Production quality early warning method and system for limestone powder building material

Abstract

The invention discloses a production quality early warning method and system for limestone powder building materials. The method comprises the following steps: firstly, obtaining a preset number of limestone powder building material production samples and historical production data, and carrying out comprehensive quality evaluation on the samples; then, mapping an evaluation result with historical production data, identifying production quality characteristics, and forming relational data between the production data and the quality characteristics; and constructing a production quality mapping model based on the self-organizing mapping network. And importing real-time production data into the model to obtain a real-time production quality network node response, comparing the real-time production quality network node response with a node response in a normal production state, and calculating a difference value. And carrying out production quality early warning according to the difference value, and formulating an early warning scheme. By constructing the production quality mapping model, real-time monitoring and early warning of the production quality of the limestone powder building material are realized, and the product quality and the production efficiency are improved.

Description

Technical Field

[0001] The invention relates to the technical field of production quality early warning, and in particular to a production quality early warning method and system for limestone powder building materials. Background Art

[0002] In the building materials production industry, limestone powder, as an important raw material, is widely used in construction projects and various building products. The quality of limestone powder building materials directly affects the quality and safety of construction projects. However, traditional quality inspection methods mainly rely on manual sampling and laboratory analysis, which is not only time-consuming and labor-intensive, but also unable to achieve real-time monitoring and early warning of the production process, resulting in quality problems in the production process, affecting the quality and reliability of the final product.

[0003] With the continuous advancement of production technology and the continuous improvement of product quality requirements, there is an urgent need for a quality control method that can achieve real-time monitoring and early warning. Some existing research and applications have shown that data-driven model methods have great potential in quality control. Among them, the self-organizing map network (SOM), as an unsupervised learning neural network model, can effectively process high-dimensional data, has good visualization and clustering capabilities, and is suitable for quality monitoring and early warning in complex production processes.

[0004] Based on the above background, the present invention proposes a limestone powder building material production quality early warning method based on a self-organizing mapping network. By constructing a production quality mapping model, real-time monitoring and early warning of quality characteristics in the production process are achieved, the quality control capability and production efficiency of the production process are improved, and the quality and reliability of the final product are guaranteed. Summary of the invention

[0005] In order to solve at least one of the above technical problems, the present invention proposes a production quality early warning method and system for limestone powder building materials.

[0006] A first aspect of the present invention provides a production quality early warning method for limestone powder building materials, comprising:

[0007] Obtaining a preset number of limestone powder building material production samples and historical production data of the limestone powder building materials, performing a comprehensive quality assessment of the appearance and performance of the limestone powder building material production samples, and obtaining a comprehensive quality assessment result;

[0008] Forming a one-to-one mapping relationship between the comprehensive quality assessment result and the historical production data, identifying the production quality characteristics of limestone powder building materials under different production data, and obtaining production data-production quality characteristic relationship data;

[0009] Constructing a production quality mapping model of limestone powder building materials based on a self-organizing mapping network, importing the production data-production quality feature relationship data into the production quality mapping model for learning and training, and obtaining a complete production quality mapping model;

[0010] Acquire the real-time production data of limestone powder building materials and import it into the production quality mapping model to perform data mapping, so as to obtain the real-time production quality network node response of limestone powder building materials;

[0011] Compare the real-time production quality network node response with the network node response in a normal production state, and calculate the difference between the real-time production quality network node response and the network node response in a normal production state;

[0012] According to the difference value, a production quality early warning is performed on the limestone powder building material to obtain a production quality early warning plan.

[0013] In this solution, a preset number of limestone powder building material production samples and historical production data of limestone powder building materials are obtained, and a comprehensive quality assessment of appearance and performance of the limestone powder building material production samples is performed to obtain a comprehensive quality assessment result, which is specifically:

[0014] Obtaining a preset number of limestone powder building material production samples from multiple preset production batches, and obtaining historical production data of the limestone powder building materials of the multiple preset production batches, wherein the historical production data includes equipment parameters, production environment temperature and humidity, and raw material ratio;

[0015] Obtaining image data of the limestone powder building material production sample and image data of defective samples with different types of appearance defects, converting the defective sample image data into grayscale images, extracting texture features of the grayscale images through a gray-level co-occurrence matrix, and identifying scratch and crack defect features of the limestone powder building material according to the texture features;

[0016] Obtaining a bubble and pothole defect type image from the defect sample image data, obtaining the brightness of each pixel in the bubble and pothole defect type image, determining the influence of the bubble and pothole defect type on the image brightness change, and obtaining the bubble and pothole defect characteristics;

[0017] Obtaining color information of qualified limestone powder building materials, constructing a defect feature database of limestone powder building materials based on the scratch and crack defect features, bubble and pothole defect features, and color information of the limestone powder building materials, extracting appearance features of image data of the limestone powder building materials production samples based on an image processing algorithm, comparing the appearance features with the defect feature database, identifying appearance defects of the limestone powder building materials production samples, performing production quality assessment on the appearance defects, and obtaining appearance production quality assessment results of the limestone powder building materials production samples;

[0018] Obtain density information of limestone powder building material production samples using a density tester, evaluate the compaction of the production samples based on the density information, and obtain data on the impact of the compaction on the compressive strength and flexural strength of the limestone powder building materials;

[0019] The compressive and flexural properties of the limestone powder building material production samples are judged according to the influencing data and the compaction degree of the production samples, and a comprehensive production quality assessment is performed according to the appearance production quality assessment results and the compressive and flexural properties to obtain a comprehensive quality assessment result for each limestone powder building material production sample.

