Automatic control system and method of aluminum profile extruding machine
Through deep learning technology, the image and sensor data of the aluminum rod extruder are extracted and analyzed, and the motor speed and pump angle are generated at the next moment, which solves the problems of high control accuracy and energy consumption of aluminum profile extruder, and improves production efficiency and Product quality.
Patent Information
- Application Number
- CN202411289089.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-14
- Publication Date
- 2024-11-26
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Existing aluminum profile extruders have shortcomings in controlling extrusion speed accuracy and energy consumption, and fixed speed and low frequency limits limit production efficiency and product quality.
Using deep learning technology, the aluminum rod collected by the camera is extruded to produce the image and the motor acquired by the sensor for multiple predetermined time points, and the optimized pixel enhancement feature vector and multimodal control feature vector are extracted, and the decoder is used to extract the optimized pixel enhancement feature vector and multimodal control feature vector. Generate the speed value and control the pump angle for the next moment.
The extrusion speed accuracy and control accuracy of the aluminum profile extruder are improved, the production efficiency and product quality are improved, and the stability and consistency of the extrusion process are optimized.
Smart Images

Figure CN119016533A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of aluminum profile extrusion machines, and more specifically, to an automatic control system and method for an aluminum profile extrusion machine. Background Art
[0002] Aluminum extruder is a device that extrude and shape aluminum bars. The extrusion speed is of great significance to product quality and efficiency. At present, in the production of aluminum manufacturing industry, the extrusion process adopts the method of using motor to drive hydraulic oil pump to provide thrust for extruding aluminum liquid. The control of hydraulic oil pressure has a significant impact on the extrusion accuracy.
[0003] For a long time, most aluminum plants have adopted fixed motor speed production, which has low precision in controlling extrusion speed and high power consumption. Some aluminum plants that use variable frequency control are also restricted by low-frequency limits, and the precision in controlling extrusion speed is limited.
[0004] Therefore, an automatic control system and method for an aluminum profile extrusion machine are 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 automatic control system and method for an aluminum profile extruder, which first obtains the aluminum bar extrusion production image collected by a camera, the control pump angle of the aluminum profile extruder motor at multiple predetermined time points collected by a sensor, and the speed value of the aluminum profile extruder motor at multiple predetermined time points collected by a sensor, and then uses deep learning technology to perform feature extraction and correlation analysis on the three, and finally obtains the aluminum profile extruder speed value and the aluminum profile extruder control pump angle at the next moment through a decoder, thereby helping to improve the accuracy of the extrusion speed of the aluminum profile extruder, further optimize the control accuracy of the extrusion process of the aluminum profile extruder, and improve production efficiency and product quality.
[0006] According to one aspect of the present application, there is provided an automatic control system for an aluminum profile extrusion machine, comprising:
[0007] Aluminum profile extruder data acquisition module, used to acquire aluminum bar extrusion production images acquired by a camera, control pump angles of the aluminum profile extruder motor at multiple predetermined time points acquired by a sensor, and speed values of the aluminum profile extruder motor at multiple predetermined time points acquired by a sensor;
[0008] An aluminum profile extruder data extraction module, used to extract optimized aluminum profile extrusion production pixel enhancement feature vectors and extruder motor multi-modal control feature vectors from the aluminum profile extrusion production images collected by the camera, the control pump angles of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, and the speed values of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor;
[0009] The aluminum profile extruder extrusion precision decoding module is used to obtain the aluminum profile extruder speed value and the aluminum profile extruder control pump angle at the next moment based on the optimized aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multi-modal control feature vector.
[0010] According to another aspect of the present application, there is provided an automatic control method for an aluminum profile extruder, comprising:
[0011] Acquire the aluminum rod extrusion production image collected by the camera, the control pump angle of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, and the speed value of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor;
[0012] Extracting optimized aluminum bar extrusion production pixel enhancement feature vectors and extruder motor multi-modal control feature vectors from the aluminum bar extrusion production images collected by the camera, the control pump angles of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, and the speed values of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor;
[0013] Based on the optimized aluminum rod extrusion production pixel enhancement feature vector and the extruder motor multi-modal control feature vector, the aluminum profile extruder speed value at the next moment and the aluminum profile extruder control pump angle at the next moment are obtained.
[0014] Compared with the prior art, the present application provides an automatic control system and method for an aluminum profile extruder, which first obtains the aluminum bar extrusion production image collected by a camera, the control pump angle of the aluminum profile extruder motor at multiple predetermined time points collected by a sensor, and the speed value of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, and then uses deep learning technology to perform feature extraction and correlation analysis on the three, and finally uses a decoder to obtain the aluminum profile extruder speed value and the aluminum profile extruder control pump angle at the next moment, thereby helping to improve the accuracy of the extrusion speed of the aluminum profile extruder, further optimize the control accuracy of the extrusion process of the aluminum profile extruder, and improve production efficiency and product quality. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] 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.
[0016] Figure 1 It is a schematic block diagram of an automatic control system of an aluminum profile extrusion machine according to an embodiment of the present application.
[0017] Figure 2 It is a block diagram of an aluminum profile extrusion machine data extraction module in an automatic control system of an aluminum profile extrusion machine according to an embodiment of the present application.
[0018] Figure 3 It is a schematic block diagram of a motor multi-modal feature extraction unit in an automatic control system of an aluminum profile extrusion machine according to an embodiment of the present application.
