Automatic control system and method for polaroid edge grinding equipment

Abstract

The invention relates to the field of intelligent control, and particularly discloses an automatic control system and method for polaroid edge grinding equipment. Grinding speeds and grinding pressures at a plurality of preset time points in a preset time period during polaroid edge grinding are collected in real time; and performing data analysis and processing on the grinding speed and the grinding pressure by adopting a deep learning technology so as to determine whether the grinding speed at the current time point needs to be increased or decreased or kept. In this way, dynamic adjustment can be conducted according to the actual situation in the polaroid grinding process, abrasion and manual intervention are reduced, the production cost is reduced, and therefore the machining quality of the polaroid is improved, and the machining efficiency and the product quality are improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent control, and more specifically, to an automatic control system and method for a polarizer edge grinding device. Background Art

[0002] Polarizer is a special optical element that can selectively pass or block light waves in a specific direction and has the function of adjusting the polarization state of light. Polarizers are widely used in optics, electronics, communications and other fields. For example, they play an important role in liquid crystal displays, camera lenses, lasers and other devices.

[0003] Polaroid edge grinding equipment is used to grind the edges of polaroids. When manufacturing polaroids, in order to ensure their performance and quality, it is usually necessary to grind the edges of polaroids to make them smooth and flat for subsequent processing and use.

[0004] However, because traditional polarizer edge grinding control is usually based on static set values, it lacks dynamic adjustment of real-time processing status. This may lead to poor results when facing different materials, workpiece conditions or processing conditions. Secondly, in traditional edge grinding control, the operator may need to constantly manually adjust the parameters of the edge grinding equipment, which is easily affected by human factors and causes inconsistent operations or errors.

[0005] Therefore, an optimized automatic control system for polarizer edge grinding equipment is desired. 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 automatic control system and method for polarizer edge grinding equipment, which collects the grinding speed and grinding pressure at multiple predetermined time points in a predetermined time period during polarizer edge grinding in real time, and uses deep learning technology to perform data analysis and processing on the grinding speed and the grinding pressure, thereby determining whether the grinding speed at the current time point should be increased, decreased, or maintained. In this way, dynamic adjustments can be made according to actual conditions during the polarizer grinding process, reducing wear and manual intervention and reducing production costs, thereby improving the processing quality of the polarizer, improving processing efficiency and product quality.

[0007] According to one aspect of the present application, an automatic control system for a polarizer edge grinding device is provided, comprising:

[0008] A polarizer edge grinding parameter acquisition module is used to acquire the grinding speed and grinding pressure at multiple predetermined time points during the predetermined time period when the polarizer is edged;

[0009] A parameter time dimension arrangement module, used for arranging the grinding speed and grinding pressure at a plurality of predetermined time points in the predetermined time period into a grinding speed time series input vector and a grinding pressure time series input vector according to the time dimension;

[0010] A grinding speed pressure feature matching and fusion module is used to perform perspective coordination matching and fusion based on topological association expression quantification on the grinding speed time series input vector and the grinding pressure time series input vector to obtain a grinding parameter time series association vector;

[0011] A grinding speed pressure characteristic association module, used for multiplying the grinding parameter time series association vector with the transposed vector of the grinding parameter time series association vector to obtain a grinding parameter time series association matrix;

[0012] A time series correlation feature extraction module is used to perform matrix segmentation on the grinding parameter time series correlation matrix to obtain a plurality of grinding parameter local time series correlation sub-matrices, and then sequentially perform local time series correlation feature extraction and feature enhancement to obtain a plurality of grinding parameter local time series spatial enhancement sub-feature matrices;

[0013] A fusion feature module, used for fusing the plurality of grinding parameter local temporal spatial reinforcement sub-feature matrices to obtain a grinding parameter temporal spatial reinforcement fusion feature matrix;

[0014] The grinding speed adjustment module is used to determine whether the grinding speed at the current time point should be increased, decreased or maintained based on the grinding parameter time-series spatial enhanced fusion feature matrix.

[0015] In the automatic control system of the above-mentioned polarizer edge grinding equipment, the grinding speed pressure feature matching fusion module is used to: construct a grinding parameter distance topological matrix between the grinding speed timing input vector and the grinding pressure timing input vector; construct a grinding parameter eigenvalue granularity association matrix between the grinding speed timing input vector and the grinding pressure timing input vector; perform topological association modulation on the grinding parameter eigenvalue granularity association matrix based on the grinding parameter distance topological matrix to obtain a grinding parameter topological modulation association matrix; use the grinding parameter topological modulation association matrix as the perspective co-projection space, project the grinding speed timing input vector and the grinding pressure timing input vector into the perspective co-projection space to obtain a perspective modulated grinding speed timing input vector and a perspective modulated grinding pressure timing input vector; fuse the perspective modulated grinding speed timing input vector and the perspective modulated grinding pressure timing input vector to obtain the grinding parameter timing association vector.

[0016] In the automatic control system of the above-mentioned polarizer edge grinding equipment, the timing correlation feature extraction module includes: a matrix segmentation unit, which is used to perform matrix segmentation on the grinding parameter timing correlation matrix to obtain the multiple grinding parameter local timing correlation sub-matrices; a local timing correlation unit, which is used to perform local timing correlation feature extraction on the multiple grinding parameter local timing correlation sub-matrices to obtain multiple grinding parameter local timing correlation sub-feature matrices; and an attention enhancement unit, which is used to perform attention feature-based enhancement on the multiple grinding parameter local timing correlation sub-feature matrices to obtain the multiple grinding parameter local timing space enhancement sub-feature matrices.