[0020] In this solution, the comprehensive quality assessment result is mapped to the historical production data in a one-to-one relationship, the production quality characteristics of limestone powder building materials under different production data are identified, and the production data-production quality characteristic relationship data is obtained, specifically:

[0021] Integrate the comprehensive quality assessment results of each limestone powder building material production sample with the corresponding historical production data to form a one-to-one mapping relationship;

[0022] The one-to-one mapping relationship is analyzed by using the Spearman correlation coefficient, and the correlation coefficient between the historical production data and the comprehensive production quality evaluation results is calculated to form a correlation coefficient matrix;

[0023] Performing a significance test on the correlation coefficient matrix by a t-test method, and selecting production data parameters related to the comprehensive quality assessment result as key production data according to the correlation coefficient matrix and the significance test result;

[0024] According to the one-to-one mapping relationship and key production data, the production quality characteristics of limestone powder building materials under different production data are identified to obtain production data-production quality characteristic relationship data.

[0025] In this solution, the production quality mapping model of limestone powder building materials is constructed based on the self-organizing mapping network, and the production data-production quality feature relationship data is imported into the production quality mapping model for learning and training to obtain a complete production quality mapping model, specifically:

[0026] Constructing a production quality mapping model of limestone powder building materials based on a self-organizing mapping network, setting the number of data input neurons of the production quality mapping model according to the data parameter items of the production data in the production data-production quality feature relationship data, and initializing the node weight of each data input neuron;

[0027] Constructing a topological structure of a production quality mapping model according to the production data-production quality feature relationship data, selecting an initial learning rate of the production quality mapping model, setting a learning rate decay function so that the learning rate gradually decreases with the training process, selecting a rectangular function as a neighborhood function of the production quality mapping model, setting an initial neighborhood radius and a neighborhood radius decay function so that the neighborhood radius gradually decreases with the training process;

[0028] Using the production data in the production data-production quality feature relationship data as data neurons of the production quality mapping model, and importing the production quality features as mapping targets into the topological structure, and learning and training the production quality mapping model according to the node weight, initial learning rate, and neighborhood function of each data input neuron;

[0029] The data neurons are clustered through a self-organizing mapping network, and the best matching unit in the topological structure after each production data is used as the data neuron of the production quality mapping model is searched according to a competition process. The data neurons after the clustering operation are mapped to the best matching unit in the topological structure according to the production data-production quality feature relationship data, so that the data neurons after the clustering operation and the production quality features in the topological structure form a mapping network to obtain a complete production quality mapping model.

[0030] In this solution, the real-time production data of limestone powder building materials is imported into the production quality mapping model for data mapping to obtain the real-time production quality network node response of limestone powder building materials, specifically:

[0031] Acquire real-time production data of limestone powder building materials in an actual production process, standardize the real-time production data, and import the standardized real-time production data as an input vector into the production quality mapping model;

[0032] Calculating the Euclidean distance between the input vector and the node weight vectors of all data neurons in the production quality mapping model, taking the data neuron node with the smallest distance as the winning node, and recording the response value of the node;

[0033] The position and distribution of the production data in the topological structure of the production quality mapping model are determined according to the response value, and the real-time production quality network node response of the limestone powder building material is obtained.

[0034] In this solution, the real-time production quality network node response is compared with the network node response in the normal production state, and the difference value between the real-time production quality network node response and the network node response in the normal production state is calculated, specifically:

[0035] Obtaining network node responses of limestone powder building materials in the production quality mapping model under normal production conditions;

[0036] The Euclidean distance between the real-time production quality network node response and the network node response in the normal production state is calculated, and the difference value between the real-time production quality network node response and the network node response in the normal production state is evaluated according to the Euclidean distance.

[0037] In this solution, the production quality warning of limestone powder building materials is carried out according to the difference value to obtain a production quality warning solution, which is specifically:

[0038] According to the production standard of limestone powder building materials, the early warning threshold of the difference between the real-time production quality network node response and the normal production state network node response is set;

[0039] When the difference value is greater than the warning threshold, the trigger time when the difference value is greater than the warning threshold and the corresponding production data item of the network node response that deviates from the normal production state are recorded;

[0040] The trigger time when the difference value is greater than the warning threshold and the corresponding production data items of the network node response that deviates from the normal production state are used to form production quality warning information, and the production quality warning information is sent to the production equipment terminal for display as a quality alarm to form a production quality warning plan.

[0041] The second aspect of the present invention further provides a production quality early warning system for limestone powder building materials, the system comprising: a memory, a processor, the memory comprising a production quality early warning method program for limestone powder building materials, and when the production quality early warning method program for limestone powder building materials is executed by the processor, the following steps are implemented:

[0042] Obtaining a preset number of limestone powder building material production samples and historical production data of the limestone powder building materials, performing a comprehensive quality assessment of the appearance and performance of the limestone powder building material production samples, and obtaining a comprehensive quality assessment result;

[0043] Forming a one-to-one mapping relationship between the comprehensive quality assessment result and the historical production data, identifying the production quality characteristics of limestone powder building materials under different production data, and obtaining production data-production quality characteristic relationship data;

[0044] Constructing a production quality mapping model of limestone powder building materials based on a self-organizing mapping network, importing the production data-production quality feature relationship data into the production quality mapping model for learning and training, and obtaining a complete production quality mapping model;

[0045] Acquire the real-time production data of limestone powder building materials and import it into the production quality mapping model to perform data mapping, so as to obtain the real-time production quality network node response of limestone powder building materials;

[0046] Compare the real-time production quality network node response with the network node response in a normal production state, and calculate the difference between the real-time production quality network node response and the network node response in a normal production state;

[0047] According to the difference value, a production quality early warning is performed on the limestone powder building material to obtain a production quality early warning plan.

[0048] In this solution, the real-time production quality network node response is compared with the network node response in the normal production state, and the difference value between the real-time production quality network node response and the network node response in the normal production state is calculated, specifically:

[0049] Obtaining network node responses of limestone powder building materials in the production quality mapping model under normal production conditions;

[0050] The Euclidean distance between the real-time production quality network node response and the network node response in the normal production state is calculated, and the difference value between the real-time production quality network node response and the network node response in the normal production state is evaluated according to the Euclidean distance.