[0019] Figure 4 It is a block diagram of an extrusion accuracy decoding module of an aluminum profile extrusion machine in an automatic control system of an aluminum profile extrusion machine according to an embodiment of the present application.
[0020] Figure 5 It is a flow chart of an automatic control method of an aluminum profile extrusion machine according to an embodiment of the present application. DETAILED DESCRIPTION
[0021] 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.
[0022] 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.
[0023] 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.
[0024] 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.
[0025] Figure 1 Schematic diagram of the automatic control system of the aluminum extrusion machine of the present application embodiment. Figure 1As shown, the automatic control system 100 of the aluminum profile extruder according to the embodiment of the present application includes: an aluminum profile extruder data acquisition module 110, which is used to obtain the aluminum bar extrusion production image collected by the camera, the control pump angle of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, and the speed value of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor; an aluminum profile extruder data extraction module 120, which is used to extract the optimized aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector from the aluminum bar extrusion production image collected by the camera, the control pump angle of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, and the speed value of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor; an aluminum profile extruder extrusion accuracy decoding module 130, which is used to obtain the aluminum profile extruder speed value and the aluminum profile extruder control pump angle at the next moment based on the optimized aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector.
[0026] In the automatic control system 100 of the aluminum extruder, the aluminum extruder data acquisition module 110 is used to obtain the aluminum bar extrusion production image collected by the camera, the control pump angle of the aluminum extruder motor at multiple predetermined time points collected by the sensor, and the speed value of the aluminum extruder motor at multiple predetermined time points collected by the sensor. It should be understood that the aluminum extruder is used to extrude aluminum bars, and its extrusion speed is crucial to product quality and production efficiency. In the modern aluminum manufacturing industry, the extrusion process is usually driven by a motor to provide thrust for the hydraulic oil pump, and the control of the hydraulic oil pressure directly affects the extrusion accuracy. However, for a long time, many aluminum plants still use fixed speed motors for production, which results in low extrusion speed control accuracy and high energy consumption. Even if some aluminum plants have adopted variable frequency control technology, these methods are still insufficient in controlling the accuracy of extrusion speed due to the limitation of low frequency operation. Therefore, in the technical solution of the present application, by acquiring the aluminum bar extrusion production image collected by the camera, the control pump angle of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, and the speed value of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, and combining the deep learning technology, the speed value of the aluminum profile extruder and the control pump angle at the next moment are obtained to improve the accuracy of the extrusion speed, thereby optimizing the control accuracy of the extrusion process, and further enhancing the production efficiency and product quality. "
[0027] Specifically, the purpose of acquiring the aluminum bar extrusion production images collected by the camera and the control pump angle and motor speed data collected by the sensor is to achieve comprehensive monitoring and precise control of the aluminum profile extrusion process. The image data provided by the camera can help identify the actual state of the aluminum bar during the extrusion process, such as surface defects or shape variations, thereby providing a visual basis for adjustments during the production process. The control pump angle and motor speed data collected by the sensor reflect the working state and operating efficiency of the extruder. The control pump angle directly affects the extrusion pressure and material flow, while the motor speed affects the extrusion speed and production rhythm. By acquiring this data at multiple predetermined time points, it is possible to fully understand the dynamic changes in the extrusion process, identify potential problems, and adjust the control strategy in a timely manner to ensure the stability of the extrusion process and the consistency of product quality.
[0028] In the automatic control system 100 of the aluminum extruder, the aluminum extruder data extraction module 120 is used to extract the optimized aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector from the aluminum bar extrusion production image collected by the camera, the control pump angle of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, and the speed value of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor. In this way, high-dimensional and complex data can be compressed into useful information and used to predict and optimize the extrusion process. This can improve the accuracy of the control system, enable it to adjust the operating parameters in a timely manner, thereby ensuring the stability of the production process, improving efficiency, and ultimately ensuring the stability and consistency of product quality.
[0029] Figure 2 FIG. 1 is a block diagram of a data extraction module for an aluminum extrusion machine in an automatic control system for an aluminum extrusion machine according to an embodiment of the present application. Figure 2 As shown, in a specific embodiment of the present application, the aluminum profile extruder data extraction module 120 includes: an aluminum bar extrusion production feature extraction unit 121, used to perform feature extraction on the aluminum bar extrusion production image collected by the camera to obtain the optimized aluminum bar extrusion production pixel enhancement feature vector; a motor multimodal feature extraction unit 122, used to perform feature extraction on the control pump angles of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor and the speed values of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor to obtain the extruder motor multimodal control feature vector.
[0030] It should be understood that the image data itself contains rich visual information, such as the surface condition, shape changes and other possible defects of the aluminum bar, but this information often exists in a high-dimensional and complex form, which is not easy to directly analyze and use. Through feature extraction technology, image data can be converted into more representative and operational feature vectors to fully understand the dynamic factors in the extrusion process, so as to make more accurate adjustments and optimization decisions.
[0031] Furthermore, feature extraction is performed on the control pump angle and speed values of the aluminum profile extruder motor collected by the sensor at multiple predetermined time points to obtain the multi-modal control feature vector of the extruder motor, in order to more comprehensively and accurately understand and optimize the operating performance of the extruder. These data reflect the key operating status of the extruder and can reveal the working characteristics of the motor and its impact on the production process. Among them, the control pump angle and motor speed are two important parameters for the operation of the extruder. The control pump angle determines the pressure output of the hydraulic system, thereby affecting the mechanical process of extrusion; the motor speed directly affects the extrusion speed and production efficiency of the aluminum profile. By collecting these data at multiple time points, dynamic changes in the extrusion process, such as load fluctuations, system response, and stability issues, can be captured. Feature extraction converts these complex raw data into meaningful feature vectors for further analysis and processing.
