result of the impact of external factors;

[0016] Based on the stress distribution results, the comprehensive evaluation report, and the evaluation result of the impact of external factors, the response data set is determined.

[0017] Optionally, the using principal component analysis and cluster analysis, dimensionality reduction and classification are performed on the predicted deterioration rate values to identify key influencing factors and similar performance groups. Based on the key influencing factors and the similar performance groups, a multi-scale modeling technique is combined to evaluate the structural changes and performance impacts of the building materials to be inspected, and a comprehensive evaluation report is generated, including:

[0018] Using principal component analysis combined with sparse coding technology, dimensionality reduction processing is performed on the predicted deterioration rate values to extract the main feature vectors;

[0019] According to the main feature vectors, a cluster analysis algorithm is applied to divide the relevant data into multiple groups with similar performance characteristics, and an association rule mining technique is used to identify the key influencing factors of each group;

[0020] Based on the key influencing factors and the clustering results, a multi-scale modeling technique is used to evaluate the structural changes and performance impacts of the building materials to be inspected, and a comprehensive evaluation report is generated.

[0021] Optionally, the according to the main feature vectors, a cluster analysis algorithm is applied to divide the relevant data into multiple groups with similar performance characteristics, and an association rule mining technique is used to identify the key influencing factors of each group, including:

[0022] According to the main feature vectors, a cluster analysis algorithm is used to divide the relevant data into multiple groups with similar performance characteristics to obtain a preliminary clustering result;

[0023] The clustering quality of the preliminary clustering result is optimized by introducing a silhouette coefficient to evaluate different clustering numbers of the preliminary clustering result to select the optimal clustering number, and based on the optimal clustering number, a target clustering result is generated;

[0024] According to the target clustering results, using association rule mining technology, by mining the frequent patterns and association rules among different variables in the group, analyze the key data of each group to determine the key influencing factors of each group, where the key data includes various factors such as the historical performance data of building materials, stress distribution characteristics, and predicted values of deterioration rates.

[0025] Optionally, based on the target stress distribution map and combining with the topology optimization algorithm, optimize the optimal material distribution pattern from the structure of the building materials to be inspected. Based on the optimal material distribution pattern and the historical performance data of the building materials to be inspected, use a deep neural network model to predict the deterioration rate of the building materials to be inspected, and combine the complex behaviors of non-linearity and time-dependence to obtain the predicted value of the deterioration rate, including:

[0026] According to the target stress distribution map, combine with the multi-objective topology optimization algorithm, set multiple optimization objectives, define the design space and constraint conditions, and use genetic algorithms and particle swarm optimization techniques to obtain the optimal material distribution pattern;

[0027] Based on the optimal material distribution pattern, combine with the historical performance data of the building materials to be inspected, use a deep neural network model for training processing, and apply data cleaning and preprocessing techniques to obtain the predicted value of the deterioration rate.

[0028] Optionally, based on the relevant data and combining with the preset multi-dimensional acceptance criteria, construct a virtual environment, and adjust the parameters of the virtual environment to obtain an optimized virtual environment, including:

[0029] Using computer simulation technology, based on the relevant data, combine with the preset multi-dimensional acceptance criteria, simulate the performance of the building materials to be inspected under actual use conditions to obtain an initial virtual environment;

[0030] Using the Bayesian optimization algorithm, adjust the parameters of the initial virtual environment to analyze the optimal parameter combination, and based on the optimal parameter combination, obtain a preliminary optimization result;

[0031] Based on the preliminary optimization result, apply sensitivity analysis to identify the key parameters of the performance in the building materials to be inspected, and generate an optimized parameter list;

[0032] According to the optimized parameter list, adjust the parameter settings of the initial virtual environment again to obtain an optimized virtual environment.

[0033] Optionally, based on the response data set, use a support vector machine classifier to determine whether the relevant data meets the multi-dimensional acceptance criteria. If it is shown that the relevant data does not meet the multi-dimensional acceptance criteria, generate an improvement opinion report, including:

[0034] Based on the response data set, use a support vector machine classifier to classify the relevant data to obtain a classification result. According to the classification result, define the decision boundary corresponding to the multi-dimensional acceptance criteria, and generate a preliminary evaluation result;

[0035] Use a fuzzy logic system to process the non-deterministic test results in the response data set to determine the uncertainty of the response data set. Based on the preliminary evaluation result, determine whether the relevant data meets the multi-dimensional acceptance criteria, and for the marginal cases and the uncertainty in the preliminary evaluation result, enable a deep belief network for fine-grained analysis to obtain a target evaluation result;

[0036] If the target evaluation result shows that the relevant data does not meet the multi-dimensional acceptance criteria, based on the response data set and the target evaluation result, analyze the reasons for the non-conformity of the relevant data, and identify the factors that cause the non-conformity of the relevant data, and generate a non-conformity cause analysis report;

[0037] Use a genetic algorithm to analyze the optimization path from the non-conformity cause analysis report to obtain the best practice plan for the performance of the building materials to be inspected, and generate an optimization path recommendation;

[0038] Based on the optimization path recommendation, generate an improvement opinion report according to the suggestions preset in the expert knowledge base.

[0039] In a second aspect, an embodiment of the present application provides a detection and acceptance system for building materials, including:

[0040] A receiving module, configured to receive relevant data of building materials to be inspected, where the relevant data includes identification information and expected performance parameters;

[0041] A construction module, configured to construct a virtual environment based on the relevant data in combination with a preset multi-dimensional acceptance criteria, and adjust the parameters of the virtual environment to obtain an optimized virtual environment, where the acceptance criteria includes physical durability, chemical stability, and ecological impact;

[0042] A calculation module, configured to, based on the optimized virtual environment, use finite element analysis to calculate the stress distribution of the building materials to be inspected, and combine a machine learning algorithm to predict the deterioration rate of the building materials to be inspected to obtain a deterioration rate prediction value. Based on the deterioration rate prediction value, combine multi-scale modeling technology and a random event triggering mechanism to adjust the relevant data to determine a response data set;

[0043] A judgment module, configured to judge whether the relevant data meets the multi-dimensional acceptance criteria based on the response data set by using a support vector machine classifier, and generate an improvement opinion report if it is shown that the relevant data does not meet the multi-dimensional acceptance criteria;

[0044] A generation module, configured to generate an inspection report based on the improvement opinion report, the response data set and the evaluation result by using blockchain technology.

[0045] In a third aspect, an embodiment of the present application provides a computing device, including a processor and a memory, where a computer program is stored in the memory, and the processor is configured to run the computer program to execute the inspection and acceptance method for a building material according to any one of the first aspects.

[0046] In a fourth aspect, an embodiment of the present application provides a computer storage medium, on which computer program instructions are stored, and when the computer program instructions are executed by a processor, the inspection and acceptance method for a building material according to any one of the first aspects is implemented.

[0047] In the embodiment of the present application, relevant data of a building material to be inspected is received, where the relevant data includes identification information and expected performance parameters; based on the relevant data, a virtual environment is constructed in combination with a preset multi-dimensional acceptance criterion, and the parameters of the virtual environment are adjusted to obtain an optimized virtual environment, where the acceptance criterion includes physical durability, chemical stability and ecological impact; based on the optimized virtual environment, the stress distribution of the building material to be inspected is calculated by using finite element analysis, and the deterioration rate of the building material to be inspected is predicted by combining a machine learning algorithm to obtain a predicted deterioration rate value, and based on the predicted deterioration rate value, the relevant data is adjusted by combining multi-scale modeling technology and a random event triggering mechanism to determine a response data set; based on the response data set, it is judged whether the relevant data meets the multi-dimensional acceptance criteria by using a support vector machine classifier, and if it is shown that the relevant data does not meet the multi-dimensional acceptance criteria, an improvement opinion report is generated; based on the improvement opinion report, the response data set and the evaluation result, and by using blockchain technology, an inspection report is generated.

[0048] The technical solution of the present application has the following beneficial effects:

[0049] This application utilizes virtual environment simulation and finite element analysis to accurately simulate the performance of building materials under different conditions, improving the accuracy and reliability of detection. By combining machine learning algorithms to predict the material degradation rate and identify potential problems in advance, it helps with preventive maintenance and extends the service life of materials. Through multi-scale modeling techniques and a random event trigger mechanism to adjust data, it provides a scientific basis for the optimal design of materials. Using blockchain technology to generate inspection reports ensures the integrity and immutability of data, enhancing trust. The systematic workflow reduces manual intervention, improves work efficiency, and lowers detection costs. Automatically generating improvement opinion reports helps manufacturers continuously optimize material formulations and production processes, promoting continuous improvement in product quality.

