Scheduling management method and system for fabric orders

Abstract

The invention relates to the field of fabric order scheduling, and particularly discloses a fabric order scheduling management method and system, and the method comprises the steps: obtaining basic information, production information and delivery information of different fabric orders, carrying out the feature extraction and correlation analysis of the basic information, the production information and the delivery information of the different fabric orders through a deep learning technology, and obtaining a fabric order scheduling result. And finally, a fabric order scheduling plan is generated through a generator, so that resource utilization is optimized, production efficiency is improved, a multi-batch and small-batch production mode is adapted, and enterprise competitiveness and customer satisfaction are improved.

Description

Technical Field

[0001] The present application relates to the field of fabric order scheduling, and more specifically, to a fabric order scheduling management method and system. Background Art

[0002] Fabric order scheduling refers to the management activities in the textile and clothing production process, which reasonably arranges the time and sequence of fabric production according to various factors to optimize resource utilization, improve production efficiency, and ensure on-time delivery.

[0003] At present, traditional fabric manufacturers are gradually turning to a multi-batch small-batch production model, which significantly increases the requirements for the connectivity of the production process. Since different types of fabrics require different production processes and raw materials, companies often need to stop the machine for inspection and waiting during the switching process, which has a significant impact on overall production efficiency.

[0004] Therefore, a scheduling management method and system for fabric orders 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 a scheduling management method and system for fabric orders, which first obtains basic information of different fabric orders, production information of different fabric orders and delivery information of different fabric orders, and then uses deep learning technology to perform feature extraction and correlation analysis on the three, and finally generates a fabric order scheduling plan through a generator, thereby optimizing resource utilization and improving production efficiency, thereby adapting to the production mode of multiple batches and small batches, and improving the competitiveness and customer satisfaction of the enterprise.

[0006] According to one aspect of the present application, a scheduling management method for fabric orders is provided, which includes: Obtain basic information of different fabric orders, production information of different fabric orders, and delivery information of different fabric orders; Extracting fabric order multimodal information association feature vectors and fabric order delivery information association feature vectors from the different fabric order basic information, the different fabric order production information, and the different fabric order delivery information; A fabric order scheduling plan is generated based on the fabric order multimodal information associated feature vector and the fabric order delivery information associated feature vector.

[0007] According to another aspect of the present application, a scheduling management system for fabric orders is provided, comprising: Fabric order scheduling management data acquisition module, used to obtain basic information of different fabric orders, production information of different fabric orders and delivery information of different fabric orders; A fabric order scheduling management data extraction module, used to extract fabric order multimodal information association feature vectors and fabric order delivery information association feature vectors from the different fabric order basic information, the different fabric order production information and the different fabric order delivery information; The fabric order scheduling plan generating module is used to generate a fabric order scheduling plan based on the fabric order multimodal information associated feature vector and the fabric order delivery information associated feature vector.

[0008] Compared with the prior art, the present application provides a fabric order scheduling management method and system, which first obtains basic information of different fabric orders, production information of different fabric orders and delivery information of different fabric orders, and then uses deep learning technology to perform feature extraction and correlation analysis on the three, and finally generates a fabric order scheduling plan through a generator, thereby optimizing resource utilization and improving production efficiency, thereby adapting to the production mode of multiple batches and small batches, and enhancing the competitiveness of the enterprise and customer satisfaction. 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 Flow chart of a method for scheduling and managing fabric orders according to an embodiment of the present application.

[0011] Figure 2 The present invention is a flowchart of extracting features from basic information of different fabric orders in a scheduling management method for fabric orders according to an embodiment of the present application to obtain a text feature vector of basic information of fabric orders.

[0012] Figure 3 The present invention is a flowchart of extracting features of different fabric order production information to obtain a semantic feature vector of fabric order production information in a fabric order scheduling management method according to an embodiment of the present application.

[0013] Figure 4 The present invention is a flowchart of extracting features of different fabric order delivery information to obtain associated feature vectors of the fabric order delivery information in a fabric order scheduling management method according to an embodiment of the present application.

[0014] Figure 5 4 is a block diagram of a scheduling management system for fabric orders according to an embodiment of the present application. DETAILED DESCRIPTION

[0015] 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.

[0016] 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.

[0017] 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.

[0018] 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.

[0019] Figure 1 FIG. 1 is a flow chart of a method for scheduling and managing fabric orders according to an embodiment of the present application. Figure 1 As shown, the scheduling management method for fabric orders according to the embodiment of the present application includes: S110, obtaining basic information of different fabric orders, production information of different fabric orders and delivery information of different fabric orders; S120, extracting fabric order multimodal information association feature vectors and fabric order delivery information association feature vectors from the different fabric order basic information, the different fabric order production information and the different fabric order delivery information; S130, generating a fabric order scheduling plan based on the fabric order multimodal information association feature vectors and the fabric order delivery information association feature vectors.