[0032] In a specific embodiment of the present application, the aluminum bar extrusion production feature extraction unit 121 includes: passing the aluminum bar extrusion production image collected by the camera through a pixel enhancement module to obtain an aluminum bar extrusion production pixel enhanced image; passing the aluminum bar extrusion production pixel enhanced image through an aluminum bar extrusion production feature extractor based on a convolutional neural network to obtain an aluminum bar extrusion production pixel enhanced feature map; pooling the aluminum bar extrusion production pixel enhanced feature map to obtain an aluminum bar extrusion production pixel enhanced feature vector; based on the extruder motor multi-modal control feature vector, performing feature manifold boundary abnormal stacking correction based on fine-grained distribution on the aluminum bar extrusion production pixel enhanced feature vector to obtain an optimized aluminum bar extrusion production pixel enhanced feature vector.
[0033] It should be understood that the image captured by the camera may contain many details and information, but this information may not be clear or difficult to identify due to factors such as lighting conditions, image noise or image resolution. The pixel enhancement module uses a series of image processing techniques to improve the original image. For example, by adjusting the contrast, brightness and color balance, or applying filtering and sharpening algorithms, the key features in the image can be highlighted. This enhancement process can improve the visualization of the image, making tiny defects, details and surface features more clearly visible, thereby improving the analysis accuracy of the image. Specifically, the aluminum rod extrusion production image captured by the camera is input into the encoder of the pixel enhancement module, wherein the encoder uses a convolution layer to explicitly spatially encode the aluminum rod extrusion production image captured by the camera to obtain image features; and the image features are input into the decoder of the pixel enhancement module, wherein the decoder uses a deconvolution layer to deconvolve the image features to obtain the aluminum rod extrusion production pixel enhanced image.
[0034] Furthermore, convolutional neural networks have significant advantages in processing image data. They can automatically learn and extract deep features in images, which are essential for accurately analyzing the extrusion production process. Pixel-enhanced images have improved the clarity and contrast of the image through preliminary processing, making the details and features more obvious. However, these enhanced images still contain a large amount of high-dimensional information, and directly analyzing this information may not be efficient or accurate enough. The introduction of convolutional neural networks (CNNs), through their multi-level convolution and pooling operations, can automatically extract key features in images and simplify image data while retaining the most useful information. Specifically, CNNs are able to identify various patterns and structures in images, such as edges, textures, and shapes, which are key features that characterize the aluminum bar extrusion process. For example, CNNs can detect tiny cracks, dimensional deviations, or uneven extrusion areas on the surface of aluminum bars, which are essential for quality control and fault diagnosis. Specifically, each layer of the aluminum bar extrusion production feature extractor based on the convolutional neural network is used to perform convolution processing, mean pooling processing based on the local feature matrix and nonlinear activation processing on the input data in the forward pass of the layer, so that the last layer of the aluminum bar extrusion production feature extractor based on the convolutional neural network outputs the aluminum bar extrusion production pixel enhanced feature map, wherein the input of the aluminum bar extrusion production feature extractor based on the convolutional neural network is the aluminum bar extrusion production pixel enhanced image.
[0035] Furthermore, after being processed by the convolutional layer, the feature map contains feature information at different levels in the image, such as edges, textures, and shapes. However, these feature maps usually have high resolution and contain a lot of detail information. The pooling operation effectively reduces the dimension of the data by downsampling the feature map while retaining the most representative features, making the representation of the feature map more compact and efficient. Among them, common pooling methods are maximum pooling and average pooling. Maximum pooling represents the features of the area by selecting the maximum value in each pooling window. This method helps to retain the salient features in the image, while average pooling calculates the average value of each pooling window to obtain a smoother feature representation.
[0036] In particular, in the technical solution of the present application, the pixel enhancement feature vectors of aluminum rod extrusion production are extracted from image data through convolutional neural networks. These image data have high-dimensional spatial information, and the image features are processed in detail through the pixel enhancement module to capture the subtle structural changes of the aluminum rod during the extrusion process. In contrast, the multimodal control feature vectors of the extruder motor process the data collected by the sensor through the time series encoder, which contains the time series information of the control pump angle and motor speed. These feature vectors mainly focus on the dynamic changes and trends in the time series data, which is completely different from the processing method of spatial image data. The time series feature vectors of the control pump angle and motor speed reflect the dynamic response characteristics of the system, rather than the specific spatial details. This difference in feature processing leads to fine-grained inconsistency in feature distribution. When trying to combine these two types of feature vectors in the fusion process, due to the significant differences in the representation and distribution characteristics of the data features of these different data sources in the feature space, the extruder control accuracy decoding feature vectors obtained after fusion will be abnormally stacked on the boundary of the feature manifold. Therefore, in the technical solution of the present application, based on the multi-modal control feature vector of the extruder motor, the aluminum rod extrusion production pixel enhancement feature vector is subjected to a feature manifold boundary anomaly stacking correction based on fine-grained distribution to obtain an optimized aluminum rod extrusion production pixel enhancement feature vector.
