Power line production and processing management system and method
By analyzing raw material status images and performance parameters using deep learning-based artificial intelligence technology, the efficiency and accuracy problems of traditional manual inspection have been solved, enabling intelligent quality control and inventory management in power cord production.
Patent Information
- Application Number
- CN202411152933.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-21
- Publication Date
- 2024-11-26
AI Technical Summary
Traditional raw material quality testing methods rely on manual sampling, which is time-consuming, labor-intensive, and susceptible to subjective factors, resulting in unstable test results and failing to meet the efficiency and accuracy requirements of modern power cord production and processing.
By employing deep learning-based artificial intelligence technology, data analysis is performed on raw material status images and raw material performance parameters. Visual features and global semantic features are extracted and embedded, and the raw materials are intelligently judged to be qualified through joint analysis.
It improves testing efficiency and accuracy, reduces human error, enables comprehensive quality control, and allows for the timely detection of potential quality problems.
Smart Images

Figure CN119027735A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of intelligent management, and more specifically, to a power cord production and processing management system and method. Background Art
[0002] With the rapid development of the power industry and the growing demand for electronic products from consumers, the market demand for power cords, as an important component of power transmission, has risen sharply. In order to meet this demand, power cord production and processing companies are faced with multiple challenges such as improving production efficiency, ensuring product quality, and optimizing inventory management.
[0003] Especially in terms of product quality, enterprises must take strict quality control measures to ensure that their products stand out in the fierce market competition. Specifically, in the process of product quality control, the quality screening of raw materials is an important factor affecting the quality of finished power cords. However, traditional raw material quality inspection methods mainly rely on manual sampling inspection, which has many limitations. First of all, manual inspection is not only time-consuming and labor-intensive, but also easily affected by the subjective factors of the inspectors, resulting in instability of the inspection results. In addition, with the expansion of production scale, the efficiency and accuracy of manual inspection are increasingly difficult to meet the needs of modern power cord production and processing.
[0004] Therefore, an intelligent power cord production and processing management system and method are expected. 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 a power cord production and processing management system and method, which uses artificial intelligence technology based on deep learning to perform data analysis on raw material state images and raw material performance parameters, respectively extracts visual feature representations of raw material states and global semantic feature embedding representations of raw material performance parameters, and jointly analyzes the two to intelligently determine whether the raw materials are qualified. In this way, the detection efficiency and accuracy can be effectively improved, human errors can be reduced, potential quality problems can be discovered in a timely manner, and comprehensive quality control can be achieved.
[0006] According to one aspect of the present application, a power cord production and processing management system is provided, which includes: The production plan generation module is used to generate a production plan based on the imported customer order data and production capacity data; The quality control module is used to determine whether the raw materials are qualified based on the entered inspection data; The inventory management module is used to monitor the inventory of the raw materials, and when the inventory of the raw materials is lower than a preset threshold, automatically generate a purchase order and notify the purchasing department.
[0007] According to another aspect of the present application, a power cord production and processing management method is provided, which includes: Generate production plans based on imported customer order data and production capacity data; Based on the entered inspection data, determine whether the raw materials are qualified; The inventory of the raw materials is monitored, and when the inventory of the raw materials is below a preset threshold, a purchase order is automatically generated and the purchasing department is notified.
[0008] Compared with the prior art, the power cord production and processing management system and method provided by the present application uses artificial intelligence technology based on deep learning to perform data analysis on raw material state images and raw material performance parameters, extracts visual feature representations of raw material states and global semantic feature embedding representations of raw material performance parameters, and conducts joint analysis of the two to intelligently determine whether the raw materials are qualified. In this way, it can effectively improve detection efficiency and accuracy, reduce human errors, and promptly discover potential quality problems, thereby achieving comprehensive quality control. BRIEF DESCRIPTION OF THE DRAWINGS
[0009] 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.
[0010] Figure 1 is a block diagram of a power cord production and processing management system according to an embodiment of the present application; Figure 2 A schematic diagram of data flow of a power cord production and processing management system according to an embodiment of the present application; Figure 3 It is a block diagram of a quality control module in a power cord production and processing management system according to an embodiment of the present application; Figure 4 It is a block diagram of a raw material quality analysis unit in a power cord production and processing management system according to an embodiment of the present application; Figure 5 The present invention is a flowchart of a power cord production and processing management method according to an embodiment of the present application. DETAILED DESCRIPTION
[0011] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.
