Intelligent control system and method for rubber product vulcanization process

Abstract

The present application relates to the field of intelligent monitoring technology, and more specifically discloses an intelligent control system and method for the vulcanization process of rubber products. The video of the vulcanization process of rubber products is monitored in real time by a camera, and the change characteristics and temperature characteristics of the vulcanization process are extracted in combination with the temperature data collected by the temperature sensor. Then, a classifier is used to determine whether the temperature of the vulcanization process should be increased or decreased, thereby achieving the purpose of intelligently controlling the vulcanization process of rubber products.

Description

Technical Field

[0001] The present application relates to the field of intelligent monitoring technology, and more specifically, to an intelligent control system and method for a rubber product vulcanization process. Background Art

[0002] Vulcanization is one of the most important links in the production process of rubber products. Only after the raw rubber macromolecules are vulcanized and cross-linked to form macromolecules with a three-dimensional network structure, can they obtain excellent high elasticity and high strength and become materials with practical use value.

[0003] With the development of the times, the types of rubber products are becoming more and more abundant, and people's requirements for product quality are also increasing. Vulcanization is an important part of rubber product production, and the quality of vulcanization seriously affects the quality of the product. At present, the vulcanization of rubber products has the problems of low automation, low vulcanization efficiency, low equipment utilization, and high labor intensity for workers.

[0004] Therefore, an intelligent control system and method for the vulcanization process of rubber products is desired. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides an intelligent control system and method for the vulcanization process of rubber products, which uses a camera and a temperature sensor to monitor the rubber vulcanization process, and realizes intelligent control by feature extraction and classifier judgment of temperature control.

[0006] Accordingly, according to one aspect of the present application, an intelligent control system for a rubber product vulcanization process is provided, comprising:

[0007] The rubber product vulcanization process acquisition module is used to acquire the video of the rubber product vulcanization process through a camera to obtain the monitoring video of the rubber product vulcanization process, and to acquire the temperature values ​​of the rubber product vulcanization process at multiple predetermined time points through a temperature sensor;

[0008] A rubber product vulcanization process processing module, used to extract a rubber product vulcanization process change feature vector and a rubber product vulcanization process temperature feature vector from the monitoring video of the rubber product vulcanization process and the temperature values ​​of the rubber product vulcanization process at the plurality of predetermined time points, respectively;

[0009] A rubber product vulcanization data fusion module is used to construct a rubber vulcanization process temperature classification feature matrix between the rubber product vulcanization process change feature vector and the rubber product vulcanization process temperature feature vector, and perform class boundary optimization on the rubber vulcanization process temperature classification feature matrix to obtain an optimized rubber vulcanization process temperature classification feature matrix;

[0010] The rubber product vulcanization process analysis module is used to pass the optimized rubber vulcanization process temperature classification feature matrix through a classifier to obtain a classification result, and the classification result is used to indicate whether the rubber product vulcanization process temperature should be increased or decreased.

[0011] According to another aspect of the present application, there is also provided an intelligent control method for a rubber product vulcanization process, which comprises:

[0012] The video of the vulcanization process of the rubber product is collected by a camera to obtain a monitoring video of the vulcanization process of the rubber product, and the temperature value of the vulcanization process of the rubber product at multiple predetermined time points is collected by a temperature sensor;

[0013] Extracting a rubber product vulcanization process change feature vector and a rubber product vulcanization process temperature feature vector from the monitoring video of the rubber product vulcanization process and the temperature values ​​of the rubber product vulcanization process at the plurality of predetermined time points, respectively;

[0014] Constructing a rubber vulcanization process temperature classification feature matrix between the rubber product vulcanization process change feature vector and the rubber product vulcanization process temperature feature vector, and performing class boundary optimization on the rubber vulcanization process temperature classification feature matrix to obtain an optimized rubber vulcanization process temperature classification feature matrix;

[0015] The optimized rubber vulcanization process temperature classification feature matrix is ​​passed through a classifier to obtain a classification result, and the classification result is used to indicate whether the vulcanization process temperature of the rubber product should be increased or decreased.

[0016] Compared with the prior art, the present application provides an intelligent control system and method for the vulcanization process of rubber products. The system monitors the video of the vulcanization process of rubber products in real time through a camera, and extracts the change characteristics and temperature characteristics of the vulcanization process in combination with the temperature data collected by the temperature sensor. Then, the classifier is used to determine whether the temperature of the vulcanization process should be increased or decreased, thereby achieving the purpose of intelligently controlling the vulcanization process of rubber products. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] 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.

[0018] Figure 1 The block diagram is a schematic diagram of an intelligent control system for the vulcanization process of rubber products according to an embodiment of the present application.

[0019] Figure 2The block diagram of the rubber product vulcanization process processing module in the intelligent control system of the rubber product vulcanization process according to the embodiment of the present application is schematic.

[0020] Figure 3 The block diagram is a schematic diagram of a monitoring video processing unit in an intelligent control system for a rubber product vulcanization process according to an embodiment of the present application.

[0021] Figure 4 The block diagram is a schematic diagram of a temperature value processing unit in an intelligent control system for a rubber product vulcanization process according to an embodiment of the present application.

[0022] Figure 5 The figure is a flow chart of an intelligent control method for a rubber product vulcanization process according to an embodiment of the present application. DETAILED DESCRIPTION

[0023] Various exemplary embodiments, features and aspects of the present application will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0024] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0025] In addition, in order to better illustrate the present application, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present application can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the subject matter of the present application.

[0026] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0027] Figure 1 The figure shows a block diagram of an intelligent control system for the vulcanization process of rubber products according to an embodiment of the present application. Figure 1As shown, the intelligent control system 100 for the rubber product vulcanization process according to the embodiment of the present application includes: a rubber product vulcanization process acquisition module 110, which is used to acquire a monitoring video of the rubber product vulcanization process by a camera, thereby obtaining a monitoring video of the rubber product vulcanization process, and to acquire a temperature value of the rubber product vulcanization process at a plurality of predetermined time points by a temperature sensor; a rubber product vulcanization process processing module 120, which is used to extract a rubber product vulcanization process change feature vector and a rubber product vulcanization process temperature feature vector from the monitoring video of the rubber product vulcanization process and the temperature values ​​of the rubber product vulcanization process at the plurality of predetermined time points, respectively; a rubber product vulcanization data fusion module 130, which is used to construct a rubber vulcanization process temperature classification feature matrix between the rubber product vulcanization process change feature vector and the rubber product vulcanization process temperature feature vector, and to perform class boundary optimization on the rubber vulcanization process temperature classification feature matrix to obtain an optimized rubber vulcanization process temperature classification feature matrix; a rubber product vulcanization process analysis module 140, which is used to pass the optimized rubber vulcanization process temperature classification feature matrix through a classifier to obtain a classification result, and the classification result is used to indicate whether the rubber product vulcanization process temperature should be increased or decreased.

