Method and system for automatically generating production plan for heat treatment furnace

Abstract

The present application relates to the field of data processing technology, and provides a method and system for automatically generating a production plan for a heat treatment furnace. The method includes obtaining the product texture corresponding to each production order for each production order, and obtaining the target heating process parameter table corresponding to the production order in a preset database based on the product texture; generating a heat treatment furnace production plan based on each production order and the target heating process parameter table corresponding to each production order. On the one hand, this method reduces manual intervention in the generation process of the heat treatment furnace production plan and improves the automation level of the heat treatment furnace production process. On the other hand, it can improve the generation efficiency and accuracy of the heat treatment furnace production plan, thereby improving the production efficiency of the heat treatment furnace.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method and system for automatically generating a production plan for a heat treatment furnace. Background Art

[0002] With the continuous development of the manufacturing industry, the importance of heat treatment furnaces in industrial production has become increasingly prominent. For heat treatment companies, the number of heat treatment orders and the types of heat treatment products are increasing day by day, which also increases the difficulty for heat treatment companies to formulate production plans. Traditional production plans are often formulated by workers analyzing production orders. However, this manual processing method has the problems of low efficiency and low accuracy. Summary of the invention

[0003] The present application provides a method and system for automatically generating a production plan for a heat treatment furnace to solve the problems raised by the above-mentioned background technology.

[0004] In a first aspect, the present application provides a method for automatically generating a production plan for a heat treatment furnace, comprising:

[0005] In response to a generation instruction of a production plan of a heat treatment furnace, a production order data set of the heat treatment furnace is obtained from a preset production order database; wherein the production order data set includes a plurality of production orders;

[0006] For each of the production orders, the product texture of the product corresponding to the production order is obtained, and based on the product texture, a target heating process parameter table corresponding to the production order is obtained in a preset database; wherein the target heating process parameter table includes heating process parameter information matching the product texture, coding information of a heat treatment furnace matching the product texture, and a blank table, wherein the blank table includes a product name column, a product quantity column, a product size column, and an order delivery date column;

[0007] For each of the production orders, obtain the product name, product quantity, product size and order delivery date of the product corresponding to the production order, and add the product name, product quantity, product size and order delivery date to the corresponding cells of the same row of the blank table of the target heating process parameter table corresponding to the production order, to obtain multiple production plan analysis tables;

[0008] A heat treatment furnace production plan is generated based on all of the production plan analysis tables.

[0009] In a second aspect, the present application provides a system for automatically generating a production plan for a heat treatment furnace, comprising:

[0010] A first acquisition module: used to respond to the generation instruction of the production plan of the heat treatment furnace, and acquire the production order data set of the heat treatment furnace in a preset production order database; wherein the production order data set includes multiple production orders;

[0011] A second acquisition module is used to acquire, for each of the production orders, the product texture of the product corresponding to the production order, and acquire, in a preset database based on the product texture, a target heating process parameter table corresponding to the production order; wherein the target heating process parameter table includes heating process parameter information matching the product texture, coding information of a heat treatment furnace matching the product texture, and a blank table, wherein the blank table includes a product name column, a product quantity column, a product size column, and an order delivery date column;

[0012] A third acquisition module is used to acquire, for each of the production orders, a product name, a product quantity, a product size, and an order delivery date of a product corresponding to the production order, and respectively add the product name, the product quantity, the product size, and the order delivery date to corresponding cells in the same row of a blank table of a target heating process parameter table corresponding to the production order, to obtain a plurality of production plan analysis tables;

[0013] A generation module generates a heat treatment furnace production plan based on all the production plan analysis tables.

