Steel mill scheduling method, scheduling system, device and medium

Abstract

The invention provides a steel mill scheduling method, system and device and a medium. The method comprises the following steps: inputting production data of a current production demand of a steel mill into a pre-trained neural network model; obtaining a steel mill scheduling scheme corresponding to the production data of the current production demand; the training of the neural network model comprises the following steps: S1, processing a structured data information table; s2, inputting the data information table into a neural network model for training to obtain a first model; s3, inputting the selected production data into a first model to obtain a first scheduling scheme; s4, correcting and marking the first scheduling scheme by using an active learning strategy and industry expert experience to obtain a second scheduling scheme; adding the second scheduling scheme into the data information table; s5, parameters of the neural network model are optimized; and S6, repeating the steps S2 to S5 until the first scheduling scheme reaches the accuracy. By means of the technical scheme, efficient processing and deep analysis of complex production data can be achieved, and the accuracy and flexibility of steel mill scheduling are improved.

Description

Technical Field

[0001] The present invention relates to the field of steel production, and particularly to a steel mill scheduling method, scheduling system, device and medium. Background Art

[0002] In the process of steel production, the links of raw material processing, production, transportation and inventory are all highly complex, and the requirements for coordination and timeliness are extremely strict. Traditional manual experience and rule-based scheduling methods are difficult to meet the needs of modern steel mills for efficiency and flexibility; these methods are easily affected by individual experience and understanding differences, further resulting in suboptimal scheduling results and prone to problems such as delays and resource waste. In addition, manual scheduling is not flexible enough in the face of emergencies (such as equipment failures and stoppages of steel mill trucks), resulting in low overall production efficiency. Especially in a production environment with multi-task parallelism and multi-link interaction, traditional scheduling methods are difficult to process a large amount of data in real time and cannot achieve fast and accurate resource allocation and adjustment. In addition, existing scheduling technologies based on mathematical models such as linear programming (LP) and mixed integer programming (MIP) can optimize resource allocation, but they rely on constraint conditions set by experts and lack the ability to dynamically adapt to rapidly changing production demands;

[0003] Natural language processing (NLP) is a method of using artificial intelligence technology to understand and process human language. Through deep learning neural network models such as recurrent neural network (RNN), long short-term memory network (LSTM) and transformer, text data is converted into numerical vectors for semantic analysis and information extraction. In the intelligent scheduling of steel mills, the application of NLP technology has high feasibility, which can intelligently process scheduling instructions and production reports, reduce the workload of schedulers and reduce human interpretation errors. In addition, the combination of NLP with big data and machine learning technologies can achieve intelligent analysis and optimization of real-time data, enhancing the adaptability and dynamic response ability of the scheduling system. However, current artificial intelligence technologies mainly rely on a large amount of data for training and optimization, lacking the constraints of domain expert experience, which may lead to poor performance of the model in the face of unseen situations and insufficient generalization ability of the generated results. Summary of the Invention

[0004] Embodiments of the present invention provide a steel mill scheduling method, scheduling system, device and medium to achieve efficient processing and in-depth analysis of complex production data, and improve the accuracy and flexibility of steel mill scheduling.

[0005] To achieve the above object, on the one hand, a steel mill scheduling method is provided, and the method includes:

[0006] Input the production data of the current production requirements of the steel plant into a pre-trained neural network model; wherein, the production data includes: ladle number, current temperature, molten iron composition, starting position, target work station, number of workers and / or number of transport vehicles;

[0007] Obtain a steel plant scheduling plan corresponding to the production data of the current production requirements;

[0008] Among them, training the neural network model includes:

[0009] S1, Classify the pre-collected historical production data and historical scheduling plans according to the production stage, equipment type, and time stamp, and process the classified historical production data and historical scheduling plans into a structured data information table according to a predetermined rule;

[0010] S2, Input the data information table as a training data set into the neural network model for network training to obtain a first model;

[0011] S3, Input the selected production data of the steel plant into the first model to obtain a first scheduling plan;

[0012] S4, Use a predetermined active learning strategy and the pre-obtained industry expert experience to correct and annotate the first scheduling plan to obtain a second scheduling plan; Add the second scheduling plan to the data information table;

[0013] S5, Calculate the loss function between the first scheduling plan and the second scheduling plan, and iteratively optimize the parameters of the neural network model according to the loss function through the backpropagation algorithm;

[0014] S6, Repeat steps S2 to S5 until the first scheduling plan reaches a predetermined accuracy rate.