[0017] In the automatic control system of the polarizer edge grinding equipment, the local timing association unit is used to: input the multiple grinding parameter local timing association sub-matrices into a convolutional neural network model including a block structure feature extraction module to obtain the multiple grinding parameter local timing association sub-feature matrices.

[0018] In the automatic control system of the polarizer edge grinding equipment mentioned above, the attention enhancement unit is used to: input the multiple local temporal correlation sub-feature matrices of grinding parameters into the temporal correlation feature extraction model based on the spatial attention mechanism to obtain the multiple local temporal spatial enhancement sub-feature matrices of grinding parameters.

[0019] In the automatic control system of the above-mentioned polarizer edge grinding equipment, the grinding speed adjustment module is used to: input the grinding parameter time-space enhanced fusion feature matrix into the classifier-based grinding speed adjuster to obtain a classification result, and the classification result is used to determine whether the grinding speed at the current time point should be increased, decreased or maintained.

[0020] In the automatic control system of the polarizer edge grinding equipment mentioned above, the grinding speed adjustment module includes: a grinding speed feature matrix expansion unit, which is used to expand the grinding parameter time-space enhancement fusion feature matrix into a grinding parameter time-space enhancement fusion classification feature vector according to a row vector or a column vector; a grinding speed feature fully connected encoding unit, which is used to use the fully connected layer of the classifier to fully connect the grinding parameter time-space enhancement fusion classification feature vector to obtain a grinding parameter time-space enhancement fusion fully connected encoding feature vector; a grinding speed feature probability acquisition unit, which is used to pass the grinding parameter time-space enhancement fusion fully connected encoding feature vector through the Softmax classification function of the classifier to obtain a first probability that the grinding speed at the current time point should be increased, a second probability that the grinding speed at the current time point should be reduced, and a third probability that the grinding speed at the current time point should be maintained; a classification result determination unit, which is used to determine the classification result based on a comparison between the first probability, the second probability and the third probability.

[0021] According to another aspect of the present application, there is provided an automatic control method for a polarizer edge grinding device, comprising:

[0022] Collecting the grinding speed and grinding pressure at multiple predetermined time points during the predetermined time period when the polarizer is edge-grinded;

[0023] Arranging the grinding speeds and grinding pressures at a plurality of predetermined time points in the predetermined time period into a grinding speed time series input vector and a grinding pressure time series input vector according to the time dimension respectively;

[0024] Performing perspective coordination matching fusion based on topological association expression quantification on the grinding speed time series input vector and the grinding pressure time series input vector to obtain a grinding parameter time series association vector;

[0025] Multiplying the grinding parameter time series association vector by the transposed vector of the grinding parameter time series association vector to obtain a grinding parameter time series association matrix;

[0026] Performing matrix segmentation on the grinding parameter temporal correlation matrix to obtain a plurality of grinding parameter local temporal correlation sub-matrices, and then sequentially performing local temporal correlation feature extraction and feature enhancement to obtain a plurality of grinding parameter local temporal spatial enhancement sub-feature matrices;

[0027] Fusing the plurality of grinding parameter local temporal spatial reinforcement sub-feature matrices to obtain a grinding parameter temporal spatial reinforcement fusion feature matrix;

[0028] Based on the grinding parameter time-series spatial enhanced fusion feature matrix, it is determined whether the grinding speed at the current time point should be increased, decreased or maintained.

[0029] In the automatic control method of the polarizer edge grinding equipment mentioned above, the grinding parameter temporal correlation matrix is ​​matrix-sliced ​​to obtain a plurality of grinding parameter local temporal correlation sub-matrices, and then local temporal correlation feature extraction and feature enhancement are performed in sequence to obtain a plurality of grinding parameter local temporal spatial enhancement sub-feature matrices, including: matrix-slicing the grinding parameter temporal correlation matrix to obtain the plurality of grinding parameter local temporal correlation sub-matrices; local temporal correlation feature extraction is performed on the plurality of grinding parameter local temporal correlation sub-matrices to obtain a plurality of grinding parameter local temporal correlation sub-feature matrices; and, attention-based feature enhancement is performed on the plurality of grinding parameter local temporal correlation sub-feature matrices to obtain the plurality of grinding parameter local temporal spatial enhancement sub-feature matrices.

[0030] In the automatic control method of the above-mentioned polarizer edge grinding equipment, based on the grinding parameter time-space enhanced fusion feature matrix, it is determined whether the grinding speed at the current time point should be increased, decreased or maintained, including: inputting the grinding parameter time-space enhanced fusion feature matrix into a grinding speed adjuster based on a classifier to obtain a classification result, and the classification result is used to determine whether the grinding speed at the current time point should be increased, decreased or maintained.

[0031] Compared with the prior art, the present application provides an automatic control system and method for polarizer edge grinding equipment, which collects the grinding speed and grinding pressure at multiple predetermined time points in a predetermined time period during polarizer edge grinding in real time, and uses deep learning technology to perform data analysis and processing on the grinding speed and the grinding pressure, thereby determining whether the grinding speed at the current time point should be increased, decreased, or maintained. In this way, dynamic adjustments can be made according to actual conditions during the polarizer grinding process, reducing wear and manual intervention and reducing production costs, thereby improving the processing quality of the polarizer, improving processing efficiency and product quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] 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.

[0033] Figure 1 It is a block diagram of an automatic control system of a polarizer edge grinding device according to an embodiment of the present application.

[0034] Figure 2 Schematic diagram of the architecture of an automatic control system for a polarizer edge grinding device according to an embodiment of the present application.

[0035] Figure 3 It is a block diagram of a timing correlation feature extraction module in an automatic control system of a polarizer edge grinding device according to an embodiment of the present application.