[0014] The present application provides a method and system for automatically generating a production plan for a heat treatment furnace, wherein the method includes responding to a generation instruction of a production plan for a heat treatment furnace, obtaining a production order data set of the heat treatment furnace in a preset production order database; wherein the production order data set includes multiple production orders; for each of the production orders, obtaining a product texture of a product corresponding to the production order, and obtaining a target heating process parameter table corresponding to the production order in a preset database based on the product texture; wherein the target heating process parameter table includes heating process parameter information matching the product texture, coding information of a heat treatment furnace matching the product texture, and a blank table, wherein the blank table includes a product name column, a product quantity column, a product size column, and an order delivery date column; for each of the production orders, obtaining a product name, a product quantity, a product size, and an order delivery date of a product corresponding to the production order, and respectively adding the product name, the product quantity, the product size, and the order delivery date to corresponding cells of the same row of the blank table of the target heating process parameter table corresponding to the production order, to obtain multiple production plan analysis tables; and generating a production plan for a heat treatment furnace based on all of the production plan analysis tables. This method, on the one hand, reduces manual intervention in the generation process of the heat treatment furnace production plan and improves the automation level of the heat treatment furnace production process. On the other hand, it can improve the generation efficiency and accuracy of the heat treatment furnace production plan, thereby improving the production efficiency of the heat treatment furnace. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying any creative work.

[0016] Figure 1 A schematic flow chart of a method for automatically generating a production plan for a heat treatment furnace provided in an embodiment of the present application;

[0017] Figure 2 A schematic block diagram of the structure of a system for automatically generating a production plan for a heat treatment furnace provided in an embodiment of the present application. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0020] It should also be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0021] It should be further understood that the term “and / or” used in the specification and appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0022] With the continuous development of the manufacturing industry, the importance of heat treatment furnaces in industrial production has become increasingly prominent. For heat treatment companies, the number of heat treatment orders and the types of heat treated products are increasing day by day, which also increases the difficulty of heat treatment companies in formulating production plans. Traditional production plans are often formulated by workers analyzing production orders. However, this manual processing method has the problems of low efficiency and low accuracy. To this end, the present application provides a method and system for automatically generating a production plan for a heat treatment furnace to solve the above problems.

[0023] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0024] See also Figure 1 , Figure 1 A flow chart of a method for automatically generating a production plan for a heat treatment furnace provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, the method for automatically generating a production plan for a heat treatment furnace provided in an embodiment of the present application includes steps S100 to S500.

[0025] Step S100, in response to an instruction to generate a production plan for a heat treatment furnace, a production order data set for the heat treatment furnace is obtained from a preset production order database; wherein the production order data set includes a plurality of production orders.

[0026] Step S200, for each of the production orders, obtain the product texture of the product corresponding to the production order, and obtain the target heating process parameter table corresponding to the production order in a preset database based on the product texture; wherein the target heating process parameter table includes heating process parameter information matching the product texture, coding information of the heat treatment furnace matching the product texture, and a blank table, wherein the blank table includes a product name column, a product quantity column, a product size column, and an order delivery date column.

[0027] Step S300: For each of the production orders, obtain the product name, product quantity, product size and order delivery date of the product corresponding to the production order, and add the product name, product quantity, product size and order delivery date to the corresponding cells in the same row of the blank table of the target heating process parameter table corresponding to the production order, so as to obtain multiple production plan analysis tables.

[0028] Step S400: Generate a heat treatment furnace production plan based on all the production plan analysis tables.

[0029] It should be noted that the executing entity of the embodiment of the present application may be a system for automatically generating a production plan for a heat treatment furnace, or may be a terminal or a server. The following description will be made using a server as an example.

[0030] In this embodiment, as described in the above step S100, the server responds to the generation instruction of the heat treatment furnace production plan, and obtains the production order data set of the heat treatment furnace in the preset production order database. Specifically, the staff inputs the generation time point of the generation instruction through the user operation interface, and when the time point is reached, the server responds to the generation instruction of the heat treatment furnace production plan, and obtains the production order data set of the heat treatment furnace in the preset production order database.

[0031] As described in step S200 above, after obtaining the production order data set, for each production order, the server first obtains the product texture of the product corresponding to the production order, and then obtains the target heating process parameter table corresponding to the production order in the preset database based on the product texture; wherein the target heating process parameter table includes heating process parameter information matching the product texture, coding information of the heat treatment furnace matching the product texture, and a blank table, the blank table includes a product name column, a product quantity column, a product size column, and an order delivery date column, the heating process parameter information includes a heating curve and a cooling time, and the product texture includes the material and mechanical properties of the product.