[0012] As shown in this application 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.
[0013] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.
[0014] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. Meanwhile, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.
[0015] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.
[0016] Traditional raw material quality inspection methods mainly rely on manual sampling inspection, which has many limitations. First of all, manual inspection is not only time-consuming and labor-intensive, but also easily affected by the subjective factors of the inspectors, resulting in instability of the inspection results. In addition, with the expansion of production scale, the efficiency and accuracy of manual inspection are increasingly difficult to meet the needs of modern power cord production and processing. Therefore, an intelligent power cord production and processing management system and method are expected.
[0017] In the technical solution of the present application, a power cord production and processing management system is proposed. Figure 1 It is a block diagram of a power cord production and processing management system according to an embodiment of the present application. Figure 2 Schematic diagram of data flow of power cord production and processing management system according to an embodiment of the present application. Figure 1 and Figure 2As shown, the power cord production and processing management system 300 according to the embodiment of the present application includes: a production plan generation module 310, which is used to generate a production plan based on imported customer order data and production capacity data; a quality control module 320, which is used to determine whether the raw materials are qualified based on the input inspection data; an inventory management module 330, which is used to monitor the inventory of the raw materials, and when the inventory of the raw materials is lower than a preset threshold, automatically generate a purchase order and notify the purchasing department.
[0018] In particular, the production plan generation module 310 is used to generate a production plan based on the imported customer order data and production capacity data. By importing customer order data and analyzing production capacity data, the production plan can help the enterprise balance customer demand and production capacity, ensure timely delivery of orders, and avoid overproduction or shortages. In this way, the production process can be optimized to maximize production efficiency and reduce waste.
[0019] In particular, the quality control module 320 is used to determine whether the raw material is qualified based on the input inspection data. The inspection data includes the raw material state image and the performance parameters of the raw material. In particular, in a specific example of the present application, Figure 3 As shown, the quality control module 320 includes: a raw material state feature extraction unit 321, used to extract image features from the raw material state image to obtain a raw material state feature map; an autocorrelation feature optimization unit 322, used to optimize the autocorrelation features of the raw material state feature map to obtain a raw material state optimization feature map; an embedded coding unit 323, used to embed the performance parameters of the raw material using a performance parameter embedding matrix to obtain a raw material performance parameter embedded coding vector; a raw material quality analysis unit 324, used to generate a quality control result based on the multi-modal joint coding features of the raw material state optimization feature map and the raw material performance parameter embedded coding vector, and the quality control result is used to indicate whether the raw material is qualified.
[0020] Specifically, the raw material state feature extraction unit 321 is used to extract image features from the raw material state image to obtain a raw material state feature map. In a specific example of the present application, the raw material state image is input into a raw material state feature extractor based on a hole convolutional neural network model to obtain the raw material state feature map. It should be understood that in order to effectively extract the visual feature representation of the raw material state, a hole convolutional neural network model is used as a raw material state feature extractor to perform image feature extraction on the raw material state image, so as to utilize the hole convolution characteristics to increase the feature receptive field without increasing the number of parameters and the amount of calculation, which can not only effectively capture the spatial context information of the raw material state image, but also retain the local detail features of the raw material state to generate a raw material state feature map.
[0021] Specifically, the autocorrelation feature optimization unit 322 is used to optimize the autocorrelation features of the raw material state feature map to obtain the raw material state optimization feature map. In a specific example of the present application, the raw material state feature map is input into the adaptive self-attention cross-channel feature space structure optimization module to obtain the raw material state optimization feature map. In order to further improve the quality of the raw material state feature representation, the present application introduces an adaptive self-attention cross-channel feature space structure optimization module to process the raw material state feature map, and through cross-channel feature interaction and feature space structure adaptive optimization, the discrimination and discrimination ability of feature representation are improved to generate the raw material state optimization feature map. Specifically, the module first performs layer normalization processing on the raw material state feature map to eliminate the scale difference between layers and ensure the stability of feature processing. Then, the channel context association and spatial structure information of the raw material state feature map are captured through multi-layer convolution operations, and the channel context association information is used as a query, and the spatial structure information is used as a key and value. Based on the self-attention mechanism, the feature cross-channel global association interaction is performed, and while retaining the spatial structure of the feature map, the channel association information between feature structures is fused in units of feature space structure to take into account the robustness and richness of feature expression.
