Automated processing method and system for quartz sand

By collecting and analyzing the temperature and thermal stress data of quartz sand, combining image data, and using deep learning technology to generate classification results, the quality problems caused by temperature changes during continuous annealing of quartz sand are solved, real-time monitoring and control are achieved, and the product is improved Quality and processing efficiency.

CN118005022BActive Publication Date: 2024-11-19凤阳昌荣粉体材料有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410154584.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-02-02
Publication Date
2024-11-19
Estimated Expiration
2044-02-02

AI Technical Summary

Technical Problem

During the continuous annealing process, quartz sand may cause particle cracking or fracture due to temperature changes and thermal stress. The lack of real-time monitoring and feedback mechanisms lead to a decline in product quality, making it difficult to evaluate and adjust the adjustment effect.

Method used

Collect the quartz sand temperature value and thermal stress data in the continuous annealing furnace, and combine the image data, feature extraction and analysis is performed through a time series encoder and a convolutional neural network model based on a mixed convolution layer to generate a classification for adjusting the annealing temperature. result.

Benefits of technology

Real-time monitoring and control of the continuous annealing process of quartz sand is achieved, processing efficiency and product quality are improved, and manual intervention and operation risks are reduced.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118005022B_ABST
    Figure CN118005022B_ABST
Patent Text Reader

Abstract

The present application relates to the field of automated processing of quartz sand, and specifically discloses a method and system for automated processing of quartz sand, which first collects the temperature value, thermal stress and corresponding multiple quartz sand image data of multiple predetermined time points in a continuous annealing furnace, then extracts the time series features of the temperature value and thermal stress of the quartz sand at multiple predetermined time points in the continuous annealing furnace through a time series encoder, and simultaneously uses a convolutional neural network model based on a mixed convolutional layer to extract and analyze the corresponding multiple quartz sand image data, and finally combines the feature information of the two to obtain a classification result for indicating whether the temperature of the continuous annealing furnace needs to be adjusted. In this way, according to the classification result, the continuous annealing process of quartz sand can be monitored and controlled in real time, thereby improving the processing efficiency and product quality of quartz sand.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of automated processing of quartz sand, and more specifically, to an automated processing method and system for quartz sand. Background Art

[0002] Quartz sand is a sedimentary rock mineral composed mainly of quartz (silicon dioxide). It is one of the most common minerals in the earth's crust and one of the most abundant minerals in nature. It is a non-metallic mineral composed mainly of quartz (silicon dioxide). Quartz sand is widely used in industry and construction. Due to its high temperature resistance, corrosion resistance and high mechanical strength, quartz sand is often used as a raw material for refractory materials, glass manufacturing, ceramic industry, casting, abrasives and grinding materials. In addition, quartz sand is also used in water treatment, oil field exploration, cosmetics, electronics industry (such as semiconductor manufacturing) and other fields.

[0003] During the processing of quartz sand, continuous annealing and water quenching are usually required. This is a common quartz sand processing method used to improve its physical and chemical properties. Continuous annealing refers to heating quartz sand to a certain temperature, maintaining it for a period of time, and then gradually cooling it. This process can be achieved through a continuous annealing furnace, in which quartz sand is continuously fed in and out of the furnace by a conveyor belt or other means. The purpose of continuous annealing is to eliminate internal stress in quartz sand and improve its thermal stability and mechanical strength.

[0004] However, during the continuous annealing process of quartz sand, due to temperature changes and thermal stress, quartz sand particles may crack or break. This may lead to quality degradation or damage of the product. To solve this problem, manufacturers usually adjust the annealing temperature and time to reduce the impact of thermal stress. However, the current adjustment methods usually lack real-time monitoring and feedback mechanisms. Manufacturers may not be able to accurately understand the temperature distribution and thermal stress state of quartz sand during the annealing process. The lack of real-time data and feedback information may result in the inability to timely evaluate and adjust the effect of the adjustment, which increases the difficulty and uncertainty of the adjustment.

[0005] Therefore, an automated processing method and system for quartz sand is expected to monitor and control the continuous annealing process of quartz sand in real time through deep learning technology, so as to improve the processing efficiency and product quality of quartz sand. Summary of the invention

[0006] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides an automated processing method and system for quartz sand, which first collects the quartz sand temperature value, thermal stress and corresponding multiple quartz sand image data at multiple predetermined time points in the continuous annealing furnace, and then extracts the time series features of the quartz sand temperature value and thermal stress at multiple predetermined time points in the continuous annealing furnace through a time series encoder, and uses a convolutional neural network model based on a mixed convolutional layer to extract and analyze the corresponding multiple quartz sand image data, and finally combines the feature information of the two to obtain a classification result for indicating whether the temperature of the continuous annealing furnace needs to be adjusted. In this way, according to the classification result, real-time monitoring and control of the continuous annealing process of quartz sand can be achieved, thereby improving the processing efficiency and product quality of quartz sand.

[0007] According to a first aspect of the present application, there is provided an automated processing method for quartz sand, comprising:

[0008] Collecting quartz sand temperature values, thermal stress and corresponding quartz sand image data at multiple predetermined time points in the continuous annealing furnace;

[0009] Extracting and analyzing the quartz sand temperature values ​​and thermal stress at multiple predetermined time points in the continuous annealing furnace and the corresponding multiple quartz sand image data to obtain a quartz sand temperature-thermal stress time series change characteristic diagram and a quartz sand state change characteristic diagram;

[0010] Based on the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram, a classification result is obtained.