[0051] In this solution, the production quality warning of limestone powder building materials is carried out according to the difference value to obtain a production quality warning solution, which is specifically:

[0052] According to the production standard of limestone powder building materials, the early warning threshold of the difference between the real-time production quality network node response and the normal production state network node response is set;

[0053] When the difference value is greater than the warning threshold, the trigger time when the difference value is greater than the warning threshold and the corresponding production data item of the network node response that deviates from the normal production state are recorded;

[0054] The trigger time when the difference value is greater than the warning threshold and the corresponding production data items of the network node response that deviates from the normal production state are used to form production quality warning information, and the production quality warning information is sent to the production equipment terminal for display as a quality alarm to form a production quality warning plan.

[0055] The present invention discloses a production quality early warning method and system for limestone powder building materials. First, a preset number of limestone powder building material production samples and historical production data are obtained, and a comprehensive quality assessment is performed on the samples. Then, the assessment results are mapped with the historical production data, the production quality characteristics are identified, and the relationship data between the production data and the quality characteristics is formed. A production quality mapping model is constructed based on a self-organizing mapping network. Real-time production data is imported into the model to obtain real-time production quality network node responses, which are compared with the node responses of normal production states to calculate the difference values. Production quality early warning is performed based on the difference values, and an early warning plan is formulated. The present invention realizes real-time monitoring and early warning of the production quality of limestone powder building materials by constructing a production quality mapping model, thereby improving product quality and production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 A flow chart showing a method for early warning of production quality of limestone powder building materials according to the present invention is shown;

[0057] Figure 2 A flow chart showing the method of obtaining a difference value of a network node response according to the present invention is shown;

[0058] Figure 3 A flow chart showing a production quality early warning scheme of the present invention is shown;

[0059] Figure 4 A block diagram of a production quality early warning system for limestone powder building materials according to the present invention is shown. DETAILED DESCRIPTION

[0060] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0061] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0062] Figure 1 The present invention shows a flow chart of a limestone powder building material production quality early warning method.

[0063] like Figure 1 As shown, the first aspect of the present invention provides a production quality early warning method for limestone powder building materials, comprising:

[0064] S102, obtaining a preset number of limestone powder building material production samples and historical production data of limestone powder building materials, and performing a comprehensive quality assessment of appearance and performance on the limestone powder building material production samples to obtain a comprehensive quality assessment result;

[0065] S104, forming a one-to-one mapping relationship between the comprehensive quality assessment result and the historical production data, identifying the production quality characteristics of limestone powder building materials under different production data, and obtaining production data-production quality characteristic relationship data;

[0066] S106, constructing a production quality mapping model of limestone powder building materials based on a self-organizing mapping network, importing the production data-production quality feature relationship data into the production quality mapping model for learning and training, and obtaining a complete production quality mapping model;

[0067] S108, obtaining real-time production data of limestone powder building materials and importing it into the production quality mapping model for data mapping, so as to obtain a real-time production quality network node response of the limestone powder building materials;

[0068] S110, comparing the real-time production quality network node response with the network node response in a normal production state, and calculating a difference value between the real-time production quality network node response and the network node response in a normal production state;

[0069] S112, performing a production quality warning on the limestone powder building material according to the difference value to obtain a production quality warning plan.

[0070] It should be noted that in the production process of limestone powder building materials, the parameter changes of production data will affect the production quality of limestone powder building materials. During the production process, limestone powder building materials are produced by setting fixed production parameters at the beginning of production. However, the voltage, current and production temperature will fluctuate due to external factors during the production process, resulting in the rotation speed, heating temperature and other production parameters of the production equipment failing to meet the requirements. Such situations cannot be detected through manual inspections. Therefore, by collecting limestone powder building material production samples and historical production data of limestone powder building materials, a comprehensive weight assessment of limestone powder building materials is conducted, and a one-to-one mapping relationship is formed between the comprehensive quality assessment results and the historical production data. The quality characteristics corresponding to different production data can be identified, and the key factors affecting production quality can be found. Traditional quality monitoring methods are often unable to efficiently and accurately analyze and warn when faced with complex data, and self-organizing The System-on-a-Synthesis Mapping Network (SOM) can effectively process and analyze multi-dimensional complex data. Through self-organizing learning and training of production data and production quality feature relationship data, it can accurately capture the impact of various factors in the production process on quality and build a highly accurate production quality mapping model. The trained production quality mapping model can intuitively display the distribution of quality features corresponding to different production data, thereby improving the accuracy of production quality prediction. By importing real-time production data into the model, the network node response of production quality can be obtained instantly, real-time monitoring of the production process can be achieved, and potential quality problems can be quickly identified. By comparing the difference values ​​of network node responses, the degree of deviation of the current production quality can be quantified. By analyzing the difference values, quality anomalies can be identified in advance, production adjustments and optimizations can be made in time, the impact of quality problems on production can be reduced, and the overall quality stability of the product can be ensured.