[0050] Furthermore, the embodiments of this application also calculate the stress distribution of the building materials to be inspected through high-precision finite element analysis combined with preset load conditions, generate a preliminary stress distribution map, and use adaptive mesh refinement technology to optimize key areas to form a target stress distribution map. Based on this map, a topology optimization algorithm is used to optimize the optimal material distribution pattern from the material structure, and combined with a deep neural network model to predict the degradation rate, considering the complex behaviors of nonlinearity and time-dependence, to obtain the predicted degradation rate value. Further, principal component analysis and cluster analysis are used to perform dimensionality reduction classification on the predicted values, identify key influencing factors and similar performance groups, evaluate the material structure changes and performance impacts, and generate a comprehensive evaluation report. At the same time, the influence of external factors is simulated through a random event trigger mechanism, a combination of external events is generated using Monte Carlo simulation, the material response data and life distribution are evaluated, and finally, the response data set is determined by combining the stress distribution results, the comprehensive evaluation report, and the evaluation results of external factor influences.

[0051] Through the above methods, the scientific nature and reliability of building material detection are improved, which is specifically reflected in the following aspects: First, by combining high-precision finite element analysis and adaptive mesh refinement technology, the accuracy of stress distribution calculation is ensured, especially the meticulous treatment of stress concentration areas, improving the accuracy of detection results; Second, using the topology optimization algorithm and deep neural network model to predict the degradation rate not only considers the static properties of materials but also introduces complex behaviors of nonlinearity and time-dependence, enhancing the prediction ability; Third, the application of principal component analysis and cluster analysis enables the effective identification of key influencing factors and similar performance groups, providing a solid foundation for multi-scale modeling, and thus more accurately evaluating the material structure changes and performance impacts; Fourth, the application of the random event trigger mechanism and Monte Carlo simulation enables the comprehensive evaluation of the influence of external factors on material performance in a virtual environment, and combining survival analysis to evaluate the material life distribution, ensuring the comprehensiveness and credibility of the evaluation results. In summary, this method realizes the efficient management of the entire process from performance evaluation to quality control by integrating a variety of advanced technologies, providing strong support for the quality assurance of building materials.

[0052] These aspects or other aspects of the present application will be more clearly understood in the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following briefly introduces the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0054] Figure 1 It is a flowchart of a method for detecting and accepting building materials provided by an embodiment of the present application;

[0055] Figure 2 It is a schematic structural diagram of a system for detecting and accepting building materials provided by an embodiment of the present application;

[0056] Figure 3 It is a schematic structural diagram of a computing device provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0057] In order to enable those skilled in the art to better understand the solution of the present application, the following clearly and completely describes the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application.

[0058] In some processes described in the specification and claims of the present application and the above-mentioned drawings, a plurality of operations appear in a specific order. However, it should be clearly understood that these operations may not be executed in the order in which they appear herein or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions such as "first" and "second" in this article are used to distinguish different messages, devices, modules, etc., and do not represent a sequence, nor do they limit that "first" and "second" are of different types.

[0059] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, rather than all embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present application.

[0060] Figure 1The figure below is a flowchart of a method for detecting and accepting building materials provided by an embodiment of the present application. As Figure 1 shown, the method includes:

[0061] Step 101: Receive relevant data of the building materials to be inspected;

[0062] In this step, the relevant data includes identification information and expected performance parameters; these data include, but are not limited to, identification information (such as material name, batch number, manufacturer information, etc.) and expected performance parameters (such as compressive strength, tensile strength, chemical composition, ecological impact assessment, etc.). These data are the basis for subsequent analysis, used to ensure the uniqueness and traceability of each sample, and provide the necessary input for constructing a virtual environment. The identification information ensures the identification of the material throughout the detection process, while the expected performance parameters provide data on the expected performance of the material under different conditions, facilitating subsequent comparison and verification.

[0063] In actual operation, the system first preprocesses the received data, including formatting, cleaning, and preliminary verification, to ensure the integrity and accuracy of the data. Then, these data are stored in a secure and easily accessible database for use in subsequent steps. For each batch of building materials, the system automatically assigns a unique identifier to track its status and results throughout the detection process.

[0064] For example, at a construction site or material supplier, quality inspection personnel can use a mobile device to scan the QR code or RFID tag of the building materials, quickly obtain and upload the relevant data to the cloud server. These data are then transmitted to the central detection system and, after preliminary processing, are used to construct a virtual environment and initiate the simulation and analysis of the next step. This not only improves work efficiency but also reduces the possibility of manual input errors.

[0065] Step 102: Based on the relevant data, combine with preset multi-dimensional acceptance criteria to construct a virtual environment, and adjust the parameters of the virtual environment to obtain an optimized virtual environment;

[0066] In this step, the acceptance criteria include physical durability, chemical stability, and ecological impact; this virtual environment not only simulates the behavior of materials under static conditions but also includes dynamic changes and long-term effects, thus more comprehensively evaluating the performance of materials. By adjusting the parameters of the virtual environment, an optimized virtual environment can be obtained to ensure that the simulation conditions are close to the actual use scenario and improve the accuracy of prediction.

[0067] In actual operation, the system automatically generates an initial virtual environment based on the received building material data and the preset multi-dimensional acceptance criteria. Then, through a series of algorithms, the parameter settings of the virtual environment are optimized to make it more in line with the actual usage conditions. For example, factors such as temperature, humidity, and stress level are adjusted to simulate different application scenarios. The optimized virtual environment will serve as the basis for subsequent finite element analysis and machine learning prediction.

[0068] For example, assume that a new type of concrete material needs to be tested for its durability under different climate conditions. The system can construct multiple virtual environments based on the historical data and expected performance parameters of the material, simulating the climate conditions in dry and hot regions, humid and cold regions, and temperate regions respectively. By adjusting the temperature, humidity, and other parameters in each virtual environment, it is ensured that the simulation conditions are as close as possible to the actual usage situation, so as to accurately evaluate the performance of the material in various environments.

[0069] Step 103: Based on the optimized virtual environment, use finite element analysis to calculate the stress distribution of the building material to be inspected, and combine machine learning algorithms to predict the deterioration rate of the building material to be inspected, obtaining a predicted value of the deterioration rate. Based on the predicted value of the deterioration rate, combine multi-scale modeling techniques and a random event triggering mechanism to adjust the relevant data to determine the response data set;

[0070] In this step, finite element analysis is a numerical method that solves complex engineering problems by decomposing a complex structure into many small elements. Combining machine learning algorithms to predict the deterioration rate of the material, a predicted value of the deterioration rate is obtained. In addition, by adjusting the relevant data through multi-scale modeling techniques and a random event triggering mechanism to determine the response data set, this process comprehensively considers the microscopic and macroscopic characteristics of the material and its responses under different external factors.

[0071] In actual operation, first, the system uses high-precision finite element analysis combined with preset load conditions to calculate the stress distribution of the material, generating a preliminary stress distribution map. Then, the stress concentration areas are identified, and adaptive mesh refinement technology is used to optimize these key areas to form a target stress distribution map. Based on this map, a topology optimization algorithm is used to optimize the optimal material distribution pattern from the material structure. Then, combined with a deep neural network model to predict the deterioration rate, considering the complex behaviors of non-linearity and time-dependence, a predicted value of the deterioration rate is obtained. Finally, principal component analysis and clustering analysis are used to reduce the dimension and classify the predicted values to identify key influencing factors and similar performance groups, and to evaluate the structural changes and performance impacts of the material.

[0072] For example, continuing with the above example of the new concrete material, the system performs finite element analysis in the optimized virtual environment to calculate the stress distribution of the concrete under different loading conditions. The analysis results show that there are high stress concentrations in certain areas. The system then applies the adaptive mesh refinement technique to optimize these areas to ensure the accuracy of the stress distribution. Next, the internal structure of the concrete is optimized through the topology optimization algorithm to reduce potential stress hotspots. At the same time, a deep neural network model is used to predict the deterioration rate of the concrete over different service years, considering complex behaviors such as hygrothermal cycles and freeze-thaw actions that are non-linear and time-dependent. Finally, principal component analysis and clustering analysis are used to identify the key influencing factors, evaluate the structural changes and performance impacts of the concrete, and provide a basis for subsequent design improvements.

[0073] Step 104: Based on the response data set, use a support vector machine classifier to determine whether the relevant data meets the multi-dimensional acceptance criteria. If it is shown that the relevant data does not meet the multi-dimensional acceptance criteria, generate an improvement opinion report.