[0011] In combination with the first aspect of the present application, in an automated processing method for quartz sand of the first aspect of the present application, the quartz sand temperature values, thermal stresses at multiple predetermined time points in the continuous annealing furnace and the corresponding multiple quartz sand image data are respectively feature extracted and analyzed to obtain a quartz sand temperature-thermal stress time series change characteristic diagram and a quartz sand state change characteristic diagram, including: performing deep convolution encoding on the quartz sand temperature values ​​and thermal stresses at multiple predetermined time points in the continuous annealing furnace to obtain the quartz sand temperature-thermal stress time series change characteristic diagram; performing deep convolution encoding on the corresponding multiple quartz sand image data to obtain the quartz sand state change characteristic diagram.

[0012] According to a second aspect of the present application, there is provided an automated processing system for quartz sand, comprising:

[0013] A quartz sand data acquisition module is used to collect quartz sand temperature values, thermal stresses and corresponding quartz sand image data at multiple predetermined time points in the continuous annealing furnace;

[0014] A quartz sand feature acquisition module, used for extracting and analyzing the features of the quartz sand temperature values ​​and thermal stress at multiple predetermined time points in the continuous annealing furnace and the corresponding multiple quartz sand image data to obtain a quartz sand temperature-thermal stress time series change feature diagram and a quartz sand state change feature diagram;

[0015] The quartz sand characteristic classification result generating module is used to obtain the classification result based on the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram.

[0016] In combination with the second aspect of the present application, in a quartz sand automated processing system of the second aspect of the present application, the quartz sand feature classification result generation module includes: performing a high-dimensional space unit manifold sub-dimension hyperconvex correlation measurement on the quartz sand temperature-thermal stress time series change feature diagram and the quartz sand state change feature diagram to obtain a quartz sand classification feature diagram; passing the quartz sand classification feature diagram through a classifier to obtain a classification result. The classification result is used to indicate whether the temperature of the continuous annealing furnace needs to be adjusted.

[0017] Compared with the prior art, the present application provides an automated processing method and system for quartz sand, which first collects the quartz sand temperature value, thermal stress and corresponding multiple quartz sand image data at multiple predetermined time points in the continuous annealing furnace, then extracts the time series features of the quartz sand temperature value and thermal stress at multiple predetermined time points in the continuous annealing furnace through a time series encoder, and uses a convolutional neural network model based on a mixed convolutional layer to extract and analyze the corresponding multiple quartz sand image data, and finally combines the feature information of the two to obtain a classification result for indicating whether the temperature of the continuous annealing furnace needs to be adjusted. In this way, according to the classification result, real-time monitoring and control of the continuous annealing process of quartz sand can be achieved, thereby improving the processing efficiency and product quality of quartz sand. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0019] Figure 1 The figure illustrates a flow chart of an automated processing method for quartz sand according to an embodiment of the present application.

[0020] Figure 2 The figure illustrates a flow chart of step S2 in the automated processing method of quartz sand according to an embodiment of the present application.

[0021] Figure 3 The figure illustrates a flow chart of step S21 in step S2 of the automated processing method for quartz sand according to an embodiment of the present application.

[0022] Figure 4 The figure illustrates a flow chart of step S22 in step S2 of the automated processing method of quartz sand according to an embodiment of the present application.

[0023] Figure 5 The figure illustrates a flow chart of step S3 in the automated processing method of quartz sand according to an embodiment of the present application.

[0024] Figure 6 A schematic block diagram of an automated processing system for quartz sand according to an embodiment of the present application is illustrated. DETAILED DESCRIPTION

[0025] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0026] Exemplary Methods

[0027] Figure 1 The figure shows a flow chart of the automated processing method of quartz sand according to an embodiment of the present application. Figure 1 As shown, the automated processing method of quartz sand according to the embodiment of the present application includes: S1, collecting quartz sand temperature values, thermal stresses and corresponding multiple quartz sand image data at multiple predetermined time points in a continuous annealing furnace; S2, respectively extracting and analyzing the quartz sand temperature values, thermal stresses and the corresponding multiple quartz sand image data at multiple predetermined time points in the continuous annealing furnace to obtain a quartz sand temperature-thermal stress time series change characteristic diagram and a quartz sand state change characteristic diagram; S3, based on the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram, to obtain a classification result.

[0028] Quartz sand is a common mineral, mainly composed of silicon dioxide (chemical formula: SiO2). It is one of the most common minerals in the earth's crust and is widely distributed in many places, including rivers, beaches, deserts, and mineral deposits. The particles of quartz sand are usually transparent or translucent, showing different colors such as white, gray, yellow, and pink. Its particle shape is usually hexagonal columnar and has good crystallinity. Quartz sand has a wide range of applications in industry. Due to its stable physical and chemical properties, high temperature resistance and corrosion resistance, it is widely used in glass manufacturing, ceramic industry, building materials, chemical raw materials, electronics industry and other fields. Quartz sand is also commonly used in water treatment, filtration materials and mortar.

[0029] The continuous annealing process of quartz sand is an important part of the quartz sand processing process. Continuous annealing refers to heating quartz sand to a certain temperature within a certain temperature range and maintaining it for a period of time, and then gradually cooling it to room temperature. This process can improve the physical properties and chemical stability of quartz sand to meet the needs of different application fields.

[0030] However, as described in the above background technology, during the continuous annealing process of quartz sand, due to temperature changes and thermal stress, quartz sand particles may crack or break. This may lead to a decline in product quality or damage. In order to solve this problem, manufacturers usually adjust the annealing temperature and time to reduce the impact of thermal stress. However, the current adjustment method usually lacks a real-time monitoring and feedback mechanism. Manufacturers may not be able to accurately understand the temperature distribution and thermal stress state of quartz sand during the annealing process. The lack of real-time data and feedback information may result in the inability to timely evaluate and adjust the effect of the adjustment, increasing the difficulty and uncertainty of the adjustment. Therefore, an automated processing method and system for quartz sand are expected to monitor and control the continuous annealing process of quartz sand in real time through deep learning technology to improve the processing efficiency and product quality of quartz sand.