[0028] In the embodiment of the present application, the rubber product vulcanization process acquisition module 110 is used to acquire the monitoring video of the rubber product vulcanization process by a camera, and to acquire the temperature value of the rubber product vulcanization process at a plurality of predetermined time points by a temperature sensor. It should be understood that the key parameters in the rubber vulcanization process are monitored and controlled in real time to ensure the stability of product quality and the controllability of the production process. Video monitoring can provide visual information to help monitor the key features such as deformation and color change in the rubber vulcanization process, while the temperature sensor can provide real-time temperature data to help monitor the temperature change in the vulcanization process, and ensure that vulcanization is performed within an appropriate temperature range, thereby ensuring product quality and production efficiency. The comprehensive use of video monitoring and temperature sensor data can provide a more comprehensive understanding of the vulcanization process and provide a more accurate basis for intelligent control.

[0029] In the embodiment of the present application, the rubber product vulcanization process processing module 120 is used to extract the rubber product vulcanization process change feature vector and the rubber product vulcanization process temperature feature vector from the monitoring video of the rubber product vulcanization process and the temperature values ​​of the rubber product vulcanization process at the multiple predetermined time points, respectively. It should be understood that visual information in the rubber vulcanization process, including deformation, color change, liquid change, etc., can be obtained through the monitoring video. These information can help understand the real-time state and change trend of the vulcanization process, and extracting key feature vectors helps to monitor and analyze the vulcanization process more carefully. Temperature is one of the important parameters affecting the rubber vulcanization process. Monitoring and analyzing the temperature change of the vulcanization process can help control the progress of the vulcanization process. Extracting the temperature feature vector can be used to describe the temperature change law in the vulcanization process, and help determine whether the vulcanization process is carried out within a suitable temperature range. The comprehensive use of video features and temperature features can more comprehensively understand the state and change trend of the rubber vulcanization process, which helps to establish a more accurate model for controlling the vulcanization process and improving production efficiency and product quality.

[0030] Specifically, in one embodiment of the present application, Figure 2 The figure shows a block diagram of a rubber product vulcanization process processing module in an intelligent control system for a rubber product vulcanization process according to an embodiment of the present application. Figure 2 As shown, in the above-mentioned intelligent control system 100 of the rubber product vulcanization process, the rubber product vulcanization process processing module 120 includes: a monitoring video processing unit 121, which is used to extract key frames from the monitoring video of the rubber product vulcanization process and obtain the rubber product vulcanization process change feature vector through convolution coding; a temperature value processing unit 122, which is used to obtain the rubber product vulcanization process temperature feature vector by a temperature extractor based on a timing encoder through the temperature values ​​of the rubber product vulcanization process at the multiple predetermined time points.

[0031] Accordingly, in a specific example of the present application, the monitoring video processing unit 121 is used to extract key frames from the monitoring video of the rubber product vulcanization process and then obtain the characteristic vector of the change of the rubber product vulcanization process through convolution coding. It should be understood that the monitoring video contains rich visual information, such as deformation, color change, liquid change, etc. of rubber products, which is very important for describing the changes in the vulcanization process. Extracting key frames helps to reduce the amount of data and retain the most representative and information-rich frames in the video, so as to more effectively characterize the entire vulcanization process. Convolutional Neural Networks (CNN) performs well in the field of image and video processing. Visual features can be extracted through convolution coding to help the system better understand the changes in video data. The extracted feature vector can be used for subsequent analysis and decision-making, such as monitoring the state of the vulcanization process, detecting abnormal conditions, optimizing control strategies, etc., so as to improve the efficiency and quality of the vulcanization process. Comprehensively utilizing the visual information in the monitoring video, extracting key frames and performing convolution coding to obtain feature vectors, helps the system to more comprehensively and accurately understand and monitor the changes in the vulcanization process of rubber products.

[0032] further, Figure 3 The figure shows a block diagram of a monitoring video processing unit in an intelligent control system for a rubber product vulcanization process according to an embodiment of the present application. Figure 3 As shown, in the rubber product vulcanization process processing module 120 of the above-mentioned intelligent control system 100 for the rubber product vulcanization process, the monitoring video processing unit 121 includes: a key frame extraction subunit 1211, which is used to extract multiple rubber product vulcanization process key frames from the monitoring video of the rubber product vulcanization process; a deep and shallow fusion subunit 1212, which is used to pass the multiple rubber product vulcanization process key frames through a first convolutional neural network model including a deep and shallow fusion module to obtain multiple rubber product vulcanization process monitoring feature matrices; a three-dimensional convolution kernel subunit 1213, which is used to aggregate the multiple rubber product vulcanization process monitoring feature matrices along the time dimension into a three-dimensional detection area feature tensor and then obtain the rubber product vulcanization process change feature vector through a second convolutional neural network model using a three-dimensional convolution kernel.

[0033] Specifically, the key frame extraction subunit 1211 is used to extract multiple key frames of the rubber product vulcanization process from the monitoring video of the rubber product vulcanization process. It should be understood that the monitoring video usually contains a large amount of information, and extracting key frames can help reduce the amount of data and retain only the most important frames, thereby saving storage space and processing time. There may be some important events or stages in the vulcanization process, and extracting key frames can highlight these events and help operators or systems better understand the progress and changes of the process. By extracting key frames, the vulcanization process can be browsed and analyzed more quickly without viewing the entire video frame by frame. Key frames usually contain key information and features in the vulcanization process, which is helpful for subsequent analysis and processing, such as for training models or feature extraction. After extracting key frames, these key frames can be used for visual display to help users understand the changes and development of the vulcanization process more intuitively.