[0020] In the above-mentioned fabric order scheduling management method, the step S110 obtains basic information of different fabric orders, production information of different fabric orders, and delivery information of different fabric orders. It should be understood that fabric order scheduling is a key activity in textile and clothing production management, which aims to reasonably arrange the production time and sequence of fabrics according to various factors to achieve the purpose of optimizing resource utilization, improving production efficiency, and ensuring on-time delivery. In recent years, traditional fabric production enterprises are gradually turning to small batch and multi-batch production methods. This change has significantly increased the requirements for production process connection. Since different types of fabrics require different production processes and raw materials, when switching fabric types, enterprises often have to suspend production to carry out necessary inspections and preparations. This downtime has a great impact on the overall production efficiency and reduces the continuity and flexibility of production. Therefore, in the technical solution of the present application, basic information of different fabric orders, production information of different fabric orders, and delivery information of different fabric orders are obtained, and combined with deep learning technology, a fabric order scheduling plan is formulated to optimize the use of resources and improve production efficiency, so as to better adapt to the small batch and multi-batch production mode. This optimization not only enhances the competitiveness of the enterprise, but also improves customer satisfaction.

[0021] Specifically, the basic information of fabric orders includes customer demand, fabric specifications, quantity, and price. The production information of fabric orders involves production progress, production equipment status, process flow, and personnel arrangements. The delivery information of fabric orders includes delivery time, logistics arrangements, transportation methods, and customer receipt confirmation. By systematically collecting and analyzing this information, companies can maintain competitiveness in a rapidly changing market environment, improve operational efficiency, and achieve sustainable development. Effective information management not only helps meet customer needs, but also promotes continuous progress in technology innovation, market expansion, and brand building.

[0022] In the above-mentioned fabric order scheduling management method, the step S120 extracts fabric order multimodal information association feature vectors and fabric order delivery information association feature vectors from the different fabric order basic information, the different fabric order production information, and the different fabric order delivery information. This helps to optimize the production process, improve efficiency, and enhance the company's advantages in customer management and market response capabilities.

[0023] In a specific embodiment of the present application, the step S120 includes: performing feature extraction on the basic information of the different fabric orders to obtain a text feature vector of the basic information of the fabric orders; performing feature extraction on the production information of the different fabric orders to obtain a semantic feature vector of the production information of the fabric orders; associating the text feature vector of the basic information of the fabric orders with the semantic feature vector of the production information of the fabric orders to obtain a feature vector associated with the multimodal information of the fabric orders; performing feature extraction on the delivery information of the different fabric orders to obtain a feature vector associated with the delivery information of the fabric orders.

[0024] It should be understood that the basic information of fabric orders usually includes multiple dimensions such as customer name, fabric type, specifications, quantity, price, delivery time, etc. This information is the basis of enterprise production and operation, but due to the diversity and complexity of information, it is often difficult to extract valuable insights by directly using raw text data for analysis and decision-making. By converting the basic information of fabric orders into text feature vectors, enterprises can use machine learning and data mining techniques for in-depth analysis. These feature vectors can capture the hidden patterns and relationships in the text, enabling enterprises to identify customer preferences, changes in demand, and potential market opportunities.

[0025] Furthermore, the production information of fabric orders usually covers multiple key elements such as production progress, production equipment status, process flow, and personnel arrangements. This information often exists in the form of unstructured text. Direct analysis of these raw data will lead to low efficiency in information extraction and difficulty in obtaining in-depth insights. Therefore, through feature extraction, these unstructured text information can be converted into structured semantic feature vectors. This conversion not only simplifies the complexity of data processing, but also makes subsequent analysis and application more efficient.

[0026] Furthermore, the basic information and production information of fabric orders provide different perspectives on customer demand and production process respectively. The basic information mainly includes customer order requirements, fabric type, quantity and delivery time, while the production information involves the actual production progress, equipment status and process flow. The two are complementary in information content, and associating these two feature vectors can effectively integrate the relationship between customer demand and production capacity. By constructing multimodal information association feature vectors, enterprises can fully understand the life cycle of orders and make more accurate production planning and scheduling decisions.

[0027] In particular, delivery information usually exists in text or unstructured form, and directly processing this information will lead to inefficient analysis. Through feature extraction, enterprises can convert this information into quantifiable feature vectors, which are convenient for in-depth mining using machine learning and data analysis techniques. This conversion can not only improve the accuracy of data processing, but also provide higher quality input data for the prediction model, thereby improving the accuracy of the prediction.