[0031] Deep learning is a branch of machine learning that uses multi-level neural networks to learn and solve complex pattern recognition and decision-making problems by simulating the structure and function of human brain neural networks. Deep learning technology has made significant breakthroughs in recent years and has demonstrated powerful capabilities and wide applications in various fields.

[0032] The core of deep learning is the Artificial Neural Network (ANN), which consists of multiple neurons (nodes) and the weights that connect them. Each neuron receives input from the previous layer of neurons, processes the input through an activation function, and then passes the output to the next layer of neurons. Through multiple levels of connections and weight adjustments, the neural network can learn the complex features and patterns of the input data.

[0033] One of the great advantages of deep learning technology is that it can automatically learn feature representations from raw data without manually designing features. This has enabled deep learning to achieve major breakthroughs in areas such as image recognition, speech recognition, and natural language processing. For example, in image recognition, deep learning can automatically learn features in images by training neural networks, thereby achieving high-precision object recognition and classification.

[0034] In this system, deep learning technology plays an important role. Deep learning technology can realize real-time monitoring and control of the continuous annealing process of quartz sand, improve processing efficiency and product quality. This automated processing method and system can reduce manual intervention, reduce operational risks, and improve processing stability and consistency. Therefore, the input data in the embodiment of this application is feature extracted and analyzed by using deep learning technology.

[0035] In step S1, the quartz sand temperature values, thermal stress and corresponding multiple quartz sand image data at multiple predetermined time points in the continuous annealing furnace are collected. It should be understood that temperature is one of the important parameters affecting the quartz sand annealing process. By collecting the quartz sand temperature values ​​in the continuous annealing furnace, the temperature distribution of the quartz sand at different positions and time points can be understood. This helps to determine whether the heating, insulation and cooling processes of the quartz sand meet the requirements, and adjust the temperature parameters in time to avoid the cracking or breaking of the quartz sand particles due to temperature changes. At the same time, during the continuous annealing process of quartz sand, the change in temperature will cause the quartz sand particles to undergo thermal expansion or thermal contraction, thereby causing the generation of thermal stress. Specifically, when the quartz sand is heated, the molecular vibration inside the particles increases, causing the particle volume to expand. On the contrary, when the quartz sand is cooled, the particle volume shrinks. This thermal expansion and thermal contraction process will cause stress to be generated inside the quartz sand particles. And the quartz sand particles may be affected by the thermal stress and crack or break. Therefore, it is also necessary to collect the thermal stress during the quartz sand annealing process. In addition, by collecting image data of quartz sand through the camera, the morphology and state of quartz sand particles can be observed and analyzed in real time. The image data of quartz sand can provide more intuitive information, such as particle shape, particle size and particle distribution. This information can be combined with temperature and thermal stress data to help determine whether there are abnormalities or problems with quartz sand particles, and take timely measures to adjust and correct them.

[0036] In particular, when collecting the temperature values ​​and thermal stress of the quartz sand at multiple predetermined time points in the continuous annealing furnace, a temperature sensor and a stress sensor can be installed in the continuous annealing furnace to monitor the temperature and stress of the quartz sand in real time. The sensor can be a thermocouple, an infrared sensor, a pressure sensor, etc. These sensors can convert the numerical values ​​of temperature and stress into electrical signals, and record and transmit them through a data acquisition device. In addition, when collecting the corresponding multiple quartz sand image data, a high-resolution industrial camera or camera can be used to capture the image of the quartz sand. The camera can be installed in an appropriate position in or near the continuous annealing furnace to obtain a clear image of the quartz sand. The camera can communicate with the data acquisition device via a cable or wireless connection.

[0037] In step S2, the quartz sand temperature values, thermal stress and the corresponding multiple quartz sand image data at multiple predetermined time points in the continuous annealing furnace are feature extracted and analyzed to obtain a quartz sand temperature-thermal stress time series change feature diagram and a quartz sand state change feature diagram. It should be understood that after obtaining these input data, the next step is to perform feature extraction and analysis on these input data.

[0038] Specifically, Figure 2 FIG. 1 is a flow chart of step S2 in the automated processing method of quartz sand according to an embodiment of the present application. Figure 2 As shown, step S2, respectively extracting and analyzing the features of the quartz sand temperature values, thermal stresses and the corresponding multiple quartz sand image data at multiple predetermined time points in the continuous annealing furnace to obtain a quartz sand temperature-thermal stress time series change characteristic diagram and a quartz sand state change characteristic diagram, including: S21, respectively performing deep convolution coding on the quartz sand temperature values ​​and thermal stresses at multiple predetermined time points in the continuous annealing furnace to obtain the quartz sand temperature-thermal stress time series change characteristic diagram; S22, performing deep convolution coding on the corresponding multiple quartz sand image data to obtain the quartz sand state change characteristic diagram.

[0039] First, the temperature values ​​and thermal stress of the quartz sand at multiple predetermined time points in the continuous annealing furnace are feature extracted and analyzed. Specifically, Figure 3 FIG. 1 is a flow chart of step S21 in step S2 of the automated processing method for quartz sand according to an embodiment of the present application. Figure 3 As shown, step S21, the quartz sand temperature values ​​and thermal stresses at multiple predetermined time points in the continuous annealing furnace are respectively deep convolutionally encoded to obtain the quartz sand temperature-thermal stress time series change characteristic diagram, including: S211, the quartz sand temperature values ​​and thermal stresses at multiple predetermined time points in the continuous annealing furnace are respectively passed through a time series feature extraction module based on a time series encoder to obtain a quartz sand temperature time series feature vector and a quartz sand thermal stress time series feature vector; S212, using a joint encoder to perform feature combination on the quartz sand temperature time series feature vector and the quartz sand thermal stress time series feature vector to obtain a quartz sand temperature-thermal stress time series feature matrix; S213, the quartz sand temperature-thermal stress time series feature matrix is ​​passed through a quartz sand parameter change feature extraction module based on a convolutional neural network model to obtain the quartz sand temperature-thermal stress time series change characteristic diagram.