[0034] Specifically, the deep and shallow fusion subunit 1212 is used to pass the multiple rubber product vulcanization process key frames through the first convolutional neural network model containing the deep and shallow fusion module to obtain multiple rubber product vulcanization process monitoring feature matrices. It should be understood that the convolutional neural network (CNN) can effectively learn features from images. Through the deep and shallow fusion module, feature information at different levels can be comprehensively utilized, which helps to more comprehensively describe the characteristics of the vulcanization process. Through the convolutional neural network model, meaningful features can be extracted from the extracted key frames, and these features can help the system better understand the changes and characteristics of the rubber product vulcanization process. The obtained monitoring feature matrix can be used for subsequent monitoring, identification and analysis, helping the system to monitor abnormal conditions or changes in the rubber product vulcanization process in real time. Through such a processing method, the system can more accurately and comprehensively understand and analyze the monitoring video of the rubber product vulcanization process, and provide more effective data support and decision-making basis for the intelligent control system.

[0035] Correspondingly, the deep and shallow fusion subunit includes: a shallow feature extraction secondary subunit, which is used to extract a shallow feature map from the i-th layer of the first convolutional neural network model, where j is greater than or equal to 1 and less than or equal to 6; a deep feature extraction secondary subunit, which is used to extract a deep feature map from the j-th layer of the first convolutional neural network model, and the ratio between the j-th layer and the i-th layer is greater than or equal to 5 and less than or equal to 10; a feature map fusion secondary subunit, which is used to use the deep and shallow feature fusion module of the first convolutional neural network model to fuse the shallow feature map and the deep feature map to obtain a fused feature map; a feature map dimensionality reduction secondary subunit, which is used to perform global pooling on the fused feature map along the channel dimension to obtain the multiple rubber product vulcanization process monitoring feature matrices.

[0036] Specifically, the three-dimensional convolutional nuclear subunit 1213 is used to aggregate the multiple rubber product vulcanization process monitoring feature matrices into three-dimensional detection area feature tensors along the time dimension, and then obtain the rubber product vulcanization process change feature vector by using the second convolutional neural network model of the three-dimensional convolutional kernel. It should be understood that by aggregating the monitoring feature matrix into a three-dimensional tensor along the time dimension, the information of the time dimension can be combined with the information of the space dimension to more comprehensively capture the changes and evolutions in the vulcanization process of the rubber product. The three-dimensional convolutional neural network can effectively learn features from three-dimensional data. By using a three-dimensional convolutional kernel, the spatiotemporal features in the vulcanization process of the rubber product can be better captured, thereby extracting richer and more accurate feature representations. Through the second convolutional neural network model, the three-dimensional detection area feature tensor can be converted into a change feature vector of the vulcanization process of the rubber product. These feature vectors can better represent the dynamic changes of the vulcanization process of the rubber product, which is helpful for subsequent analysis and prediction. In summary, this method can effectively combine time and space information to extract the change characteristics of the vulcanization process of the rubber product, thereby achieving more accurate and comprehensive monitoring and analysis.

[0037] Correspondingly, the three-dimensional convolution kernel unit includes: a rubber product vulcanization process encoding secondary subunit, which is used to use the second convolutional neural network model to perform three-dimensional convolution encoding on the three-dimensional rubber product vulcanization process feature tensor to obtain a rubber product vulcanization process change feature map; a rubber product vulcanization process dimensionality reduction secondary subunit, which is used to perform global mean pooling on each feature matrix of the rubber product vulcanization process change feature map along the channel dimension to obtain the rubber product vulcanization process change feature vector.

[0038] Specifically, the rubber product vulcanization process encoding secondary subunit is used to perform three-dimensional convolution encoding on the three-dimensional rubber product vulcanization process feature tensor using the second convolutional neural network model to obtain a rubber product vulcanization process change feature map. It should be understood that a higher level feature representation is extracted from the time series data to capture the complex time and space relationship in the rubber product vulcanization process. Through the three-dimensional convolution operation, the model can automatically learn the features in the time and space dimensions, so as to better understand the changes and patterns in the rubber product vulcanization process. The three-dimensional convolution operation can take into account the relationship between adjacent frames, so as to capture continuous feature patterns in the time dimension. This method can help the model better understand the dynamic changes in the rubber product vulcanization process, improve the ability to extract process features, and thus more accurately capture the changes and characteristics of the rubber product vulcanization process.

[0039] Specifically, the rubber product vulcanization process dimension reduction secondary subunit is used to perform global mean pooling on each feature matrix along the channel dimension of the rubber product vulcanization process change feature map to obtain the rubber product vulcanization process change feature vector. It should be understood that global mean pooling can summarize the information of each feature channel into a single value, thereby greatly reducing the dimension of the data. This helps to reduce the complexity and computational cost of the model while avoiding overfitting. By taking the average value of each feature channel, global mean pooling helps to retain the most important feature information. This ensures that key information will not be lost while reducing the dimension, thereby maintaining the characterization ability of the model. Global mean pooling can reduce noise and unnecessary details in the feature map, making the model more robust and easier to generalize to new data sets. After converting the feature map to a feature vector, it can be more easily input into a subsequent classifier or regressor to make a final prediction or decision. Therefore, converting the rubber product vulcanization process change feature map to a feature vector by global mean pooling helps to simplify data representation, extract key information, reduce computational burden, and help improve the generalization ability and robustness of the model.

[0040] Accordingly, in a specific example of the present application, the temperature value processing unit 122 is used to obtain the temperature feature vector of the rubber product vulcanization process through a temperature extractor based on a timing encoder through the temperature values ​​of the rubber product vulcanization process at the multiple predetermined time points. It should be understood that the temperature extractor based on the timing encoder can effectively capture the timing information in the time series data and help identify the temperature change trend and periodicity. Through the timing encoder, the complex temperature change sequence can be converted into a more representative and high-level feature vector, which helps the model to better understand the impact of temperature on the rubber vulcanization process. Extracting the temperature feature vector can reduce the dimension and abstract the original temperature data, reduce the complexity of the data, and retain key information, which helps the model to better process and analyze the data. Therefore, extracting the temperature feature vector of the rubber product vulcanization process through the temperature extractor based on the timing encoder can help the deep learning model better understand and utilize the temperature information in the vulcanization process, thereby improving the modeling and prediction capabilities of the vulcanization process.