[0036] Figure 4 The present invention is a flowchart of an automatic control method of a polarizer edge grinding device according to an embodiment of the present application. DETAILED DESCRIPTION

[0037] The embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, but rather these embodiments are provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0038] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0039] It should be noted that the terms "first\second\third" involved in the embodiments of the present application are only used to distinguish similar objects, and do not represent a specific order for the objects. It is understandable that the specific order or sequence of "first\second\third" can be interchanged where permitted. It should be understood that the objects distinguished by "first\second\third" can be interchanged where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein.

[0040] As shown in the present disclosure and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0041] Various exemplary embodiments, features and aspects of the present disclosure 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.

[0042] Polarizer is a special optical element that can selectively pass or block light waves in a specific direction, thereby adjusting the polarization state of light. This element has important applications in a variety of devices such as liquid crystal displays, camera lenses, lasers, etc. For example, in liquid crystal displays, polarizers help control the transmission of light on the display screen to create clear images. Therefore, polarizers are a very critical component in the field of optics.

[0043] Polaroid edge grinding equipment is used to perform precise edge grinding on polaroids to ensure that their edges are smooth and the quality meets the standards, so as to facilitate subsequent processing and application. Polaroid edge grinding equipment usually includes edge grinding machines, automatic control systems, sensors, grinding tools and other components. The edge grinding process usually requires precise control to ensure that the edge quality of the polaroid after edge grinding meets the requirements. The automatic control system can realize automatic monitoring and adjustment of the edge grinding process to improve production efficiency and product quality. In general, polaroid edge grinding equipment is a device used to perform edge grinding on polaroids, and is one of the important equipment in the optical component manufacturing process.

[0044] However, because traditional polarizer edge grinding control is usually based on static set values, it lacks dynamic adjustment of real-time processing status. This may lead to poor results when facing different materials, workpiece conditions or processing conditions. Secondly, in traditional edge grinding control, the operator may need to constantly manually adjust the parameters of the edge grinding equipment, which is easily affected by human factors and causes inconsistent operations or errors.

[0045] Therefore, an optimized automatic control system for polarizer edge grinding equipment is desired, which collects the grinding speed and grinding pressure at multiple predetermined time points in a predetermined time period during polarizer edge grinding in real time, and uses deep learning technology to perform data analysis and processing on the grinding speed and the grinding pressure, thereby determining whether the grinding speed at the current time point should be increased, decreased, or maintained. In this way, dynamic adjustments can be made according to actual conditions during the polarizer grinding process, reducing wear and manual intervention and reducing production costs, thereby improving the processing quality of the polarizer, improving processing efficiency and product quality.

[0046] Figure 1 It is a block diagram of an automatic control system of a polarizer edge grinding device according to an embodiment of the present application. Figure 2 FIG. 1 is a schematic diagram of the architecture of an automatic control system of a polarizer edge grinding device according to an embodiment of the present application. Figure 1 and Figure 2As shown, according to the automatic control system 100 of the polarizer edge grinding equipment of the embodiment of the present application, it includes: a polarizer edge grinding parameter acquisition module 110, which is used to collect the grinding speed and grinding pressure at multiple predetermined time points in a predetermined time period during polarizer edge grinding; a parameter time dimension arrangement module 120, which is used to arrange the grinding speed and grinding pressure at multiple predetermined time points in the predetermined time period according to the time dimension as a grinding speed time series input vector and a grinding pressure time series input vector respectively; a grinding speed pressure feature matching and fusion module 130, which is used to perform perspective coordinated matching and fusion based on topological association expression quantification on the grinding speed time series input vector and the grinding pressure time series input vector to obtain a grinding parameter time series association vector; a grinding speed pressure feature association module 140, which is used to The grinding parameter time series association vector is multiplied by the transposed vector of the grinding parameter time series association vector to obtain a grinding parameter time series association matrix; a time series association feature extraction module 150 is used to perform matrix segmentation on the grinding parameter time series association matrix to obtain multiple grinding parameter local time series association sub-matrices, and then perform local time series association feature extraction and feature enhancement in sequence to obtain multiple grinding parameter local time series space enhancement sub-feature matrices; a fusion feature module 160 is used to fuse the multiple grinding parameter local time series space enhancement sub-feature matrices to obtain a grinding parameter time series space enhancement fusion feature matrix; and a grinding speed adjustment module 170 is used to determine whether the grinding speed at the current time point should be increased, decreased or maintained based on the grinding parameter time series space enhancement fusion feature matrix.

[0047] In an embodiment of the present application, the polarizer edge grinding parameter acquisition module 110 is used to collect the grinding speed and grinding pressure at multiple predetermined time points in a predetermined time period when the polarizer is edged. It should be understood that considering that the grinding speed and grinding pressure are two key processing parameters in the polarizer edge grinding process, they have an important impact on the processing quality and efficiency. In detail, the grinding speed refers to the speed at which the grinding tool grinds the surface of the workpiece per unit time, usually expressed as the grinding length per unit time. The grinding speed directly affects the grinding efficiency and processing quality, and a suitable grinding speed can ensure processing accuracy, surface finish and processing efficiency. The grinding pressure is the force applied by the grinding tool to the workpiece, which can also be understood as pressure per unit area. The grinding pressure affects wear, thermal deformation and processing quality during the grinding process. Appropriate grinding pressure can ensure that the grinding process is stable and the processing accuracy is high. Based on this, in the technical solution of the present application, the grinding speed and grinding pressure at multiple predetermined time points in a predetermined time period when the polarizer is edged are collected. That is, grinding speed and grinding pressure are important factors affecting processing quality. By monitoring and controlling these two parameters, the processing quality can be guaranteed to be stable and consistent. The timing change law and mutual correlation of grinding speed and grinding pressure can be understood, which provides a basis and support for the precise control of subsequent grinding speed, thereby optimizing the processing technology and improving the efficiency of polarizer edge grinding and product quality.