[0022] In an embodiment of the present application, the raw material state feature map is input into an adaptive self-attention cross-channel feature space structure optimization module to obtain the raw material state optimization feature map, including: performing layer normalization on the raw material state feature map to obtain a normalized raw material state feature map; performing point convolution processing on the normalized raw material state feature map to obtain a raw material state channel context association representation feature map; performing convolution encoding on the raw material state channel context association representation feature map to obtain a raw material state space context association representation feature map; copying the raw material state space context association representation feature map to obtain a backup raw material state space context association representation feature map; and performing cross-channel feature optimization on the raw material state channel context association representation feature map, the raw material state space context association representation feature map and the backup raw material state space context association representation feature map. Reshape the features to obtain a raw material state channel context association representation feature matrix, a raw material state space context association representation feature matrix and a backup raw material state space context association representation feature matrix; calculate the cross-channel cross covariance matrix between the raw material state channel context association representation feature matrix and the raw material state space context association representation feature matrix; use the Softmax function to activate the cross-channel cross covariance matrix to obtain a raw material state feature global interactive attention matrix; calculate the product between the backup raw material state space context association representation feature matrix and the raw material state feature global interactive attention matrix to obtain an attention-enhanced raw material state feature representation matrix; reshape the attention-enhanced raw material state feature representation matrix to obtain the raw material state optimization feature map.
[0023] In summary, in the above embodiment, the raw material state feature map is input into the adaptive self-attention cross-channel feature space structure optimization module to obtain the raw material state optimization feature map, including: processing the raw material state feature map with the following autocorrelation attention enhancement formula to obtain the raw material state optimization feature map, wherein the autocorrelation attention enhancement formula is:
[0024]
[0025]
[0026]
[0027]
[0028]
[0029]
[0030]
[0031]
[0032] in, A characteristic diagram showing the state of the raw material, Representation layer normalization operation, represents the normalized raw material state characteristic diagram, represents point convolution, A context-related representation feature graph representing the raw material state channel, Represents convolution processing based on 3×3 convolution kernel, represents the raw material state space context association representation feature map, Indicates a copy operation. represents the backup raw material state space context association representation feature map, represents the feature shape reshaping, , and They represent the raw material state channel context association representation feature matrix, the raw material state space context association representation feature matrix and the backup raw material state space context association representation feature matrix respectively. represents the cross-channel cross covariance matrix, is the scaling factor, represents the normalized exponential function, represents the matrix multiplication operation, Represents the optimized characteristic diagram of the raw material state.
[0033] Specifically, the embedded coding unit 323 is used to embed the performance parameters of the raw material using a performance parameter embedding matrix to obtain a raw material performance parameter embedded coding vector. Considering that the performance parameters of raw materials often contain multiple dimensions, such as hardness, strength, density, etc., and there are usually complex relationships between the performance parameters of each dimension, they together constitute a comprehensive description of the performance of the raw material. Therefore, in order to fully understand the global semantic characteristics of the raw material performance parameters, the present application further uses a performance parameter embedding matrix to embed the performance parameters of the raw material, and maps the multidimensional performance parameters into a unified semantic feature space, thereby capturing the correlation and inherent laws between different performance parameters, and generating a raw material performance parameter embedded coding vector, so as to facilitate subsequent joint analysis and quality judgment.
[0034] Specifically, the raw material quality analysis unit 324 is used to generate a quality control result based on the raw material state optimization feature map and the multi-modal joint coding feature of the raw material performance parameter embedded coding vector, and the quality control result is used to indicate whether the raw material is qualified. In particular, in a specific example of the present application, Figure 4As shown, the raw material quality analysis unit 324 includes: a joint encoding subunit 3241, which is used to embed the raw material state optimization feature map and the raw material performance parameter into the encoding vector and input it into the heterogeneous data-driven multimodal Transformer joint encoder to obtain the multimodal fusion feature map of the raw material state under the performance parameter constraint; a quality control result generation subunit 3242, which is used to input the multimodal fusion feature map of the raw material state under the performance parameter constraint into the classifier-based quality control result generator to obtain the quality control result.