[0040] It should be understood that temperature and thermal stress usually change over time during the continuous annealing process. By using a time series feature extraction module, important change patterns and trends can be captured from the time series data of temperature and thermal stress. These features can provide useful information about the quartz sand annealing process. Therefore, in an embodiment of the present application, the quartz sand temperature values ​​and thermal stresses at multiple predetermined time points in the continuous annealing furnace are respectively deep convolutionally encoded by a time series feature extraction module based on a time series encoder to extract their time series features.

[0041] Specifically, step S211, the quartz sand temperature values ​​and thermal stresses at multiple predetermined time points in the continuous annealing furnace are respectively passed through a timing feature extraction module based on a timing encoder to obtain a quartz sand temperature timing feature vector and a quartz sand thermal stress timing feature vector, which are used to: arrange the quartz sand temperature values ​​and thermal stresses at multiple predetermined time points in the continuous annealing furnace as a temperature input vector and a thermal stress input vector, respectively; use the fully connected layer of the timing encoder to fully connect the temperature input vector and the thermal stress input vector to extract high-dimensional implicit features of the eigenvalues ​​at each position in the temperature input vector and the thermal stress input vector; and use the one-dimensional convolution layer of the timing encoder to perform one-dimensional convolution encoding on the temperature input vector and the thermal stress input vector to extract high-dimensional implicit association features of the association between the eigenvalues ​​at each position in the temperature input vector and the thermal stress input vector.

[0042] Next, it should be understood that in the continuous annealing process of quartz sand, temperature and thermal stress are two related parameters, and there is a certain correlation between them. For example, an increase in temperature will cause thermal expansion of the material, that is, the volume of the object will increase with the increase in temperature. This thermal expansion will cause stress changes inside the material, resulting in the generation of thermal stress. Therefore, in order to better capture the correlation characteristics between the two, the quartz sand temperature time series feature vector and the quartz sand thermal stress time series feature vector are feature-joined through a joint encoder. In this way, the correlation information between them can be captured, further improving the monitoring and control capabilities of the quartz sand annealing process.

[0043] Specifically, S212, using a joint encoder to perform feature combination on the quartz sand temperature time series feature vector and the quartz sand thermal stress time series feature vector to obtain a quartz sand temperature-thermal stress time series feature matrix, for: using the joint encoder to jointly encode the quartz sand temperature time series feature vector and the quartz sand thermal stress time series feature vector using the following formula to generate the quartz sand temperature-thermal stress time series feature matrix; wherein the formula is: in represents vector multiplication, M represents the quartz sand temperature-thermal stress time series characteristic matrix, V a1represents the time series characteristic vector of the quartz sand temperature, V a2 represents the time series characteristic vector of thermal stress of quartz sand, represents the transpose of the quartz sand thermal stress time series eigenvector.

[0044] Further, it should be understood that convolutional neural networks (CNNs) have excellent performance in the field of image processing and can effectively extract local features. In the temperature-thermal stress time series feature matrix of quartz sand, features at different positions may have different effects on the annealing process of quartz sand. Therefore, in order to better capture the local patterns and laws of the time series changes of the temperature-thermal stress of quartz sand, the quartz sand temperature-thermal stress time series feature matrix is ​​deeply convolutionally encoded by a quartz sand parameter change feature extraction module based on a convolutional neural network model.

[0045] Specifically, S213, the quartz sand temperature-thermal stress time series feature matrix is ​​passed through the quartz sand parameter change feature extraction module based on the convolutional neural network model to obtain the quartz sand temperature-thermal stress time series change feature map, and the specific feature extraction steps are: 1. Data preparation: The time series data of quartz sand temperature and thermal stress are sorted into a feature matrix, in which each row represents a time step and each column represents a feature. Ensure that the format of the data is suitable for the input of the convolutional neural network. 2. Feature extraction module construction: Design a quartz sand parameter change feature extraction module based on a convolutional neural network (CNN). The module should include components such as convolutional layers, pooling layers, and activation functions for learning and extracting the characteristics of the quartz sand temperature-thermal stress time series changes. 3. Feature extraction process: The quartz sand temperature-thermal stress time series feature matrix is ​​input into the feature extraction module. Through convolution operations and pooling operations, the model will extract key time series change features from the time series feature matrix. 4. Feature map generation: According to the output of the feature extraction module, a quartz sand temperature-thermal stress time series change feature map is generated. A feature map can be a two-dimensional matrix where each element represents the strength or importance of a feature.

[0046] It should be noted that the specific feature extraction steps may vary depending on the specific application scenario and data characteristics. When designing the feature extraction module, you can choose the appropriate convolutional neural network structure, activation function and other parameters according to the actual situation. In addition, you can also optimize the effect of feature extraction by adjusting the model's hyperparameters and training.

[0047] Then, feature extraction and analysis are performed on the corresponding multiple quartz sand image data. Specifically, Figure 4 FIG. 1 is a flowchart of step S22 in step S2 of the automated processing method for quartz sand according to an embodiment of the present application. Figure 4As shown, step S22, performing deep convolution encoding on the corresponding multiple quartz sand image data to obtain the quartz sand state change characteristic diagram, includes: S221, passing the corresponding multiple quartz sand image data through a quartz sand state feature extraction module based on a hybrid convolution layer to obtain multiple quartz sand state feature matrices; S222, arranging the multiple quartz sand state feature matrices according to the time dimension to obtain a three-dimensional tensor of quartz sand state change, and then passing the quartz sand state change feature extraction module based on a three-dimensional convolutional neural network model to obtain a quartz sand state change characteristic diagram.