[0048] It is worth mentioning that the grinding speed and the grinding pressure are usually collected through special sensors and instruments. Specifically, the force sensor is used to measure the force applied to the grinding tool or workpiece during the grinding process, that is, the grinding pressure. In particular, the force sensor can convert the force into an electrical signal output, and the grinding pressure data can be obtained by measuring these electrical signals. Common force sensors include strain gauge sensors, piezoelectric sensors, etc. The speed sensor is used to measure the movement speed of the grinding tool during the grinding process, that is, the grinding speed. In particular, the speed sensor can detect the rotation speed or linear speed of the grinding tool to obtain the grinding speed data. Common speed sensors include photoelectric encoders, Hall sensors, etc. The above instruments can accurately collect the data of grinding speed and grinding pressure in real time, and provide important support for the monitoring, control and optimization of the processing process.

[0049] In an embodiment of the present application, the parameter time dimension arrangement module 120 is used to arrange the grinding speed and grinding pressure at multiple predetermined time points in the predetermined time period according to the time dimension as a grinding speed timing input vector and a grinding pressure timing input vector. Accordingly, considering that in the grinding process, the grinding speed and the grinding pressure are often related to time, that is, the grinding parameters at the previous moment will affect the subsequent processing conditions. In other words, the grinding speed and the grinding pressure have certain timing characteristic information in the time dimension. Therefore, in order to retain the timing correlation of the grinding speed and the grinding pressure, it helps the model to better understand the time characteristic relationship between the grinding speed and the grinding pressure. In the technical solution of the present application, the grinding speed and the grinding pressure at multiple predetermined time points in the predetermined time period are arranged according to the time dimension as a grinding speed timing input vector and a grinding pressure timing input vector. Specifically, during the grinding process, the grinding speed and the grinding pressure tend to change over time. By arranging the grinding speed and the grinding pressure according to the time dimension, the dynamic change characteristics of the grinding speed and the grinding pressure in time series can be captured, so as to subsequently predict the future grinding speed and perform accurate control.

[0050] In an embodiment of the present application, the grinding speed pressure feature matching and fusion module 130 is used to perform perspective coordination matching and fusion based on topological association expression quantification on the grinding speed time series input vector and the grinding pressure time series input vector to obtain a grinding parameter time series correlation vector. In particular, in the technical solution of the present application, considering that the grinding speed time series input vector is extracted from the time series data of the grinding speed, the main focus is on the changing trend of the grinding speed. The grinding pressure time series input vector is extracted from the time series data of the grinding pressure, and the main focus is on the changing trend of the grinding pressure. The grinding speed time series input vector and the grinding pressure time series input vector are extracted from the time series data of the grinding speed and the grinding pressure, respectively, and they represent different physical properties in the edge grinding process of the polarizer. Due to the essential differences in these features and the different extraction methods, these two feature vectors have significant feature heterogeneity. Due to the feature heterogeneity, the grinding speed time series input vector and the grinding pressure time series input vector may exist on different manifolds, which means that they have different distribution laws in their respective spaces. Directly splicing them together may result in the final feature vector not having a clear manifold in the new space, that is, the feature manifold smoothness is insufficient. Since the two feature vectors come from different physical properties, they may not be well mapped to each other in the fused feature space, that is, one type of feature may not accurately reflect the feature changes of another type, resulting in increased uncertainty in feature expression. Based on this, in the technical solution of the present application, the grinding speed time series input vector and the grinding pressure time series input vector are fused by perspective coordination matching based on topological association expression quantization.

[0051] Specifically, in an embodiment of the present application, the grinding speed pressure feature matching fusion module is used to: construct a grinding parameter distance topological matrix between the grinding speed time series input vector and the grinding pressure time series input vector; construct a grinding parameter eigenvalue granularity association matrix between the grinding speed time series input vector and the grinding pressure time series input vector; perform topological association modulation on the grinding parameter eigenvalue granularity association matrix based on the grinding parameter distance topological matrix to obtain a grinding parameter topological modulation association matrix; use the grinding parameter topological modulation association matrix as a perspective co-projection space, project the grinding speed time series input vector and the grinding pressure time series input vector to the perspective co-projection space to obtain a perspective modulated grinding speed time series input vector and a perspective modulated grinding pressure time series input vector; fuse the perspective modulated grinding speed time series input vector and the perspective modulated grinding pressure time series input vector to obtain the grinding parameter time series association vector.

[0052] More specifically, in the embodiment of the present application, the grinding speed time series input vector and the grinding pressure time series input vector are subjected to perspective coordination matching fusion based on topological association expression quantification to obtain a grinding parameter time series association vector according to the following formula; wherein the formula is:

[0053] M t =D(V1,V2)

[0054]

[0055] V c =α·V1'+β·V2'

[0056] Wherein, V1 represents the grinding speed time series input vector, V2 represents the grinding pressure time series input vector, and D(V1, V2) is the distance topological matrix between the grinding speed time series input vector and the grinding pressure time series input vector, that is, d i,j =d(v 1i ,v 2j ), d i,j is the eigenvalue of the position (i, j) in the distance topology matrix, d(v 1i ,v 2j ) represents the distance between the eigenvalue of the i-th position of the grinding speed time series input vector and the eigenvalue of the j-th position of the grinding pressure time series input vector, and V1 and V2 are both column vectors, Mt represents the distance topology matrix, T represents the transpose of the vector, Represents matrix multiplication, M v represents the eigenvalue granularity association matrix, M c represents the topological modulation association matrix, V1' represents the time series input vector of the viewing angle modulation grinding speed, V2' represents the time series input vector of the viewing angle modulation grinding pressure, α and β represent weighted hyperparameters, V c Represents the fused feature vector.