[0071] According to an embodiment of the present invention, a preset number of limestone powder building material production samples and historical production data of limestone powder building materials are obtained, and a comprehensive quality evaluation of appearance and performance of the limestone powder building material production samples is performed to obtain a comprehensive quality evaluation result, which is specifically:

[0072] Obtaining a preset number of limestone powder building material production samples from multiple preset production batches, and obtaining historical production data of the limestone powder building materials of the multiple preset production batches, wherein the historical production data includes equipment parameters, production environment temperature and humidity, and raw material ratio;

[0073] Obtaining image data of the limestone powder building material production sample and image data of defective samples with different types of appearance defects, converting the defective sample image data into grayscale images, extracting texture features of the grayscale images through a gray-level co-occurrence matrix, and identifying scratch and crack defect features of the limestone powder building material according to the texture features;

[0074] Obtaining a bubble and pothole defect type image from the defect sample image data, obtaining the brightness of each pixel in the bubble and pothole defect type image, determining the influence of the bubble and pothole defect type on the image brightness change, and obtaining the bubble and pothole defect characteristics;

[0075] Obtaining color information of qualified limestone powder building materials, constructing a defect feature database of limestone powder building materials based on the scratch and crack defect features, bubble and pothole defect features, and color information of the limestone powder building materials, extracting appearance features of image data of the limestone powder building materials production samples based on an image processing algorithm, comparing the appearance features with the defect feature database, identifying appearance defects of the limestone powder building materials production samples, performing production quality assessment on the appearance defects, and obtaining appearance production quality assessment results of the limestone powder building materials production samples;

[0076] Obtain density information of limestone powder building material production samples using a density tester, evaluate the compaction of the production samples based on the density information, and obtain data on the impact of the compaction on the compressive strength and flexural strength of the limestone powder building materials;

[0077] The compressive and flexural properties of the limestone powder building material production samples are judged according to the influencing data and the compaction degree of the production samples, and a comprehensive production quality assessment is performed according to the appearance production quality assessment results and the compressive and flexural properties to obtain a comprehensive quality assessment result for each limestone powder building material production sample.

[0078] It should be noted that by obtaining image data of limestone powder building materials production samples and image data of defective samples with different types of appearance defects, and using the grayscale co-occurrence matrix to extract the texture features of grayscale images, appearance defect features such as scratches and cracks can be accurately identified. At the same time, detecting bubbles and potholes through changes in image brightness helps to comprehensively and accurately evaluate the product appearance quality and improve detection accuracy; constructing various appearance defect features and color information of qualified products into a detailed defect feature database can provide comprehensive data support for subsequent quality assessments. This database can significantly improve the automation and accuracy of appearance defect detection, making the detection results more reliable; by combining the appearance production quality assessment results with the compression and flexural properties, comprehensive production quality assessment can comprehensively reflect the overall quality level of limestone powder building materials; the equipment parameters include the current, voltage, heating temperature, building material compaction pressure, limestone powder grinding speed, mixing time, curing time, processing environment humidity, etc. of the limestone powder building materials production equipment; the appearance defect types include cracks, potholes, bubbles, color difference, surface impurities, and scratches; the comprehensive quality assessment result is the total score obtained by scoring each quality assessment item and multiplying it by the corresponding weight.

[0079] According to an embodiment of the present invention, the comprehensive quality assessment result is mapped to the historical production data in a one-to-one relationship, the production quality characteristics of limestone powder building materials under different production data are identified, and the production data-production quality characteristic relationship data is obtained, specifically:

[0080] Integrate the comprehensive quality assessment results of each limestone powder building material production sample with the corresponding historical production data to form a one-to-one mapping relationship;

[0081] The one-to-one mapping relationship is analyzed by using the Spearman correlation coefficient, and the correlation coefficient between the historical production data and the comprehensive production quality evaluation results is calculated to form a correlation coefficient matrix;

[0082] Performing a significance test on the correlation coefficient matrix by a t-test method, and selecting production data parameters related to the comprehensive quality assessment result as key production data according to the correlation coefficient matrix and the significance test result;

[0083] According to the one-to-one mapping relationship and key production data, the production quality characteristics of limestone powder building materials under different production data are identified to obtain production data-production quality characteristic relationship data.

[0084] It should be noted that, through the Spearman correlation coefficient analysis and t-test method, the correlation between the historical production data and the comprehensive quality assessment results is quantitatively analyzed, and the production data parameters that are significantly correlated with the quality assessment results are screened out to form key production data, which is helpful to accurately identify the main production factors affecting the quality of limestone powder building materials and optimize production control; the significance test is used to determine whether the production data and the production quality of limestone powder building materials are statistically significant, that is, whether these production qualities may be caused by random factors. In other words, the significance test helps us determine whether a result has sufficient evidence to support a hypothesis, or whether it may be just the result of random errors; the production quality characteristics of limestone powder building materials under different production data are the production qualities of limestone powder building materials corresponding to different production data. For example, if cracks will appear in limestone powder at a certain production temperature, it can be considered that crack defect characteristics will appear in limestone powder building materials at a certain production temperature.

[0085] According to an embodiment of the present invention, the production quality mapping model of limestone powder building materials is constructed based on the self-organizing mapping network, and the production data-production quality feature relationship data is imported into the production quality mapping model for learning and training to obtain a complete production quality mapping model, specifically:

[0086] Constructing a production quality mapping model of limestone powder building materials based on a self-organizing mapping network, setting the number of data input neurons of the production quality mapping model according to the data parameter items of the production data in the production data-production quality feature relationship data, and initializing the node weight of each data input neuron;

[0087] Constructing a topological structure of a production quality mapping model according to the production data-production quality feature relationship data, selecting an initial learning rate of the production quality mapping model, setting a learning rate decay function so that the learning rate gradually decreases with the training process, selecting a rectangular function as a neighborhood function of the production quality mapping model, setting an initial neighborhood radius and a neighborhood radius decay function so that the neighborhood radius gradually decreases with the training process;

[0088] Using the production data in the production data-production quality feature relationship data as data neurons of the production quality mapping model, and importing the production quality features as mapping targets into the topological structure, and learning and training the production quality mapping model according to the node weight, initial learning rate, and neighborhood function of each data input neuron;

[0089] The data neurons are clustered through a self-organizing mapping network, and the best matching unit in the topological structure after each production data is used as the data neuron of the production quality mapping model is searched according to a competition process. The data neurons after the clustering operation are mapped to the best matching unit in the topological structure according to the production data-production quality feature relationship data, so that the data neurons after the clustering operation and the production quality features in the topological structure form a mapping network to obtain a complete production quality mapping model.