[0032] As described in the above step S300, after obtaining the target heating process parameter table corresponding to each of the production orders, the server obtains the product name, product quantity, product size and order delivery date of the product corresponding to the production order, and adds the product name, product quantity, product size and order delivery date to the corresponding cells in the same row of the blank table of the target heating process parameter table corresponding to the production order, respectively, to obtain multiple production plan analysis tables.

[0033] As described in the above step S400, after obtaining multiple production plan analysis tables, a heat treatment furnace production plan is generated based on all the production plan analysis tables. Specifically, first, for each of the production plan analysis tables, based on the order of order delivery dates corresponding to each row of information in the production plan analysis table, each row of information in the production plan analysis table is arranged in sequence to obtain an intermediate production plan table corresponding to the production plan analysis table. Then, for each of the intermediate production plan tables, based on the heating process parameter information corresponding to the intermediate production plan table, the coding information of the heat treatment furnace, and the product quantity, product size and order delivery date corresponding to each row of information in the intermediate production plan table, a production plan for the product corresponding to each row of information in the intermediate production plan table is generated to obtain a target production plan table corresponding to the intermediate production plan table. Finally, a heat treatment furnace production plan is generated based on all the target production plan tables.

[0034] The method provided in this embodiment, on the one hand, reduces the manual intervention in the generation process of the heat treatment furnace production plan and improves the automation level of the heat treatment furnace production process; on the other hand, it can improve the generation efficiency and accuracy of the heat treatment furnace production plan, thereby improving the production efficiency of the heat treatment furnace.

[0035] In some embodiments, the obtaining of the product texture of the product corresponding to the production order comprises the following steps:

[0036] Obtaining the product name of the product corresponding to the production order through the production order;

[0037] The product name is input into a preset product texture classification model to obtain the product texture of the product corresponding to the order.

[0038] Specifically, the server first obtains the product name of the product corresponding to the production order through the production order, and then inputs the product name into a preset product texture classification model to obtain the product texture of the product corresponding to the order.

[0039] The method provided in this embodiment can quickly obtain the product texture of the product, which helps to improve the efficiency of generating a heat treatment furnace production plan.

[0040] In some embodiments, the method for training the product texture classification model comprises the following steps:

[0041] Acquire a training data set and an initial model, and send the training data set and the initial model to a plurality of terminal devices, so that each of the terminal devices respectively trains the initial model based on the training data set to obtain a plurality of intermediate product texture classification models; wherein the training data set includes a plurality of mapping relationships, the mapping relationships being mapping relationships between product names and product textures corresponding to the product names, and each of the terminal devices is different from the other;

[0042] Receive the intermediate product texture classification model sent by each of the terminal devices, fuse the model parameters of each of the intermediate product texture classification models to obtain fused model parameters, and use the fused model parameters to replace the model parameters of the initial model to obtain the product texture classification model.

[0043] Specifically, the server first obtains a training data set and an initial model, and sends the training data set and the initial model to multiple terminal devices, so that each terminal device trains the initial model based on the training data set to obtain multiple intermediate product texture classification models, and then receives the intermediate product texture classification models sent by each terminal device, and fuses the model parameters of each intermediate product texture classification model to obtain a fused model parameter, and uses the fused model parameter to replace the model parameter of the initial model to obtain the product texture classification model. The method for fusing the model parameters of each intermediate product texture classification model is the parameter averaging method.

[0044] The method provided in this embodiment obtains the product texture classification model by fusing the model parameters of each of the intermediate product texture classification models to obtain the fused model parameters, and uses the fused model parameters to replace the model parameters of the initial model, thereby improving the training effect of the product texture classification model and further improving the prediction accuracy of the product texture classification model.