[0041] further, Figure 4 The figure shows a block diagram of a temperature value processing unit in an intelligent control system for a rubber product vulcanization process according to an embodiment of the present application. Figure 4As shown, in the rubber product vulcanization process processing module 120 of the above-mentioned intelligent control system 100 for the vulcanization process of rubber products, the temperature value processing unit 122 includes: a temperature value arrangement subunit 1221, which is used to arrange the temperature values ​​of the rubber product vulcanization process at the plurality of predetermined time points into a rubber product vulcanization process temperature input vector according to the time dimension; a normalization processing subunit 1222, which is used to perform a maximum value-based normalization processing on the rubber product vulcanization process temperature input vector to obtain a normalized rubber product vulcanization process temperature input vector; a fully connected layer subunit 1221, which is used to perform a maximum value-based normalization processing on the rubber product vulcanization process temperature input vector to obtain a normalized rubber product vulcanization process temperature input vector; a fully connected layer subunit 12 ... Element 1223 is used to use the fully connected layer of the time series encoder to fully connect the input vector of the normalized temperature of the rubber product vulcanization process to extract the high-dimensional implicit features of the eigenvalues ​​at each position in the input vector of the normalized temperature of the rubber product vulcanization process; one-dimensional convolution layer subunit 1224 is used to use the one-dimensional convolution layer of the time series encoder to one-dimensionally encode the input vector of the normalized temperature of the rubber product vulcanization process to extract the high-dimensional implicit correlation features of the correlation between the eigenvalues ​​at each position in the input vector of the normalized temperature of the rubber product vulcanization process.

[0042] Specifically, the temperature value arrangement subunit 1221 is used to arrange the temperature values ​​of the rubber product vulcanization process at the multiple predetermined time points into the temperature input vector of the rubber product vulcanization process according to the time dimension. It should be understood that in order to retain the time information and order relationship in the time series. By arranging these temperature values ​​according to the time dimension, the law and trend of temperature change over time in the vulcanization process can be better captured, so that subsequent data processing and analysis can more effectively utilize the information of the time dimension. In the vulcanization process of rubber products, the change of temperature is an important parameter, which directly affects the progress of the vulcanization reaction and the quality of the final product. Therefore, arranging these temperature values ​​into vectors according to the time dimension helps the system to better understand and analyze the temperature change in the vulcanization process, thereby providing more accurate information and guidance for the intelligent control system. By arranging the temperature values ​​in the time dimension, the characteristics of the time series data can be better utilized, such as capturing the dynamic change of temperature over time, identifying the periodicity or trend of temperature change, etc. Doing so helps to improve the system's understanding and prediction capabilities of the vulcanization process, thereby achieving the goal of more accurate control and optimization of the vulcanization process.

[0043] Specifically, the normalization processing subunit 1222 is used to perform normalization processing on the temperature input vector of the rubber product vulcanization process based on the maximum value to obtain the normalized temperature input vector of the rubber product vulcanization process. It should be understood that the temperature value ranges at different time points are unified to a similar scale. This helps to eliminate the absolute differences between the temperature values ​​at different time points, so that the model can better learn the relative relationship and change law between the temperature values ​​at each time point. Through normalization processing, the training effect of the model can be improved, so that the model has more generalization ability and can better adapt to temperature input data of different ranges.

[0044] Specifically, the fully connected layer subunit 1223 is used to use the fully connected layer of the temporal encoder to perform fully connected encoding on the normalized temperature input vector of the rubber product vulcanization process to extract the high-dimensional implicit features of the eigenvalues ​​at each position in the normalized temperature input vector of the rubber product vulcanization process. It should be understood that the fully connected layer can help the model learn the complex nonlinear features in the input vector, so as to better represent the eigenvalues ​​at different positions in the temperature sequence. This helps to extract the high-dimensional implicit features of the temperature change during the vulcanization process of the rubber product. The parameters in the fully connected layer can be learned and optimized through the training process to adapt to specific data sets and tasks. In this way, the model can automatically learn how to best extract and utilize the features in the temperature sequence according to the characteristics of the data. The fully connected layer can increase the expressive power of the model, so that it can better capture the complex relationships in the input vector. This helps the model to more accurately understand the changing laws of temperature during the vulcanization process of the rubber product. Therefore, the fully connected layer is used to perform fully connected encoding on the normalized temperature input vector of the rubber product vulcanization process, which helps to extract the high-dimensional implicit features of the eigenvalues ​​at each position in the temperature sequence, so as to better understand and utilize the temperature information during the vulcanization process of the rubber product.

[0045] Specifically, the one-dimensional convolutional layer subunit 1224 is used to use the one-dimensional convolutional layer of the time series encoder to perform one-dimensional encoding on the normalized rubber product vulcanization process temperature input vector to extract the high-dimensional implicit correlation features of the correlation between the feature values ​​of each position in the normalized rubber product vulcanization process temperature input vector. It should be understood that in the time series data processing, the purpose of using the one-dimensional convolutional layer to perform one-dimensional encoding on the normalized rubber product vulcanization process temperature input vector is to extract the feature association between different positions and capture the local pattern in the sequence and the dependency between the features. The one-dimensional convolutional layer can effectively learn the feature interaction between different positions in the input sequence, thereby helping the model to better understand the time series relationship and pattern in the sequence data. Through the processing of the one-dimensional convolutional layer, the model can learn the correlation between the features between different positions, thereby extracting richer high-dimensional implicit correlation features, helping the model to better understand and utilize the information in the input sequence. This method helps to improve the model's modeling ability for sequence data, thereby improving the performance and effect of the model when processing the rubber product vulcanization process temperature input sequence.