[0028] Figure 2 The flowchart is a method for scheduling and managing fabric orders according to an embodiment of the present application, in which the basic information of different fabric orders is extracted to obtain a text feature vector of the basic information of the fabric order. Figure 2 As shown, in a specific embodiment of the present application, feature extraction is performed on the different basic information of fabric orders to obtain a fabric order basic information text feature vector, including: S210, the different basic information of fabric orders are passed through a fabric order basic information text word embedding model to obtain a sequence of fabric order basic information text word vectors; S220, the sequence of fabric order basic information text word vectors is passed through a fabric order basic information text word semantic encoder to obtain the fabric order basic information text feature vector.

[0029] It should be understood that the word embedding model can map words with similar meanings to similar vector spaces by learning contextual information in text data. This means that key words in fabric orders, such as "cotton", "silk", etc., can reflect the similarity and correlation between them in the word vector space. This feature representation method can help companies understand customer needs and market trends more deeply, so as to make more accurate production decisions and marketing strategies. Specifically, the basic information of the different fabric orders is segmented to obtain a basic information word sequence; the embedding layer of the basic information text word embedding model of the fabric order basic information is used to map each basic information word in the basic information word sequence to a word embedding vector to obtain a sequence of basic information text word vectors of the fabric order.

[0030] Furthermore, the basic information of fabric orders usually contains multiple levels of semantic content, such as customer information, fabric types, order quantity and delivery time. This information exists in the form of word vectors. Although it has been converted into numerical form, it still lacks understanding of context and deep semantics. Through the introduction of word semantic encoders, these word vector sequences can be further processed to extract more representative feature vectors. This conversion not only improves the expressiveness of text information, but also captures more complex relationships between information. Among them, word semantic encoders usually use deep learning models (such as LSTM, Transformer, etc.) to process sequence data, which can consider the contextual relationship between words. By encoding the word vector sequence, enterprises can obtain more accurate and dynamic text feature vectors, so as to better understand customer needs and market dynamics. Specifically, the sequence of the basic information text word vectors of the fabric order is encoded based on global context semantics using the transformer-based Bert model of the basic information text word semantic encoder of the fabric order to obtain multiple basic information feature vectors; and the multiple basic information feature vectors are cascaded to obtain the basic information text feature vector of the fabric order.

[0031] Figure 3 The flowchart is a method for scheduling and managing fabric orders according to an embodiment of the present application, wherein the feature extraction of the fabric order production information is performed to obtain a semantic feature vector of the fabric order production information. Figure 3 As shown, in a specific embodiment of the present application, feature extraction is performed on the different fabric order production information to obtain a semantic feature vector of the fabric order production information, including: S310, segmenting the different fabric order production information to obtain a fabric order production information word sequence; S320, encoding the fabric order production information word sequence through the fabric order production information context to obtain the fabric order production information semantic feature vector.

[0032] It should be understood that the production information of fabric orders usually contains multiple types of data, including production progress, production equipment status, process parameters, and production quality. This information usually exists in text form and may contain a large number of industry terms and professional terms. When conducting data analysis and model building, the structuring and standardization of original text information is particularly important. Through word segmentation, these unstructured text data can be converted into independent words or phrases to form a word sequence. This conversion makes subsequent feature extraction and analysis more efficient and systematic. Specifically, different words carry different semantics in fabric order production information, and word segmentation can express this information clearly. By segmenting production information, enterprises can provide rich feature information for machine learning models, enabling them to capture potential patterns and trends in the production process to a greater extent.

[0033] Furthermore, context encoding usually adopts a neural network structure, such as long short-term memory (LSTM), gated recurrent unit (GRU) or Transformer. These models can effectively process sequence data, taking into account the dependency relationship and context information between words. For example, the Transformer model can dynamically pay attention to the importance of other words when processing each word through the self-attention mechanism. This mechanism ensures that the generated feature vector not only reflects the meaning of a single word, but also comprehensively reflects the semantic information of the entire sequence, thereby improving the accuracy of text analysis. Specifically, the embedding layer of the fabric order production information context encoding is used to map each production information word in the fabric order production information word sequence to a word embedding vector to obtain a sequence of production information word embedding vectors; the Bert model based on the transformer of the fabric order production information context encoding is used to perform global context semantic encoding on the sequence of production information word embedding vectors to obtain multiple production information feature vectors; and the multiple production information feature vectors are cascaded to obtain the fabric order production information semantic feature vector.