[0090] It should be noted that by constructing a production quality mapping model based on a self-organizing map network and setting the initial learning rate, learning rate decay function, neighborhood function and neighborhood radius decay function, the model can be effectively initialized and optimized, ensuring efficient learning and training of the model, so that the model can accurately reflect the relationship between production data and production quality characteristics, thereby improving the prediction accuracy and reliability of the model; the self-organizing map network clusters the production data, searches for the best matching units through a competitive process, clusters similar production data together, better identifies and classifies patterns and features in production data, facilitates the analysis and processing of quality characteristics under different production conditions, and improves the interpretability and utilization value of the data; due to the self-organizing map network The network has the characteristics of nonlinear dimensionality reduction, clustering function, competitive learning mechanism, topological structure maintenance, adaptive learning and data-driven. SOM realizes the visualization and understanding of complex data by reducing high-dimensional data to low-dimensional space and maintaining the topological structure of the data; similar data points are classified into the same neuron through clustering function, revealing the natural classification and characteristic distribution of data; through competitive learning mechanism and adaptive learning, the network gradually learns and adapts to the distribution of input data, forming a stable data and feature mapping relationship. Therefore, the production quality mapping model constructed by the self-organizing mapping network can well reflect the relationship between production data and production quality characteristics; the topological structure is the corresponding relationship structure formed by production data and production quality.

[0091] According to an embodiment of the present invention, the real-time production data of limestone powder building materials is imported into the production quality mapping model for data mapping to obtain a real-time production quality network node response of limestone powder building materials, specifically:

[0092] Acquire real-time production data of limestone powder building materials in an actual production process, standardize the real-time production data, and import the standardized real-time production data as an input vector into the production quality mapping model;

[0093] Calculating the Euclidean distance between the input vector and the node weight vectors of all data neurons in the production quality mapping model, taking the data neuron node with the smallest distance as the winning node, and recording the response value of the node;

[0094] The position and distribution of the production data in the topological structure of the production quality mapping model are determined according to the response value, and the real-time production quality network node response of the limestone powder building material is obtained.

[0095] It should be noted that by calculating the Euclidean distance between the input vector and the node weight vector of all data neurons in the production quality mapping model, and determining the winning node with the smallest distance, the position and distribution of the real-time production data of limestone powder building materials in the topological structure of the production quality mapping model can be accurately located, which is helpful to accurately identify the production stage and position where the quality problem is located. By recording the response value of the winning node, the real-time production quality situation can be quickly understood; the winning node refers to the best matching unit found by the input vector in the production quality mapping model, and the best matching unit is the most similar neuron corresponding to the topological structure of the production quality mapping model, that is, the historical production data that is most similar to the topological structure of the production quality mapping model; the response value is the similarity between the input vector and the data neuron node; the real-time production quality network node response is the position and distribution of the production data in the topological structure of the production quality mapping model, and different positions of the production data in the topological structure of the production quality mapping model correspond to different production quality results.

[0096] Figure 2 The flowchart of the present invention for obtaining the difference value of the network node response is shown.

[0097] According to an embodiment of the present invention, comparing the real-time production quality network node response with the network node response in a normal production state, and calculating the difference between the real-time production quality network node response and the network node response in a normal production state, is specifically:

[0098] S202, obtaining network node responses of limestone powder building materials in the production quality mapping model under normal production conditions;

[0099] S204, calculating the Euclidean distance between the real-time production quality network node response and the network node response in the normal production state, and evaluating the difference value between the real-time production quality network node response and the network node response in the normal production state according to the Euclidean distance.

[0100] It should be noted that by comparing the difference between the real-time production quality network node response and the normal production status network node response, it is possible to detect whether there are abnormal conditions in the real-time production process. If the Euclidean distance exceeds the preset threshold range, it means that the production data deviates from the production data under the normal production status.

[0101] Figure 3 A flow chart of the production quality early warning scheme obtained by the present invention is shown.

[0102] According to an embodiment of the present invention, the production quality warning of limestone powder building materials is performed according to the difference value to obtain a production quality warning plan, which is specifically:

[0103] S302, setting an early warning threshold of the difference between the real-time production quality network node response and the normal production state network node response according to the production standard of limestone powder building materials;

[0104] S304, when the difference value is greater than the warning threshold, record the trigger time when the difference value is greater than the warning threshold and the corresponding production data item of the network node response that deviates from the normal production state;

[0105] S306, the trigger time when the difference value is greater than the warning threshold and the corresponding production data items of the network node response that deviates from the normal production state are formed into production quality warning information, and the production quality warning information is sent to the production equipment terminal for display as a quality alarm to form a production quality warning plan.

[0106] It should be noted that according to the production standards and quality requirements of limestone powder building materials, an early warning threshold of the difference between the real-time production quality network node response and the network node response in the normal production status is set. This threshold can be set based on historical data, and the difference between the real-time production quality network node response and the network node response in the normal production status is continuously monitored. When the difference value exceeds the early warning threshold, it indicates that quality problems or abnormal conditions may have occurred in the production process. When the difference value is greater than the early warning threshold, the trigger time and the production data items corresponding to the network node response that deviates from the normal production status are recorded, which is helpful for further analysis and troubleshooting of problems, reminding production personnel to pay attention and take timely measures to prevent quality problems from further expanding or affecting product quality.