[0046] In the embodiment of the present application, the rubber product vulcanization data fusion module 130 is used to construct a rubber vulcanization process temperature classification feature matrix between the rubber product vulcanization process change feature vector and the rubber product vulcanization process temperature feature vector, and perform class boundary optimization on the rubber vulcanization process temperature classification feature matrix to obtain an optimized rubber vulcanization process temperature classification feature matrix. It should be understood that the fusion of the change feature vector and the temperature feature vector of the rubber vulcanization process can comprehensively consider the temperature change and other related features in the vulcanization process, thereby more comprehensively describing the characteristics of the vulcanization process and helping to improve the accuracy of classification. The change feature vector and the temperature feature vector often contain information at different levels. By fusing these two types of features, the classifier can simultaneously consider the dynamic changes of the vulcanization process and the influence of temperature characteristics on classification, thereby improving the generalization ability of the classifier. Fusion of different types of feature vectors can reduce the risk of model overfitting. If only a single type of feature vector is used, the model may be overly dependent on a certain type of feature during training, and fusion of different types of features can make the model more robust. By fusing multiple features, the classification model can improve the temperature classification performance of the rubber vulcanization process. Comprehensively considering the influence of different features helps the model to more accurately judge the temperature classification of the rubber vulcanization process. Therefore, integrating the change characteristic vector and temperature characteristic vector of the rubber product vulcanization process can provide more comprehensive and multi-angle information, which helps to obtain a more accurate and reliable temperature classification characteristic matrix of the rubber vulcanization process.

[0047] Specifically, in one embodiment of the present application, the rubber product vulcanization data fusion module includes: a fusion matrix unit, used to fuse the rubber product vulcanization process change feature vector and the rubber product vulcanization process temperature feature vector to obtain a rubber vulcanization process associated feature matrix; a spatial attention unit, used to obtain the rubber vulcanization process temperature classification feature matrix by using a third convolutional neural network model with a spatial attention mechanism for the rubber vulcanization process associated feature matrix; an optimization unit, used to perform class boundary optimization on the rubber vulcanization process temperature classification feature matrix to obtain the optimized rubber vulcanization process temperature classification feature matrix.

[0048] Accordingly, in a specific example of the present application, the fusion matrix unit is used to fuse the rubber product vulcanization process change feature vector and the rubber product vulcanization process temperature feature vector to obtain the rubber vulcanization process associated feature matrix. It should be understood that there may be a certain correlation between the temperature change and other features in the vulcanization process. By fusing these two different types of feature vectors, the characteristics of the vulcanization process can be more comprehensively represented, and the comprehensiveness and accuracy of feature characterization can be improved. Temperature is one of the important factors affecting the vulcanization process, and the change feature vector can capture the dynamic change information in the vulcanization process. Fusing these two aspects of information together can make the feature matrix more information-rich. By fusing different types of feature vectors, the accuracy and generalization ability of the classification model can be improved. Comprehensive consideration of the characteristics of multiple aspects helps to better distinguish different vulcanization process types or states. Fusion of multiple feature vectors can help reduce the overfitting risk of the model, making the model more generalizable and more suitable for different vulcanization process data sets.

[0049] Accordingly, in a specific example of the present application, the spatial attention unit is used to obtain the rubber vulcanization process temperature classification feature matrix by using the third convolutional neural network model of the spatial attention mechanism. It should be understood that through the spatial attention mechanism, the model can better focus on the important features and patterns in the rubber vulcanization process, thereby improving the accuracy and efficiency of classification. The third convolutional neural network model usually has a deeper network structure and a more powerful feature extraction capability, which can better capture the complex features in the rubber vulcanization process, and then achieve more accurate extraction and classification of temperature classification features. Through such a design, the accuracy and reliability of the temperature classification of the rubber vulcanization process can be improved.

[0050] Specifically, the spatial attention unit is used to: use each layer of the third convolutional neural network model to perform the following on the input data in the forward pass of the layer: convolution processing on the input data based on the convolution kernel to obtain a convolution feature map; pass the convolution feature map through the spatial attention module to obtain the spatial attention score matrix; multiply the spatial attention score matrix and each feature matrix of the convolution feature map along the channel dimension by the spatial attention feature map at each position point; pool the spatial attention feature map along the channel dimension to obtain a pooled feature map; perform nonlinear activation on the pooled feature map to obtain an activated feature map; wherein the input of the first layer of the third convolutional neural network model is the rubber vulcanization process association feature matrix, and the output of the last layer of the third convolutional neural network model is the rubber vulcanization process temperature classification feature matrix.

[0051] Accordingly, in a specific example of the present application, the optimization unit is used to optimize the class boundary of the rubber vulcanization process temperature classification feature matrix to obtain the optimized rubber vulcanization process temperature classification feature matrix. It should be understood that, in particular, in the technical solution of the present application, considering that the intensive data collection at multiple predetermined time points by the camera and the temperature sensor may cause the data points adjacent in time to have a high correlation, thereby generating redundancy in the spatial dimension. The key frames extracted from the surveillance video may be visually similar, especially when the rubber product does not change much at certain stages of the vulcanization process. This visual similarity will cause redundancy in the feature matrix. The temperature extractor based on the temporal encoder may not be able to fully distinguish the temperature changes at all time points, especially when the temperature changes are subtle or the change patterns are similar, which may lead to the introduction of redundant information in the feature vector. Although the spatial attention mechanism is designed to highlight important features and suppress unimportant features, if the attention mechanism fails to accurately capture all important spatial information, or overemphasizes certain features, the rubber vulcanization process temperature classification feature matrix has feature redundancy in the spatial dimension. When there are redundant features in the rubber vulcanization process temperature classification feature matrix, the model needs to find the decision boundary in a higher dimensional space. This may cause the decision boundary to become too complex, making it difficult to capture the true distribution of the data, resulting in sudden spatial folding between different categories. That is, the rubber vulcanization process temperature classification feature matrix has feature redundancy in the spatial dimension, which causes the class boundary of the rubber vulcanization process temperature classification feature matrix to undergo sudden spatial nonlinear folding, affecting the accuracy of its classification judgment. In order to solve this problem, in the technical solution of the present application, the class boundary of the rubber vulcanization process temperature classification feature matrix is ​​optimized.

[0052] Specifically, the optimization unit includes: a normalization processing subunit, which is used to perform maximum value-based normalization processing on the rubber vulcanization process temperature classification feature matrix to obtain a normalized rubber vulcanization process temperature classification feature matrix; and a class boundary optimization subunit, which is used to perform class boundary optimization on the normalized rubber vulcanization process temperature classification feature matrix to obtain the optimized rubber vulcanization process temperature classification feature matrix.