[0048] It should be understood that in order to better understand and analyze the state changes of quartz sand, it is necessary to extract key features from the quartz sand image data. The feature extraction module based on the hybrid convolution layer can learn and extract important features at different levels in the image data, such as the texture, particle distribution, temperature distribution, etc. of quartz sand. Therefore, in order to more comprehensively understand the state changes of quartz sand, the corresponding multiple quartz sand image data are respectively feature extracted and analyzed by the quartz sand state feature extraction module based on the hybrid convolution layer.

[0049] Specifically, step S221, the corresponding multiple quartz sand image data are passed through a quartz sand state feature extraction module based on a hybrid convolutional layer to obtain multiple quartz sand state feature matrices, which are used to: use each hybrid convolutional layer of the quartz sand state feature extraction module to respectively perform on the input data in the forward transfer of the layer: use a first convolution kernel with a first size to perform convolution encoding on each quartz sand image data in the corresponding multiple quartz sand image data to obtain a first scale feature map; use a second convolution kernel with a first void rate to perform convolution encoding on each quartz sand image data in the corresponding multiple quartz sand image data to obtain a second scale feature map; use a third convolution kernel with a second void rate to perform convolution encoding on each quartz sand image data in the corresponding multiple quartz sand image data to obtain a third scale feature map; use a third convolution kernel with a third void rate to perform convolution encoding on each quartz sand image data in the corresponding multiple quartz sand image data to obtain a third scale feature map; Four convolution kernels perform convolution encoding on each of the corresponding multiple quartz sand image data to obtain a fourth scale feature map, wherein the first convolution kernel, the second convolution kernel, the third convolution kernel and the fourth convolution kernel have the same size, and the second convolution kernel, the third convolution kernel and the fourth convolution kernel have different void rates; the first scale feature map, the second scale feature map, the third scale feature map and the fourth scale feature map are aggregated along the channel dimension to obtain an aggregated feature map; the aggregated feature map is pooled to generate a pooled feature map; and the pooled feature map is activated to generate an activated feature map; wherein the output of the last layer of the quartz sand state feature extraction module based on the hybrid convolution layer is each quartz sand state feature matrix in the multiple quartz sand state feature matrices.

[0050] Next, considering that the state of quartz sand changes over time during the continuous annealing process, multiple quartz sand state feature matrices are arranged into a three-dimensional tensor according to the time dimension, which can retain the information of the change of quartz sand state over time. Through the feature extraction module based on the three-dimensional convolutional neural network model, the key features of the quartz sand state change can be extracted from this three-dimensional tensor. Therefore, firstly, multiple quartz sand state feature matrices are arranged into a three-dimensional tensor of quartz sand state change according to the time dimension.

[0051] Further, it should be understood that the quartz sand state change characteristic diagram can provide more intuitive and explainable information. By performing convolution operations and pooling operations on the three-dimensional tensor of the quartz sand state change, it can be converted into a two-dimensional feature map. These feature maps can show the spatial distribution and intensity of the quartz sand state change, which helps to understand and analyze the annealing process of quartz sand. Therefore, the quartz sand state change three-dimensional tensor is further deep convolutionally encoded by a quartz sand state change feature extraction module based on a three-dimensional convolutional neural network model to obtain a quartz sand state change characteristic diagram. This helps to optimize processing parameters, improve processing efficiency, and ensure the quality of quartz sand products.

[0052] Specifically, step S222, after arranging the multiple quartz sand state feature matrices according to the time dimension to obtain a three-dimensional tensor of quartz sand state change, a quartz sand state change feature extraction module based on a three-dimensional convolutional neural network model is used to obtain a quartz sand state change feature map, which is used to: use each layer of the three-dimensional convolutional neural network model to perform the following on the input data in the forward transfer of the layer: perform convolution processing based on a three-dimensional convolution kernel on the input data to obtain a convolution feature map; perform pooling processing based on a local feature matrix on the convolution feature map to obtain a pooling feature map; perform nonlinear activation on the pooling feature map to obtain an activation feature map; wherein the input of the first layer of the three-dimensional convolutional neural network model is the three-dimensional tensor of quartz sand state change, and the output of the last layer of the three-dimensional convolutional neural network model is the quartz sand state change feature map.

[0053] In step S3, a classification result is obtained based on the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram.

[0054] Specifically, Figure 5 FIG. 1 is a flow chart of step S3 in the automated processing method of quartz sand according to an embodiment of the present application. Figure 5As shown, step S3, based on the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram, to obtain a classification result, includes: S31, performing a high-dimensional space unit manifold sub-dimension hyperconvex correlation measurement on the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram to obtain a quartz sand classification characteristic diagram; S32, passing the quartz sand classification characteristic diagram through a classifier to obtain a classification result. Wherein, the classification result is used to indicate whether the temperature of the continuous annealing furnace needs to be adjusted.

[0055] In particular, in the technical solution of the present application, the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram respectively represent the temperature, thermal stress and state information of the quartz sand in the continuous annealing furnace. The quartz sand temperature-thermal stress time series change characteristic diagram is obtained by extracting the time series features and combining the features of the quartz sand temperature and thermal stress, and is used to represent the change of the temperature and thermal stress of the quartz sand over time. The quartz sand state change characteristic diagram is obtained by extracting the features and performing three-dimensional convolution processing on the quartz sand image data, and is used to represent the state change of the quartz sand. Since the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram are derived from different data sources and processing methods, they may have differences in dimension and scale in the high-dimensional feature space unit. Specifically, the quartz sand temperature-thermal stress time series change characteristic diagram is obtained by extracting the time series features and combining the encoding of the temperature and thermal stress, and may have a higher dimension and scale. The quartz sand state change characteristic diagram is obtained by extracting the features and performing three-dimensional convolution processing on the quartz sand image data, and may have a lower dimension and scale. When the two feature maps are fused, some technical problems may arise due to the differences in dimension and scale. For example, during the fusion process, due to the difference in dimension, it may be difficult to align the features, that is, how to align the two feature maps in a high-dimensional space so that they can meaningfully influence each other. In addition, due to the difference in scale, the weight distribution during feature fusion may be uneven, thus affecting the final classification result. This difference may also lead to local structural collapse in the feature space, that is, during the fusion process, the structural information of some local features may be lost or become unstable. In order to solve these problems, in the technical solution of the present application, the present application performs a high-dimensional space unit manifold sub-dimensional hyperconvex correlation measurement on the quartz sand temperature-thermal stress time series change feature map and the quartz sand state change feature map to obtain a quartz sand classification feature map.