[0074] In this step, a support vector machine (SVM) classifier is used to evaluate whether the building materials to be inspected meet the preset multi-dimensional acceptance criteria based on the response data set. SVM is a supervised learning model suitable for classification and regression analysis, especially good at dealing with data in high-dimensional spaces. If the test results show that the materials do not meet the acceptance criteria, the system will generate an improvement opinion report, putting forward specific improvement suggestions to help manufacturers optimize the production process and formula.

[0075] In actual operation, the system inputs the response data set generated in the previous step into the trained SVM classifier for classification and judgment. The classifier evaluates whether the performance indicators of the materials meet the standards according to the preset multi-dimensional acceptance criteria. If any indicator is found not to meet the requirements, the system will record the specific problems and generate a detailed improvement opinion report. This report not only points out the non-conforming points but also proposes possible solutions and technical guidance to help manufacturers take actions quickly.

[0076] For example, continuing with the previous example, when the predicted value of the deterioration rate of the concrete material shows that it may not reach the expected service life under specific climatic conditions, the SVM classifier will determine that the material does not meet the acceptance criteria. The system will then generate an improvement opinion report, suggesting that the manufacturer adjust the mix ratio of the concrete, add certain additives to improve its durability, or improve the curing process to enhance the overall performance of the material. This report not only helps the manufacturer understand the problem but also provides specific improvement suggestions, promoting the continuous improvement of product quality.

[0077] Step 105: Based on the improvement opinion report, the response data set, and the evaluation result, and using blockchain technology, generate a detection report.

[0078] In this step, blockchain, as a distributed ledger technology, ensures the transparency, security, and immutability of data. The generated detection report not only includes all the detection results of the materials but also records the improvement opinions and follow-up actions, providing users with a comprehensive and trustworthy quality assurance document.

[0079] In actual operation, the system aggregates all relevant information (including the improvement opinion report, the response data set, and the evaluation result) and encrypts and stores it through blockchain technology. Each block contains a set of transaction records, which are part of the content of the detection report, ensuring the authenticity and integrity of the data. Users can view and verify these reports through a blockchain browser without worrying about data tampering or loss. In addition, blockchain technology can also enable information sharing between upstream and downstream of the supply chain, promoting collaborative work.

[0080] For example, in the above concrete material case, the detection report generated by the system not only details all the detection results of the materials but also includes specific improvement suggestions for non-conforming items. This report is encrypted and stored through blockchain technology to ensure its security and immutability. Construction companies, supervision units, and regulatory departments can all view this report through a blockchain browser to confirm the quality status of the materials. In addition, the manufacturer can quickly adjust the production process according to the suggestions in the report to ensure that the next batch of products meets all acceptance standards. This not only improves the material quality but also enhances the trust among all parties.

[0081] Through the above five steps, this method realizes the efficient management of the entire process from receiving the relevant data of the building materials to be inspected to generating an immutable detection report. It not only improves the scientificity and reliability of building material inspection but also ensures the comprehensiveness and credibility of the inspection results by integrating multiple advanced technologies, such as high-precision finite element analysis, machine learning, principal component analysis, clustering analysis, support vector machine classifier, and blockchain technology. This method significantly enhances the quality control ability of building materials, promotes the continuous improvement of product quality, and at the same time improves the efficiency and transparency of the entire construction industry.

[0082] To address the limitations of traditional stress distribution calculation and deterioration rate prediction methods, in some embodiments, in step 103, based on the optimized virtual environment, use finite element analysis to calculate the stress distribution of the building materials to be inspected, and combine machine learning algorithms to predict the deterioration rate of the building materials to be inspected to obtain a deterioration rate prediction value. Based on the deterioration rate prediction value, combine multi-scale modeling technology and a random event triggering mechanism to adjust the relevant data to determine the response data set, including:

[0083] Using high-precision finite element analysis combined with preset load conditions, calculate the stress distribution of the relevant data to obtain a preliminary stress distribution map, identify the key areas of stress concentration from the preliminary stress distribution map, and use the adaptive mesh refinement technique to adjust the key areas to generate a target stress distribution map; based on the target stress distribution map combined with the topology optimization algorithm, optimize the optimal material distribution pattern from the structure of the building material to be inspected, and based on the optimal material distribution pattern and the historical performance data of the building material to be inspected, use a deep neural network model to predict the deterioration rate of the building material to be inspected, and combine the complex behaviors of non-linearity and time-dependence to obtain a deterioration rate prediction value; use principal component analysis and cluster analysis to reduce the dimension and classify the deterioration rate prediction value to identify the key influencing factors and similar performance groups, and based on the key influencing factors and the similar performance groups combined with multi-scale modeling techniques, evaluate the structural changes and performance impacts of the building material to be inspected to generate a comprehensive evaluation report; according to the random event trigger mechanism, simulate the preset external factors in the virtual environment, evaluate the performance of the building material to be inspected, and use Monte Carlo simulation to generate a combination of external events, evaluate the response data of the building material to be inspected, and combine survival analysis to evaluate the life distribution of the building material to be inspected to generate an external factor impact evaluation result; based on the stress distribution result, the comprehensive evaluation report, and the external factor impact evaluation result, determine the response data set.

[0084] In this embodiment, using high-precision finite element analysis combined with preset load conditions, calculate the stress distribution of the relevant data. Here, "high-precision finite element analysis" refers to an advanced numerical simulation method that can accurately calculate the stress distribution in complex structures. By dividing the material into multiple small units (i.e., finite elements), the behavior of each unit under different load conditions can be detailedly simulated, thereby generating a preliminary stress distribution map. Identify the key areas of stress concentration from the preliminary stress distribution map, and use the adaptive mesh refinement technique to adjust the key areas to generate a target stress distribution map. The adaptive mesh refinement technique can increase the mesh density where higher resolution is required to ensure more accurate simulation of the stress concentration area.

[0085] In the embodiments of the present application, first, the system calculates the stress distribution of the material by using high-precision finite element analysis combined with preset load conditions, and generates a preliminary stress distribution map. Then, the stress concentration area is optimized through the adaptive mesh refinement technology to form a target stress distribution map. Next, based on this map, a topology optimization algorithm is used to optimize the optimal material distribution pattern from the material structure to ensure the rationality and strength of the material structure. Subsequently, the degradation rate is predicted by combining a deep neural network model, considering the complex behaviors of nonlinearity and time-dependence, and a degradation rate prediction value is obtained. The deep neural network model can more accurately predict the degradation of the material during long-term use by learning a large amount of historical performance data. Finally, principal component analysis and cluster analysis are used to perform dimensionality reduction classification on the prediction values, identify key influencing factors and similar performance groups, evaluate the structural changes and performance impacts of the material, and generate a comprehensive evaluation report. In addition, according to the random event trigger mechanism, preset external factors are simulated in a virtual environment to evaluate the material performance, and a Monte Carlo simulation is used to generate a combination of external events. Combined with survival analysis, the material life distribution is evaluated to generate an evaluation result of the impact of external factors. Finally, based on all the above results, a response data set is determined.

[0086] The following is a specific embodiment:

[0087] Suppose a new type of high-strength concrete material needs to be comprehensively tested. The system first performs high-precision finite element analysis in an optimized virtual environment to simulate the stress distribution of the concrete under different loading conditions. The analysis results show that there are high stress concentrations in some areas, and the system immediately applies the adaptive mesh refinement technology to optimize these areas to ensure the accuracy of the stress distribution. Then, the internal structure of the concrete is optimized through a topology optimization algorithm to reduce potential stress hotspots. At the same time, a deep neural network model is used to predict the degradation rate of the concrete over different service years, considering complex behaviors such as hygrothermal cycling and freeze-thaw action that are nonlinear and time-dependent. The prediction results show that the concrete may experience accelerated degradation under specific climatic conditions. To further evaluate the impact of this degradation on the structure, the system uses principal component analysis and cluster analysis to identify the key influencing factors and finds that humidity and temperature fluctuations are the main inducements. Subsequently, the system simulates various extreme weather conditions in the virtual environment and generates a combination of external events through Monte Carlo simulation to evaluate the response data of the concrete under these conditions. Combining survival analysis to evaluate the life distribution of the concrete, the results show that under specific conditions, the expected life of the concrete may be shortened. Finally, the system integrates all the information to generate a detailed comprehensive evaluation report and an evaluation result of the impact of external factors, helping the manufacturer understand the problems and propose improvement measures to ensure that the next batch of products meet all acceptance criteria.