[0107] According to an embodiment of the present invention, it also includes:

[0108] Obtaining a limestone powder raw material sample of a preset mass before entering the production line, performing a dyeing operation on the limestone powder raw material sample, injecting water of the same mass into the dyed limestone powder raw material sample to dilute and stir evenly, to obtain a primary diluted sample;

[0109] Performing density analysis on the primary diluted sample by sedimentation analysis to determine density data of the primary diluted sample;

[0110] Obtaining optimal observed density data of limestone powder raw material sample particles, determining a secondary dilution multiple of the limestone powder raw material sample according to the density data of the primary dilution sample and the optimal observed density data, and performing secondary dilution on the limestone powder raw material sample according to the secondary dilution multiple to obtain a secondary dilution sample;

[0111] Obtaining high-definition image data of the secondary dilution sample through a microscope, dividing the high-definition image data into N small areas, randomly selecting a preset number of small areas for magnification by a preset multiple, estimating the enlarged diameter of each particle in the enlarged small area according to an image recognition algorithm, performing original diameter analysis on the enlarged diameter according to the preset magnification multiple, and counting the number of particles of each diameter in the enlarged small area, and determining the particle size distribution of the limestone powder raw material sample according to the number of particles of each diameter;

[0112] Obtaining data on the effect of the particle size distribution of limestone powder raw materials on the compactness of limestone powder building materials, and analyzing the optimal particle size distribution of the limestone powder raw materials based on the effect data;

[0113] Obtaining the relationship between the grinding time of the limestone powder production equipment and the change in the particle size of the limestone powder raw material, calculating the deviation between the particle size distribution of the limestone powder raw material sample and the optimal particle size distribution, and when the deviation is greater than a first preset value, performing a grinding operation on the limestone powder raw material, and determining the grinding time according to the particle size distribution of the limestone powder raw material sample and the relationship between the grinding time and the change in the particle size of the limestone powder raw material;

[0114] When the deviation value is less than the second preset value, the compaction time and compaction strength of the limestone powder building materials in the production process are increased; when the deviation value is between the first preset value and the second preset value, the limestone powder raw material is not ground, and a grinding control plan for the limestone powder raw material is obtained.

[0115] It should be noted that since the particle size distribution of the limestone powder raw material will affect the production quality of limestone powder building materials, the limestone powder raw material is diluted. Since the dilution multiple cannot be accurately determined, it is necessary to first determine the particle density of the limestone powder raw material sample through a primary dilution, and then determine the dilution density and perform a secondary dilution, so that the limestone powder raw material sample can be accurately diluted to an optimal particle size observation state; the dyeing operation can improve the subsequent particle observation effect; through microscope high-definition image analysis and image recognition algorithm, the particle size distribution of the raw material is accurately determined, the deviation value between the particle size distribution and the optimal particle size distribution is analyzed, and the grinding time and compaction parameters are dynamically adjusted to ensure that the particle size distribution in the production process meets the quality requirements, thereby improving product quality; through the grinding control plan, unnecessary grinding operations can be avoided, energy and resource waste can be reduced, and production efficiency can be improved; the production equipment of limestone powder building materials includes heating equipment, compaction equipment, grinding equipment, molding equipment, curing equipment, mixing equipment, etc.

[0116] According to an embodiment of the present invention, it also includes:

[0117] Acquire production quality change data of multiple limestone powder building materials production lines in a target factory in a preset time period through a production quality mapping model, and construct production quality time series data from the production quality change data in the preset time period;

[0118] Constructing a production quality prediction model based on the LSTM algorithm, setting the output gate, input gate, and forget gate of the production quality prediction model, and importing the production quality time series data into the production quality prediction model for learning and training;

[0119] Acquire the quality change data of each production line in the current time period according to the production quality mapping model, import the quality change data of each production line in the current time period into the production quality prediction model, predict the production quality change of each production line in the future preset time period, and obtain the prediction result;

[0120] Obtain the order volume of limestone powder building materials of the target factory and the historical unit time production volume of each production line, determine the time point when the production quality of each production line is lower than the preset production quality threshold according to the prediction result, and determine the estimated production capacity of each production line before the production quality problem occurs according to the time point and the historical unit time point;

[0121] The production volume of each production line is allocated according to the estimated production capacity, and the production line having production quality problems is inspected and repaired according to the time point when the production quality of each production line is lower than a preset production quality threshold.

[0122] It should be noted that the production quality mapping model is used to obtain the production quality change data of multiple limestone powder building materials production lines in the target factory in a preset time period, and a production quality prediction model based on LSTM (long short-term memory network) is constructed to predict the production quality changes of each limestone powder building materials production line in the target factory in the future time period. The time point when the production quality of each production line is lower than the preset threshold is determined according to the prediction results, and combined with the historical unit time production data, the estimated production capacity of each production line before quality problems occur can be calculated. By allocating production volume to each production line and dynamically adjusting it based on the prediction results, the production process can be optimized to ensure that the production strategy is adjusted before quality problems occur, maximize the efficient operation of the production line and the stability of product quality, optimize production planning and resource allocation, and improve production efficiency and resource utilization; identify the time point when quality problems may occur in the production line in advance and arrange maintenance operations, so that preventive maintenance can be carried out before quality problems expand, reduce shutdowns and production losses caused by quality problems, and improve the stability and reliability of the production line.

[0123] Figure 4 A block diagram of a production quality early warning system for limestone powder building materials according to the present invention is shown.

[0124] The second aspect of the present invention further provides a limestone powder building material production quality early warning system 4, the system comprising: a memory 41, a processor 42, the memory comprising a limestone powder building material production quality early warning method program, the limestone powder building material production quality early warning method program when executed by the processor, the following steps are implemented:

[0125] Obtaining a preset number of limestone powder building material production samples and historical production data of the limestone powder building materials, performing a comprehensive quality assessment of the appearance and performance of the limestone powder building material production samples, and obtaining a comprehensive quality assessment result;

[0126] Forming a one-to-one mapping relationship between the comprehensive quality assessment result and the historical production data, identifying the production quality characteristics of limestone powder building materials under different production data, and obtaining production data-production quality characteristic relationship data;

[0127] Constructing a production quality mapping model of limestone powder building materials based on a self-organizing mapping network, importing the production data-production quality feature relationship data into the production quality mapping model for learning and training, and obtaining a complete production quality mapping model;

[0128] Acquire the real-time production data of limestone powder building materials and import it into the production quality mapping model to perform data mapping, so as to obtain the real-time production quality network node response of limestone powder building materials;

[0129] Compare the real-time production quality network node response with the network node response in a normal production state, and calculate the difference between the real-time production quality network node response and the network node response in a normal production state;

[0130] According to the difference value, a production quality early warning is performed on the limestone powder building material to obtain a production quality early warning plan.