[0057] Based on this, in the technical solution of the present application, the grinding speed time series input vector and the grinding pressure time series input vector are subjected to perspective coordinated matching fusion based on the quantification of topological association expression, which first constructs the grinding parameter distance topological matrix between the grinding speed time series input vector and the grinding pressure time series input vector. This step utilizes the idea of ​​graph theory, takes the characteristic values ​​of each position in the grinding speed time series input vector and the grinding pressure time series input vector as nodes, and uses the distance metric function to determine the low-dimensional embedding expression of the edges between nodes, so as to construct a characteristic expression that can quantify the topological association relationship between the grinding speed time series input vector and the grinding pressure time series input vector. This step is crucial to understanding the spatial distribution of data points and their mutual connections, and also provides basic structural information for subsequent steps.

[0058] Next, a grinding parameter eigenvalue granularity association matrix between the grinding speed time series input vector and the grinding pressure time series input vector is constructed. This step focuses on exploring the correlation between the internal attributes of the eigenvectors. In a specific example, the product between the transposed vectors of the grinding speed time series input vector and the grinding pressure time series input vector is calculated to obtain the grinding parameter eigenvalue granularity association matrix. Furthermore, the grinding parameter eigenvalue granularity association matrix is ​​topologically modulated based on the grinding parameter distance topological matrix to obtain a grinding parameter topological modulation association matrix. That is, the eigenvalue granularity association information of the grinding speed time series input vector and the grinding pressure time series input vector is driven to wander on the distance topological map to construct a low-dimensional modulation perspective co-projection space domain for mapping the grinding speed time series input vector and the grinding pressure time series input vector to a continuous high-dimensional regression space attribute.

[0059] Next, the grinding parameter topological modulation correlation matrix is ​​used as the perspective co-projection space, and the grinding speed time series input vector and the grinding pressure time series input vector are projected into the perspective co-projection space to obtain the perspective modulated grinding speed time series input vector and the perspective modulated grinding pressure time series input vector. That is, the key to this technical solution is to find an implicit third space with modulation capability, in which as much information of the original high-dimensional data as possible is retained, so that the converted new feature vector not only reduces the dimension, but also maintains the relative position relationship between the original features.

[0060] Finally, the perspective modulated grinding speed time series input vector and the perspective modulated grinding pressure time series input vector are fused to obtain the grinding parameter time series association vector. Fusion can be achieved in various forms, ranging from simple arithmetic average to complex ensemble learning algorithms, depending on the application scenario and the goals pursued. The final grinding parameter time series association vector combines the main features of the two sets of input data and provides a more comprehensive data representation, which is not only conducive to improving the performance of machine learning models, but also promotes the ability of cross-modal or multi-source data analysis, thereby significantly enhancing the understanding and utilization efficiency of complex data structures.

[0061] In an embodiment of the present application, the grinding speed pressure feature association module 140 is used to multiply the grinding parameter timing association vector with the transposed vector of the grinding parameter timing association vector to obtain a grinding parameter timing association matrix. It should be understood that the grinding parameter timing association vector contains the mutual influence relationship between the grinding speed and the grinding pressure in the timing. And considering that the grinding speed and the grinding pressure have a mutual influence relationship at different time points. Therefore, in order to better understand the correlation between the grinding speed timing input vector and the grinding pressure timing input vector at different time points, so as to conduct a more accurate analysis later, in the technical solution of the present application, the grinding parameter timing association vector is multiplied with the transposed vector of the grinding parameter timing association vector to obtain a grinding parameter timing association matrix, so that the timing effects and influences between them can be better understood, and more in-depth data support is provided for the monitoring, optimization and grinding speed prediction of the polarizer processing process.

[0062] In an embodiment of the present application, the timing correlation feature extraction module 150 is used to perform matrix segmentation on the grinding parameter timing correlation matrix to obtain multiple grinding parameter local timing correlation sub-matrices, and then perform local timing correlation feature extraction and feature enhancement in sequence to obtain multiple grinding parameter local timing space enhancement sub-feature matrices. Figure 3 is a block diagram of a timing correlation feature extraction module in an automatic control system of a polarizer edge grinding device according to an embodiment of the present application. Specifically, in the embodiment of the present application, Figure 3 As shown, the timing correlation feature extraction module 150 includes: a matrix segmentation unit 151, used for performing matrix segmentation on the grinding parameter timing correlation matrix to obtain the multiple grinding parameter local timing correlation sub-matrices; a local timing correlation unit 152, used for performing local timing correlation feature extraction on the multiple grinding parameter local timing correlation sub-matrices to obtain multiple grinding parameter local timing correlation sub-feature matrices; and an attention enhancement unit 153, used for performing attention feature enhancement on the multiple grinding parameter local timing correlation sub-feature matrices to obtain the multiple grinding parameter local timing space enhancement sub-feature matrices.