[0035] More specifically, the joint encoding subunit 3241 is used to input the raw material state optimization feature map and the raw material performance parameter embedding coding vector into the heterogeneous data driven multimodal Transformer joint encoder to obtain the multimodal fusion feature map of the raw material state under the performance parameter constraint. It should be understood that in order to realize the joint analysis of the visual features and performance parameter features of the raw materials, the present application further introduces a heterogeneous data driven multimodal Transformer joint encoder to process the raw material state optimization feature map and the raw material performance parameter embedding coding vector, and introduces an attention mechanism in the joint encoding process to capture the correlation between visual features and performance parameters, thereby realizing the deep fusion of multimodal features and enhancing the comprehensive perception of the raw material state. Specifically, in the joint encoding process, the raw material state optimization feature map and the raw material performance parameter embedded coding vector are first linearly transformed and feature-shaped to adapt to the input requirements of the Transformer structure, and the raw material performance parameter characteristics are used as queries, the visual characteristics of the raw material state are used as keys and values, and the performance parameter characteristics of the raw material are used as auxiliary information to guide the cross-domain combination of visual characteristics and performance parameter characteristics, thereby generating a multimodal fusion feature map of the raw material state under performance parameter constraints. In this way, a more accurate description of the raw material state is achieved, providing more accurate decision support for subsequent production processes.
[0036] In an embodiment of the present application, the raw material state optimization feature map and the raw material performance parameter embedding coding vector are input into a heterogeneous data-driven multimodal Transformer joint encoder to obtain a multimodal fusion feature map of the raw material state under performance parameter constraints, including: linearly transforming and reshaping the feature shape of the raw material performance parameter embedding coding vector to obtain a raw material performance parameter semantic feature auxiliary modal modulation matrix; performing a first linear transformation and a second linear transformation on the raw material state optimization feature map to obtain a first linear transformation raw material state optimization feature map and a second linear transformation raw material state optimization feature map; reshaping the feature shape of the first linear transformation raw material state optimization feature map and the second linear transformation raw material state optimization feature map to obtain a first linear transformation raw material state optimization feature map. state optimization feature matrix and a second linear transformation raw material state optimization feature matrix; using the first linear transformation raw material state optimization feature matrix and the second linear transformation raw material state optimization feature matrix as a key matrix and a value matrix, and using the auxiliary modal modulation matrix of the raw material performance parameter semantic features as a query matrix, inputting the query matrix, the key matrix and the value matrix into the Transformer structure to obtain an auxiliary modal attention query result matrix; reshaping and point convolution encoding are performed on the auxiliary modal attention query result matrix to obtain an auxiliary modal attention query result feature map; calculating the position point multiplication of the auxiliary modal attention query result feature map and the raw material state optimization feature map to obtain a multimodal fusion feature map of the raw material state under the constraints of the performance parameters.
[0037] In summary, in the above embodiment, the raw material state optimization feature map and the raw material performance parameter embedding coding vector are input into the heterogeneous data-driven multimodal Transformer joint encoder to obtain the multimodal fusion feature map of the raw material state under the performance parameter constraint, including: processing the raw material state optimization feature map and the raw material performance parameter embedding coding vector with the following cross-domain joint coding formula to obtain the multimodal fusion feature map of the raw material state under the performance parameter constraint, wherein the cross-domain joint coding formula is:
[0038]
[0039]
[0040]
[0041]
[0042]
[0043]
[0044] in, represents the raw material performance parameter embedded coding vector, represents the optimized characteristic diagram of the raw material state, represents the feature shape reshaping, , and Represent different linear transformation weight matrices, , and represents different bias matrices, represents the Sigmoid activation function, represents the first linear transformation raw material state optimization characteristic diagram, represents the second linear transformation raw material state optimization characteristic graph, represents the raw material performance parameter semantic feature auxiliary modal modulation matrix, represents the first linear transformation raw material state optimization characteristic matrix, represents the second linear transformation raw material state optimization characteristic matrix, represents the transpose of a matrix, represents matrix multiplication, represents the normalized exponential function, represents the auxiliary modality attention query result matrix, represents point convolution, represents the auxiliary modality attention query result feature map, It means point multiplication by position. Represents a multimodal fusion feature graph of raw material status under the constraints of the performance parameters.