[0088] To address the limitations of traditional dimensionality reduction and classification methods when dealing with complex data, in some embodiments, in step 103, using principal component analysis and clustering analysis to perform dimensionality reduction and classification on the predicted deterioration rate values to identify key influencing factors and similar performance groups, and based on the key influencing factors and the similar performance groups, combining with multi-scale modeling techniques to evaluate the structural changes and performance impacts of the building materials to be inspected and generate a comprehensive evaluation report, further includes:

[0089] Using principal component analysis combined with sparse coding technology to perform dimensionality reduction processing on the predicted deterioration rate values to extract the main feature vectors; according to the main feature vectors, applying a clustering analysis algorithm to divide the relevant data into multiple groups with similar performance characteristics, and using association rule mining technology to identify the key influencing factors of each group; based on the key influencing factors and the clustering results, using multi-scale modeling techniques to evaluate the structural changes and performance impacts of the building materials to be inspected and generate a comprehensive evaluation report. Optionally, the step of applying a clustering analysis algorithm to divide the relevant data into multiple groups with similar performance characteristics according to the main feature vectors and using association rule mining technology to identify the key influencing factors of each group includes: according to the main feature vectors, using a clustering analysis algorithm to divide the relevant data into multiple groups with similar performance characteristics to obtain a preliminary clustering result; optimizing the clustering quality of the preliminary clustering result, evaluating different clustering numbers of the preliminary clustering result by introducing a silhouette coefficient to select the optimal clustering number, and generating a target clustering result based on the optimal clustering number; according to the target clustering result, using association rule mining technology to analyze the key data of each group by mining the frequent patterns and association rules between different variables in the group to determine the key influencing factors of each group, where the key data includes various factors such as the historical performance data of the building materials, stress distribution characteristics, and predicted deterioration rate values.

[0090] In this embodiment, using principal component analysis combined with sparse coding technology to perform dimensionality reduction processing on the predicted deterioration rate values to extract the main feature vectors. Principal component analysis is a statistical method used to transform high-dimensional data into low-dimensional data while retaining as much information as possible. Sparse coding technology further enhances the effect of dimensionality reduction. By learning a set of sparse representations, the dimensionality-reduced data is not only more concise but also better able to capture the important features in the data. According to the main feature vectors, applying a clustering analysis algorithm to divide the relevant data into multiple groups with similar performance characteristics, and using association rule mining technology to identify the key influencing factors of each group. This step helps to discover the potential relationships between different groups, providing a basis for subsequent evaluation.

[0091] In the embodiments of the present application, first, the system uses principal component analysis combined with sparse coding technology to perform dimensionality reduction processing on the predicted values of the deterioration rate, and extracts the main feature vectors, which represent the main change trends of the material deterioration rate. Then, according to the extracted main feature vectors, a clustering analysis algorithm is applied to divide the relevant data into multiple groups with similar performance characteristics. Next, in order to optimize the quality of the preliminary clustering results, the silhouette coefficient is introduced to evaluate the advantages and disadvantages of different numbers of clusters, and the optimal number of clusters is selected to generate the target clustering result. Finally, according to the target clustering result, the association rule mining technology is used to analyze the key data of each group by mining the frequent patterns and association rules between different variables in the group, so as to determine the key influencing factors of each group. These key data include the historical performance data of building materials, the stress distribution characteristics, and each factor in the predicted value of the deterioration rate, so as to comprehensively evaluate the structural changes and performance impacts of the materials and generate a comprehensive evaluation report.

[0092] The following is a specific example:

[0093] Suppose a new composite material needs to be tested for its deterioration under different environmental conditions. First, the system uses principal component analysis combined with sparse coding technology to perform dimensionality reduction processing on the predicted values of the deterioration rate and extracts the main feature vectors. For example, for a specific composite material, the system may extract humidity sensitivity, temperature sensitivity, and mechanical stress as the main feature vectors. Then, based on these main feature vectors, the system applies a clustering analysis algorithm to divide the relevant data into multiple groups with similar performance characteristics. For example, some groups may be particularly sensitive to humidity changes, while others are more affected by temperature fluctuations. To optimize the clustering quality, the system introduces the silhouette coefficient to evaluate the advantages and disadvantages of different numbers of clusters, and finally selects the optimal number of clusters to generate the target clustering result. Subsequently, based on the target clustering result, the system uses the association rule mining technology to discover the frequent patterns and association rules between different variables. For example, a certain group may show that the combined action of humidity and temperature will significantly accelerate the deterioration of the material. Finally, the system integrates all the information to generate a detailed comprehensive evaluation report to help the manufacturer understand the problem and propose improvement measures to ensure that the next batch of products meet all acceptance criteria.

[0094] In order to solve the limitations of traditional material structure optimization methods under complex conditions and to further improve the accuracy of deterioration rate prediction, in some embodiments, based on the target stress distribution map and the topology optimization algorithm in step 103, the optimal material distribution pattern is optimized from the structure of the building material to be inspected. Based on the optimal material distribution pattern and the historical performance data of the building material to be inspected, a deep neural network model is used to predict the deterioration rate of the building material to be inspected, and combined with the complex behaviors of non-linearity and time-dependence, the predicted value of the deterioration rate is obtained, and further includes:

[0095] According to the target stress distribution map, combined with the multi-objective topology optimization algorithm, multiple optimization objectives are set, and the design space and constraint conditions are defined. Using the genetic algorithm and particle swarm optimization technology, the optimal material distribution pattern is obtained; based on the optimal material distribution pattern, combined with the historical performance data of the building materials to be inspected, training processing is carried out using a deep neural network model, and data cleaning and preprocessing technologies are applied to obtain the predicted value of the deterioration rate.

[0096] In this embodiment, according to the target stress distribution map, combined with the multi-objective topology optimization algorithm, multiple optimization objectives are set, and the design space and constraint conditions are defined. Here, the "multi-objective topology optimization algorithm" refers to an algorithm that can optimize multiple performance indicators simultaneously. By setting multiple optimization objectives, it can be ensured that the optimized material distribution pattern can not only meet a single performance requirement but also take into account other important attributes. In addition, defining the design space and constraint conditions is to ensure the feasibility and practicality of the optimization results. For example, it is to avoid the material thickness exceeding the range allowed by the manufacturing process. Using the genetic algorithm and particle swarm optimization technology, the optimal material distribution pattern is obtained. The genetic algorithm is a global search algorithm that simulates natural selection and genetic mechanisms, while the particle swarm optimization technology imitates the behavior of bird flocks to find the optimal solution. The combination of the two can efficiently explore the design space and find the optimal material distribution pattern.

[0097] In the embodiment of the present application, first, the system sets multiple optimization objectives according to the target stress distribution map, combined with the multi-objective topology optimization algorithm, such as minimizing weight, maximizing strength and stiffness, etc., and defines the design space and constraint conditions to ensure the feasibility and practicality of the optimization results. Then, using the genetic algorithm and particle swarm optimization technology, global search is carried out within the design space, and the optimization process is gradually iterated until the optimal material distribution pattern is finally obtained. Next, based on this optimal material distribution pattern, combined with the historical performance data of the building materials to be inspected, training processing is carried out using a deep neural network model. During the training process, data cleaning and preprocessing technologies are applied to remove outliers and noise, ensuring the quality of the input data, thereby improving the prediction accuracy of the model. The deep neural network model can more accurately predict the deterioration of materials during long-term use through learning a large amount of historical performance data, especially combined with the complex behaviors of non-linearity and time-dependence, to obtain a more accurate predicted value of the deterioration rate.

[0098] The following is a specific example:

[0099] Suppose a new type of steel structure material needs to be comprehensively tested and optimized. First, based on the target stress distribution map generated in the optimized virtual environment and combined with the multi-objective topology optimization algorithm, the system sets the optimization objectives as minimizing weight, maximizing strength and stiffness, and defines the design space and constraint conditions to ensure that the optimization results meet the requirements of actual manufacturing processes. Then, the system uses genetic algorithms and particle swarm optimization techniques to conduct a global search within the design space, gradually iterating the optimization process, and finally obtaining the optimal material distribution pattern. For example, for some key parts, the optimization results show that increasing the local thickness can significantly improve the strength and stiffness of the overall structure. Next, based on this optimal material distribution pattern and combined with the historical performance data of the steel structure material (such as strength, stiffness, and corrosion conditions changing over time under different environmental conditions), the system uses a deep neural network model for training. During the training process, data cleaning and preprocessing techniques are applied to remove outliers and noise to ensure the quality of the input data. Through learning a large amount of historical performance data, the deep neural network model can more accurately predict the deterioration of the steel structure material during long-term use, especially in combination with complex behaviors such as hygrothermal cycling and freeze-thaw effects, to obtain more accurate deterioration rate prediction values. Finally, the system integrates all the information to generate a detailed comprehensive evaluation report to help manufacturers understand the problems and propose improvement measures to ensure that the next batch of products meet all acceptance criteria.