[0034] Figure 4 This is a flowchart of extracting features of different fabric order delivery information to obtain the fabric order delivery information-related feature vector in the fabric order scheduling management method according to an embodiment of the present application. Figure 4 As shown, in a specific embodiment of the present application, feature extraction is performed on the different fabric order delivery information to obtain the fabric order delivery information associated feature vector, including: passing the different fabric order delivery information through a fabric order delivery information semantic understanding model to obtain a fabric order delivery information text feature vector; multiplying the fabric order delivery information text feature vector by its transpose to obtain a fabric order delivery information association matrix; passing the fabric order delivery information association matrix through a fabric order delivery information feature encoder based on a convolutional neural network to obtain the fabric order delivery information associated feature vector.

[0035] It should be understood that semantic understanding models usually use deep learning technology, such as Transformer and BERT in natural language processing. These models can analyze contextual information in the text and dynamically understand the meaning of words in a specific context. Through training, the model can identify and learn potential patterns and trends in fabric order delivery information. The generated text feature vector can not only reflect the characteristics of a single piece of information, but also comprehensively consider all relevant information to improve the accuracy and effectiveness of the overall analysis. Specifically, the Bert model of the fabric order delivery information semantic understanding model is used to perform context semantic encoding on each delivery information in the different fabric order delivery information to obtain multiple delivery information text feature vectors; and, the multiple delivery information text feature vectors are cascaded to obtain the fabric order delivery information text feature vector.

[0036] Furthermore, the text feature vectors are generated by a natural language processing model, which can capture the multi-dimensional semantic features of fabric order delivery information. These features not only represent the unique information of each order, but also reflect the potential relationship between them. When these feature vectors are multiplied by their transpose, the similarity or association between each pair of orders is actually calculated. In this way, each element in the generated association matrix represents the strength of the relationship between the corresponding two orders in the feature space. This feature enables enterprises to quickly identify similar orders, thus providing a basis for subsequent analysis and decision-making. Specifically, the rows and columns in the matrix correspond to different orders, and the values ​​of the elements represent the similarities between these orders. This visual representation method can help enterprises identify which orders have similar delivery conditions, customer needs or other characteristics, thereby providing support for tasks such as classification, clustering or predictive analysis. For example, by analyzing similar orders, enterprises can more effectively optimize inventory management, production planning and logistics arrangements, and improve overall operational efficiency.

[0037] Furthermore, the feature encoder based on convolutional neural network (CNN) can abstract the input data layer by layer through multi-layer convolution and pooling operations. Each layer of the network can learn different levels of feature representation, from simple local patterns to complex global structures. This hierarchical feature learning method enables the encoder to deeply understand the correlation between fabric orders and discover potential patterns and regularities. For example, some fabric orders may be clustered together due to similarities in delivery time, transportation methods or customer needs. Through the processing of CNN, these associations will be effectively extracted and converted into high-dimensional feature vectors. Specifically, the association matrix of fabric order delivery information reflects the similarities and relationships between different orders. In this matrix, each element represents the degree of association between two orders in the feature space, constructing a complex relationship network. Convolutional neural network has excellent local feature extraction capabilities and can effectively identify and learn important patterns in data through convolution operations. Inputting the association matrix into CNN can fully explore the complex relationships between orders, thereby extracting more representative features. These features can not only capture local information, but also integrate global context, thereby improving the expressiveness of features. Specifically, each layer of the fabric order delivery information feature encoder based on convolutional neural network is used to perform convolution processing, mean pooling processing based on 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 fabric order delivery information feature encoder based on convolutional neural network outputs the fabric order delivery information associated feature vector, wherein the input of the fabric order delivery information feature encoder based on convolutional neural network is the fabric order delivery information association matrix.

[0038] In a specific embodiment of the present application, the fabric order delivery information text feature vector is multiplied by its transpose to obtain a fabric order delivery information association matrix, including: obtaining a fabric order delivery information text feature vector feature_vectors; performing a transposition operation on the fabric order delivery information text feature vector feature_vectors to obtain a transposed fabric order delivery information text feature vector feature_vectors_transposed; performing a matrix multiplication operation on the fabric order delivery information text feature vector feature_vectors and the transposed fabric order delivery information text feature vector feature_vectors_transposed to generate a fabric order delivery information association matrix correlation_matrix; storing the fabric order delivery information association matrix correlation_matrix in a database, the database being connected via SQLAlchemy, and storing each element of the fabric order delivery information association matrix in a database table correlation_matrix, the table comprising a row index row_index, a column index col_index and an association value value.

[0039] It should be understood that ResultSet is an object used in Java to represent a database query result set. It provides a way to access and operate query results row by row. Through ResultSet, you can traverse each row of the query result and obtain the value of each column, thereby realizing flexible processing and analysis of database query results. Specifically, ResultSet provides an efficient way to process and traverse database query results in Java databases, so that data can be accessed row by row without loading all data into memory at once, thereby saving memory resources. Secondly, ResultSet supports dynamic operations on query results, such as sorting, filtering, and paging, which enhances the flexibility of data processing. In addition, ResultSet also supports update operations on query results, so that data can be modified directly on the result set, simplifying the process of database operations.