[0056] Specifically, step S31, performs high-dimensional space unit manifold sub-dimension hyperconvex correlation measurement on the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram to obtain a quartz sand classification characteristic diagram, including: calculating the position mean characteristic diagram between the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram; calculating the position difference characteristic diagram between the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram; respectively calculating the difference characteristic diagram between the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram and the position mean characteristic diagram. To obtain a first differential off-center feature map and a second differential off-center feature map; calculate the logarithmic function value with base 2 of the feature value of each position in the first differential off-center feature map and the second differential off-center feature map to obtain the first logarithmic differential off-center feature map and the second logarithmic differential off-center feature map; divide the first logarithmic differential off-center feature map by the position-based sum feature map between the first logarithmic differential off-center feature map and the second logarithmic differential off-center feature map to obtain a correction feature map; and calculate the position-based weighted sum between the correction feature map and the logarithmic position-based differential feature map of the position-based differential feature map to obtain the quartz sand classification feature map.

[0057] That is, considering that there are differences in dimension and scale in the characteristic manifolds of the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram in the high-dimensional feature space unit, in the process of fusing the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram, there will be technical problems such as local structural collapse or pathological alignment of the fused quartz sand temperature-thermal stress time series change characteristic diagram due to the differences in dimension and scale.

[0058] In response to the above technical problems, in the technical solution of the present application, the quartz sand temperature-thermal stress time series change characteristic graph and the quartz sand state change characteristic graph are measured by hyperconvex correlation of the sub-dimension of the unit manifold in high-dimensional space. The mean characteristic graph of the quartz sand temperature-thermal stress time series change characteristic graph and the quartz sand state change characteristic graph are used as the pseudo-clustering center of the characteristic manifold, and a hyperconvex correlation measurement function of the characteristic manifold based on the characteristic graph is constructed, so that the eigenvalue of each position between the characteristic graphs can maintain consistency with the pseudo-clustering center of the characteristic manifold in its sub-dimension, thereby achieving hyperconvex correlation matching of the characteristic manifold of the characteristic graph, which effectively measures the similarity and difference of the characteristic manifolds between the feature graphs, enhances the hyperconvex correlation of the characteristic manifolds of the feature graphs, and improves the robustness and accuracy of the characteristic manifolds of the feature graphs.

[0059] Furthermore, considering that the classifier is a machine learning model, it can classify and judge the input data according to its characteristics. By training the classifier, it can learn the characteristic representation and distinguishing ability of different types of quartz sand, thereby realizing the classification of the quartz sand state characteristic diagram. Therefore, the quartz sand classification characteristic diagram is further characterized by the classifier. That is, by using the classifier, the quartz sand is classified according to the characteristics of the quartz sand state characteristic diagram, and the corresponding classification results are obtained. In this way, according to the state of the quartz sand, the annealing parameters can be adjusted in time, thereby improving the processing efficiency and product quality of the quartz sand.

[0060] Specifically, step S32, passing the quartz sand classification feature map through a classifier to obtain a classification result, is used to: use the classifier to process the quartz sand classification feature map using the following formula to obtain the classification result; wherein the formula is: O = softmax{(W c ,B c )|Project(F b )}, where Project(F b ) represents the projection of the quartz sand classification feature map into a vector, W c is the weight matrix, B c represents a bias vector, softmax represents a normalized exponential function, and O represents the classification result.

[0061] It is worth mentioning that in addition to using a classifier to classify the quartz sand classification feature map, a clustering algorithm can also be used. Clustering is an unsupervised learning method that can divide samples in a data set into different groups with similar characteristics. In the automated processing method of quartz sand, the quartz sand state feature map can be used as input data, and a clustering algorithm can be used to perform cluster analysis on it to discover potential patterns and similarities therein.

[0062] The specific steps are as follows: 1. Feature extraction: First, use deep learning technology to extract the feature representation of the quartz sand state feature map. This can be achieved through methods such as convolutional neural networks to convert the feature map into a feature vector with a fixed dimension. 2. Clustering algorithm: Select an appropriate clustering algorithm, such as K-means clustering, hierarchical clustering, DBSCAN, etc. These algorithms can group the quartz sand state feature maps according to the similarity measure between samples. 3. Clustering result analysis: Analyze and interpret the clustering results to identify different clusters. Each cluster represents a specific type of quartz sand state feature map. The clustering results can be understood by the sample similarity within the cluster and the sample difference between the clusters. 4. Monitoring and control: According to the clustering results, corresponding monitoring and control strategies can be adopted for different types of quartz sand state feature maps. For example, for the quartz sand in a certain cluster, the parameters in the annealing process can be adjusted or other control measures can be taken to optimize the processing efficiency and product quality of quartz sand.

[0063] Using clustering algorithms to cluster the quartz sand state characteristic graphs can reveal potential patterns and similarities, helping to understand and analyze the state changes of quartz sand. This implementation method can provide additional information for real-time monitoring and control of quartz sand and support more refined processing management and optimization decisions.