[0131] The present invention discloses a production quality early warning method and system for limestone powder building materials. First, a preset number of limestone powder building material production samples and historical production data are obtained, and a comprehensive quality assessment is performed on the samples. Then, the assessment results are mapped with the historical production data, the production quality characteristics are identified, and the relationship data between the production data and the quality characteristics is formed. A production quality mapping model is constructed based on a self-organizing mapping network. Real-time production data is imported into the model to obtain real-time production quality network node responses, which are compared with the node responses of normal production states to calculate the difference values. Production quality early warning is performed based on the difference values, and an early warning plan is formulated. The present invention realizes real-time monitoring and early warning of the production quality of limestone powder building materials by constructing a production quality mapping model, thereby improving product quality and production efficiency.

[0132] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0133] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0134] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0135] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiments; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks or optical disks, and other media that can store program codes.

[0136] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0137] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A production quality early warning method for limestone powder building materials, characterized in that: The following steps are involved: Obtaining a preset number of limestone powder building material production samples and historical production data of the limestone powder building materials, performing a comprehensive quality assessment of the appearance and performance of the limestone powder building material production samples, and obtaining a comprehensive quality assessment result; Forming a one-to-one mapping relationship between the comprehensive quality assessment result and the historical production data, identifying the production quality characteristics of limestone powder building materials under different production data, and obtaining production data-production quality characteristic relationship data; Constructing a production quality mapping model of limestone powder building materials based on a self-organizing mapping network, importing the production data-production quality feature relationship data into the production quality mapping model for learning and training, and obtaining a complete production quality mapping model; Acquire the real-time production data of limestone powder building materials and import it into the production quality mapping model to perform data mapping, so as to obtain the real-time production quality network node response of limestone powder building materials; Compare the real-time production quality network node response with the network node response in a normal production state, and calculate the difference between the real-time production quality network node response and the network node response in a normal production state; According to the difference value, a production quality early warning is performed on the limestone powder building material to obtain a production quality early warning plan.

2. The method for early warning of production quality of limestone powder building materials according to claim 1, characterized in that: The obtaining of a preset number of limestone powder building material production samples and historical production data of limestone powder building materials, and performing a comprehensive quality assessment of the appearance and performance of the limestone powder building material production samples to obtain a comprehensive quality assessment result is specifically: Obtaining a preset number of limestone powder building material production samples from multiple preset production batches, and obtaining historical production data of the limestone powder building materials of the multiple preset production batches, wherein the historical production data includes equipment parameters, production environment temperature and humidity, and raw material ratio; Obtaining image data of the limestone powder building material production sample and image data of defective samples with different types of appearance defects, converting the defective sample image data into grayscale images, extracting texture features of the grayscale images through a gray-level co-occurrence matrix, and identifying scratch and crack defect features of the limestone powder building material according to the texture features; Obtaining a bubble and pothole defect type image from the defect sample image data, obtaining the brightness of each pixel in the bubble and pothole defect type image, determining the influence of the bubble and pothole defect type on the image brightness change, and obtaining the bubble and pothole defect characteristics; Obtaining color information of qualified limestone powder building materials, constructing a defect feature database of limestone powder building materials based on the scratch and crack defect features, bubble and pothole defect features, and color information of the limestone powder building materials, extracting appearance features of image data of the limestone powder building materials production samples based on an image processing algorithm, comparing the appearance features with the defect feature database, identifying appearance defects of the limestone powder building materials production samples, performing production quality assessment on the appearance defects, and obtaining appearance production quality assessment results of the limestone powder building materials production samples; Obtain density information of limestone powder building material production samples using a density tester, evaluate the compaction of the production samples based on the density information, and obtain data on the impact of the compaction on the compressive strength and flexural strength of the limestone powder building materials; The compressive and flexural properties of the limestone powder building material production samples are judged according to the influencing data and the compaction degree of the production samples, and a comprehensive production quality assessment is performed according to the appearance production quality assessment results and the compressive and flexural properties to obtain a comprehensive quality assessment result for each limestone powder building material production sample.

3. The production quality early warning method of a limestone powder building material according to claim 1 is characterized in that: The comprehensive quality assessment result is mapped to the historical production data in a one-to-one relationship, the production quality characteristics of limestone powder building materials under different production data are identified, and the production data-production quality characteristic relationship data is obtained, specifically: Integrate the comprehensive quality assessment results of each limestone powder building material production sample with the corresponding historical production data to form a one-to-one mapping relationship; The one-to-one mapping relationship is analyzed by using the Spearman correlation coefficient, and the correlation coefficient between the historical production data and the comprehensive production quality evaluation results is calculated to form a correlation coefficient matrix; Performing a significance test on the correlation coefficient matrix by a t-test method, and selecting production data parameters related to the comprehensive quality assessment result as key production data according to the correlation coefficient matrix and the significance test result; According to the one-to-one mapping relationship and key production data, the production quality characteristics of limestone powder building materials under different production data are identified to obtain production data-production quality characteristic relationship data.