[0015] Preferably, in the method for generating a steel plant scheduling plan, wherein, the pre-obtained industry expert experience is obtained by a predetermined person.

[0016] Preferably, in the method for generating a steel plant scheduling plan, wherein, the neural network model is: a neural network model built by combining the TransFormer architecture and the depthwise separable convolution architecture.

[0017] Preferably, in the method for generating a steel plant scheduling plan, wherein, the fully connected feedforward neural network of the TransFormer architecture is based on two layers of depthwise separable convolution networks, and combines the ReLU activation function and regularization techniques.

[0018] Preferably, in the method for generating a steel mill scheduling plan, in the Transformer architecture, the multiple stacked Transformer modules generate the scheduling plan through a multi-head attention mechanism.

[0019] Preferably, in the method for generating a steel mill scheduling plan, the fully connected feedforward neural network includes: a forward fully connected feedforward neural network module and a reverse fully connected feedforward neural network module; wherein, the forward fully connected feedforward neural network module and the reverse fully connected feedforward neural network module perform bidirectional conversion between the text information in the data information table and the computer vector representation of the neural network model.

[0020] On the other hand, an embodiment of the present invention provides a steel mill scheduling system, wherein the steel mill scheduling system is used to implement any one of the above steel mill scheduling methods, including:

[0021] A visualization page that runs the neural network model in Claim 1 by calling the Python algorithm interface through MATLAB.

[0022] On yet another aspect, an embodiment of the present invention provides a steel mill scheduling device, which includes a memory and a processor. The memory stores at least one segment of program, and the at least one segment of program is executed by the processor to implement any one of the above steel mill scheduling methods.

[0023] On yet another aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the storage medium stores at least one segment of program, and the at least one segment of program is executed by the processor to implement any one of the above steel mill scheduling methods.

[0024] The above technical solutions have the following technical effects:

[0025] The technical solution of the embodiment of the present invention inputs the production data of the current production requirements of the steel plant into a pre-trained neural network model; obtains a steel plant scheduling plan corresponding to the production data of the current production requirements; wherein, training the neural network model includes: collecting and structurally processing the historical production data and historical scheduling plans of the steel plant to construct a rich training data set; training the pre-built neural network model through the training data set, and this model can generate a preliminary first scheduling plan according to the production data; through an active learning strategy, according to the industry expert experience obtained in advance, the first scheduling plan is carefully corrected and labeled to generate a better second scheduling plan, and it is fed back into the training data set to enrich and improve the decision-making ability of the model; by calculating the loss function between the two plans and using the backpropagation algorithm to iteratively optimize the model parameters, the model can learn how to reduce errors and improve the quality of the plan; this iterative process is continuously repeated until the scheduling plan generated by the model reaches a predetermined accuracy rate, ensuring the reliability and effectiveness of the model output, thereby realizing the automation and intelligence of steel plant scheduling, improving the scheduling efficiency and response speed, reducing resource waste, and ultimately enhancing the overall production efficiency of the steel plant;

[0026] In a further embodiment, by integrating the Transformer architecture and the depthwise separable convolutional network architecture, a powerful neural network model is constructed. This model uses two layers of depthwise separable convolutional networks as the basis to strengthen the capture of local features, and introduces non-linearity through the ReLU activation function to enhance the expression ability of the model. The addition of regularization technology effectively prevents overfitting, improves the generalization performance of the model, realizes in-depth understanding and efficient processing of complex data patterns, optimizes the computational efficiency of the model, and enhances the accuracy and adaptability in actual scheduling tasks.