[0045] More specifically, the quality control result generation subunit 3242 is used to input the multimodal fusion feature map of the raw material state under the performance parameter constraint into a quality control result generator based on a classifier to obtain the quality control result. In a specific example of the present application, the multimodal fusion feature map of the raw material state under the performance parameter constraint is input into a quality control result generator based on a classifier to obtain a quality control result, and the quality control result is used to indicate whether the raw material is qualified. In one example, first, the multimodal fusion feature map of the raw material state under the performance parameter constraint is expanded into a classification feature vector based on a row vector or a column vector; then, the classification feature vector is fully connected encoded using multiple fully connected layers of the classifier to obtain an encoded classification feature vector; then, the encoded classification feature vector is passed through the Softmax classification function of the classifier to obtain the quality control result.
[0046] In a preferred example, when the raw material state optimization feature map and the raw material performance parameter embedded coding vector respectively express the channel dimension cross-distribution self-reinforcement image semantic features of the raw material state image and the embedded coding semantic features of the performance parameters of the raw material, when heterogeneous data-driven multimodal Transformer joint encoding is performed on them, the obtained multimodal fusion feature map of the raw material state under the performance parameter constraint will also cause the overall distribution regression mode to shift due to the difference in the heterogeneous data feature constraint on the channel dimension, that is, it will cause the overflow of the quality control result obtained by the multimodal fusion feature map of the raw material state under the performance parameter constraint input into the classifier-based quality control result generator, affecting its accuracy.
[0047] Based on this, when the applicant of the present application inputs the multimodal fusion feature graph of the raw material state under the performance parameter constraint into the quality control result generator based on the classifier to obtain the quality control result, the multimodal fusion feature graph of the raw material state under the performance parameter constraint is optimized, and the optimization process includes the steps of: Determine the qualified probability value corresponding to the qualified raw material obtained by inputting the multimodal fusion feature map of the raw material state under the constraints of the performance parameters into the quality control result generator based on the classifier; The multimodal fusion feature graph of the raw material state under the performance parameter constraint is passed through a probabilistic activation function to obtain a probabilistic multimodal fusion feature graph of the raw material state under the performance parameter constraint; Subtract the multimodal fusion feature map of the raw material state under the probabilistic performance parameter constraint from the unit feature map, and use the feature map resulting from the point subtraction as a base to calculate a power function with the qualified probability value as an exponent; Performing a dot multiplication of the power function feature graph and the probabilistic material state multimodal fusion feature graph under performance parameter constraints to obtain a material state multimodal fusion intermediate feature graph under performance parameter constraints; Taking the absolute value of each eigenvalue of the multimodal fusion feature map of the raw material state under the probabilistic performance parameter constraint as the base, calculating a power function with the reciprocal of the qualified probability value as the exponent, and summing all the eigenvalues to obtain the multimodal fusion bias eigenvalue of the raw material state under the performance parameter constraint; and After multiplying the bias eigenvalue of the multimodal fusion of the raw material state under the performance parameter constraint by the weight serving as a hyperparameter, the bias eigenvalue is interpolated with the intermediate feature map of the multimodal fusion of the raw material state under the performance parameter constraint to obtain an optimized multimodal fusion feature map of the raw material state under the performance parameter constraint.
[0048] Wherein, the multimodal fusion feature graph of raw material state under the probabilistic performance parameter constraint is recorded as , then the multimodal fusion feature map of raw material state under the optimized performance parameter constraints
[0049] It is expressed as: in, is the multimodal fusion feature map of the raw material state under the constraints of the probabilistic performance parameters, is a unit eigenmap with all eigenvalues being one, and , is the characteristic value of the multimodal fusion feature map of the raw material state under the constraints of the probabilistic performance parameters, represents the qualified probability value, represents a power function that calculates the eigenvalues of each position in the feature map with the specified probability value as an exponent, Indicates point addition, Indicates point reduction, represents dot product, represents the weight as a hyperparameter, It is a multimodal fusion feature map of raw material status under the constraints of the optimized performance parameters.