[0053] Correspondingly, the class boundary optimization subunit includes: a first exponential function secondary subunit, which is used to calculate the natural exponential function value raised to the power of the eigenvalue of the predetermined position of the normalized rubber vulcanization process temperature classification feature matrix to obtain a first exponential function value; a first absolute value secondary subunit, which is used to calculate the first exponential function value minus one and divided by the eigenvalue of the predetermined position of the normalized rubber vulcanization process temperature classification feature matrix to obtain a first quotient value, and take the absolute value of the first quotient value to obtain a first absolute value; a second exponential function secondary subunit, which is used to calculate the opposite number of the eigenvalue of the predetermined position of the normalized rubber vulcanization process temperature classification feature matrix to obtain a first absolute value. to obtain a second exponential function value; a second absolute value secondary subunit, used to calculate the eigenvalue of a predetermined position of the normalized rubber vulcanization process temperature classification feature matrix and divide it by the second exponential function value to obtain a second quotient value, and take the absolute value of the second quotient value to obtain a second absolute value; a sum calculation secondary subunit, used to calculate the sum of the first absolute value and the second absolute value, and calculate a logarithmic function value with base 2 for the sum, and then subtract the calculated logarithmic function value from the second norm of the normalized rubber vulcanization process temperature classification feature matrix to obtain the eigenvalue of the predetermined position of the optimized rubber vulcanization process temperature classification feature matrix.

[0054] Further, class boundary optimization is performed on the temperature classification feature matrix of the rubber vulcanization process to obtain an optimized temperature classification feature matrix of the rubber vulcanization process, including: performing a maximum value-based normalization process on the temperature classification feature matrix of the rubber vulcanization process to obtain a normalized temperature classification feature matrix of the rubber vulcanization process; class boundary optimization is performed on the normalized temperature classification feature matrix of the rubber vulcanization process according to the following formula, wherein the formula is:

[0055]

[0056] in, represents the normalized temperature classification feature matrix of the rubber vulcanization process, Represents the normalized rubber vulcanization process temperature classification feature matrix The eigenvalues ​​of the positions, represents the bi-norm of the normalized temperature classification feature matrix of the rubber vulcanization process, represents the logarithmic function value with base 2, Represents the temperature classification feature matrix of the optimized rubber vulcanization process The eigenvalue of the location.

[0057] In view of the above technical problems, the class boundary optimization is performed on the temperature classification feature matrix of the rubber vulcanization process, which uses the sparsity principle to reconstruct the temperature classification feature matrix of the rubber vulcanization process, reduce the redundancy between features, and make the class boundary clearer and more stable. Specifically, in the process of implementing the class boundary optimization, firstly, by analyzing the feature distribution of the temperature classification feature matrix of the rubber vulcanization process, those features that contribute less or are unstable to the classification decision are identified, and their weights are reduced or removed; then, the group sparsity regularization term is introduced to consider not only the sparsity of a single feature, but also the relationship and synergy between the feature groups in the temperature classification feature matrix of the rubber vulcanization process, so as to optimize the feature distribution structure of the entire temperature classification feature matrix of the rubber vulcanization process. In this way, the temperature classification feature matrix of the rubber vulcanization process converges in the decision domain under the inductive constraint of the classification problem for the class probability limit, thereby improving the certainty of the inductive result under the target problem.

[0058] In the embodiment of the present application, the rubber product vulcanization process analysis module 140 is used to pass the optimized rubber vulcanization process temperature classification feature matrix through a classifier to obtain a classification result, and the classification result is used to indicate that the temperature of the rubber product vulcanization process should increase or decrease. It should be understood that vulcanization is a crucial step in the production process of rubber products, and temperature is one of the key parameters that affect the vulcanization effect. Classifying and predicting the temperature in the vulcanization process through a model can help optimize the vulcanization process and improve product quality and production efficiency. According to the prediction results of the classifier, the temperature control parameters in the vulcanization process can be adjusted in time to make the vulcanization effect more stable and controllable. If the model predicts that the temperature should increase, the operator can adjust the temperature setting value accordingly; conversely, if the predicted temperature should decrease, adjustments can also be made in time.

[0059] Specifically, in one embodiment of the present application, the rubber product vulcanization process analysis module 140 includes: a matrix expansion unit, which is used to expand the optimized rubber vulcanization process temperature classification feature matrix into an optimized rubber vulcanization process temperature classification feature vector according to a row vector or a column vector; a fully connected encoding unit, which is used to use the fully connected layer of the classifier to fully connect the optimized rubber vulcanization process temperature classification feature vector to obtain a fully connected encoded optimized rubber vulcanization process temperature classification feature vector; a probability acquisition unit, which is used to pass the fully connected encoded optimized rubber vulcanization process temperature classification feature vector through the Softmax classification function of the classifier to obtain a first probability that the rubber product vulcanization process temperature should be increased and a second probability that the rubber product vulcanization process temperature should be decreased; a classification result determination unit, which is used to determine the classification result based on a comparison between the first probability and the second probability.

[0060] In summary, based on the intelligent control system and method of the rubber product vulcanization process of the embodiment of the present application, the video of the rubber product vulcanization process is monitored in real time by a camera, and the temperature data collected by the temperature sensor is combined to extract the change characteristics and temperature characteristics of the vulcanization process, and then the classifier is used to determine whether the temperature of the vulcanization process should be increased or decreased, thereby achieving the purpose of intelligently controlling the vulcanization process of the rubber product.

[0061] As described above, the intelligent control system 100 for the vulcanization process of rubber products according to the embodiment of the present application can be implemented in various terminal devices, such as a server of the intelligent control system for the vulcanization process of rubber products. In one example, the intelligent control system 100 for the vulcanization process of rubber products can be integrated into the terminal device as a software module and / or a hardware module. For example, the intelligent control system 100 for the vulcanization process of rubber products can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the intelligent control system 100 for the vulcanization process of rubber products can also be one of the many hardware modules of the terminal device.

[0062] Alternatively, in another example, the intelligent control system 100 for the rubber product vulcanization process and the terminal device may also be separate devices, and the intelligent control system 100 for the rubber product vulcanization process may be connected to the terminal device via a wired and / or wireless network, and transmit interactive information in accordance with an agreed data format.