[0063] Specifically, in an embodiment of the present application, the matrix segmentation unit 151 is used to perform matrix segmentation on the grinding parameter time series correlation matrix to obtain the multiple grinding parameter local time series correlation sub-matrices. It should be understood that, considering that the grinding parameter time series correlation matrix contains a complex time series correlation relationship between the grinding speed and the grinding pressure, and the grinding speed and the grinding pressure have different time series correlation characteristics between different local time scales. That is, the correlation between different parameters may be local, that is, there is a strong correlation between some parameters, and a weaker relationship with other parameters. Therefore, in order to understand and analyze different local time series correlations separately, and to better characterize the subsequent grinding speed, in the technical solution of the present application, the grinding parameter time series correlation matrix is ​​matrix segmented to obtain the multiple grinding parameter local time series correlation sub-matrices. That is, by splitting the matrix, the overall complex grinding parameter timing correlation matrix can be decomposed into multiple grinding parameter local timing correlation sub-matrices. Each grinding parameter local timing correlation sub-matrix represents the grinding parameter timing correlation feature information of a local time period. In this way, we can focus on the local relationship of grinding parameters and study these local correlation relationships of grinding parameters in more depth to achieve better grinding effects and processing quality.

[0064] Specifically, in an embodiment of the present application, the local temporal association unit 152 is used to extract local temporal association features from the multiple local temporal association sub-matrices of grinding parameters to obtain multiple local temporal association sub-feature matrices of grinding parameters. More specifically, in an embodiment of the present application, the local temporal association unit is used to: input the multiple local temporal association sub-matrices of grinding parameters into a convolutional neural network model including a block structure feature extraction module to obtain the multiple local temporal association sub-feature matrices of grinding parameters. In particular, the convolutional neural network model including the block structure feature extraction module includes: a first two-dimensional convolution kernel and a second two-dimensional convolution kernel, wherein the first two-dimensional convolution kernel and the second two-dimensional convolution kernel are transposed to each other. Further, considering that the multiple local temporal association sub-matrices of grinding parameters all contain local implicit feature information, and the convolutional neural network is good at extracting implicit features from data, especially for structured data such as sequences, the feature extraction effect is remarkable. Therefore, in the technical solution of the present application, the multiple local time-series correlation submatrices of grinding parameters are input into a convolutional neural network model including a block structure feature extraction module to obtain the multiple local time-series correlation sub-feature matrices of grinding parameters. In particular, the convolutional neural network can automatically learn the correlation features between parameters, which helps to discover the laws and patterns hidden in the data. For data containing block structures, such as the local time-series correlation submatrix of grinding parameters, the convolutional neural network can effectively capture the local characteristics and time-series information about the correlation of grinding parameters in the local time-series correlation submatrix of grinding parameters, so as to better understand the time-series correlation relationship between parameters.

[0065] Specifically, in the embodiment of the present application, the attention strengthening unit 153 is used to strengthen the multiple grinding parameter local temporal association sub-feature matrices based on the attention feature to obtain the multiple grinding parameter local temporal spatial reinforcement sub-feature matrices. More specifically, in the embodiment of the present application, the attention strengthening unit is used to: input the multiple grinding parameter local temporal association sub-feature matrices into the temporal association feature extraction model based on the spatial attention mechanism to obtain the multiple grinding parameter local temporal spatial reinforcement sub-feature matrices. Accordingly, considering that each of the multiple grinding parameter local temporal association sub-feature matrices in the multiple grinding parameter local temporal association sub-feature matrices has key feature information about the local temporal association features of the grinding parameters, the spatial attention mechanism can help the model focus on important features, thereby better capturing the correlation between parameters. Therefore, in the technical solution of the present application, the multiple grinding parameter local temporal association sub-feature matrices are input into the temporal association feature extraction model based on the spatial attention mechanism to obtain the multiple grinding parameter local temporal spatial reinforcement sub-feature matrices. It is worth mentioning that the spatial attention mechanism can help the model to be more flexible and precise when learning the correlation between parameters. That is, the model can dynamically adjust the weights according to the importance of the local temporal correlation of grinding parameters between the sub-feature matrices of the local temporal correlation of grinding parameters, so as to better understand and utilize the temporal correlation information between the local temporal correlation features of grinding parameters.

[0066] In the embodiment of the present application, the fusion feature module 160 is used to fuse the multiple grinding parameter local temporal space reinforcement sub-feature matrices to obtain a grinding parameter temporal space reinforcement fusion feature matrix. Accordingly, considering that each grinding parameter local temporal space reinforcement sub-feature matrix of the multiple grinding parameter local temporal space reinforcement sub-feature matrices represents the grinding parameter association feature information under a local time dimension, therefore, in order to integrate the different grinding parameter local temporal space reinforcement sub-feature information, in the technical solution of the present application, the multiple grinding parameter local temporal space reinforcement sub-feature matrices are fused to obtain a grinding parameter temporal space reinforcement fusion feature matrix, which can form a more comprehensive and richer feature representation, which helps to improve the model's understanding and representation capabilities of data.

[0067] In an embodiment of the present application, the grinding speed adjustment module 170 is used to determine whether the grinding speed at the current time point should be increased, decreased, or maintained based on the grinding parameter time-series space enhanced fusion feature matrix. Specifically, in an embodiment of the present application, the grinding speed adjustment module is used to: input the grinding parameter time-series space enhanced fusion feature matrix into a grinding speed adjuster based on a classifier to obtain a classification result, and the classification result is used to determine whether the grinding speed at the current time point should be increased, decreased, or maintained. That is, classification processing is performed through the grinding parameter time-series space enhanced fusion feature matrix to determine whether the grinding speed at the current time point should be increased, decreased, or maintained. In this way, dynamic adjustments can be made according to actual conditions during the polarizer grinding process, reducing wear and manual intervention and reducing production costs, thereby improving the processing quality of the polarizer, improving processing efficiency and product quality.