[0040] Part of the deployment code is shown below.

[0041] import numpy as np from sqlalchemy import create_engine # Data preparation feature_vectors = np.array([ [0.1, 0.2, 0.3], [0.4, 0.5, 0.6], [0.7, 0.8, 0.9] ]) # Transpose eigenvector feature_vectors_transposed = feature_vectors.T Matrix multiplication correlation_matrix = np.dot(feature_vectors, feature_vectors_transposed) # Result storage engine = create_engine('sqlite: / / / correlation_matrix.db') with engine.connect() as connection: for i, row in enumerate(correlation_matrix): for j, value in enumerate(row): connection.execute( "INSERT INTO correlation_matrix (row_index, col_index, value) VALUES (a,b,c)", (i, j, value) ) print("The association matrix has been successfully stored in the database.") Here, a computer-implemented method for generating a fabric order delivery information correlation matrix is ​​implemented, which includes data preparation, feature vector transposition, matrix multiplication and result storage. In the data preparation process, the feature vector of the fabric order delivery information text is extracted. The feature vector of the fabric order delivery information text is represented as a feature_vectors. In this way, the original text data can be converted into a numerical form that can be processed by the computer, providing basic data for subsequent matrix operations. Next, the feature vector feature_vectors of the fabric order delivery information text is transposed. The transposition operation converts the dimension of the feature vector from a row vector to a column vector, so that the matrix multiplication operation can proceed smoothly. Next, the matrix multiplication operation is performed to generate the fabric order delivery information text feature vector correlation_matrix. It should be understood that the result of the matrix multiplication is a square matrix, each element of which represents the degree of correlation between two different elements in the original feature vector. Finally, the generated fabric order delivery information text feature vector correlation_matrix is ​​stored in the database. Specifically, SQLAlchemy is used to connect to the SQLite database, and each element of the correlation matrix is ​​stored in the database table correlation_matrix. The table contains row index row_index, column index col_index and association value value. In this way, the text feature vector of fabric order delivery information can be effectively converted into an association matrix and stored in the database. This process not only reveals the intrinsic relationship between feature vectors. Specifically, the generated association matrix can help understand the key features and their interrelationships in fabric order delivery information, thereby optimizing supply chain management, improving order processing efficiency, and providing strong support for future data-driven decision-making. In addition, by storing the association matrix in the database, data query and analysis can be easily performed to further explore the potential value in the data.

[0042] In the above-mentioned fabric order scheduling management method, the step S130 generates a fabric order scheduling plan based on the fabric order multimodal information association feature vector and the fabric order delivery information association feature vector. It should be understood that the fabric order multimodal information association feature vector integrates data from different sources, including customer demand, fabric characteristics, production capacity, market trends and other information. These data may show their own characteristics and patterns in different modes. For example, historical data on customer demand can reveal order peaks in certain seasons or holidays, while the characteristics of fabric characteristics may affect the production time and process requirements. By associating these multimodal information, the generated feature vector can fully reflect the complexity and diversity of orders. At the same time, the association feature vector of fabric order delivery information focuses on the relationship between orders, including delivery time, priority, mode of transportation, etc. These feature vectors can effectively capture the similarities and dependencies between different orders. For example, if some orders overlap in delivery time, or their production processes are similar, then these associations need to be considered during scheduling to optimize the utilization efficiency of the production line. Combining these two feature vectors to generate a scheduling plan can achieve more accurate and efficient production scheduling. Specifically, the scheduling system can use the multimodal information association feature vector to identify the order priority and resource requirements, so as to arrange the production tasks reasonably. At the same time, through the delivery information association feature vector, orders with similar delivery time and conditions can be processed centrally, reducing the switching cost and time loss in the production process.

[0043] In a specific embodiment of the present application, a fabric order scheduling plan is generated based on the fabric order multimodal information associated feature vector and the fabric order delivery information associated feature vector, including: fusing the fabric order multimodal information associated feature vector and the fabric order delivery information associated feature vector to obtain an order scheduling generation feature vector; performing information confusion correction on the order scheduling generation feature vector based on the learning parameter's own structure to obtain a sparse optimized order scheduling generation feature vector; and passing the sparse optimized order scheduling generation feature vector through a generator to generate a fabric order scheduling plan.

[0044] It should be understood that by fusing these two types of feature vectors, a more comprehensive order scheduling feature vector can be generated. Such a feature vector not only contains detailed information about the order itself, but also captures the interdependencies between orders and changes in the overall production environment. This comprehensive information enables the scheduling system to perform more effective analysis in complex decision-making environments. For example, when a new order is received, the system can quickly evaluate its impact on existing orders and make corresponding scheduling adjustments to ensure the continuity and efficiency of the production process.