[0064] In summary, the automated processing method of the quartz sand according to the embodiment of the present application is explained, which first collects the quartz sand temperature value, thermal stress and corresponding multiple quartz sand image data at multiple predetermined time points in the continuous annealing furnace, and then extracts the time series features of the quartz sand temperature value and thermal stress at multiple predetermined time points in the continuous annealing furnace through a time series encoder, and uses a convolutional neural network model based on a mixed convolutional layer to extract and analyze the corresponding multiple quartz sand image data, and finally combines the feature information of the two to obtain a classification result for indicating whether the temperature of the continuous annealing furnace needs to be adjusted. In this way, according to the classification results, real-time monitoring and control of the continuous annealing process of quartz sand can be achieved, thereby improving the processing efficiency and product quality of quartz sand.

[0065] Exemplary Systems

[0066] Figure 6 FIG. 1 is a schematic block diagram of an automated processing system for quartz sand according to an embodiment of the present application. Figure 6As shown, the automated processing system 100 of quartz sand according to the embodiment of the present application includes: a quartz sand data acquisition module 110, which is used to collect quartz sand temperature values, thermal stresses and corresponding multiple quartz sand image data at multiple predetermined time points in the continuous annealing furnace; a quartz sand feature acquisition module 120, which is used to extract and analyze the quartz sand temperature values, thermal stresses and the corresponding multiple quartz sand image data at multiple predetermined time points in the continuous annealing furnace to obtain a quartz sand temperature-thermal stress time series change characteristic diagram and a quartz sand state change characteristic diagram; a quartz sand feature classification result generation module 130, which is used to obtain a classification result based on the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram.

[0067] As described above, the automated processing system 100 for quartz sand according to the embodiment of the present application can be implemented in various wireless terminals, such as a server for automated processing of quartz sand. In one example, the automated processing system 100 for quartz sand according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the automated processing system 100 for quartz sand can be a software module in the operating system of the wireless terminal, or can be an application developed for the wireless terminal; of course, the automated processing system 100 for quartz sand can also be one of the many hardware modules of the wireless terminal.

[0068] Alternatively, in another example, the automated quartz sand processing system 100 and the wireless terminal may also be separate devices, and the automated quartz sand processing system 100 may be connected to the wireless terminal via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0069] Here, those skilled in the art will appreciate that the specific functions and operations of each step in the above-mentioned automated processing system for quartz sand have been described in detail above. Figure 1 The description of the automated processing method of quartz sand has been described in detail, and therefore, its repeated description will be omitted.

[0070] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0071] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0072] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the system 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 integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0073] 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 on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0074] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0075] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0076] 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. An automated processing method for quartz sand, characterized in that: include: Collecting quartz sand temperature values, thermal stress and corresponding quartz sand image data at multiple predetermined time points in the continuous annealing furnace; Extracting and analyzing the quartz sand temperature values ​​and thermal stress at multiple predetermined time points in the continuous annealing furnace and the corresponding multiple quartz sand image data to obtain a quartz sand temperature-thermal stress time series change characteristic diagram and a quartz sand state change characteristic diagram; Based on the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram, a classification result is obtained; Wherein, based on the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram, the classification result is obtained, including: Performing high-dimensional space unit manifold sub-dimension hyperconvex correlation measurement on the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram to obtain a quartz sand classification characteristic diagram; Passing the quartz sand classification characteristic diagram through a classifier to obtain a classification result; Among them, the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram are subjected to high-dimensional space unit manifold sub-dimension hyperconvex correlation measurement to obtain a quartz sand classification characteristic diagram, including: Calculating a position mean characteristic diagram between the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram; Calculating a position-based differential characteristic diagram between the quartz sand temperature-thermal stress time series variation characteristic diagram and the quartz sand state variation characteristic diagram; Respectively calculating the quartz sand temperature-thermal stress time series change characteristic diagram and the difference characteristic diagram between the quartz sand state change characteristic diagram and the position mean characteristic diagram to obtain a first difference off-center characteristic diagram and a second difference off-center characteristic diagram; Calculating a logarithmic function value with base 2 of the feature value at each position in the first differential off-center feature map and the second differential off-center feature map to obtain a first logarithmic difference off-center feature map and a second logarithmic difference off-center feature map; Dividing the first logarithmic difference off-center feature map by the position-based sum feature map between the first logarithmic difference off-center feature map and the second logarithmic difference off-center feature map to obtain a correction feature map; and A position-based weighted sum between the correction feature map and the logarithmic position-based differential feature map of the position-based differential feature map is calculated to obtain the quartz sand classification feature map.

2. The automated processing method for quartz sand according to claim 1, characterized in that: The quartz sand temperature values, thermal stress and the corresponding multiple quartz sand image data at multiple predetermined time points in the continuous annealing furnace are respectively subjected to feature extraction and analysis to obtain a quartz sand temperature-thermal stress time series change characteristic diagram and a quartz sand state change characteristic diagram, including: The quartz sand temperature values ​​and thermal stress at a plurality of predetermined time points in the continuous annealing furnace are respectively subjected to deep convolution coding to obtain the quartz sand temperature-thermal stress time series variation characteristic diagram; The corresponding multiple quartz sand image data are subjected to deep convolution coding to obtain the quartz sand state change characteristic diagram.