[0063] Figure 5 FIG. 1 is a flow chart of an intelligent control method for a rubber product vulcanization process according to an embodiment of the present application. Figure 5As shown, the intelligent control method for the vulcanization process of the rubber product according to the embodiment of the present application includes the following steps: S110, collecting video of the vulcanization process of the rubber product through a camera to obtain a monitoring video of the vulcanization process of the rubber product, and collecting temperature values ​​of the vulcanization process of the rubber product at multiple predetermined time points through a temperature sensor; S120, extracting a rubber product vulcanization process change feature vector and a rubber product vulcanization process temperature feature vector from the monitoring video of the vulcanization process of the rubber product and the temperature values ​​of the vulcanization process of the rubber product at the multiple predetermined time points, respectively; S130, constructing a rubber vulcanization process temperature classification feature matrix between the rubber product vulcanization process change feature vector and the rubber product vulcanization process temperature feature vector, and performing class boundary optimization on the rubber vulcanization process temperature classification feature matrix to obtain an optimized rubber vulcanization process temperature classification feature matrix; S140, passing the optimized rubber vulcanization process temperature classification feature matrix through a classifier to obtain a classification result, and the classification result is used to indicate whether the vulcanization process temperature of the rubber product should be increased or decreased.

[0064] Here, those skilled in the art will appreciate that the specific operations of each step in the above-mentioned intelligent control method for the vulcanization process of rubber products have been described in the above reference. Figures 1 to 4 The description of the intelligent control system for the rubber product vulcanization process has been introduced in detail, and therefore, its repeated description will be omitted.

[0065] 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.

[0066] 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.

[0067] 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 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 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 through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0068] 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.

[0069] 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.

[0070] 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 enabling 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: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0071] 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 intelligent control system for the vulcanization process of rubber products, characterized in that: include: The rubber product vulcanization process acquisition module is used to acquire the video of the rubber product vulcanization process through a camera to obtain the monitoring video of the rubber product vulcanization process, and to acquire the temperature values ​​of the rubber product vulcanization process at multiple predetermined time points through a temperature sensor; A rubber product vulcanization process processing module, used to extract a rubber product vulcanization process change feature vector and a rubber product vulcanization process temperature feature vector from the monitoring video of the rubber product vulcanization process and the temperature values ​​of the rubber product vulcanization process at the plurality of predetermined time points, respectively; A rubber product vulcanization data fusion module is used to construct a rubber vulcanization process temperature classification feature matrix between the rubber product vulcanization process change feature vector and the rubber product vulcanization process temperature feature vector, and perform class boundary optimization on the rubber vulcanization process temperature classification feature matrix to obtain an optimized rubber vulcanization process temperature classification feature matrix; A rubber product vulcanization process analysis module, used for passing the optimized rubber vulcanization process temperature classification feature matrix through a classifier to obtain a classification result, wherein the classification result is used to indicate whether the rubber product vulcanization process temperature should be increased or decreased; Wherein, the rubber product vulcanization data fusion module includes: A fusion matrix unit, used for fusing the rubber product vulcanization process change feature vector and the rubber product vulcanization process temperature feature vector to obtain a rubber vulcanization process correlation feature matrix; A spatial attention unit is used to obtain a temperature classification feature matrix of the rubber vulcanization process by using a third convolutional neural network model using a spatial attention mechanism to obtain a temperature classification feature matrix of the rubber vulcanization process; An optimization unit, used for performing class boundary optimization on the temperature classification feature matrix of the rubber vulcanization process to obtain the optimized temperature classification feature matrix of the rubber vulcanization process; Wherein, the optimization unit comprises: A normalization processing subunit, used for performing a normalization processing on the rubber vulcanization process temperature classification feature matrix based on a maximum value to obtain a normalized rubber vulcanization process temperature classification feature matrix; A class boundary optimization subunit, used for performing class boundary optimization on the normalized rubber vulcanization process temperature classification feature matrix to obtain the optimized rubber vulcanization process temperature classification feature matrix; Wherein, the class boundary optimization subunit includes: A first exponential function secondary subunit is used to calculate a natural exponential function value raised to the power of the eigenvalue of a predetermined position of the normalized rubber vulcanization process temperature classification characteristic matrix to obtain a first exponential function value; A first absolute value secondary subunit is used to calculate the first exponential function value minus one and divide it by the eigenvalue of a predetermined position of the normalized rubber vulcanization process temperature classification feature matrix to obtain a first quotient value, and take an absolute value of the first quotient value to obtain a first absolute value; A second exponential function secondary subunit, used for calculating a natural exponential function value raised to a power of the inverse of the eigenvalue at a predetermined position of the normalized rubber vulcanization process temperature classification characteristic matrix to obtain a second exponential function value; A second absolute value secondary subunit is used to calculate the eigenvalue of a predetermined position of the normalized rubber vulcanization process temperature classification feature matrix and divide it by the second exponential function value to obtain a second quotient value, and subtract one from the second quotient value and then take the absolute value to obtain a second absolute value; The second-level subunit for calculating the sum value is used to calculate the sum value of the first absolute value and the second absolute value, and calculate the logarithmic function value with base 2 for the sum value, and then subtract the second norm of the normalized rubber vulcanization process temperature classification feature matrix from the calculated logarithmic function value to obtain the eigenvalue of the predetermined position of the optimized rubber vulcanization process temperature classification feature matrix.

2. The intelligent control system for the vulcanization process of rubber products according to claim 1, characterized in that: The rubber product vulcanization process processing module comprises: A monitoring video processing unit, used for extracting key frames from the monitoring video of the rubber product vulcanization process and performing convolution coding to obtain a change feature vector of the rubber product vulcanization process; The temperature value processing unit is used to obtain a temperature feature vector of the rubber product vulcanization process by using a temperature extractor based on a timing encoder to process the temperature values ​​of the rubber product vulcanization process at the plurality of predetermined time points.

3. The intelligent control system for the vulcanization process of rubber products according to claim 2, characterized in that: The monitoring video processing unit comprises: A key frame extraction subunit is used to extract a plurality of key frames of the rubber product vulcanization process from the monitoring video of the rubber product vulcanization process; A deep-shallow fusion subunit, used for respectively passing the plurality of rubber product vulcanization process key frames through a first convolutional neural network model including a deep-shallow fusion module to obtain a plurality of rubber product vulcanization process monitoring feature matrices; The three-dimensional convolution kernel unit is used to aggregate the multiple rubber product vulcanization process monitoring feature matrices into a three-dimensional detection area feature tensor along the time dimension and then obtain the rubber product vulcanization process change feature vector by using a second convolution neural network model with a three-dimensional convolution kernel.