[0045] In some embodiments, the initial model includes a feature extraction layer, a feature vector generation layer, a feature vector linear transformation layer and a classification layer, and the terminal device trains the initial model based on the training data set, including the following steps:

[0046] Extracting feature information of products corresponding to product names corresponding to the mapping relationships respectively through the feature extraction layer;

[0047] Generate feature vectors corresponding to each of the feature information respectively through the feature vector generation layer;

[0048] Mapping each of the feature vectors to a high-dimensional feature space through the feature vector linear transformation layer to obtain multiple linear transformation results;

[0049] Performing nonlinear transformation on each of the linear transformation results respectively through the activation function of the classification layer to obtain multiple nonlinear transformation results;

[0050] Performing deep learning on each of the nonlinear transformation results through the global average pooling layer of the classification layer, respectively, to obtain the predicted product texture of the product corresponding to the product name corresponding to each of the mapping relationships;

[0051] Constructing a predicted product texture-actual product texture mapping relationship table based on the predicted product texture and the actual product texture of the product corresponding to the product name corresponding to each of the mapping relationships;

[0052] Calculating the FocalLoss loss value of the initial model based on the predicted product texture-actual product texture mapping relationship table;

[0053] The model parameters of the initial model are optimized based on the Focal Loss value to obtain an intermediate product texture classification model.

[0054] Specifically, first, for each of the mapping relationships, the product name corresponding to the mapping relationship is input into the initial model, the feature extraction layer extracts the feature information of the product corresponding to the product name corresponding to the mapping relationship, the feature vector generation layer generates the feature vector corresponding to the mapping relationship based on a preset encoding method, each neuron of the feature vector transformation layer is connected to each element of the feature vector, the feature vector is mapped to a high-dimensional space, and the linear transformation result corresponding to the mapping relationship is obtained, the activation function of the classification layer performs nonlinear transformation on the linear transformation result to obtain the nonlinear transformation result corresponding to the mapping relationship, the global average pooling layer of the classification layer performs deep learning on the nonlinear transformation result to obtain the predicted product texture corresponding to the mapping relationship, and then a predicted product texture-actual product texture mapping relationship table is constructed based on the predicted product texture and actual product texture of the product corresponding to each of the product names corresponding to the mapping relationship, and the Focal Loss loss value of the initial model is calculated based on the predicted product texture-actual product texture mapping relationship table, and the model parameters of the initial model are optimized based on the Focal Loss loss value to obtain an intermediate product texture classification model. Wherein, the activation function is a softmax activation function.

[0055] The method provided in this embodiment maps each of the feature vectors to a high-dimensional feature space through the feature vector linear transformation layer to obtain multiple linear transformation results, and performs nonlinear transformation on each of the linear transformation results through the activation function of the classification layer to obtain multiple nonlinear transformation results, which helps to improve the expressive power of the product texture classification model, and further improve the prediction accuracy of the product texture classification model.

[0056] In some embodiments, the calculating of the Focal Loss value of the initial model based on the predicted product texture-actual product texture mapping relationship table comprises the following steps:

[0057] respectively calculating the Euclidean distances corresponding to the respective predicted product texture-actual product texture mapping relationships in the predicted product texture-actual product texture mapping relationship table;

[0058] Calculating the initial Focal Loss loss value corresponding to each of the predicted product texture-actual product texture mapping relationships based on a preset Focal Loss loss function and the Euclidean distance corresponding to each of the predicted product texture-actual product texture mapping relationships;

[0059] The sum of the initial Focal Loss loss values ​​is taken as the Focal Loss loss value.

[0060] Specifically, first according to the Euclidean distance = sqrt((x1- x2) 2 + (y1 - y2) 2) This formula calculates the Euclidean distance corresponding to each predicted product texture-actual product texture mapping relationship respectively, wherein, for any of the predicted product texture-actual product texture mapping relationships, (x1, y1) represents the coordinates of the predicted product texture corresponding to the predicted product texture-actual product texture mapping relationship in the feature space, and (x2, y2) represents the coordinates of the actual product texture corresponding to the predicted product texture-actual product texture mapping relationship in the feature space.