4. The method for early warning of production quality of limestone powder building materials according to claim 1, characterized in that: The production quality mapping model of limestone powder building materials is constructed based on the self-organizing mapping network, and the production data-production quality feature relationship data is imported into the production quality mapping model for learning and training to obtain a complete production quality mapping model, specifically: Constructing a production quality mapping model of limestone powder building materials based on a self-organizing mapping network, setting the number of data input neurons of the production quality mapping model according to the data parameter items of the production data in the production data-production quality feature relationship data, and initializing the node weight of each data input neuron; Constructing a topological structure of a production quality mapping model according to the production data-production quality feature relationship data, selecting an initial learning rate of the production quality mapping model, setting a learning rate decay function so that the learning rate gradually decreases with the training process, selecting a rectangular function as a neighborhood function of the production quality mapping model, setting an initial neighborhood radius and a neighborhood radius decay function so that the neighborhood radius gradually decreases with the training process; Using the production data in the production data-production quality feature relationship data as data neurons of the production quality mapping model, importing the production quality features as mapping targets into the topological structure, and learning and training the production quality mapping model according to the node weight, initial learning rate, and neighborhood function of each data input neuron; The data neurons are clustered through a self-organizing mapping network, and the best matching unit in the topological structure after each production data is used as the data neuron of the production quality mapping model is searched according to a competition process. The data neurons after the clustering operation are mapped to the best matching unit in the topological structure according to the production data-production quality feature relationship data, so that the data neurons after the clustering operation and the production quality features in the topological structure form a mapping network to obtain a complete production quality mapping model.

5. The method for early warning of production quality of limestone powder building materials according to claim 1, characterized in that: The real-time production data of limestone powder building materials is imported into the production quality mapping model for data mapping to obtain the real-time production quality network node response of limestone powder building materials, specifically: Acquire real-time production data of limestone powder building materials in an actual production process, standardize the real-time production data, and import the standardized real-time production data as an input vector into the production quality mapping model; Calculating the Euclidean distance between the input vector and the node weight vectors of all data neurons in the production quality mapping model, taking the data neuron node with the smallest distance as the winning node, and recording the response value of the node; The position and distribution of the production data in the topological structure of the production quality mapping model are determined according to the response value, and the real-time production quality network node response of the limestone powder building material is obtained.

6. The method for early warning of production quality of limestone powder building materials according to claim 1, characterized in that: The comparing the real-time production quality network node response with the network node response in the normal production state, and calculating the difference between the real-time production quality network node response and the network node response in the normal production state, is specifically: Obtaining network node responses of limestone powder building materials in the production quality mapping model under normal production conditions; The Euclidean distance between the real-time production quality network node response and the network node response in the normal production state is calculated, and the difference value between the real-time production quality network node response and the network node response in the normal production state is evaluated according to the Euclidean distance.

7. The method for early warning of production quality of limestone powder building materials according to claim 1, characterized in that: The production quality warning of limestone powder building materials is performed according to the difference value to obtain a production quality warning plan, which is specifically: According to the production standard of limestone powder building materials, the early warning threshold of the difference between the real-time production quality network node response and the normal production state network node response is set; When the difference value is greater than the warning threshold, the trigger time when the difference value is greater than the warning threshold and the corresponding production data item of the network node response that deviates from the normal production state are recorded; The trigger time when the difference value is greater than the warning threshold and the corresponding production data items of the network node response that deviates from the normal production state are used to form production quality warning information, and the production quality warning information is sent to the production equipment terminal for display as a quality alarm to form a production quality warning plan.

8. A production quality early warning system for limestone powder building materials, characterized in that: The production quality early warning system of limestone powder building materials comprises a storage device and a processor, wherein the storage device comprises a production quality early warning method program of limestone powder building materials, and when the production quality early warning method program of limestone powder building materials is executed by the processor, the following steps are implemented: Obtaining a preset number of limestone powder building material production samples and historical production data of the limestone powder building materials, performing a comprehensive quality assessment of the appearance and performance of the limestone powder building material production samples, and obtaining a comprehensive quality assessment result; Forming a one-to-one mapping relationship between the comprehensive quality assessment result and the historical production data, identifying the production quality characteristics of limestone powder building materials under different production data, and obtaining production data-production quality characteristic relationship data; Constructing a production quality mapping model of limestone powder building materials based on a self-organizing mapping network, importing the production data-production quality feature relationship data into the production quality mapping model for learning and training, and obtaining a complete production quality mapping model; Acquire the real-time production data of limestone powder building materials and import it into the production quality mapping model to perform data mapping, so as to obtain the real-time production quality network node response of limestone powder building materials; Compare the real-time production quality network node response with the network node response in a normal production state, and calculate the difference between the real-time production quality network node response and the network node response in a normal production state; According to the difference value, a production quality early warning is performed on the limestone powder building material to obtain a production quality early warning plan.

9. A production quality early warning system for limestone powder building materials according to claim 8, characterized in that: The comparing the real-time production quality network node response with the network node response in the normal production state, and calculating the difference between the real-time production quality network node response and the network node response in the normal production state, is specifically: Obtaining network node responses of limestone powder building materials in the production quality mapping model under normal production conditions; The Euclidean distance between the real-time production quality network node response and the network node response in the normal production state is calculated, and the difference value between the real-time production quality network node response and the network node response in the normal production state is evaluated according to the Euclidean distance.

10. A production quality early warning system for limestone powder building materials according to claim 8, characterized in that: The production quality warning of limestone powder building materials is performed according to the difference value to obtain a production quality warning plan, which is specifically: According to the production standard of limestone powder building materials, the early warning threshold of the difference between the real-time production quality network node response and the normal production state network node response is set; When the difference value is greater than the warning threshold, the trigger time when the difference value is greater than the warning threshold and the corresponding production data item of the network node response that deviates from the normal production state are recorded; The trigger time when the difference value is greater than the warning threshold and the corresponding production data items of the network node response that deviates from the normal production state are used to form production quality warning information, and the production quality warning information is sent to the production equipment terminal for display as a quality alarm to form a production quality warning plan.

Images