[0045] In particular, in the technical solution of the present application, the order scheduling generates a feature vector by fusing multiple information (including basic information, production information, and delivery information), which contains rich semantic information and contextual features. However, if the dimension, shape, or data distribution of the learning parameter matrix is ​​inconsistent with the feature vector, structural redundancy will occur. For example, the learning parameter matrix may contain too many redundant parameters, which makes it unable to effectively capture the core information of the feature vector in model training, but increases the computational complexity and unnecessary complexity of the model. Further, if the learning parameter matrix fails to clearly distinguish the relationship between different features, it may mix irrelevant or minor features with the main features, thereby reducing the generalization ability of the model. This information confusion makes it difficult for the model to form an accurate feature mapping during the learning process, resulting in unnecessary complexity and noise in the final scheduling plan. Redundant features cause the model to repeatedly focus on similar information during learning, thereby reducing learning efficiency and effectiveness. At the same time, noise may guide the model to learn the wrong pattern, causing the model's prediction performance to decline when facing new data. These factors combined often lead to overfitting of the model, ultimately affecting the training and effect of the model. Therefore, in the technical solution of the present application, the order scheduling generation feature vector is corrected by information confusion based on the learning parameter's own structure to obtain a sparse optimized order scheduling generation feature vector.

[0046] Among them, the order scheduling generation feature vector is corrected by information confusion based on the learning parameter's own structure to obtain a sparse optimized order scheduling generation feature vector, including: extracting each row vector of the learning parameter matrix in the generator as a set of order scheduling learning parameter row vectors; calculating the order scheduling entropy degradation maximization simulation value between the order scheduling generation feature vector and each order scheduling learning parameter row vector in the set of order scheduling learning parameter row vectors to obtain a set of order scheduling entropy degradation maximization simulation values; based on the comparison between each order scheduling entropy degradation maximization simulation value in the set of order scheduling entropy degradation maximization simulation values ​​and a preset threshold, the learning parameter matrix is ​​gated and pruned to obtain an order scheduling learning parameter entropy degradation optimization matrix; the order scheduling generation feature vector is matrix-multiplied with the order scheduling learning parameter entropy degradation optimization matrix to obtain the sparse optimized order scheduling generation feature vector.

[0047] The optimization steps are specifically expressed as follows: ;in, is the set of row vectors of learning parameters for order scheduling, is a set operation, , , , are the first, second, and third order scheduling learning parameter row vectors respectively. , The eigenvalues ​​of the row vector of order scheduling learning parameters, The first value in the set of simulation values ​​that maximizes the order scheduling entropy degradation The order scheduling entropy degradation maximization simulation value, is the maximum value function, Represents the calculation vector and The Euclidean distance between Generate feature vectors for order scheduling, For the The transposed vector of the row vector of order scheduling learning parameters, is the length of the feature vector, is matrix multiplication, , , , The first, second, and , learning parameter mask values, For mask operation, Indicates the preset threshold value, The optimization matrix of entropy degradation for order scheduling learning parameters, Generate feature vectors for sparse optimal order scheduling.

[0048] In the technical solution of the present application, it is taken into consideration that in the process of inputting the order scheduling generated feature vector into the learning parameter matrix of the generator to obtain the fabric order scheduling plan, there is a mismatch between the structure of the learning parameter matrix itself and the order scheduling generated feature vector, that is, the learning parameter matrix has structural redundancy and information confusion relative to the order scheduling generated feature vector, which may cause feature redundancy or noise in the fabric order scheduling plan, resulting in model overfitting.

[0049] Based on this, in the technical solution of the present application, the order scheduling feature vector is subjected to information confusion correction based on the structure of the learning parameter itself, and the order scheduling feature vector is used as a wandering node, and the feature distribution between the order scheduling feature vector and each row vector in the learning parameter matrix is ​​used to simulate the wandering pattern, and the distance topology between the order scheduling feature vector and each row vector in the learning parameter matrix is ​​used as a wandering map. In this way, the order scheduling entropy degradation maximization simulation value is obtained by the wandering result of the wandering node in the specific wandering pattern on the wandering map. Furthermore, based on the comparison between each order scheduling entropy degradation maximization simulation value in the set of the order scheduling entropy degradation maximization simulation value and the preset threshold, the part of the learning parameter matrix that does not contribute significantly to the input data representation and may even introduce noise or redundant information is screened out. Specifically, based on the comparison result, the learning parameter matrix is ​​gated and pruned to obtain the order scheduling learning parameter entropy degradation optimization matrix. Finally, the structure-optimized learning parameter matrix is ​​used as a feature modulation unit to linearly modulate the order scheduling generation feature vector to obtain the sparse optimized order scheduling generation feature vector.