[0061] Then, the initial Focal Loss loss value corresponding to each of the predicted product texture-actual product texture mapping relationships is calculated based on the preset Focal Loss loss function and the Euclidean distance corresponding to each of the predicted product texture-actual product texture mapping relationships. Wherein, the Focal Loss loss function is FL(p_t) = -\alpha_t(1-p_t)^\gamma \log(p_t), where, for any of the predicted product texture-actual product texture mapping relationships, (p_t) is the Euclidean distance corresponding to the predicted product texture-actual product texture mapping relationship, FL(p_t) is the initial Focal Loss loss value corresponding to the predicted product texture-actual product texture mapping relationship, \alpha_t and \gamma are both preset parameters, and the \alpha_t parameter is usually used to balance the class imbalance problem, so that the loss function has a greater penalty for samples of the minority class. Usually, the \alpha_t value can be set according to the proportion of samples of different categories in the data set. If the sample ratio of a certain category is small, the corresponding alpha value can be set larger to increase the weight of the samples in this category. For categories with a large sample ratio, the alpha value can be set smaller. The gamma parameter is usually used to adjust the distinction between easy-to-classify samples and difficult-to-classify samples to enhance the model's attention to difficult-to-classify samples. Generally speaking, the larger the gamma value, the more attention the model pays to difficult-to-classify samples. In practical applications, the gamma value can be set according to the difficulty of the task and the distribution of samples in the data set. Usually, the gamma value can be selected between 0 and 5.

[0062] Finally, the sum of the initial Focal Loss values ​​is calculated.

[0063] The method provided in this embodiment combines the Euclidean distance and the Focal Loss loss function to provide comprehensive evaluation and guidance for the optimization of the initial model.

[0064] In some embodiments, the optimizing the model parameters of the initial model based on the Focal Loss value to obtain the intermediate product texture classification model includes:

[0065] Back-propagating the Focal Loss value in the initial model to obtain the gradients of each model parameter of the initial model;

[0066] Based on each of the gradients, each model parameter of the initial model is optimized by a preset gradient descent algorithm to obtain an intermediate product texture classification model.

[0067] Specifically, first, the Focal Loss loss value is back-propagated in the initial model starting from the output layer of the initial model, and the gradients of the various parameters of the initial model are calculated layer by layer from the output layer to the input layer according to the chain rule. Then, based on the various gradients, the various model parameters of the initial model are continuously optimized through a preset gradient descent algorithm until the Focal Loss loss value of the initial model is less than the preset Focal Loss loss value, thereby obtaining an intermediate product texture classification model.

[0068] The method provided in this embodiment can effectively optimize the model parameters of the initial model, which helps to improve the prediction accuracy of the product texture classification model, thereby improving the accuracy of the heat treatment furnace production plan.

[0069] In some embodiments, the step of obtaining a target heating process parameter table corresponding to the production order in a preset database based on the product texture includes the following steps:

[0070] Encoding the product texture based on a preset encoding rule to obtain a product texture code;

[0071] For each coding character of the product texture code, the coding character is used as a first coding character, and the remaining coding characters except the first coding character in the product texture code are used as second coding characters, and the similarities between the first coding character and each of the second coding characters are respectively obtained, and based on the order of the second coding characters corresponding to each of the similarities in the product texture code, each of the similarities is arranged in sequence to obtain a similarity sequence corresponding to the coding characters;

[0072] Based on the order of the first coding characters corresponding to each of the similarity sequences in the product texture code, the similarity sequences are arranged in order from top to bottom to obtain a target matrix;

[0073] The matrix elements on the diagonal line formed by the upper left corner and the lower right corner of the target matrix are used as target vector elements, and based on the order of each target vector element in the target matrix, each target vector element is arranged in sequence to obtain an initial vector;

[0074] Multiplying the target matrix by the initial vector to obtain a target vector;

[0075] Deleting non-numeric symbols in the target vector to obtain an identification sequence;

[0076] Based on the identification sequence, a target heating process parameter table corresponding to the production order is obtained in a preset database; wherein a plurality of heating process parameter tables are provided in the preset database, and each of the heating process parameter tables is provided with a corresponding identification sequence.

[0077] The method provided in this embodiment generates an identification sequence through the product texture, and obtains the target heating process parameter table corresponding to the production order in the preset database based on the identification sequence, so that the target heating process parameter table corresponding to the production order can be accurately obtained in the preset database.