4. The intelligent control system for the vulcanization process of rubber products according to claim 3, characterized in that: The deep and shallow fusion subunit comprises: A shallow feature extraction secondary subunit, used to extract a shallow feature map from the i-th layer of the first convolutional neural network model, where j is greater than or equal to 1 and less than or equal to 6; A deep feature extraction secondary subunit, used to extract a deep feature map from the jth layer of the first convolutional neural network model, wherein a ratio between the jth layer and the i-th layer is greater than or equal to 5 and less than or equal to 10; A feature map fusion secondary subunit, used to fuse the shallow feature map and the deep feature map using the deep and shallow feature fusion module of the first convolutional neural network model to obtain a fused feature map; The feature graph dimensionality reduction secondary subunit is used to perform global pooling on the fused feature graph along the channel dimension to obtain the plurality of rubber product vulcanization process monitoring feature matrices.

5. The intelligent control system for the vulcanization process of rubber products according to claim 4, characterized in that: The three-dimensional convolution kernel unit comprises: A rubber product vulcanization process encoding secondary subunit, used for performing three-dimensional convolution encoding on the three-dimensional rubber product vulcanization process feature tensor using the second convolutional neural network model to obtain a rubber product vulcanization process change feature map; The secondary subunit of the rubber product vulcanization process dimensionality reduction is used to perform global mean pooling on each feature matrix along the channel dimension of the rubber product vulcanization process change feature graph to obtain the rubber product vulcanization process change feature vector.

6. The intelligent control system for the rubber product vulcanization process according to claim 5, characterized in that: The temperature value processing unit comprises: A temperature value arrangement subunit, used for arranging the temperature values ​​of the rubber product vulcanization process at the plurality of predetermined time points according to the time dimension into a temperature input vector of the rubber product vulcanization process; A normalization processing subunit, used for performing normalization processing on the temperature input vector of the rubber product vulcanization process based on a maximum value to obtain a normalized temperature input vector of the rubber product vulcanization process; A fully connected layer subunit, used for performing fully connected encoding on the normalized temperature input vector of the vulcanization process of the rubber product using the fully connected layer of the temporal encoder to extract high-dimensional implicit features of the eigenvalues ​​of each position in the normalized temperature input vector of the vulcanization process of the rubber product; The one-dimensional convolutional layer subunit is used to use the one-dimensional convolutional layer of the temporal encoder to perform one-dimensional encoding on the normalized temperature input vector of the rubber product vulcanization process to extract high-dimensional implicit correlation features of the correlation between the eigenvalues ​​at each position in the normalized temperature input vector of the rubber product vulcanization process.

7. An intelligent control method for the vulcanization process of rubber products, characterized in that: include: The video of the vulcanization process of the rubber product is collected by a camera to obtain a monitoring video of the vulcanization process of the rubber product, and the temperature value of the vulcanization process of the rubber product at multiple predetermined time points is collected by a temperature sensor; Extracting a rubber product vulcanization process change feature vector and a rubber product vulcanization process temperature feature vector from the monitoring video of the rubber product vulcanization process and the temperature values ​​of the rubber product vulcanization process at the plurality of predetermined time points, respectively; Constructing a rubber vulcanization process temperature classification feature matrix between the rubber product vulcanization process change feature vector and the rubber product vulcanization process temperature feature vector, and performing class boundary optimization on the rubber vulcanization process temperature classification feature matrix to obtain an optimized rubber vulcanization process temperature classification feature matrix; Passing the optimized rubber vulcanization process temperature classification feature matrix through a classifier to obtain a classification result, wherein the classification result is used to indicate whether the rubber product vulcanization process temperature should be increased or decreased; Among them, constructing a rubber vulcanization process temperature classification feature matrix between the rubber product vulcanization process change feature vector and the rubber product vulcanization process temperature feature vector, and performing class boundary optimization on the rubber vulcanization process temperature classification feature matrix to obtain an optimized rubber vulcanization process temperature classification feature matrix, including: Merging the rubber product vulcanization process change feature vector and the rubber product vulcanization process temperature feature vector to obtain a rubber vulcanization process correlation feature matrix; The rubber vulcanization process correlation feature matrix is ​​passed through a third convolutional neural network model using a spatial attention mechanism to obtain the rubber vulcanization process temperature classification feature matrix; Performing class boundary optimization on the temperature classification feature matrix of the rubber vulcanization process to obtain the optimized temperature classification feature matrix of the rubber vulcanization process; The step of performing class boundary optimization on the temperature classification feature matrix of the rubber vulcanization process to obtain the optimized temperature classification feature matrix of the rubber vulcanization process comprises: Performing a normalization process based on the maximum value on the temperature classification feature matrix of the rubber vulcanization process to obtain a normalized temperature classification feature matrix of the rubber vulcanization process; Performing class boundary optimization on the normalized rubber vulcanization process temperature classification feature matrix to obtain the optimized rubber vulcanization process temperature classification feature matrix; The normalized rubber vulcanization process temperature classification feature matrix is ​​optimized by class boundary optimization to obtain the optimized rubber vulcanization process temperature classification feature matrix, including: Calculating a natural exponential function value raised to the power of the eigenvalue of a predetermined position of the normalized rubber vulcanization process temperature classification characteristic matrix to obtain a first exponential function value; Calculate the first exponential function value minus one and divide it by the eigenvalue of a predetermined position of the normalized rubber vulcanization process temperature classification feature matrix to obtain a first quotient value, and take an absolute value of the first quotient value to obtain a first absolute value; Calculating a natural exponential function value raised to a power of the inverse of the eigenvalue at a predetermined position of the normalized rubber vulcanization process temperature classification characteristic matrix to obtain a second exponential function value; Calculating the eigenvalue of a predetermined position of the normalized rubber vulcanization process temperature classification feature matrix and dividing it by the second exponential function value to obtain a second quotient value, and subtracting one from the second quotient value and taking the absolute value to obtain a second absolute value; The sum of the first absolute value and the second absolute value is calculated, and a logarithmic function value with base 2 is calculated for the sum, and then the second norm of the normalized rubber vulcanization process temperature classification feature matrix is ​​subtracted from the calculated logarithmic function value to obtain the eigenvalue of the predetermined position of the optimized rubber vulcanization process temperature classification feature matrix.

Images