[0027] In a further embodiment, an intuitive visual operation interface is built through MATLAB, enabling users to quickly input production information and generate a scheduling plan with one key; at the same time, the adaptability and user-friendliness of the system are enhanced, effectively promoting the intelligent transformation of the steel industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a schematic flowchart of the steel plant scheduling method according to an embodiment of the present invention;

[0029] Figure 2 It is a flowchart of the steel plant scheduling method according to an embodiment of the present invention;

[0030] Figure 3 It is a schematic diagram of the model training process in the steel plant scheduling method according to an embodiment of the present invention;

[0031] Figure 4 It is a schematic diagram of the visual page in the steel plant scheduling system according to an embodiment of the present invention;

[0032] Figure 5 In the steel plant scheduling system according to an embodiment of the present invention, it is a schematic diagram of inputting production data of current production requirements in a visualization page;

[0033] Figure 6 In the steel plant scheduling system according to an embodiment of the present invention, it is a schematic diagram of generating a scheduling plan in a visualization page;

[0034] Figure 7 It is a schematic structural diagram of a steel plant scheduling device according to an embodiment of the present invention. Detailed implementation manners

[0035] To further illustrate each embodiment, the present invention provides accompanying drawings. These drawings are part of the disclosure of the present invention, mainly used to illustrate the embodiments, and can be combined with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, those of ordinary skill in the art should be able to understand other possible implementation manners and the advantages of the present invention. The components in the figures are not drawn to scale, and similar component symbols are usually used to represent similar components.

[0036] Now, the present invention will be further described in conjunction with the accompanying drawings and specific implementation manners.

[0037] Embodiment 1:

[0038] In order to achieve efficient processing and in-depth analysis of complex production data, and improve the accuracy and flexibility of steel plant scheduling, an embodiment of the present invention provides a steel plant scheduling method. Figure 1 It is a schematic flowchart of the steel plant scheduling method according to an embodiment of the present invention. As Figure 1 shown, the method includes:

[0039] Input the production data of the current production requirements of the steel plant into a pre-trained neural network model; wherein, the production data includes: ladle number, current temperature, molten iron composition, starting position, target work station, number of workers, and / or number of transport vehicles;

[0040] Obtain a steel plant scheduling plan corresponding to the production data of the current production requirements;

[0041] Among them, training the neural network model includes:

[0042] S1, classify the pre-collected historical production data and historical scheduling plans according to the production stage, equipment type, and time stamp, and process the classified historical production data and historical scheduling plans into a structured data information table according to a predetermined rule;

[0043] S2, input the data information table as a training data set into the neural network model for network training to obtain a first model;

[0044] S3. Input the selected production data of the steel mill into the first model to obtain the first scheduling plan;

[0045] S4. Use a predetermined active learning strategy and the pre-acquired industry expert experience to correct and annotate the first scheduling plan to obtain the second scheduling plan; add the second scheduling plan to the data information table;

[0046] S5. Calculate the loss function between the first scheduling plan and the second scheduling plan, and iteratively optimize the parameters of the neural network model according to the loss function through the backpropagation algorithm;

[0047] S6. Repeat steps S2 to S5 until the first scheduling plan reaches the predetermined accuracy.

[0048] Embodiment 2:

[0049] Traditional steel mill scheduling methods are mainly based on manual experience and rule guidance, with strong subjectivity, difficult to quickly adapt to changes in the production environment, and limited support for complex decisions. In addition, current intelligent scheduling dominated by artificial intelligence technology mostly relies on data-driven methods, lacking the constraint of domain expert knowledge, which may lead to understanding deviations of specific industry terms and scenarios. In order to make full use of historical data while integrating industry model knowledge into scheduling decisions, the embodiments of the present invention provide a steel mill scheduling method. Through this innovative method, the accuracy and flexibility of steel mill scheduling are improved, promoting the intelligent transformation of the steel industry. Figure 2 It is a flowchart of the steel mill scheduling method according to an embodiment of the present invention. As Figure 2 shown, the method is as follows:

[0050] Input the production data of the current production demand of the steel mill, that is, the steel mill data, into the pre-trained neural network (Trans-CNN) model;

[0051] Preferably, the production data includes: ladle number, current temperature, molten iron composition, starting position, target work station, number of workers, and / or number of transport vehicles, etc.;

[0052] Preferably, the neural network model is a neural network model built by combining the Transformer architecture and the depthwise separable convolution architecture.

[0053] Obtain the steel mill scheduling plan corresponding to the production data of the current production demand, that is, the intelligent scheduling plan.

[0054] Among them, Figure 3 It is a schematic diagram of the model training process in the steel mill scheduling method according to an embodiment of the present invention. As Figure 3 shown, training the neural network model includes:

[0055] (1) Collect the past information of the steel plant and perform preprocessing, namely text information extraction, to obtain a data information table, i.e., text data;

[0056] Specifically, extract and classify the collected historical production data and historical scheduling plans according to the production stage, equipment type, and timestamp, and process the classified historical production data and historical scheduling plans into a structured data information table according to predetermined rules;

[0057] Preferably, the past information includes: the number of idle locomotives, the locomotive transportation path, the material arrival time, etc.