[0078] In some embodiments, generating a heat treatment furnace production plan based on all of the production plan analysis tables includes:

[0079] For each of the production plan analysis tables, based on the order of order delivery dates corresponding to each row of information in the production plan analysis table, each row of information in the production plan analysis table is arranged in sequence to obtain an intermediate production plan table corresponding to the production plan analysis table;

[0080] For each of the intermediate production schedules, based on the heating process parameter information corresponding to the intermediate production schedule, the coding information of the heat treatment furnace, the product quantity, product size and order delivery date corresponding to each row of information in the intermediate production schedule, a production plan for the product corresponding to each row of information in the intermediate production schedule is generated to obtain a target production schedule corresponding to the intermediate production schedule;

[0081] A heat treatment furnace production plan is generated based on all of the target production schedules.

[0082] Specifically, first, for each of the production plan analysis tables, based on the order of order delivery dates corresponding to each row of information in the production plan analysis table, each row of information in the production plan analysis table is arranged in sequence to obtain an intermediate production plan table corresponding to the production plan analysis table.

[0083] Then, for each of the intermediate production schedules, based on the heating process parameter information corresponding to the intermediate production schedule, the coding information of the heat treatment furnace, and the product quantity, product size and order delivery date corresponding to each row of information in the intermediate production schedule, a production plan for the product corresponding to each row of information in the intermediate production schedule is generated to obtain a target production schedule corresponding to the intermediate production schedule. Specifically, first, the size and quantity of the heat treatment furnace are obtained based on the coding information of the heat treatment furnace. Then, based on the product quantity, product size, order delivery date, the size of the heat treatment furnace and the number of the heat treatment furnaces corresponding to each row of information in the intermediate production schedule, the heat treatment batches of all products in the intermediate production schedule and the processing time corresponding to each heat treatment batch are determined. Finally, the products, product quantities and processing times corresponding to each heat treatment batch are integrated to obtain an initial target production schedule, and the heating process parameter information and the coding information of the heat treatment furnace are marked in the initial target production schedule to obtain a target production schedule.

[0084] Finally, a heat treatment furnace production plan is generated based on all the target production schedules.

[0085] The method provided in this embodiment helps to optimize the production process of the heat treatment furnace, thereby effectively utilizing the production resources of the heat treatment furnace and reducing the waste of the production resources of the heat treatment furnace.

[0086] See also Figure 2 , Figure 2 A schematic block diagram of the structure of a system 100 for automatically generating a production plan for a heat treatment furnace provided in an embodiment of the present application is shown in FIG. Figure 2 As shown, the system 100 for automatically generating a production plan for a heat treatment furnace comprises:

[0087] The first acquisition module 110 is used to respond to the generation instruction of the production plan of the heat treatment furnace, and obtain the production order data set of the heat treatment furnace in a preset production order database; wherein the production order data set includes multiple production orders.

[0088] The second acquisition module 120 is used to acquire the product texture of the product corresponding to each production order, and acquire the target heating process parameter table corresponding to the production order in a preset database based on the product texture; wherein the target heating process parameter table includes heating process parameter information matching the product texture, coding information of the heat treatment furnace matching the product texture, and a blank table, wherein the blank table includes a product name column, a product quantity column, a product size column, and an order delivery date column.

[0089] The third acquisition module 130 is used to obtain the product name, product quantity, product size and order delivery date of the product corresponding to each production order, and add the product name, product quantity, product size and order delivery date to the corresponding cells in the same row of the blank table of the target heating process parameter table corresponding to the production order, so as to obtain multiple production plan analysis tables.

[0090] The generation module 140 generates a heat treatment furnace production plan based on all the production plan analysis tables.

[0091] It should be noted that, technical personnel in the relevant technical field can clearly understand that, for the convenience and conciseness of description, the specific working process of the system and each module described above can refer to the corresponding process in the aforementioned method embodiment of automatically generating a heat treatment furnace production plan, and will not be repeated here.