[0037] Among them, based on the extruder motor multimodal control feature vector, the aluminum rod extrusion production pixel enhancement feature vector is corrected for feature manifold boundary abnormal stacking based on fine-grained distribution to obtain an optimized aluminum rod extrusion production pixel enhancement feature vector, including: calculating the position difference between the aluminum rod extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector to obtain an extruder production control shift position difference representation vector; after cascading the aluminum rod extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector into an extruder production control joint feature vector, the extruder production control joint feature vector is input into a one-dimensional convolution layer to obtain an extruder production control shift convolution coding representation vector; after using the first extruder production control linear embedding weight matrix to perform extruder production control linear embedding modulation coding on the aluminum rod extrusion production pixel enhancement feature vector into an aluminum rod extrusion production pixel enhancement feature vector linear modulation vector, calculate The position difference between the linear modulation vector of the aluminum rod extrusion production pixel enhancement feature vector and the multi-modal control feature vector of the extruder motor is obtained to obtain the extruder production control shift modulation differential representation vector; the extruder production control shift position difference representation vector is used as the value vector, the extruder production control shift convolutional coding representation vector and the extruder production control shift modulation differential representation vector are used as the query vector and the key vector, and the extruder production control shift position difference representation vector, the extruder production control shift convolutional coding representation vector and the extruder production control shift modulation differential representation vector are input into the shift corrector based on the simulated converter structure to obtain the extruder production control multi-scale hierarchical fusion shift modulation weight vector; the position point multiplication between the extruder production control multi-scale hierarchical fusion shift modulation weight vector and the aluminum rod extrusion production pixel enhancement feature vector is calculated to obtain the optimized aluminum rod extrusion production pixel enhancement feature vector.
[0038] The following optimization formula is used to express that the aluminum rod extrusion production pixel enhancement feature vector is subjected to the feature manifold boundary abnormal stacking correction based on fine-grained distribution based on the extruder motor multi-modal control feature vector to obtain the optimized aluminum rod extrusion production pixel enhancement feature vector;
[0039] Wherein, the optimization formula is:
[0040]
[0041] V 2d =Conv 1D ([V1; V2])
[0042]
[0043] V1′=V1⊙V d
[0044] Among them, V 1d represents the position difference representation vector of the extruder production control shift, V1 represents the pixel enhancement feature vector of aluminum bar extrusion production, V2 represents the multi-modal control feature vector of the extruder motor, V 2d represents the extruder production control shift convolutional coding representation vector, V 3d represents the extruder production control shift modulation differential representation vector, W1 represents the first extruder production control linear embedding weight matrix, b1 represents the extruder production control modulation bias vector, V d represents the multi-scale hierarchical fusion shift modulation weight vector for extruder production control, V 3d T represents the transposed vector of the extruder production control shift modulation differential representation vector, d represents the length of the extruder production control shift convolutional coding representation vector, Indicates position difference, Conv 1D represents a one-dimensional convolutional layer, [V1; V2] represents the cascade of the aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multi-modal control feature vector, Relu represents a linear rectification function, represents matrix multiplication, Softmax represents the normalized exponential function, ⊙ represents the positional point multiplication, and V1' represents the optimized aluminum rod extrusion production pixel enhancement feature vector.
[0045] In the technical solution of the present application, in the process of fusing the aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector, the fine-grained inconsistency of the feature distribution between the aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector is taken into account, which will cause the fused extruder control precision decoding feature vector to have abnormal stacking on the boundary of the feature manifold. Based on this, in the technical solution of the present application, before fusing the aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector to obtain the extruder control precision decoding feature vector, the aluminum bar extrusion production pixel enhancement feature vector is corrected for abnormal stacking of the feature manifold boundary based on fine-grained distribution based on the extruder motor multimodal control feature vector.
[0046] In the technical solution of the present application, a method for correcting abnormal stacking of feature manifold boundaries based on fine-grained distribution is proposed to address the problem of fine-grained inconsistency in feature distribution that occurs in the process of fusing pixel enhancement feature vectors and multi-modal control feature vectors of extruder motor in aluminum rod extrusion production. The method aims to reduce the abnormal stacking of extruder control precision decoding feature vectors on the feature manifold boundaries, so as to improve the robustness and classification performance of the model.
[0047] Specifically, the method first calculates the element-by-element difference between the aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector to obtain the extruder production control shift element-by-element differential representation vector, thereby capturing the difference between the two feature vectors. Then, the aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector are cascaded into an extruder production control joint feature vector and input into a one-dimensional convolution layer to generate an extruder production control shift convolution coding representation vector. That is, the deep-level shift implicit information between the aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector is captured by one-dimensional convolution coding. At the same time, the aluminum bar extrusion production pixel enhancement feature vector is subjected to extruder production control linear embedding modulation coding using the first extruder production control linear embedding weight matrix to be a linear modulation vector of the aluminum bar extrusion production pixel enhancement feature vector, and the element-by-element difference between the aluminum bar extrusion production pixel enhancement feature vector linear modulation vector and the extruder motor multimodal control feature vector feature vector is calculated to obtain the extruder production control shift modulation differential representation vector. The extruder production control shift modulation differential representation vector is used to represent the positional expression of the displacement information of the aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector in the linear space. That is, in the technical solution of the present application, the extruder production control shift position difference, the extruder production control shift convolutional coding representation vector and the extruder production control shift modulation differential representation vector respectively represent the explicit expression of the fine-grained distribution difference displacement information between the aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector in different feature spaces.