[0058] Preferably, the fully connected feedforward (MLP) neural network of the TransFormer architecture is based on a two-layer depthwise separable convolutional network and combines the ReLU activation function and regularization techniques;

[0059] Preferably, the multi-layer stacked Transformer modules in the Transformer architecture generate a scheduling plan through the multi-head attention mechanism;

[0060] Preferably, the fully connected feedforward neural network includes: a forward fully connected feedforward neural network module and a reverse fully connected feedforward neural network module; among them, the forward fully connected feedforward neural network module and the reverse fully connected feedforward neural network module realize the bidirectional conversion between the text information in the data information table and the computer vector representation of the neural network model; while the multi-layer stacked Transformer modules capture the deep association between the input information and the scheduling plan through the multi-head attention mechanism, and then generate a scheduling plan that meets the intelligent requirements.

[0061] (2) Input the data information table into the neural network model to obtain a first scheduling plan, i.e., a preliminary scheduling plan;

[0062] Specifically, use the data information table as a training data set to input into the neural network model for network training to obtain a first model; input the selected production data of the steel plant into the first model to obtain a first scheduling plan;

[0063] (3) Judge the first scheduling plan through industry experts, i.e., modify and annotate it, to obtain a second scheduling plan;

[0064] Specifically, use a predetermined active learning strategy and the pre-acquired industry expert experience to correct and annotate the first scheduling plan to obtain a second scheduling plan; add the second scheduling plan to the data information table to enhance the learning effect of the neural network model; adopt the active learning idea, judge the first scheduling plan through the pre-acquired industry expert experience, and feedback the evaluation index;

[0065] Preferably, the industry expert experience obtained in advance is obtained by a predetermined person; preferably, the predetermined person is an expert in the industry field;

[0066] In a specific embodiment, by calculating the loss function between the first scheduling scheme and the second scheduling scheme, the parameters of the neural network model are iteratively optimized according to the loss function through the backpropagation algorithm, further improving the model accuracy;

[0067] In a specific embodiment, during the training process of the neural network model, through an active learning mechanism, the industry expert randomly judges the generated scheduling scheme according to the expert experience, and selects the scheduling scheme with the largest deviation from the expert experience knowledge or the most valuable annotation recognized by the expert for re-modification, that is, expert model guidance; the modified scheduling scheme is added to the data information table to participate in the next round of model training, and the neural network model iterates the model parameters according to the modified scheduling scheme, thereby dynamically introducing expert knowledge for constraint, and realizing the organic combination of data-driven and expert experience model knowledge.

[0068] (4) Repeat (2) to (3) until the first scheduling scheme reaches the predetermined accuracy rate;

[0069] Preferably, the predetermined accuracy rate is the accuracy rate set according to the industry expert experience;

[0070] Specifically, repeat the iterative process until the first scheduling scheme output by the neural network model reaches the expected accuracy or completes the fitting; on this basis, retain the parameters of the best neural network model, that is, retain the parameter weights, and use the production data of the current production demand of the steel mill to intelligently generate the scheduling scheme.

[0071] The steel mill scheduling method of the embodiment of the present invention combines the Transformer model architecture with the depthwise separable convolution architecture to achieve efficient processing and in-depth analysis of complex production data. In addition, by using the active learning mechanism to dynamically introduce industry expert experience knowledge, it not only improves the adaptability of the model to domain-specific knowledge, but also effectively guarantees the accuracy and stability of the model, showing significant advantages in terms of accuracy and usability, and providing a scientific and efficient intelligent solution for the scheduling optimization in the steel mill production process.

[0072] Embodiment 3:

[0073] An embodiment of the present invention also provides a steel plant scheduling system, which is used to implement the steel plant scheduling method in Embodiment 1 of the present invention. During the construction of the steel plant scheduling system in the embodiment of the present invention, MATLAB is used as the front-end development tool to encapsulate and visualize the trained neural network model in Embodiment 1. The powerful graphics processing and visualization functions of MATLAB enable the steel plant scheduling system to more intuitively display the scheduling decision-making process and results, facilitating users to observe and adjust. In addition, the good compatibility between MATLAB and Python provides convenience for model implementation. By calling the Python algorithm interface in MATLAB, not only can the Trans-CNN network model be efficiently run, but also data transfer and functional cooperation between the two can be achieved. The construction of this system ensures the operability and practicality of the Trans-CNN neural network model, providing an intuitive and intelligent support tool for steel plant scheduling. The steel plant scheduling system includes: a visualization page.