[0092] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A method for automatically generating a production plan for a heat treatment furnace, characterized in that: include: In response to a generation instruction of a production plan of a heat treatment furnace, a production order data set of the heat treatment furnace is obtained from a preset production order database; wherein the production order data set includes a plurality of production orders; For each of the production orders, the product texture of the product corresponding to the production order is obtained, and based on the product texture, a target heating process parameter table corresponding to the production order is obtained in a preset database; wherein the target heating process parameter table includes heating process parameter information matching the product texture, coding information of a heat treatment furnace matching the product texture, and a blank table, wherein the blank table includes a product name column, a product quantity column, a product size column, and an order delivery date column; For each of the production orders, obtain the product name, product quantity, product size and order delivery date of the product corresponding to the production order, and add the product name, product quantity, product size and order delivery date to the corresponding cells of the same row of the blank table of the target heating process parameter table corresponding to the production order, to obtain multiple production plan analysis tables; Generate a heat treatment furnace production plan based on all of the production plan analysis tables; Wherein, obtaining a target heating process parameter table corresponding to the production order in a preset database based on the product texture includes: Encoding the product texture based on a preset encoding rule to obtain a product texture code; For each coding character of the product texture code, the coding character is used as a first coding character, and the remaining coding characters except the first coding character in the product texture code are used as second coding characters, and the similarities between the first coding character and each of the second coding characters are respectively obtained, and based on the order of the second coding characters corresponding to each of the similarities in the product texture code, each of the similarities is arranged in sequence to obtain a similarity sequence corresponding to the coding characters; Based on the order of the first coding characters corresponding to each of the similarity sequences in the product texture code, the similarity sequences are arranged in order from top to bottom to obtain a target matrix; The matrix elements on the diagonal line formed by the upper left corner and the lower right corner of the target matrix are used as target vector elements, and based on the order of each target vector element in the target matrix, each target vector element is arranged in sequence to obtain an initial vector; Multiplying the target matrix by the initial vector to obtain a target vector; Deleting non-numeric symbols in the target vector to obtain an identification sequence; A target heating process parameter table corresponding to the production order is obtained in a preset database based on the identification sequence.

2. The method for automatically generating a production plan for a heat treatment furnace according to claim 1, characterized in that: The obtaining of the product texture of the product corresponding to the production order includes: Obtaining the product name of the product corresponding to the production order through the production order; The product name is input into a preset product texture classification model to obtain the product texture of the product corresponding to the order.

3. The method for automatically generating a production plan for a heat treatment furnace according to claim 2, characterized in that: The training method of the product texture classification model comprises: Acquire a training data set and an initial model, and send the training data set and the initial model to a plurality of terminal devices, so that each of the terminal devices respectively trains the initial model based on the training data set to obtain a plurality of intermediate product texture classification models; wherein the training data set includes a plurality of mapping relationships, the mapping relationships being mapping relationships between product names and product textures corresponding to the product names, and each of the terminal devices is different from the other; Receive the intermediate product texture classification model sent by each of the terminal devices, fuse the model parameters of each of the intermediate product texture classification models to obtain fused model parameters, and use the fused model parameters to replace the model parameters of the initial model to obtain the product texture classification model.

4. The method for automatically generating a production plan for a heat treatment furnace according to claim 3, characterized in that: The initial model includes a feature extraction layer, a feature vector generation layer, a feature vector linear transformation layer and a classification layer, and the terminal device trains the initial model based on the training data set, including: Extracting feature information of products corresponding to product names corresponding to the mapping relationships respectively through the feature extraction layer; Generate feature vectors corresponding to each of the feature information respectively through the feature vector generation layer; Mapping each of the feature vectors to a high-dimensional feature space through the feature vector linear transformation layer to obtain multiple linear transformation results; Performing nonlinear transformation on each of the linear transformation results respectively through the activation function of the classification layer to obtain multiple nonlinear transformation results; Performing deep learning on each of the nonlinear transformation results through the global average pooling layer of the classification layer, respectively, to obtain the predicted product texture of the product corresponding to the product name corresponding to each of the mapping relationships; Constructing a predicted product texture-actual product texture mapping relationship table based on the predicted product texture and the actual product texture of the product corresponding to the product name corresponding to each of the mapping relationships; Calculating the FocalLoss loss value of the initial model based on the predicted product texture-actual product texture mapping relationship table; The model parameters of the initial model are optimized based on the FocalLoss loss value to obtain an intermediate product texture classification model.