[0048] Next, the extruder production control shift element-by-element differential representation vector is used as the value vector, the extruder production control shift convolutional coding representation vector and the extruder production control shift modulation differential representation vector are used as the query vector and the key vector respectively, and these vectors are input into the shift corrector based on the simulated transformer structure to generate the extruder production control multi-scale hierarchical fusion shift modulation weight vector. In this way, the attention mechanism is used to evaluate the importance of different scale features, that is, the attention mechanism is used to simulate the hierarchical precipitation expression of the fine-grained difference feature distribution between the aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector in different feature spaces. Finally, the element-by-element product of the extruder production control multi-scale hierarchical fusion shift modulation weight vector and the aluminum bar extrusion production pixel enhancement feature vector is calculated to obtain the optimized aluminum bar extrusion production pixel enhancement feature vector, that is, the weighted operation is used to assign weights to each element in the aluminum bar extrusion production pixel enhancement feature vector, which reduces the influence of local disturbances and improves the quality of the feature vector.
[0049] In summary, in the method, the position difference of the extruder production control shift, the extruder production control shift convolutional coding representation vector and the extruder production control shift modulation difference representation vector respectively represent the explicit expression of the fine-grained distribution difference shift information between the aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector in different feature spaces. The generation of the multi-scale hierarchical fusion shift modulation weight vector of the extruder production control uses the attention mechanism to evaluate the importance of features at different scales and provide weight information for feature shift correction. Finally, the generation of the optimized aluminum bar extrusion production pixel enhancement feature vector reduces the impact of disturbances, improves the robustness of the model to local disturbances in the input data, enhances the model's adaptability to noisy or variant data sets, improves the fusion quality of feature vectors, and improves the overall classification performance and training efficiency of the model.
[0050] Figure 3 FIG. 1 is a block diagram of a motor multi-modal feature extraction unit in an automatic control system of an aluminum profile extruder according to an embodiment of the present application. Figure 3 As shown, in a specific embodiment of the present application, the motor multimodal feature extraction unit 122 includes: an extruder control pump angle feature encoding subunit 1221, which is used to feature encode the control pump angle of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor to obtain a control pump angle feature vector; an extruder speed feature encoding subunit 1222, which is used to feature encode the speed values of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor to obtain an extruder motor speed feature vector; an extruder multimodal feature association subunit 1223, which is used to associate the control pump angle feature vector with the extruder motor speed feature vector to obtain the extruder motor multimodal control feature vector.
[0051] It should be understood that the control pump angle is a key parameter of the aluminum extruder, which determines the pressure output of the hydraulic system, and thus affects the mechanical properties and production efficiency of the extrusion process. The data collected by the sensor is usually a continuous time series. Directly using these raw data for analysis may encounter problems such as high processing complexity and low computational efficiency. Through feature encoding, these time series data are converted into feature vectors, which can effectively simplify the data representation and retain the key information.
[0052] Furthermore, the speed values of the aluminum profile extruder motor collected by the sensor at multiple predetermined time points are feature encoded to obtain the extruder motor speed feature vector, in order to convert the complex time series data into a structured feature representation, so as to more effectively perform data analysis, pattern recognition and optimization decisions. The speed value is an important dynamic parameter in the operation of the extruder, which directly affects the stability of the extrusion process and the quality of the product, but its data itself exists in the form of a time series, which is often complex and difficult to use directly. Feature encoding helps to simplify the complexity of data analysis. The original time series data may be very large and contain a lot of redundant information, making it difficult to calculate and analyze efficiently. By encoding features, these data can be compressed into a fixed-dimensional vector, reducing the dimension of the data while retaining important dynamic features and trend information. This conversion not only reduces the burden of storage and calculation, but also makes subsequent data processing and model training more efficient.
[0053] Furthermore, the purpose of associating the control pump angle eigenvector and the extruder motor speed eigenvector to obtain the extruder motor multimodal control eigenvector is to integrate key information from different sources to obtain a comprehensive and in-depth understanding of the extruder operating status. This multimodal eigenvector combines data from different sensors to provide a more accurate description of the extruder performance, which helps to optimize the control strategy, improve production efficiency and predict potential problems.
[0054] In a specific embodiment of the present application, the extruder control pump angle characteristic encoding subunit 1221 includes: constructing the control pump angle of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor into a control pump angle input vector; passing the control pump angle input vector through a control pump angle timing encoder to obtain the control pump angle characteristic vector.
[0055] It should be understood that the control pump angle is a key parameter that affects the pressure output of the hydraulic system, which is directly related to the mechanical properties and product quality during the extrusion process. The control pump angle data collected by the sensor usually exists in the form of a time series, which reflects the change of the pump angle over time. Since the original time series data may contain a lot of details and noise, directly using these data for analysis may result in high computational complexity and difficulty in extracting effective information. Therefore, it is necessary to convert these data into an input vector with a fixed structure for subsequent processing and analysis. By constructing the control pump angle data at multiple time points as an input vector, the key information in the time series data can be systematically represented. This input vector usually includes the pump angle value within a certain time window, so that the timing characteristics and dynamic changes of the data can be retained.