[0100] To address the limitations of traditional virtual environment construction methods in simulating actual usage conditions and to further improve the efficiency and accuracy of virtual environment parameter optimization, in some embodiments, constructing a virtual environment based on the relevant data in combination with a preset multi-dimensional acceptance standard and adjusting the parameters of the virtual environment to obtain an optimized virtual environment in step 102 includes:

[0101] Using computer simulation technology, based on the relevant data and in combination with a preset multi-dimensional acceptance standard, simulating the performance of the building material to be inspected under actual usage conditions to obtain an initial virtual environment; using the Bayesian optimization algorithm to adjust the parameters of the initial virtual environment to analyze the optimal parameter combination, and based on the optimal parameter combination, obtaining a preliminary optimization result; based on the preliminary optimization result, applying sensitivity analysis to identify the key parameters of the performance in the building material to be inspected and generating an optimized parameter list; according to the optimized parameter list, adjusting the parameter settings of the initial virtual environment again to obtain an optimized virtual environment.

[0102] In this embodiment, using computer simulation technology, based on the relevant data (such as identification information, expected performance parameters, etc.), combined with preset multi-dimensional acceptance criteria (such as physical durability, chemical stability, and ecological impact), the performance of the building material to be inspected under actual use conditions is simulated to obtain an initial virtual environment. Here, "computer simulation technology" refers to simulating the behavior of materials under different conditions through software tools to evaluate their performance. The simulation technology can simulate various factors such as temperature, humidity, and stress to ensure that the virtual environment is as close as possible to the actual use scenario. Next, using the Bayesian optimization algorithm, the parameters of the initial virtual environment are adjusted to analyze the optimal parameter combination. Based on the optimal parameter combination, a preliminary optimization result is obtained. Bayesian optimization is an efficient global optimization method, especially suitable for complex problems that require a large amount of computing resources. It constructs a probability model to predict the best parameter combination, reduces unnecessary computational effort, and improves the optimization efficiency. Based on the preliminary optimization result, sensitivity analysis is applied to identify the key parameters of the performance of the building material to be inspected and generate an optimized parameter list. Sensitivity analysis is used to evaluate the degree of influence of each parameter on the material performance, so as to find out the key parameters that have a significant impact on the performance. These key parameters will be given priority consideration to ensure that the optimized virtual environment is more in line with the actual requirements. Finally, according to the optimized parameter list, the parameter settings of the initial virtual environment are adjusted again to obtain an optimized virtual environment. This process ensures that the final virtual environment can not only accurately reflect the actual use situation of the material, but also highlight the influence of key parameters, providing a solid foundation for subsequent analysis.

[0103] In the embodiment of the present application, first, the system uses computer simulation technology to construct an initial virtual environment based on the received relevant data and preset multi-dimensional acceptance criteria, simulating the performance of the building material to be inspected under actual use conditions. Then, the system uses the Bayesian optimization algorithm to adjust the parameters of the initial virtual environment to find the optimal parameter combination and obtain a preliminary optimization result. On this basis, the system applies sensitivity analysis to identify the key parameters that have a significant impact on the material performance and generate an optimized parameter list. Finally, the system adjusts the parameter settings of the virtual environment again according to the optimized parameter list to obtain an optimized virtual environment, ensuring that its parameter configuration not only meets the acceptance criteria but also accurately reflects the performance of the material in actual use.

[0104] The following is a specific embodiment:

[0105] Suppose a new type of waterproof coating needs to be tested for its durability and environmental friendliness under different climate conditions. First, the system uses computer simulation technology to construct an initial virtual environment based on the identification information of the coating and the expected performance parameters (such as water resistance, UV resistance, volatile organic compound content, etc.), combined with preset multi-dimensional acceptance criteria (such as physical durability, chemical stability, and ecological impact), to simulate the performance of the coating under the climate conditions of dry and hot regions, humid and cold regions, and temperate regions. Then, the system uses the Bayesian optimization algorithm to adjust the parameters of the virtual environment (such as temperature, humidity, UV intensity, etc.) to find the optimal parameter combination and obtain a preliminary optimization result. For example, the optimization result shows that the durability of the coating decreases significantly in high-temperature and high-humidity environments. Based on this preliminary optimization result, the system applies sensitivity analysis to identify temperature and humidity as the key parameters affecting the durability of the coating and generates a list of optimized parameters. Finally, the system adjusts the parameter settings of the virtual environment again according to the list of optimized parameters to obtain an optimized virtual environment, ensuring that its parameter configuration can accurately reflect the performance of the coating in actual use and meet the multi-dimensional acceptance criteria. This not only improves the accuracy of the detection but also provides a scientific basis for the improvement of the coating formula.

[0106] In order to solve the limitations of traditional classification methods in dealing with complex and uncertain data and to further improve the accuracy and reliability of the detection results, in some embodiments, in step 104, based on the response data set, using a support vector machine classifier, it is determined whether the relevant data meets the multi-dimensional acceptance criteria. If it is shown that the relevant data does not meet the multi-dimensional acceptance criteria, an improvement opinion report is generated, including:

[0107] Based on the response data set, the relevant data is classified using a support vector machine classifier to obtain a classification result. According to the classification result, the decision boundary corresponding to the multi-dimensional acceptance criteria is defined to generate a preliminary evaluation result; the fuzzy logic system is used to process the non-deterministic test results in the response data set to determine the uncertainty of the response data set. Based on the preliminary evaluation result, it is determined whether the relevant data meets the multi-dimensional acceptance criteria, and for the marginal cases and the uncertainty in the preliminary evaluation result, a deep belief network is enabled for fine-grained analysis to obtain a target evaluation result; if the target evaluation result shows that the relevant data does not meet the multi-dimensional acceptance criteria, based on the response data set and the target evaluation result, the reasons for the non-compliance of the relevant data are analyzed, and the factors causing the non-compliance of the relevant data are identified to generate a non-compliance reason analysis report; using a genetic algorithm, the optimization path is analyzed from the non-compliance reason analysis report to obtain the best practice plan for the performance of the building material to be tested, and an optimization path suggestion is generated; based on the optimization path suggestion, according to the suggestions preset in the expert knowledge base, an improvement opinion report is generated.

[0108] In this embodiment, based on the response dataset, a support vector machine classifier is used to classify the relevant data to obtain a classification result. Here, the support vector machine is a supervised learning model suitable for classification problems in high-dimensional spaces. By training the SVM classifier, the system can define the decision boundary corresponding to the multi-dimensional acceptance criteria, thereby generating a preliminary evaluation result. Next, a fuzzy logic system is used to process the non-deterministic test results in the response dataset to determine the uncertain situations in the response dataset. The fuzzy logic system can handle data with uncertainty or ambiguity, ensuring that the evaluation result is more comprehensive and reliable. Based on the preliminary evaluation result, it is judged whether the relevant data meets the multi-dimensional acceptance criteria, and for the marginal cases and the uncertain situations in the preliminary evaluation result, a deep belief network is enabled for fine-grained analysis to obtain the target evaluation result. The deep belief network is a deep neural network structure that can provide more accurate predictions in complex situations.

[0109] In the embodiment of the present application, first, based on the response dataset, the system uses a support vector machine classifier to classify the relevant data to obtain a classification result, and defines the decision boundary corresponding to the multi-dimensional acceptance criteria according to the classification result to generate a preliminary evaluation result. Then, a fuzzy logic system is used to process the non-deterministic test results in the response dataset to identify the uncertain situations. Next, based on the preliminary evaluation result, the system judges whether the relevant data meets the multi-dimensional acceptance criteria, and for the marginal cases and the uncertain situations in the preliminary evaluation result, a deep belief network is enabled for fine-grained analysis, and finally the target evaluation result is obtained. If the target evaluation result shows that the relevant data does not meet the multi-dimensional acceptance criteria, the system will analyze the reasons for non-compliance of the relevant data based on the response dataset and the target evaluation result, identify the factors leading to non-compliance, and generate an analysis report on the reasons for non-compliance. Subsequently, a genetic algorithm is used to analyze the optimization path from the analysis report on the reasons for non-compliance, obtain the best practice solution for the performance of the building materials to be inspected, and generate a suggestion on the optimization path. Finally, based on the suggestion on the optimization path and combined with the suggestions preset in the expert knowledge base, an improvement opinion report is generated to help the manufacturer adjust the production process and formula and improve the product quality.