5. The method for automatically generating a production plan for a heat treatment furnace according to claim 4, characterized in that: The calculating the FocalLoss loss value of the initial model based on the predicted product texture-actual product texture mapping relationship table includes: respectively calculating the Euclidean distances corresponding to the respective predicted product texture-actual product texture mapping relationships in the predicted product texture-actual product texture mapping relationship table; Calculating the initial FocalLoss loss value corresponding to each of the predicted product texture-actual product texture mapping relationships based on a preset Focal Loss loss function and the Euclidean distance corresponding to each of the predicted product texture-actual product texture mapping relationships; The sum of the initial FocalLoss loss values ​​is taken as the FocalLoss loss value.

6. The method for automatically generating a production plan for a heat treatment furnace according to claim 4, characterized in that: The method of optimizing the model parameters of the initial model based on the FocalLoss loss value to obtain an intermediate product texture classification model includes: Back-propagating the FocalLoss loss value in the initial model to obtain the gradients of each model parameter of the initial model; Based on each of the gradients, each model parameter of the initial model is optimized by a preset gradient descent algorithm to obtain an intermediate product texture classification model.

7. The method for automatically generating a production plan for a heat treatment furnace according to claim 1, characterized in that: The generating of the heat treatment furnace production plan based on all the production plan analysis tables comprises: For each of the production plan analysis tables, based on the order of order delivery dates corresponding to each row of information in the production plan analysis table, each row of information in the production plan analysis table is arranged in sequence to obtain an intermediate production plan table corresponding to the production plan analysis table; For each of the intermediate production schedules, based on the heating process parameter information corresponding to the intermediate production schedule, the coding information of the heat treatment furnace, the product quantity, product size and order delivery date corresponding to each row of information in the intermediate production schedule, a production plan for the product corresponding to each row of information in the intermediate production schedule is generated to obtain a target production schedule corresponding to the intermediate production schedule; A heat treatment furnace production plan is generated based on all of the target production schedules.

8. A system for automatically generating a production plan for a heat treatment furnace, characterized in that: include: A first acquisition module: used to respond to the generation instruction of the production plan of the heat treatment furnace, and acquire the production order data set of the heat treatment furnace in a preset production order database; wherein the production order data set includes multiple production orders; A second acquisition module is used to acquire, for each of the production orders, the product texture of the product corresponding to the production order, and acquire, in a preset database based on the product texture, a target heating process parameter table corresponding to the production order; wherein the target heating process parameter table includes heating process parameter information matching the product texture, coding information of a heat treatment furnace matching the product texture, and a blank table, wherein the blank table includes a product name column, a product quantity column, a product size column, and an order delivery date column; A third acquisition module is used to acquire, for each of the production orders, a product name, a product quantity, a product size, and an order delivery date of a product corresponding to the production order, and respectively add the product name, the product quantity, the product size, and the order delivery date to corresponding cells in the same row of a blank table of a target heating process parameter table corresponding to the production order, to obtain a plurality of production plan analysis tables; A generation module, generating a heat treatment furnace production plan based on all the production plan analysis tables; Wherein, obtaining a target heating process parameter table corresponding to the production order in a preset database based on the product texture includes: Encoding the product texture based on a preset encoding rule to obtain a product texture code; For each coding character of the product texture code, the coding character is used as a first coding character, and the remaining coding characters except the first coding character in the product texture code are used as second coding characters, and the similarities between the first coding character and each of the second coding characters are respectively obtained, and based on the order of the second coding characters corresponding to each of the similarities in the product texture code, each of the similarities is arranged in sequence to obtain a similarity sequence corresponding to the coding characters; Based on the order of the first coding characters corresponding to each of the similarity sequences in the product texture code, the similarity sequences are arranged in order from top to bottom to obtain a target matrix; The matrix elements on the diagonal line formed by the upper left corner and the lower right corner of the target matrix are used as target vector elements, and based on the order of each target vector element in the target matrix, each target vector element is arranged in sequence to obtain an initial vector; Multiplying the target matrix by the initial vector to obtain a target vector; Deleting non-numeric symbols in the target vector to obtain an identification sequence; A target heating process parameter table corresponding to the production order is obtained in a preset database based on the identification sequence.

Images