[0056] Furthermore, the control pump angle input vector usually contains angle data at multiple time points, and it may be difficult to capture the dynamic changes and complex relationships in time by directly using these data for analysis. The role of the control pump angle timing encoder is to perform deep feature learning on the time series data and transform the original time series data into a more compact and information-rich feature vector. By considering the sequential relationship and dependency in the time series, the timing encoder can extract the temporal patterns and laws, which is crucial for understanding the changing trend of the control pump angle and its impact on the performance of the extruder. The timing encoder usually adopts model structures such as recurrent neural network (RNN), long short-term memory network (LSTM) or gated recurrent unit (GRU), which are particularly suitable for processing time series data because they can capture the long-term dependencies and temporal dynamic characteristics in the data. Through these models, the timing encoder can learn the trend and periodic pattern of the change of the control pump angle over time, thereby generating a comprehensive and informative control pump angle feature vector. This feature vector not only contains the statistical information of the time series, but also reflects the temporal change pattern and potential dynamic relationship. Specifically, the control pump angle input vector is fully-connectedly encoded using the fully-connected layer of the control pump angle timing encoder to extract high-dimensional implicit features of the eigenvalues at each position in the control pump angle input vector; and the control pump angle input vector is one-dimensionally encoded using the one-dimensional convolutional layer of the control pump angle timing encoder to extract high-dimensional implicit correlation features of the correlation between the eigenvalues at each position in the control pump angle input vector.
[0057] In a specific embodiment of the present application, the extruder speed characteristic encoding subunit 1222 includes: constructing the speed values of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor into an extruder motor speed input vector; passing the extruder motor speed input vector through an extruder motor speed sequence encoder to obtain the extruder motor speed characteristic vector.
[0058] It should be understood that the speed data of the aluminum extruder motor is usually presented in the form of a time series, which reflects the speed change of the motor at a specific time point. The original time series data may contain a lot of details and noise. Directly using this data for analysis may make the calculation complicated and difficult to extract effective information. Therefore, it is necessary to convert this data into an input vector for subsequent processing and analysis. The process of constructing the extruder motor speed input vector usually involves integrating the speed values at multiple time points into a vector, which can systematically represent the change of the motor speed. This input vector can contain the speed values within a certain time window. The design of this time window makes the vector reflect not only the speed at the current moment, but also the speed trend over a period of time in the past. In this way, the input vector can retain the dynamic characteristics of the time series, such as the fluctuation pattern, change trend and possible periodic characteristics of the speed.
[0059] Furthermore, the extruder motor speed input vector is processed by the extruder motor speed sequence encoder to obtain the extruder motor speed feature vector in order to extract more valuable features from the time series data, thereby improving the understanding of motor performance and its impact on the extrusion process. This process can not only effectively capture the dynamic changes of motor speed over time, but also identify potential patterns and trends, thereby providing support for subsequent data analysis and decision-making. The extruder motor speed input vector usually contains speed data at multiple time points, which can reflect the operating status of the motor under different operating conditions. Since the change of motor speed may be affected by many factors, including load changes, control strategies, and equipment wear, directly using the original time series data for analysis may lead to information redundancy and noise interference. Therefore, processing these input vectors through a sequence encoder can more effectively extract relevant features, reduce the impact of noise, and enhance data availability. By encoding the input vector, the sequence encoder can automatically learn the temporal pattern of the motor speed, such as periodic changes, mutation points, or trend fluctuations in the speed. These learned features are summarized and converted into extruder motor speed feature vectors, making the information more concentrated and efficient. Specifically, the extruder motor speed input vector is fully-connectedly encoded using the fully-connected layer of the extruder motor speed sequence encoder to extract high-dimensional implicit features of the eigenvalues at each position in the extruder motor speed input vector; and the extruder motor speed input vector is one-dimensionally encoded using the one-dimensional convolutional layer of the extruder motor speed sequence encoder to extract high-dimensional implicit correlation features of the correlation between the eigenvalues at each position in the extruder motor speed input vector.
[0060] In the automatic control system 100 of the aluminum extruder, the aluminum extruder extrusion precision decoding module 130 is used to obtain the speed value of the aluminum extruder at the next moment and the control pump angle of the aluminum extruder at the next moment based on the optimized aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector. It should be understood that the pixel enhancement feature vector in the aluminum bar extrusion production is usually derived from image data processing, such as analyzing the surface characteristics, shape changes and defects of the aluminum profile by computer vision technology. On the other hand, the multimodal control feature vector of the extruder motor integrates data from different sensors, such as speed sensors, pressure sensors and temperature sensors. Through these data, information such as motor operation status, load changes and control strategies can be obtained. Combining these two feature vectors for analysis can achieve accurate prediction of the speed value and control pump angle of the aluminum extruder at the next moment, thereby optimizing the production process, improving product quality and reducing scrap rate. This prediction method based on multi-source information not only enhances the control ability of the extrusion process, but also improves the response speed and accuracy of the system.
[0061] Figure 4 FIG. 1 is a block diagram of an aluminum profile extrusion machine extrusion precision decoding module in an automatic control system of an aluminum profile extrusion machine according to an embodiment of the present application. Figure 4 As shown, in a specific embodiment of the present application, the aluminum profile extruder extrusion accuracy decoding module 130 includes: an extruder control feature fusion unit 131, which is used to fuse the optimized aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector to obtain an extruder control accuracy decoding feature vector; an extruder control accuracy decoding unit 132, which is used to pass the extruder control accuracy decoding feature vector through a decoder to obtain a decoding value, and the decoding value is used to represent the aluminum profile extruder speed value and the aluminum profile extruder control pump angle at the next moment.