[0110] The following is a specific embodiment:

[0111] Suppose a new type of environmentally friendly brick needs to be comprehensively tested to ensure that it meets multi-dimensional acceptance criteria (such as physical durability, chemical stability, and ecological impact). First, based on the response dataset, the system uses a support vector machine classifier to classify the relevant data, obtains the classification results, and defines the decision boundaries corresponding to the multi-dimensional acceptance criteria according to the classification results to generate a preliminary evaluation result. For example, some bricks show abnormal fluctuations in the compressive strength test. The system uses a fuzzy logic system to process these non-deterministic test results and identify uncertain situations. Based on the preliminary evaluation results, the system determines whether these bricks meet the acceptance criteria, and for the marginal cases (such as data points close to the upper or lower limits of the acceptance criteria) and uncertain situations in the preliminary evaluation results, a deep belief network is enabled for fine-grained analysis to finally obtain the target evaluation result. The results show that some bricks do not meet the physical durability standard. Immediately, based on the response dataset and the target evaluation result, the system analyzes the reasons for non-compliance, finds that the firing temperature and time are the main influencing factors, and generates an analysis report on the reasons for non-compliance. Next, the system uses a genetic algorithm to analyze the optimization path from the analysis report on the reasons for non-compliance, recommends adjusting the firing process parameters, obtains the best practice plan, and generates a recommendation on the optimization path. Finally, based on the recommendation on the optimization path and combined with the suggestions preset in the expert knowledge base, the system generates an improvement opinion report to guide the manufacturer to optimize the production process and ensure that the next batch of products meets all acceptance criteria.

[0112] This application considers that in the detection and classification of building materials, traditional clustering methods often fail to effectively handle the complexity and non-linear relationships of high-dimensional data, resulting in inaccurate classification results and limitations in identifying the key influencing factors of different groups, making it difficult to provide targeted improvement suggestions. Therefore, a new alternative solution is proposed, which includes:

[0113] According to the main feature vector, applying a clustering analysis algorithm to divide the relevant data into multiple groups with similar performance characteristics, and using association rule mining technology to identify the key influencing factors of each group, including:

[0114] According to the main feature vector, using a clustering analysis algorithm to classify the building material-related data, dividing the relevant data into multiple groups with similar performance characteristics to obtain a preliminary clustering result, where d(x i ,x j ) is used to calculate the distance metric between the relevant data and is the basis of the clustering analysis. The calculation formula is:

[0115]

[0116] where d(x i ,x j ) represents the relevant data x iand x j The distance metric between them, n is the dimension of the feature vector, x ik and x jk are the eigenvalues of the relevant data x i and x j on the k-th dimension respectively, w k is the weight of the k-th dimensional feature, λ k is the decay factor, represents the exponential decay function;

[0117] The following is a detailed explanation of each parameter:

[0118] d(x i , x j ) represents the distance metric between the relevant data x i and x j . It is obtained through the calculation formula and is used to measure the similarity or difference between two samples.

[0119] n represents the dimension of the feature vector, that is, the number of features each sample has. It is determined according to the number of features contained in each sample in the actual application scenario. For example, building materials may include features such as compressive strength, tensile strength, chemical composition, etc.

[0120] x ik and x jk are the eigenvalues of the samples x i and x j on the k-th dimension respectively. They are directly extracted from the sample data, such as experimental measurements or historical records.

[0121] w k represents the weight of the k-th dimensional feature, reflecting the importance of this feature in clustering. It can be set through expert knowledge or automatically determined using machine learning methods. The higher the weight, the greater the impact of this feature on the classification result.

[0122] λ k represents the decay factor, which is used to adjust the influence degree of feature differences. It is usually set according to experience or optimized through cross-validation. A larger λ k value will cause the distance to decrease rapidly as the feature difference increases, and vice versa more gently. This helps to handle the importance differences and non-linear relationships of different features.

[0123] represents the exponential decay function, which is used to adjust the influence of feature differences on the distance. It is calculated based on λ k and the feature difference |x ik - x jk |. As the feature difference increases, the exponential term gradually decreases, thereby reducing the influence of large-difference features on the total distance.

[0124] The reasons for each sub - design are introduced as follows:

[0125] (x ik -x jk ) 2 is designed to use the difference of squares to measure the difference between two samples in a certain feature. The squaring operation amplifies the difference, making significantly different features contribute more to the distance, while ensuring that the distance is non - negative.

[0126] w k is designed to introduce the weight w k to reflect the importance of different features. Some features may have a greater impact on the classification result, and assigning higher weights can enhance the role of these features.

[0127] is designed to use the exponential decay function to adjust the impact of feature differences on the distance. It can effectively handle extreme values or outliers and prevent them from overly affecting the clustering result. By introducing the decay factor λ k , the impact degree of feature differences can be flexibly adjusted.

[0128] By multiplying each sub - item, multiplying the difference of squares with the weight and the exponential decay function is to comprehensively consider the importance of features (weights) and the impact of feature differences (exponential decay). This combination method ensures that the distance metric not only considers the absolute differences between features but also takes into account the importance of features and the impact of differences on the overall distance.

[0129] Through the above - mentioned distance metric formula, a preliminary clustering result is obtained. Based on the preliminary clustering result, the advantages and disadvantages of different numbers of clusters are evaluated, and the optimal number of clusters is selected;

[0130] Based on the preliminary clustering result, the silhouette coefficient is introduced to evaluate the advantages and disadvantages of different numbers of clusters in the preliminary clustering result in order to select the optimal number of clusters. Among them, the calculation formula of the silhouette coefficient s(i) is:

[0131]

[0132] where s(i) represents the silhouette coefficient of the relevant data i, a(i) is the average distance between the relevant data i and other relevant data within the same group, b(i) is the average minimum distance between the relevant data i and all relevant data in the nearest different group, η is the smoothing parameter, and σ(a(i), b(i)) is the standard deviation;

[0133] The following gives a detailed explanation of each parameter:

[0134] s(i) represents the silhouette coefficient of the relevant data point i, which is used to evaluate the suitability of this data point in clustering. It is obtained through the calculation formula and reflects the compactness of data point i within its affiliated group and the separation from other groups.

[0135] a(i) represents the average distance between data point i and other data points within the same group. It is obtained by calculating the average of the distances between data point i and all other data points within the same group. A smaller a(i) indicates that data point i is closer to other data points within its group.

[0136] b(i) represents the average minimum distance between data point i and all data points in the nearest different group. It is obtained by calculating the minimum of the distances between data point i and all data points in the nearest different group, and then taking the average of these minimum values. A larger b(i) indicates that data point i is more separated from data points in different groups.

[0137] α(i) and β(i) are the powers of b(i) and a(i) respectively, which are used to adjust the degree of their influence on the silhouette coefficient. Usually, they are set according to experience or optimized through cross-validation. Different powers can enhance or weaken the influence of a(i) and b(i) on the silhouette coefficient, so as to better adapt to different data distribution situations.

[0138] η represents the smoothing parameter, which is used to adjust the degree of influence of the standard deviation term. Usually, it is set according to experience or optimized through cross-validation. A larger η value will make the standard deviation term have a greater influence on the denominator, increasing the robustness of the formula.

[0139] σ(a(i),b(i)) represents the standard deviation of a(i) and b(i), which is used to measure their degree of fluctuation. It is obtained by calculating the standard deviation of a(i) and b(i). The standard deviation term helps to handle the differences and outliers between data points, making the silhouette coefficient more stable and reliable.

[0140] The following introduces the design reasons for each item:

[0141] b(i) α(i) is designed to raise b(i) to the power of α(i), which can enhance or weaken the influence of b(i) on the silhouette coefficient. A larger α(i) can amplify the role of b(i), emphasizing the separation of data point i from different groups.

[0142] a(i) β(i) is designed to raise a(i) to the power of β(i), which can enhance or weaken the influence of a(i) on the silhouette coefficient. A larger β(i) can amplify the role of a(i), emphasizing the compactness of data point i with other data points within its group.

[0143] max{a(i) β(i),b(i) α(i) is designed to take a(i) β(i) and b(i) α(i) The maximum value in is used as part of the denominator to ensure that the denominator will not be zero and to keep the range of the silhouette coefficient within [-1, 1]. The choice of the maximum value reflects the most significant influencing factors.

[0144] η·σ(a(i), b(i)) is designed to introduce the smoothing parameter η and the standard deviation term σ(a(i), b(i)) to adjust the stability of the silhouette coefficient. The standard deviation term takes into account the degree of fluctuation of a(i) and b(i) and prevents extreme values or outliers from having too much impact on the silhouette coefficient.

[0145] Subtracting each term is to subtract b(i) α(i) and a(i) β(i) Subtracting is to measure the compactness (represented by a negative sign) of data point i with other data points within its group and the separation from other groups (represented by a positive sign). This subtraction operation intuitively reflects the suitability of data point i in clustering.