[0050] Furthermore, during the scheduling process, multiple factors may affect the decision, such as fabric type, delivery time, production capacity, and machine scheduling. By inputting the sparse optimized order scheduling feature vector into the generator, the focus can be placed on the most critical features, making the generated scheduling plan more accurate, thereby improving resource utilization. Specifically, the generator can generate the corresponding scheduling plan in real time based on the input sparse optimized order scheduling feature vector. This dynamic generation capability enables the system to respond quickly to new orders and market changes without having to recalculate the entire scheduling model. Through the mechanism of the generator, scheduling requirements in different scenarios can be effectively handled, and real-time optimization can be achieved, thereby improving overall production efficiency.

[0051] In summary, the embodiment of the present application first obtains the basic information of different fabric orders, the production information of different fabric orders, and the delivery information of different fabric orders, and then uses deep learning technology to perform feature extraction and correlation analysis on the three, and finally generates a fabric order scheduling plan through a generator, thereby optimizing resource utilization and improving production efficiency, thereby adapting to the production mode of multiple batches and small batches, and enhancing the competitiveness of the enterprise and customer satisfaction.

[0052] Figure 5 FIG. 1 is a block diagram of a scheduling management system for fabric orders according to an embodiment of the present application. Figure 5As shown, according to the fabric order scheduling management system 100 of the embodiment of the present application, it includes: a fabric order scheduling management data acquisition module 110, which is used to obtain different fabric order basic information, different fabric order production information and different fabric order delivery information; a fabric order scheduling management data extraction module 120, which is used to extract fabric order multimodal information association feature vectors and fabric order delivery information association feature vectors from the different fabric order basic information, the different fabric order production information and the different fabric order delivery information; a fabric order scheduling plan generation module 130, which is used to generate a fabric order scheduling plan based on the fabric order multimodal information association feature vector and the fabric order delivery information association feature vector.

[0053] Here, those skilled in the art will appreciate that the specific operations of each step in the above-mentioned fabric order scheduling management system have been described in the above reference. Figures 1 to 4 The scheduling management method of fabric orders has been described in detail, and therefore, its repeated description will be omitted.

[0054] As described above, the fabric order scheduling management system 100 according to the embodiment of the present application can be implemented in various terminal devices. In one example, the fabric order scheduling management system 100 can be integrated into the terminal device as a software module and / or a hardware module. For example, the fabric order scheduling management system 100 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 fabric order scheduling management system 100 can also be one of the many hardware modules of the terminal device.

[0055] Alternatively, in another example, the fabric order scheduling management system 100 and the terminal device may also be separate devices, and the fabric order scheduling management system 100 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.

[0056] 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.

[0057] 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.

[0058] 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.

[0059] 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.

[0060] 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.

[0061] 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. A scheduling management method for fabric orders, characterized in that: include: Obtain basic information of different fabric orders, production information of different fabric orders, and delivery information of different fabric orders; Extracting fabric order multimodal information association feature vectors and fabric order delivery information association feature vectors from the different fabric order basic information, the different fabric order production information, and the different fabric order delivery information; A fabric order scheduling plan is generated based on the fabric order multimodal information associated feature vector and the fabric order delivery information associated feature vector.

2. The fabric order scheduling management method according to claim 1, characterized in that: Extracting fabric order multimodal information association feature vectors and fabric order delivery information association feature vectors from the different fabric order basic information, the different fabric order production information, and the different fabric order delivery information, including: Performing feature extraction on the basic information of the different fabric orders to obtain a text feature vector of the basic information of the fabric orders; Extracting features of the different fabric order production information to obtain a semantic feature vector of the fabric order production information; Associating the basic information text feature vector of the fabric order with the semantic feature vector of the fabric order production information to obtain the multimodal information association feature vector of the fabric order; Feature extraction is performed on the different fabric order delivery information to obtain a feature vector associated with the fabric order delivery information.

3. The fabric order scheduling management method according to claim 2 is characterized in that: Feature extraction is performed on the basic information of the different fabric orders to obtain a text feature vector of the basic information of the fabric order, including: The basic information of different fabric orders is embedded into a fabric order basic information text word embedding model to obtain a sequence of fabric order basic information text word vectors; The sequence of the word vectors of the basic information text of the fabric order is passed through a basic information text word semantic encoder of the fabric order to obtain the basic information text feature vector of the fabric order.