[0050] That is, for the multimodal fusion feature graph of raw material state under the performance parameter constraint, based on the convergence limit of the Cauchy-Hadamard form power series probability distribution, the feature set of the multimodal fusion feature graph of raw material state under the performance parameter constraint is represented by a formal power series based on the convergence probability, and a set convergence bounded bias represented by the convergence radius based on each eigenvalue of the multimodal fusion feature graph of raw material state under the performance parameter constraint is added, so as to avoid the invalid overflow of the classification result caused by the deviation of the distribution mode of the multimodal fusion feature graph of raw material state under the performance parameter constraint during the regression process of the quality control result generator based on the classifier due to the gradient propagation of the model parameters based on the convergence probability under a unified probability convergence limit standard, thereby improving the accuracy of the quality control result obtained by the quality control result generator based on the classifier.
[0051] In particular, the inventory management module 330 is used to monitor the inventory of the raw materials, and when the inventory of the raw materials is lower than a preset threshold, automatically generate a purchase order and notify the purchasing department. It should be understood that by monitoring the inventory and automatically generating purchase orders, the enterprise can ensure the continuous supply of raw materials and avoid production interruptions or delays. In addition, adaptive generation of purchase orders can help enterprises optimize inventory management, avoid excess or shortage of inventory, thereby reducing costs and improving efficiency.
[0052] As described above, the power cord production and processing management system 300 according to the embodiment of the present application can be implemented in various wireless terminals, such as a server with a power cord production and processing management algorithm. In a possible implementation, the power cord production and processing management system 300 according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the power cord production and processing management system 300 can be a software module in the operating system of the wireless terminal, or it can be an application developed for the wireless terminal; of course, the power cord production and processing management system 300 can also be one of the many hardware modules of the wireless terminal.
[0053] Alternatively, in another example, the power cord production and processing management system 300 and the wireless terminal may also be separate devices, and the power cord production and processing management system 300 may be connected to the wireless terminal via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.
[0054] Furthermore, a power cord production and processing management method is also provided.
[0055] Figure 5 Flow chart of the power cord production and processing management method according to an embodiment of the present application. Figure 5 As shown, the power cord production and processing management method according to the embodiment of the present application includes the following steps: S1, generating a production plan based on imported customer order data and production capacity data; S2, judging whether the raw materials are qualified based on the entered inspection data; S3, monitoring the inventory of the raw materials, and when the inventory of the raw materials is lower than a preset threshold, automatically generating a purchase order and notifying the purchasing department.
[0056] In summary, the power cord production and processing management method according to the embodiment of the present application is explained, which uses artificial intelligence technology based on deep learning to perform data analysis on the raw material state image and the performance parameters of the raw materials, respectively extracts the visual feature representation of the raw material state and the global semantic feature embedding representation of the raw material performance parameters, and conducts a joint analysis of the two to intelligently determine whether the raw materials are qualified. In this way, the detection efficiency and accuracy can be effectively improved, human errors can be reduced, potential quality problems can be discovered in a timely manner, and comprehensive quality control can be achieved.
[0057] 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. A power cord production and processing management system, characterized in that: include: The production plan generation module is used to generate a production plan based on the imported customer order data and production capacity data; The quality control module is used to determine whether the raw materials are qualified based on the entered inspection data; The inventory management module is used to monitor the inventory of the raw materials, and when the inventory of the raw materials is lower than a preset threshold, automatically generate a purchase order and notify the purchasing department.
2. The power cord production and processing management system according to claim 1, characterized in that: The inspection data includes raw material state images and performance parameters of the raw materials.
3. The power cord production and processing management system according to claim 2, characterized in that: The quality control module comprises: A raw material state feature extraction unit, used for performing image feature extraction on the raw material state image to obtain a raw material state feature map; An autocorrelation feature optimization unit, used for performing autocorrelation feature optimization on the raw material state feature graph to obtain a raw material state optimized feature graph; An embedding coding unit, used for embedding coding the performance parameters of the raw material using a performance parameter embedding matrix to obtain a raw material performance parameter embedding coding vector; The raw material quality analysis unit is used to generate a quality control result based on the raw material state optimization characteristic diagram and the multimodal joint coding feature of the raw material performance parameter embedded coding vector, and the quality control result is used to indicate whether the raw material is qualified.
4. The power cord production and processing management system according to claim 3, characterized in that: The raw material state feature extraction unit is used to: The raw material state image is input into a raw material state feature extractor based on a hole convolutional neural network model to obtain the raw material state feature map.