[0146] Adding each term is to add the maximum value and the standard deviation term in the denominator to ensure that the denominator will not be zero and to make the silhouette coefficient more stable and reliable. The addition of the standard deviation term increases the robustness of the formula and prevents extreme values or outliers from affecting the result.

[0147] By calculating the average silhouette coefficient of the said relevant data, the number of clusters that maximizes the average silhouette coefficient is selected as the optimal number of clusters to generate the target clustering result.

[0148] The following is a specific example:

[0149] Suppose a new type of concrete material needs to be comprehensively tested to ensure its compliance with multi-dimensional acceptance criteria (such as physical durability, chemical stability, and ecological impact). The system first receives relevant data of a batch of concrete materials, including compressive strength, tensile strength, chemical composition, etc. This data will be used to construct a virtual environment for further analysis.

[0150] Step 1: Distance metric calculation

[0151] Suppose there are data points x of three concrete samples 1 , x 2 and x 3 , and each sample has three features (compressive strength, tensile strength, chemical composition), that is, n = 3. Let the feature weight w 1 = 0.5, w 2 = 0.3, w 3 = 0.2, and the attenuation factor λ 1= 0.1, λ 2 = 0.2, λ 3 = 0.15. For x 1 and x 2 , assume that their values on three features are respectively:

[0152] x 11 = 40, x 12 = 30, x 13 = 60, x 21 = 45, x 22 = 35, x 23 = 65, calculate d(x 1 , x 2 ):

[0153]

[0154] Step 2: Optimization of the preliminary clustering result

[0155] Cluster all samples using the above distance metric formula to obtain the preliminary clustering result. Assume that different numbers of clusters k (such as k = 2 and k = 3) are tried, and the average silhouette coefficient for each number of clusters is calculated.

[0156] For k = 2, calculate the silhouette coefficient s(i) for each sample and find the average value.

[0157] For k = 3, similarly calculate the silhouette coefficient s(i) for each sample and find the average value.

[0158] Assume that the average silhouette coefficient for k = 2 is 0.65, while the average silhouette coefficient for k = 3 is 0.75. Since the average silhouette coefficient for k = 3 is larger, k = 3 is selected as the optimal number of clusters to generate the target clustering result.

[0159] It can be seen from the calculation results that the optimal number of clusters k = 3 is selected, which means that the concrete samples are divided into three groups with similar performance characteristics. This not only improves the classification accuracy but also provides a solid foundation for subsequent analysis. By identifying the key influencing factors for each group, the manufacturer can adjust the production process and formula according to the analysis results to improve product quality.

[0160] Figure 2 The structure diagram of a detection and acceptance system for building materials is provided for the embodiments of this application, as Figure 2 shown. This system includes:

[0161] A receiving module 21, configured to receive relevant data of the building materials to be inspected, where the relevant data includes identification information and expected performance parameters;

[0162] A construction module 22, configured to construct a virtual environment based on the relevant data in combination with a preset multi-dimensional acceptance standard, adjust the parameters of the virtual environment to obtain an optimized virtual environment, where the acceptance standard includes physical durability, chemical stability, and ecological impact;

[0163] A calculation module 23, configured to calculate the stress distribution of the building material to be inspected based on the optimized virtual environment by using finite element analysis, and predict the deterioration rate of the building material to be inspected in combination with a machine learning algorithm to obtain a predicted deterioration rate value. Based on the predicted deterioration rate value, adjust the relevant data in combination with a multi-scale modeling technique and a random event triggering mechanism to determine a response data set;

[0164] A judgment module 24, configured to judge whether the relevant data meets the multi-dimensional acceptance standard by using a support vector machine classifier based on the response data set. If it is shown that the relevant data does not meet the multi-dimensional acceptance standard, generate an improvement opinion report;

[0165] A generation module 25, configured to generate a detection report based on the improvement opinion report, the response data set, and the evaluation result, and by using blockchain technology.

[0166] Figure 2 The described detection and acceptance system for building materials can execute Figure 1 The detection and acceptance method for building materials in the illustrated embodiment, the implementation principle and technical effects will not be elaborated. For the detection and acceptance system for building materials in the above embodiment, the specific manners in which each module and unit perform operations have been described in detail in the embodiment related to the method, and will not be elaborated here.

[0167] In a possible design, Figure 2 The detection and acceptance system for building materials in the illustrated embodiment can be implemented as a computing device, such as Figure 3 shown, and this computing device can include a storage component 31 and a processing component 32;

[0168] The storage component 31 stores one or more computer instructions, where the one or more computer instructions are called and executed by the processing component 32.

[0169] The processing component 32 is configured to: receive relevant data of the building materials to be inspected, where the relevant data includes identification information and expected performance parameters; based on the relevant data, combine with a preset multi-dimensional acceptance standard to construct a virtual environment, adjust the parameters of the virtual environment to obtain an optimized virtual environment, where the acceptance standard includes physical durability, chemical stability, and ecological impact; based on the optimized virtual environment, use finite element analysis to calculate the stress distribution of the building materials to be inspected, and combine with a machine learning algorithm to predict the deterioration rate of the building materials to be inspected to obtain a predicted deterioration rate value, and based on the predicted deterioration rate value, combine with multi-scale modeling technology and a random event triggering mechanism to adjust the relevant data to determine a response data set; based on the response data set, use a support vector machine classifier to determine whether the relevant data meets the multi-dimensional acceptance standard, and if it shows that the relevant data does not meet the multi-dimensional acceptance standard, generate an improvement opinion report; based on the improvement opinion report, the response data set, and the evaluation result, and use blockchain technology to generate a detection report.

[0170] Among them, the processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components for executing the above method.

[0171] The storage component 31 is configured to store various types of data to support the operation of the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk, or optical disk.

[0172] Of course, the computing device may also necessarily include other components, such as input / output interfaces, display components, communication components, etc.

[0173] The input / output interface provides an interface between the processing component and the peripheral interface module, and the above peripheral interface module may be an output device, an input device, etc.

[0174] The communication component is configured to facilitate wired or wireless communication between the computing device and other devices, etc.

[0175] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform. In this case, the computing device can refer to a cloud server. The above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from a cloud computing platform.

[0176] An embodiment of the present application also provides a computer storage medium storing a computer program, which when executed by a computer can implement the above Figure 1 inspection and acceptance method of a building material shown in the embodiment.

[0177] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0178] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative labor.

[0179] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0180] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit them. Although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features. These modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of each embodiment of the present application.

Claims

1. A method for testing and accepting building materials, characterized in that: include: Receiving relevant data of the building material to be inspected, wherein the relevant data includes identification information and expected performance parameters; Based on the relevant data and in combination with preset multi-dimensional acceptance criteria, a virtual environment is constructed, and parameters of the virtual environment are adjusted to obtain an optimized virtual environment, wherein the acceptance criteria include physical durability, chemical stability, and ecological impact; Based on the optimized virtual environment, the stress distribution of the building material to be inspected is calculated using finite element analysis, and the degradation rate of the building material to be inspected is predicted in combination with a machine learning algorithm to obtain a degradation rate prediction value, and based on the degradation rate prediction value combined with a multi-scale modeling technology and a random event trigger mechanism, the relevant data is adjusted to determine a response data set; Based on the response data set, using a support vector machine classifier, determining whether the relevant data meets the multi-dimensional acceptance criteria, and generating an improvement suggestion report if it is shown that the relevant data does not meet the multi-dimensional acceptance criteria; Based on the improvement suggestion report, the response data set and the evaluation results, and using blockchain technology, a test report is generated.

2. The method according to claim 1, characterized in that The method comprises: calculating the stress distribution of the building material to be inspected by using finite element analysis based on the optimized virtual environment, predicting the degradation rate of the building material to be inspected by combining a machine learning algorithm, obtaining a degradation rate prediction value, and adjusting the relevant data based on the degradation rate prediction value in combination with a multi-scale modeling technology and a random event trigger mechanism to determine a response data set, including: Using high-precision finite element analysis combined with preset load conditions, the stress distribution of the relevant data is calculated to obtain a preliminary stress distribution map, key areas of stress concentration are identified from the preliminary stress distribution map, and the key areas are adjusted using adaptive mesh refinement technology to generate a target stress distribution map; Based on the target stress distribution diagram combined with a topology optimization algorithm, an optimal material distribution pattern is optimized from the structure of the building material to be inspected, and based on the optimal material distribution pattern and the historical performance data of the building material to be inspected, a deep neural network model is used to predict the degradation rate of the building material to be inspected, and a predicted value of the degradation rate is obtained by combining the complex behaviors of nonlinearity and time dependence; Using principal component analysis and cluster analysis, the predicted values ​​of the degradation rate are reduced in dimension and classified to identify key influencing factors and similar performance groups, and based on the key influencing factors and the similar performance groups combined with multi-scale modeling technology, the structural changes and performance impacts of the building materials to be inspected are evaluated to generate a comprehensive evaluation report; According to the random event triggering mechanism, preset external factors are simulated in the virtual environment to evaluate the performance of the building material to be inspected, and external event combinations are generated by Monte Carlo simulation to evaluate the response data of the building material to be inspected, and the life distribution of the building material to be inspected is evaluated in combination with survival analysis to generate an external factor impact evaluation result; The response data set is determined based on the stress distribution result, the comprehensive evaluation report and the external factor impact evaluation result.