[0068] More specifically, in the embodiment of the present application, the grinding speed adjustment module includes: a grinding speed feature matrix expansion unit, which is used to expand the grinding parameter time-space enhancement fusion feature matrix into a grinding parameter time-space enhancement fusion classification feature vector according to a row vector or a column vector; a grinding speed feature fully connected encoding unit, which is used to use the fully connected layer of the classifier to fully connect the grinding parameter time-space enhancement fusion classification feature vector to obtain a grinding parameter time-space enhancement fusion fully connected encoding feature vector; a grinding speed feature probability acquisition unit, which is used to pass the grinding parameter time-space enhancement fusion fully connected encoding feature vector through the Softmax classification function of the classifier to obtain a first probability that the grinding speed at the current time point should be increased, a second probability that the grinding speed at the current time point should be reduced, and a third probability that the grinding speed at the current time point should be maintained; a classification result determination unit, which is used to determine the classification result based on a comparison between the first probability, the second probability and the third probability.

[0069] In summary, the automatic control system 100 of the polarizer edge grinding equipment based on the embodiment of the present application is explained, which collects the grinding speed and grinding pressure at multiple predetermined time points in a predetermined time period during polarizer edge grinding in real time, and uses deep learning technology to perform data analysis and processing on the grinding speed and the grinding pressure, thereby determining whether the grinding speed at the current time point should be increased, decreased, or maintained. In this way, dynamic adjustments can be made according to actual conditions during the polarizer grinding process, reducing wear and manual intervention and reducing production costs, thereby improving the processing quality of the polarizer, improving processing efficiency and product quality.

[0070] Figure 4 Flow chart of the automatic control method of the polarizer edge grinding device according to the embodiment of the present application. Figure 4As shown, according to the embodiment of the present application, the automatic control method of the polarizer edge grinding device includes: S110, collecting the grinding speed and grinding pressure at multiple predetermined time points in a predetermined time period when the polarizer is edged; S120, arranging the grinding speed and grinding pressure at multiple predetermined time points in the predetermined time period into a grinding speed time series input vector and a grinding pressure time series input vector according to the time dimension; S130, performing perspective coordination matching and fusion based on topological association expression quantification on the grinding speed time series input vector and the grinding pressure time series input vector to obtain a grinding parameter time series association vector; S140, comparing the grinding parameter time series association vector with the The transposed vectors of the grinding parameter time series association vectors are multiplied to obtain a grinding parameter time series association matrix; S150, the grinding parameter time series association matrix is ​​matrix-sliced ​​to obtain a plurality of grinding parameter local time series association sub-matrices, and then local time series association feature extraction and feature enhancement are performed in sequence to obtain a plurality of grinding parameter local time series space enhancement sub-feature matrices; S160, the plurality of grinding parameter local time series space enhancement sub-feature matrices are fused to obtain a grinding parameter time series space enhancement fusion feature matrix; and, S170, based on the grinding parameter time series space enhancement fusion feature matrix, it is determined whether the grinding speed at the current time point should be increased, decreased or maintained.

[0071] Here, those skilled in the art can understand that the specific operations of each step in the automatic control method of the polarizer edge grinding device have been referred to above. Figures 1 to 3 The automatic control system of the polarizer edge grinding device has been described in detail, and therefore, its repeated description will be omitted.

[0072] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. An automatic control system for a polarizer edge grinding device, characterized in that: include: A polarizer edge grinding parameter acquisition module is used to acquire the grinding speed and grinding pressure at multiple predetermined time points during the predetermined time period when the polarizer is edged; A parameter time dimension arrangement module, used for arranging the grinding speed and grinding pressure at a plurality of predetermined time points in the predetermined time period into a grinding speed time series input vector and a grinding pressure time series input vector according to the time dimension; A grinding speed pressure feature matching and fusion module is used to perform perspective coordination matching and fusion based on topological association expression quantification on the grinding speed time series input vector and the grinding pressure time series input vector to obtain a grinding parameter time series association vector; A grinding speed pressure characteristic association module, used for multiplying the grinding parameter time series association vector with the transposed vector of the grinding parameter time series association vector to obtain a grinding parameter time series association matrix; A time series correlation feature extraction module is used to perform matrix segmentation on the grinding parameter time series correlation matrix to obtain a plurality of grinding parameter local time series correlation sub-matrices, and then sequentially perform local time series correlation feature extraction and feature enhancement to obtain a plurality of grinding parameter local time series spatial enhancement sub-feature matrices; A fusion feature module, used for fusing the plurality of grinding parameter local temporal spatial reinforcement sub-feature matrices to obtain a grinding parameter temporal spatial reinforcement fusion feature matrix; The grinding speed adjustment module is used to determine whether the grinding speed at the current time point should be increased, decreased or maintained based on the grinding parameter time-series spatial enhanced fusion feature matrix.

2. The automatic control system of the polarizer edge grinding equipment according to claim 1, characterized in that: The grinding speed and pressure feature matching and fusion module is used to: Constructing a grinding parameter distance topological matrix between the grinding speed time series input vector and the grinding pressure time series input vector; Constructing a grinding parameter characteristic value particle size correlation matrix between the grinding speed time series input vector and the grinding pressure time series input vector; Performing topological correlation modulation on the grinding parameter eigenvalue particle size correlation matrix based on the grinding parameter distance topological matrix to obtain a grinding parameter topological modulation correlation matrix; Taking the grinding parameter topological modulation correlation matrix as the perspective co-projection space, projecting the grinding speed time series input vector and the grinding pressure time series input vector into the perspective co-projection space to obtain the perspective modulation grinding speed time series input vector and the perspective modulation grinding pressure time series input vector; The viewing angle modulated grinding speed time series input vector and the viewing angle modulated grinding pressure time series input vector are fused to obtain the grinding parameter time series association vector.