3. The automated processing method for quartz sand according to claim 2, characterized in that: The quartz sand temperature values ​​and thermal stresses at a plurality of predetermined time points in the continuous annealing furnace are respectively subjected to deep convolution coding to obtain the quartz sand temperature-thermal stress time series variation characteristic diagram, including: The quartz sand temperature values ​​and thermal stress at multiple predetermined time points in the continuous annealing furnace are respectively passed through a time series feature extraction module based on a time series encoder to obtain a quartz sand temperature time series feature vector and a quartz sand thermal stress time series feature vector; Using a joint encoder, performing feature combination on the quartz sand temperature time series feature vector and the quartz sand thermal stress time series feature vector to obtain a quartz sand temperature-thermal stress time series feature matrix; The quartz sand temperature-thermal stress time series feature matrix is ​​passed through a quartz sand parameter change feature extraction module based on a convolutional neural network model to obtain the quartz sand temperature-thermal stress time series change feature diagram.

4. The automated processing method for quartz sand according to claim 3, characterized in that: Using a joint encoder to perform feature combination on the quartz sand temperature time series feature vector and the quartz sand thermal stress time series feature vector to obtain a quartz sand temperature-thermal stress time series feature matrix, specifically using the joint encoder to jointly encode the quartz sand temperature time series feature vector and the quartz sand thermal stress time series feature vector using the following formula to generate the quartz sand temperature-thermal stress time series feature matrix; Wherein, the formula is: in represents vector multiplication, represents the quartz sand temperature-thermal stress time series characteristic matrix, represents the time series characteristic vector of the quartz sand temperature, represents the time series characteristic vector of thermal stress of quartz sand, represents the transpose of the quartz sand thermal stress time series eigenvector.

5. The automated processing method for quartz sand according to claim 4, characterized in that: The corresponding multiple quartz sand image data are subjected to deep convolution coding to obtain the quartz sand state change characteristic diagram, including: The corresponding multiple quartz sand image data are passed through a quartz sand state feature extraction module based on a hybrid convolutional layer to obtain multiple quartz sand state feature matrices; The multiple quartz sand state feature matrices are arranged according to the time dimension to obtain a three-dimensional tensor of quartz sand state change, and then a quartz sand state change feature extraction module based on a three-dimensional convolutional neural network model is used to obtain a quartz sand state change feature graph.

6. The automated processing method for quartz sand according to claim 5, characterized in that: The corresponding multiple quartz sand image data are passed through a quartz sand state feature extraction module based on a hybrid convolutional layer to obtain multiple quartz sand state feature matrices, specifically, each hybrid convolutional layer of the quartz sand state feature extraction module is used to perform the following operations on the input data in the forward pass of the layer: Using a first convolution kernel having a first size, convolution encoding is performed on each of the corresponding plurality of quartz sand image data to obtain a first scale feature map; Using a second convolution kernel with a first void ratio, convolution encoding is performed on each of the corresponding multiple quartz sand image data to obtain a second scale feature map; Using a third convolution kernel with a second void ratio, convolution encoding is performed on each of the corresponding multiple quartz sand image data to obtain a third scale feature map; Using a fourth convolution kernel with a third void rate to perform convolution encoding on each of the corresponding multiple quartz sand image data to obtain a fourth scale feature map, wherein the first convolution kernel, the second convolution kernel, the third convolution kernel and the fourth convolution kernel have the same size, and the second convolution kernel, the third convolution kernel and the fourth convolution kernel have different void rates; Aggregating the first scale feature map, the second scale feature map, the third scale feature map, and the fourth scale feature map along the channel dimension to obtain an aggregated feature map; Performing pooling processing on the aggregated feature map to generate a pooled feature map; and Performing activation processing on the pooled feature map to generate an activation feature map; Among them, the output of the last layer of the quartz sand state feature extraction module based on the mixed convolutional layer is each quartz sand state feature matrix in the multiple quartz sand state feature matrices.

7. The automated processing method for quartz sand according to claim 6, characterized in that: The classification result is used to indicate whether the temperature of the continuous annealing furnace needs to be adjusted.

8. An automated processing system for quartz sand, characterized in that ,include: A quartz sand data acquisition module is used to collect quartz sand temperature values, thermal stresses and corresponding quartz sand image data at multiple predetermined time points in the continuous annealing furnace; A quartz sand feature acquisition module, used for extracting and analyzing the features of the quartz sand temperature values ​​and thermal stress at multiple predetermined time points in the continuous annealing furnace and the corresponding multiple quartz sand image data to obtain a quartz sand temperature-thermal stress time series change feature diagram and a quartz sand state change feature diagram; A quartz sand characteristic classification result generation module is used to obtain a classification result based on the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram; Wherein, based on the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram, the classification result is obtained, including: Performing high-dimensional space unit manifold sub-dimension hyperconvex correlation measurement on the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram to obtain a quartz sand classification characteristic diagram; Passing the quartz sand classification characteristic diagram through a classifier to obtain a classification result; Among them, the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram are subjected to high-dimensional space unit manifold sub-dimension hyperconvex correlation measurement to obtain a quartz sand classification characteristic diagram, including: Calculating a position mean characteristic diagram between the quartz sand temperature-thermal stress time series change characteristic diagram and the quartz sand state change characteristic diagram; Calculating a position-based differential characteristic diagram between the quartz sand temperature-thermal stress time series variation characteristic diagram and the quartz sand state variation characteristic diagram; Respectively calculating the quartz sand temperature-thermal stress time series change characteristic diagram and the difference characteristic diagram between the quartz sand state change characteristic diagram and the position mean characteristic diagram to obtain a first difference off-center characteristic diagram and a second difference off-center characteristic diagram; Calculating a logarithmic function value with base 2 of the feature value at each position in the first differential off-center feature map and the second differential off-center feature map to obtain a first logarithmic difference off-center feature map and a second logarithmic difference off-center feature map; Dividing the first logarithmic difference off-center feature map by the position-based sum feature map between the first logarithmic difference off-center feature map and the second logarithmic difference off-center feature map to obtain a correction feature map; and A position-based weighted sum between the correction feature map and the logarithmic position-based differential feature map of the position-based differential feature map is calculated to obtain the quartz sand classification feature map.