[0062] It should be understood that fusing these two feature vectors to form a decoding feature vector for extruder control accuracy can achieve more comprehensive and sophisticated control of the extrusion process. The fusion process integrates visual information and control data to generate a comprehensive feature representation that can more accurately describe the overall state of the extruder during the production process. This decoded feature vector can not only reflect the operation of the motor, but also combine the visual quality information of the aluminum profile to optimize the control strategy of the extruder, adjust the parameter settings, and improve production accuracy and product quality. Through this comprehensive information analysis, it is possible to better cope with complex changes in production, such as load fluctuations, material non-uniformity, etc., thereby achieving more efficient and reliable extrusion control.
[0063] Furthermore, a decoder is a model or algorithm that aims to map high-dimensional feature vectors to specific control values, such as the motor speed value and the control pump angle at the next moment. The decoder usually includes a trained neural network or other type of model that can generate appropriate output values based on the input feature vector. In this way, the decoder converts complex control precision characteristics into actual operating parameters, allowing the control system to understand and apply this information. This intelligent control method not only improves the stability of the production process, but also promotes the autonomous optimization and fault warning capabilities of production equipment.
[0064] In summary, the embodiment of the present application first obtains the aluminum rod extrusion production image collected by the camera, the control pump angle of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, and the speed value of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, and then uses deep learning technology to perform feature extraction and correlation analysis on the three, and finally uses a decoder to obtain the aluminum profile extruder speed value and the aluminum profile extruder control pump angle at the next moment, thereby helping to improve the accuracy of the extrusion speed of the aluminum profile extruder, further optimize the control accuracy of the extrusion process of the aluminum profile extruder, and improve production efficiency and product quality.
[0065] As described above, the automatic control system 100 of the aluminum extrusion machine according to the embodiment of the present application can be implemented in various terminal devices. In one example, the automatic control system 100 of the aluminum extrusion machine can be integrated into the terminal device as a software module and / or a hardware module. For example, the automatic control system 100 of the aluminum extrusion machine 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 automatic control system 100 of the aluminum extrusion machine can also be one of the many hardware modules of the terminal device.
[0066] Alternatively, in another example, the automatic control system 100 of the aluminum profile extrusion machine and the terminal device may also be separate devices, and the automatic control system 100 of the aluminum profile extrusion machine 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.
[0067] Figure 5 FIG. 1 is a flow chart of an automatic control method for an aluminum profile extruder according to an embodiment of the present application. Figure 5As shown, according to the automatic control method of the aluminum profile extruder of the embodiment of the present application, it includes: S110, acquiring the aluminum bar extrusion production image collected by the camera, the control pump angle of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, and the speed value of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor; S120, extracting the optimized aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector from the aluminum bar extrusion production image collected by the camera, the control pump angle of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, and the speed value of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor; S130, based on the optimized aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multimodal control feature vector, obtaining the aluminum profile extruder speed value and the aluminum profile extruder control pump angle at the next moment.
[0068] Here, those skilled in the art can understand that the specific operations of each step in the automatic control method of the aluminum profile extruder have been described in the above reference. Figures 1 to 4 The description of the automatic control system of the aluminum profile extrusion machine has been described in detail, and therefore, its repeated description will be omitted.
[0069] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.
[0070] The modules described as separate components may or may not be physically separated, and the components shown as modules 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 modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0071] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.
[0072] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.
[0073] Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is limited by the appended claims rather than the above description, so it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any attached figure mark in the claims should not be regarded as limiting the claims involved.
[0074] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the system claim can also be implemented by one unit or device through software or hardware. The second and other words are used to indicate names, but not to indicate any particular order.
[0075] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit of the technical solution of the present invention.
Claims
1. An automatic control system for an aluminum profile extruder, characterized in that: include: Aluminum profile extruder data acquisition module, used to acquire aluminum bar extrusion production images acquired by a camera, control pump angles of the aluminum profile extruder motor at multiple predetermined time points acquired by a sensor, and speed values of the aluminum profile extruder motor at multiple predetermined time points acquired by a sensor; An aluminum profile extruder data extraction module, used to extract optimized aluminum profile extrusion production pixel enhancement feature vectors and extruder motor multi-modal control feature vectors from the aluminum profile extrusion production images collected by the camera, the control pump angles of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, and the speed values of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor; The aluminum profile extruder extrusion precision decoding module is used to obtain the aluminum profile extruder speed value and the aluminum profile extruder control pump angle at the next moment based on the optimized aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multi-modal control feature vector.
2. The automatic control system of the aluminum profile extruder according to claim 1, characterized in that: The aluminum profile extruder data extraction module comprises: An aluminum bar extrusion production feature extraction unit, used for performing feature extraction on the aluminum bar extrusion production image collected by the camera to obtain the optimized aluminum bar extrusion production pixel enhancement feature vector; The motor multimodal feature extraction unit is used to extract features of the control pump angles of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor and the speed values of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor to obtain the multimodal control feature vector of the extruder motor.
3. The automatic control system of the aluminum profile extruder according to claim 2, characterized in that: The aluminum bar extrusion production feature extraction unit comprises: Passing the aluminum rod extrusion production image collected by the camera through a pixel enhancement module to obtain an aluminum rod extrusion production pixel enhanced image; The aluminum bar extrusion production pixel enhanced image is passed through an aluminum bar extrusion production feature extractor based on a convolutional neural network to obtain an aluminum bar extrusion production pixel enhanced feature map; Pooling the aluminum bar extrusion production pixel enhancement feature map to obtain an aluminum bar extrusion production pixel enhancement feature vector; Based on the multi-modal control feature vector of the extruder motor, the aluminum rod extrusion production pixel enhancement feature vector is corrected for abnormal stacking of feature manifold boundaries based on fine-grained distribution to obtain an optimized aluminum rod extrusion production pixel enhancement feature vector.