5. The power cord production and processing management system according to claim 4, characterized in that: The autocorrelation feature optimization unit is used to: The raw material state feature map is input into an adaptive self-attention cross-channel feature space structure optimization module to obtain the raw material state optimized feature map.
6. The power cord production and processing management system according to claim 5, characterized in that: The autocorrelation feature optimization unit is used to: Performing layer normalization on the raw material state characteristic graph to obtain a normalized raw material state characteristic graph; Performing point convolution processing on the normalized raw material state feature map to obtain a raw material state channel context association representation feature map; Convolutionally encoding the raw material state channel context association representation feature map to obtain a raw material state space context association representation feature map; Copying the raw material state space context association representation feature map to obtain a backup raw material state space context association representation feature map; Reshaping the raw material state channel context association representation feature map, the raw material state space context association representation feature map and the backup raw material state space context association representation feature map to obtain a raw material state channel context association representation feature matrix, a raw material state space context association representation feature matrix and a backup raw material state space context association representation feature matrix; Calculating a cross-channel cross covariance matrix between the raw material state channel context association representation feature matrix and the raw material state space context association representation feature matrix; Using a Softmax function to activate the cross-channel cross covariance matrix to obtain a global interactive attention matrix of raw material state features; Calculating the product of the backup raw material state space context association representation feature matrix and the raw material state feature global interaction attention matrix to obtain an attention-enhanced raw material state feature representation matrix; The attention-enhanced raw material state feature representation matrix is feature-reshaped to obtain the raw material state optimized feature map.
7. The power cord production and processing management system according to claim 6, characterized in that: The raw material quality analysis unit comprises: A joint encoding subunit, used for embedding the raw material state optimization feature map and the raw material performance parameter into a coding vector and inputting the encoding vector into a heterogeneous data driven multimodal Transformer joint encoder to obtain a multimodal fusion feature map of the raw material state under the constraints of the performance parameters; The quality control result generating subunit is used to input the multimodal fusion feature map of the raw material state under the performance parameter constraint into the classifier-based quality control result generator to obtain the quality control result.
8. The power cord production and processing management system according to claim 7, characterized in that: The joint encoding subunit is used for: Performing linear transformation and feature shape reshaping on the raw material performance parameter embedding coding vector to obtain a raw material performance parameter semantic feature auxiliary modal modulation matrix; Perform a first linear transformation and a second linear transformation on the raw material state optimization characteristic graph to obtain a first linear transformation raw material state optimization characteristic graph and a second linear transformation raw material state optimization characteristic graph; Reshaping the first linear transformation raw material state optimization characteristic graph and the second linear transformation raw material state optimization characteristic graph to obtain a first linear transformation raw material state optimization characteristic matrix and a second linear transformation raw material state optimization characteristic matrix; Taking the first linear transformation raw material state optimization feature matrix and the second linear transformation raw material state optimization feature matrix as a key matrix and a value matrix, and taking the raw material performance parameter semantic feature auxiliary modal modulation matrix as a query matrix, inputting the query matrix, the key matrix and the value matrix into a Transformer structure to obtain an auxiliary modal attention query result matrix; Reshape and perform point convolution encoding on the auxiliary modality attention query result matrix to obtain an auxiliary modality attention query result feature map; The auxiliary modal attention query result feature map and the raw material state optimization feature map are multiplied by the position points to obtain a multimodal fusion feature map of the raw material state under the constraints of the performance parameters.
9. A power cord production and processing management method, characterized in that: include: Generate production plans based on imported customer order data and production capacity data; Based on the entered inspection data, determine whether the raw materials are qualified; The inventory of the raw materials is monitored, and when the inventory of the raw materials is below a preset threshold, a purchase order is automatically generated and the purchasing department is notified.
Citation Information
Patent Citations
Power transmission cable production quality management feedback system based on industrial internet
CN111445103A
Composite material performance prediction model and method based on microstructure and material component multi-modal fusion
CN115985420A
Cleansing wet tissue manufacturing whole process tracking management system and method
CN118469264A
System arrangement for monitoring and / or checking at least one quality parameter of sample materials containing particles
DE202016004993U1
Cited By
Fault positioning method, device and system for optical fiber line
CN119276354A