3. The automatic control system of the polarizer edge grinding equipment according to claim 2, characterized in that: The temporal correlation feature extraction module comprises: A matrix segmentation unit, used for performing matrix segmentation on the grinding parameter time series correlation matrix to obtain the plurality of grinding parameter local time series correlation sub-matrices; A local time series association unit, used for extracting local time series association features from the plurality of grinding parameter local time series association sub-matrices to obtain a plurality of grinding parameter local time series association sub-feature matrices; The attention enhancement unit is used to enhance the plurality of grinding parameter local temporal correlation sub-feature matrices based on attention features to obtain the plurality of grinding parameter local temporal spatial enhancement sub-feature matrices.

4. The automatic control system of the polarizer edge grinding equipment according to claim 3, characterized in that: The local timing association unit is used to: The multiple grinding parameter local temporal correlation sub-matrices are input into a convolutional neural network model including a block structure feature extraction module to obtain the multiple grinding parameter local temporal correlation sub-feature matrices.

5. The automatic control system of the polarizer edge grinding equipment according to claim 4, characterized in that: The attention strengthening unit is used to: The plurality of local temporal correlation sub-feature matrices of grinding parameters are input into a temporal correlation feature extraction model based on a spatial attention mechanism to obtain the plurality of local temporal spatial reinforcement sub-feature matrices of grinding parameters.

6. The automatic control system of the polarizer edge grinding equipment according to claim 5, characterized in that: The grinding speed adjustment module is used for: The grinding parameter time-series spatial enhanced fusion feature matrix is ​​input into a grinding speed adjuster based on a classifier to obtain a classification result, and the classification result is used to determine whether the grinding speed at the current time point should be increased, decreased or maintained.

7. The automatic control system of the polarizer edge grinding equipment according to claim 6, characterized in that: The grinding speed adjustment module comprises: A grinding speed feature matrix expansion unit is used to expand the grinding parameter time-series space enhanced fusion feature matrix into a grinding parameter time-series space enhanced fusion classification feature vector according to row vectors or column vectors; A grinding speed feature fully connected encoding unit, used for performing fully connected encoding on the grinding parameter time-series spatial enhanced fusion classification feature vector using the fully connected layer of the classifier to obtain a grinding parameter time-series spatial enhanced fusion fully connected encoding feature vector; a grinding speed characteristic probability obtaining unit, for passing the grinding parameter time-series spatial enhanced fusion fully connected encoding feature vector through the Softmax classification function of the classifier to obtain a first probability that the grinding speed at the current time point should be increased, a second probability that the grinding speed at the current time point should be decreased, and a third probability that the grinding speed at the current time point should be maintained; A classification result determination unit is used to determine the classification result based on a comparison among the first probability, the second probability and the third probability.

8. An automatic control method for a polarizer edge grinding device, characterized in that: include: Collecting the grinding speed and grinding pressure at multiple predetermined time points during the predetermined time period when the polarizer is edge-grinded; Arranging the grinding speeds and grinding pressures at a plurality of predetermined time points in the predetermined time period into a grinding speed time series input vector and a grinding pressure time series input vector according to the time dimension respectively; Performing perspective coordination matching fusion based on topological association expression quantification on the grinding speed time series input vector and the grinding pressure time series input vector to obtain a grinding parameter time series association vector; Multiplying the grinding parameter time series association vector by the transposed vector of the grinding parameter time series association vector to obtain a grinding parameter time series association matrix; Performing matrix segmentation on the grinding parameter temporal correlation matrix to obtain a plurality of grinding parameter local temporal correlation sub-matrices, and then sequentially performing local temporal correlation feature extraction and feature enhancement to obtain a plurality of grinding parameter local temporal spatial enhancement sub-feature matrices; Fusing the plurality of grinding parameter local temporal spatial reinforcement sub-feature matrices to obtain a grinding parameter temporal spatial reinforcement fusion feature matrix; Based on the grinding parameter time-series spatial enhanced fusion feature matrix, it is determined whether the grinding speed at the current time point should be increased, decreased or maintained.

9. The automatic control method of the polarizer edge grinding equipment according to claim 8, characterized in that: The grinding parameter temporal correlation matrix is ​​matrix-sliced ​​to obtain a plurality of grinding parameter local temporal correlation sub-matrices, and then local temporal correlation feature extraction and feature enhancement are sequentially performed to obtain a plurality of grinding parameter local temporal spatial enhancement sub-feature matrices, including: The grinding parameter time series correlation matrix is ​​matrix-sliced ​​to obtain the plurality of grinding parameter local time series correlation sub-matrices; Extract local time series correlation features from the plurality of grinding parameter local time series correlation sub-matrices to obtain a plurality of grinding parameter local time series correlation sub-feature matrices; The plurality of grinding parameter local temporal correlation sub-feature matrices are enhanced based on the attention feature to obtain the plurality of grinding parameter local temporal spatial enhancement sub-feature matrices.

10. The automatic control method of the polarizer edge grinding equipment according to claim 9, characterized in that: Based on the grinding parameter time-series spatial enhanced fusion feature matrix, determining whether the grinding speed at the current time point should be increased, decreased, or maintained includes: The grinding parameter time-series spatial enhanced fusion feature matrix is ​​input into a grinding speed adjuster based on a classifier to obtain a classification result, and the classification result is used to determine whether the grinding speed at the current time point should be increased, decreased or maintained.

Images