4. The automatic control system of the aluminum profile extruder according to claim 3, characterized in that: Based on the multi-modal control feature vector of the extruder motor, the aluminum rod extrusion production pixel enhancement feature vector is corrected for abnormal stacking of feature manifold boundaries based on fine-grained distribution to obtain an optimized aluminum rod extrusion production pixel enhancement feature vector, including: Calculating the position difference between the aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multi-modal control feature vector to obtain the extruder production control displacement position difference representation vector; After cascading the aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multi-modal control feature vector into an extruder production control joint feature vector, the extruder production control joint feature vector is input into a one-dimensional convolution layer to obtain an extruder production control shift convolution coding representation vector; After performing extruder production control linear embedding modulation encoding on the aluminum bar extrusion production pixel enhancement feature vector into an aluminum bar extrusion production pixel enhancement feature vector linear modulation vector using a first extruder production control linear embedding weight matrix, calculating the position difference between the aluminum bar extrusion production pixel enhancement feature vector linear modulation vector and the extruder motor multi-modal control feature vector to obtain an extruder production control shift modulation difference representation vector; Taking the position differential representation vector of the extruder production control shift as a value vector, the extruder production control shift convolutional coding representation vector and the extruder production control shift modulation differential representation vector as a query vector and a key vector, the extruder production control shift position differential representation vector, the extruder production control shift convolutional coding representation vector and the extruder production control shift modulation differential representation vector are input into a shift corrector based on an imitation transformer structure to obtain a multi-scale hierarchical fusion shift modulation weight vector for extruder production control; The optimized aluminum rod extrusion production pixel enhancement feature vector is obtained by calculating the position point multiplication between the multi-scale hierarchical fusion shift modulation weight vector of the extrusion machine production control and the aluminum rod extrusion production pixel enhancement feature vector.
5. The automatic control system of the aluminum profile extruder according to claim 4, characterized in that: The motor multimodal feature extraction unit comprises: An extruder control pump angle feature encoding subunit, used for feature encoding the control pump angles of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor to obtain a control pump angle feature vector; An extruder speed characteristic encoding subunit, used for characteristic encoding the speed values of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor to obtain a speed characteristic vector of the extruder motor; The extruder multi-modal feature association subunit is used to associate the control pump angle feature vector with the extruder motor speed feature vector to obtain the extruder motor multi-modal control feature vector.
6. The automatic control system of the aluminum profile extruder according to claim 5, characterized in that: The extruder control pump angle feature encoding subunit comprises: constructing the control pump angles of the aluminum profile extruder motor at a plurality of predetermined time points collected by the sensor as a control pump angle input vector; The control pump angle input vector is passed through a control pump angle timing encoder to obtain the control pump angle characteristic vector.
7. The automatic control system of the aluminum profile extruder according to claim 6, characterized in that: The extruder speed characteristic encoding subunit comprises: Constructing the speed values of the aluminum profile extruder motor at a plurality of predetermined time points collected by the sensor as an extruder motor speed input vector; The extruder motor speed input vector is passed through an extruder motor speed sequence encoder to obtain the extruder motor speed characteristic vector.
8. The automatic control system of the aluminum profile extruder according to claim 7, characterized in that: The extrusion accuracy decoding module of the aluminum profile extruder comprises: An extruder control feature fusion unit, used for fusing the optimized aluminum bar extrusion production pixel enhancement feature vector and the extruder motor multi-modal control feature vector to obtain an extruder control accuracy decoding feature vector; The extruder control accuracy decoding unit is used to pass the extruder control accuracy decoding feature vector through a decoder to obtain a decoding value, and the decoding value is used to represent the aluminum profile extruder speed value and the aluminum profile extruder control pump angle at the next moment.
9. An automatic control method for an aluminum profile extruder, characterized in that: include: Acquire the aluminum rod extrusion production image collected by the camera, the control pump angle of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, and the speed value of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor; Extracting optimized aluminum bar extrusion production pixel enhancement feature vectors and extruder motor multi-modal control feature vectors from the aluminum bar extrusion production images collected by the camera, the control pump angles of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, and the speed values of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor; Based on the optimized aluminum rod extrusion production pixel enhancement feature vector and the extruder motor multi-modal control feature vector, the aluminum profile extruder speed value at the next moment and the aluminum profile extruder control pump angle at the next moment are obtained.
10. The automatic control method of the aluminum profile extruder according to claim 9, characterized in that: Extracting optimized aluminum bar extrusion production pixel enhancement feature vectors and extruder motor multi-modal control feature vectors from the aluminum bar extrusion production images collected by the camera, the control pump angles of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, and the speed values of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor, including: Performing feature extraction on the aluminum bar extrusion production image captured by the camera to obtain the optimized aluminum bar extrusion production pixel enhancement feature vector; Feature extraction is performed on the control pump angles of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor and the speed values of the aluminum profile extruder motor at multiple predetermined time points collected by the sensor to obtain the multi-modal control feature vector of the extruder motor.
Citation Information
Cited By
Optimization method and system for aluminum alloy extrusion production
CN120372917A