3. The method according to claim 2, characterized in that The principal component analysis and cluster analysis are used to reduce the dimension and classify the degradation rate prediction value to identify key influencing factors and similar performance groups. Based on the key influencing factors and the similar performance groups combined with multi-scale modeling technology, the structural changes and performance impacts of the building materials to be inspected are evaluated to generate a comprehensive evaluation report, including: Using principal component analysis combined with sparse coding technology, the degradation rate prediction value is subjected to dimensionality reduction processing to extract main feature vectors; According to the main feature vectors, a cluster analysis algorithm is applied to divide the relevant data into a plurality of groups with similar performance characteristics, and an association rule mining technique is used to identify key influencing factors of each group; Based on the key influencing factors and the clustering results, multi-scale modeling technology is used to evaluate the structural changes and performance impacts of the building materials to be inspected, and a comprehensive evaluation report is generated.

4. The method according to claim 3, characterized in that The method of applying a cluster analysis algorithm based on the main feature vector to divide the relevant data into a plurality of groups with similar performance characteristics, and using association rule mining technology to identify key influencing factors of each group includes: According to the main feature vectors, using a cluster analysis algorithm, the relevant data is divided into a plurality of groups with similar performance characteristics to obtain a preliminary clustering result; Optimizing the clustering quality of the preliminary clustering result, evaluating different clustering numbers of the preliminary clustering result by introducing a silhouette coefficient to select an optimal clustering number, and generating a target clustering result based on the optimal clustering number; According to the target clustering results, association rule mining technology is used to analyze the key data of each group by mining the frequent patterns and association rules between different variables in the group to determine the key influencing factors of each group, wherein the key data include historical performance data of building materials, stress distribution characteristics and various factors in the degradation rate prediction value.

5. The method according to claim 2, characterized in that: The method optimizes the structure of the building material to be inspected based on the target stress distribution diagram in combination with a topology optimization algorithm, and predicts the degradation rate of the building material to be inspected using a deep neural network model based on the optimal material distribution pattern and the historical performance data of the building material to be inspected, and obtains a degradation rate prediction value by combining nonlinear and time-dependent complex behaviors, including: According to the target stress distribution diagram, combined with a multi-objective topology optimization algorithm, multiple optimization objectives are set, and the design space and constraint conditions are defined, and the optimal material distribution pattern is obtained by using a genetic algorithm and a particle swarm optimization technology; Based on the optimal material distribution pattern, combined with the historical performance data of the building materials to be inspected, a deep neural network model is used for training and processing, and data cleaning and preprocessing techniques are applied to obtain a degradation rate prediction value.

6. The method according to claim 1, characterized in that The step of constructing a virtual environment based on the relevant data in combination with a preset multi-dimensional acceptance standard, and adjusting parameters of the virtual environment to obtain an optimized virtual environment includes: Using computer simulation technology, based on the relevant data and in combination with preset multi-dimensional acceptance criteria, the performance of the building material to be inspected under actual use conditions is simulated to obtain an initial virtual environment; Using a Bayesian optimization algorithm, various parameters of the initial virtual environment are adjusted to analyze an optimal parameter combination, and a preliminary optimization result is obtained based on the optimal parameter combination; Based on the preliminary optimization results, applying sensitivity analysis, identifying key parameters of the performance of the building material to be tested, and generating an optimization parameter list; According to the optimization parameter list, the parameter settings of the initial virtual environment are adjusted again to obtain an optimized virtual environment.

7. The method according to claim 1, characterized in that Based on the response data set, a support vector machine classifier is used to determine whether the relevant data meets the multi-dimensional acceptance criteria. If it is shown that the relevant data does not meet the multi-dimensional acceptance criteria, an improvement suggestion report is generated, including: Based on the response data set, the relevant data is classified by using a support vector machine classifier to obtain a classification result, and according to the classification result, a decision boundary corresponding to a multi-dimensional acceptance criterion is defined to generate a preliminary evaluation result; Using a fuzzy logic system to process the non-deterministic test results in the response data set to determine the uncertainty of the response data set, judging whether the relevant data meets the multi-dimensional acceptance criteria based on the preliminary evaluation results, and enabling a deep belief network to perform fine-grained analysis on edge cases and the uncertainty in the preliminary evaluation results to obtain a target evaluation result; If the target evaluation result shows that the relevant data does not meet the multi-dimensional acceptance criteria, based on the response data set and the target evaluation result, analyze the reasons for the failure of the relevant data, identify the factors that lead to the failure of the relevant data, and generate a failure reason analysis report; Utilizing a genetic algorithm, analyzing an optimization path from the non-conformity cause analysis report to obtain a best practice solution for the performance of the building material to be inspected, and generating an optimization path suggestion; Based on the optimization path suggestion, an improvement suggestion report is generated according to the suggestions preset in the expert knowledge base. A. The method according to claim 4 is characterized in that the clustering analysis algorithm is applied according to the main feature vector to divide the relevant data into multiple groups with similar performance characteristics, and the association rule mining technology is used to identify the key influencing factors of each group, including: According to the main feature vector, the building material related data is classified and processed using a cluster analysis algorithm, and the related data is divided into multiple groups with similar performance characteristics to obtain a preliminary clustering result, where d(x i , x j ) is used to calculate the distance metric between the related data, which is the basis of cluster analysis. The calculation formula is: Among them, d(x i , x j ) represents the relevant data x i and x j The distance measure between them, n is the dimension of the feature vector, x ik and x jk The relevant data x are i and x j The eigenvalue in the kth dimension, w k is the weight of the k-th dimension feature, λ k is the attenuation factor, represents an exponential decay function; Through the above distance measurement formula, a preliminary clustering result is obtained. Based on the preliminary clustering result, the advantages and disadvantages of different cluster numbers are evaluated, and the optimal cluster number is selected; Based on the preliminary clustering results, the profile coefficient is introduced to evaluate the pros and cons of different cluster numbers in the preliminary clustering results to select the optimal cluster number, wherein the calculation formula of the profile sampling number s(i) is: Where s(i) represents the rotation coefficient of relevant data i, a(i) is the average distance between relevant data i and other relevant data in the same group, b(i) is the average minimum distance between relevant data i and all relevant data in the nearest different groups, η is the smoothing parameter, and σ(a(i), b(i)) is the standard deviation; The target clustering result is generated by calculating the average silhouette coefficient of the relevant data and selecting the clustering number that maximizes the average silhouette coefficient as the optimal clustering number.

8. A building material inspection and acceptance system, characterized in that: include: A receiving module, used for receiving relevant data of the building material to be inspected, wherein the relevant data includes identification information and expected performance parameters; A construction module, configured to construct a virtual environment based on the relevant data in combination with preset multi-dimensional acceptance criteria, and adjust parameters of the virtual environment to obtain an optimized virtual environment, wherein the acceptance criteria include physical durability, chemical stability, and ecological impact; A calculation module is used to calculate the stress distribution of the building material to be inspected by using finite element analysis based on the optimized virtual environment, and predict the degradation rate of the building material to be inspected in combination with a machine learning algorithm to obtain a degradation rate prediction value, and adjust the relevant data based on the degradation rate prediction value in combination with a multi-scale modeling technology and a random event trigger mechanism to determine a response data set; A judgment module, used to judge whether the relevant data meets the multi-dimensional acceptance criteria based on the response data set by using a support vector machine classifier, and generate an improvement suggestion report if it is shown that the relevant data does not meet the multi-dimensional acceptance criteria; A generation module is used to generate a detection report based on the improvement suggestion report, the response data set and the evaluation result, and using blockchain technology.

9. A computing device, characterized in that It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a building material inspection and acceptance method as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a computer, a method for inspecting and accepting building materials as claimed in any one of claims 1 to 7 is implemented.

Images