4. The fabric order scheduling management method according to claim 3 is characterized in that: Feature extraction is performed on the different fabric order production information to obtain a semantic feature vector of the fabric order production information, including: Segmenting the different fabric order production information to obtain a fabric order production information word sequence; The fabric order production information word sequence is encoded through the fabric order production information context to obtain the fabric order production information semantic feature vector.

5. The fabric order scheduling management method according to claim 4 is characterized in that: Extracting features of the different fabric order delivery information to obtain the fabric order delivery information associated feature vectors includes: The different fabric order delivery information is passed through a fabric order delivery information semantic understanding model to obtain a fabric order delivery information text feature vector; Multiplying the fabric order delivery information text feature vector by its transpose to obtain a fabric order delivery information association matrix; The fabric order delivery information association matrix is ​​passed through a fabric order delivery information feature encoder based on a convolutional neural network to obtain the fabric order delivery information association feature vector.

6. The fabric order scheduling management method according to claim 5 is characterized in that: The fabric order delivery information text feature vector is multiplied by its transpose to obtain a fabric order delivery information association matrix, including: Get the fabric order delivery information text feature vector feature_vectors; Performing a transposition operation on the fabric order delivery information text feature vector feature_vectors to obtain a transposed fabric order delivery information text feature vector feature_vectors_transposed; Performing a matrix multiplication operation on the fabric order delivery information text feature vector feature_vectors and the transposed fabric order delivery information text feature vector feature_vectors_transposed to generate a fabric order delivery information correlation matrix correlation_matrix; The fabric order delivery information correlation matrix correlation_matrix is ​​stored in a database, the database is connected through SQLAlchemy, and each element of the fabric order delivery information correlation matrix is ​​stored in a database table correlation_matrix, the table includes a row index row_index, a column index col_index and a correlation value value.

7. The fabric order scheduling management method according to claim 6, characterized in that: Generating a fabric order scheduling plan based on the fabric order multimodal information association feature vector and the fabric order delivery information association feature vector includes: Fusion of the fabric order multimodal information associated feature vector and the fabric order delivery information associated feature vector to obtain an order scheduling generation feature vector; Performing information confusion correction on the order scheduling generation feature vector based on the learning parameter's own structure to obtain a sparse optimized order scheduling generation feature vector; The sparse optimization order scheduling generates a feature vector through a generator to generate a fabric order scheduling plan.

8. The fabric order scheduling management method according to claim 7, characterized in that: The order scheduling feature vector is corrected by information confusion based on the learning parameter structure to obtain a sparse optimized order scheduling feature vector, including: Extract each row vector of the learning parameter matrix in the generator as a set of order scheduling learning parameter row vectors; Calculating the order scheduling entropy degradation maximization simulation value between the order scheduling generated feature vector and each order scheduling learning parameter row vector in the set of order scheduling learning parameter row vectors to obtain a set of order scheduling entropy degradation maximization simulation values; Based on the comparison between each order scheduling entropy degradation maximization simulation value in the set of order scheduling entropy degradation maximization simulation values ​​and a preset threshold, the learning parameter matrix is ​​gated and pruned to obtain an order scheduling learning parameter entropy degradation optimization matrix; The order scheduling generated feature vector is matrix multiplied with the order scheduling learning parameter entropy degradation optimization matrix to obtain the sparse optimized order scheduling generated feature vector.

9. A scheduling management system for fabric orders, characterized in that: include: Fabric order scheduling management data acquisition module, used to obtain basic information of different fabric orders, production information of different fabric orders and delivery information of different fabric orders; A fabric order scheduling management data extraction module, used to extract fabric order multimodal information association feature vectors and fabric order delivery information association feature vectors from the different fabric order basic information, the different fabric order production information and the different fabric order delivery information; The fabric order scheduling plan generating module is used to generate a fabric order scheduling plan based on the fabric order multimodal information associated feature vector and the fabric order delivery information associated feature vector.

10. The fabric order scheduling management system according to claim 9, characterized in that: The fabric order scheduling management data extraction module is used to extract fabric order multimodal information association feature vectors and fabric order delivery information association feature vectors from the different fabric order basic information, the different fabric order production information and the different fabric order delivery information, including: Extracting fabric order multimodal information association feature vectors and fabric order delivery information association feature vectors from the different fabric order basic information, the different fabric order production information, and the different fabric order delivery information, including: Performing feature extraction on the basic information of the different fabric orders to obtain a text feature vector of the basic information of the fabric orders; Extracting features of the different fabric order production information to obtain a semantic feature vector of the fabric order production information; Associating the basic information text feature vector of the fabric order with the semantic feature vector of the fabric order production information to obtain the multimodal information association feature vector of the fabric order; Feature extraction is performed on the different fabric order delivery information to obtain a feature vector associated with the fabric order delivery information.

Images