[0074] Figure 4 It is a schematic diagram of the visualization page in the steel plant scheduling system according to an embodiment of the present invention; Figure 5 It is a schematic diagram of inputting production data of current production requirements in the visualization page of the steel plant scheduling system according to an embodiment of the present invention; Figure 6 It is a schematic diagram of generating a scheduling plan in the visualization page of the steel plant scheduling system according to an embodiment of the present invention. As Figure 4 shown, the visualization page includes: a production plan, real-time data, hot metal scheduling, scheduling analysis, data analysis, and early warning monitoring modules; among them, the production plan module includes: daily material consumption report, overall material plan, order management, monthly work order trend, monthly energy consumption ranking, requirements for various raw materials, and production plan, etc.

[0075] As Figure 5 shown, the hot metal scheduling module of the visualization page includes: equipment status overview, current task scheduling, hot metal ladle status overview, and equipment online status; in the current task scheduling column, users can input relevant information to be scheduled, such as key data such as hot metal ladle number, current temperature, and hot metal composition; after the input is completed, the system will perform intelligent analysis on the input information through the above-mentioned trained Trans-CNN model;

[0076] As Figure 6 shown, the scheduling analysis module of the visualization page includes: scheduling detailed information, scheduling history records, and scheduling analysis and optimization suggestions; in this module, the system generates a hot metal ladle scheduling route adapted to the current production requirements, scheduling arrangements for transportation vehicles, etc. with one key, realizing the full-process automation from information input to plan output, and the visualization page is designed simply and intuitively, facilitating users to quickly operate and schedule management.

[0077] The design of the front-end visualization page based on MATLAB makes the scheduling process more intuitive, facilitating real-time monitoring and adjustment. The function of the system to generate a scheduling plan with one key also enables users to complete scheduling tasks quickly and accurately, reducing the need for manual intervention.

[0078] Embodiment 4:

[0079] The present invention also provides a steel mill scheduling device, as Figure 7 shown. The device includes a processor 701, a memory 702, a bus 703, and a computer program stored in the memory 702 and executable on the processor 701. The processor 701 includes one or more processing cores. The memory 702 is connected to the processor 701 through the bus 703. The memory 702 is used to store program instructions. When the processor 701 executes the computer program, it implements the steps in the above method embodiment of Embodiment 1 of the present invention.

[0080] Furthermore, as an executable solution, the steel mill scheduling device may be a computer unit, and this computer unit may be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The computer unit may include, but is not limited to, a processor and a memory. Those skilled in the art can understand that the above composition structure of the computer unit is only an example of the computer unit and does not constitute a limitation on the computer unit. It may include more or fewer components than the above, or combine certain components, or different components. For example, the computer unit may further include input and output devices, network access devices, a bus, etc. The embodiments of the present invention do not make limitations in this regard.

[0081] Furthermore, as an executable solution, the so-called processor may be a central processing unit (CPU), or may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The processor is the control center of the computer unit, connecting various parts of the entire computer unit through various interfaces and lines.

[0082] The memory can be used to store the computer program and / or modules. By running or executing the computer program and / or modules stored in the memory, and invoking the data stored in the memory, the processor realizes various functions of the computer unit. The memory mainly includes a program storage area and a data storage area. Among them, the program storage area can store the operating system and application programs required for at least one function; the data storage area can store data created according to the use of the mobile phone, etc. In addition, the memory can include high-speed random access memory, and can also include non-volatile memory, such as hard disks, memory, plug-in hard disks, smart media cards (SMC), secure digital (SD) cards, flash cards, at least one magnetic disk storage device, flash memory device, or other volatile solid-state storage devices.

[0083] Embodiment 5:

[0084] The present invention also provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the steps of the method in the above embodiments of the present invention are realized.

[0085] If the modules / units integrated in the computer unit are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, to implement all or part of the processes in the above embodiment methods of the present invention, it can also be completed by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above various method embodiments can be realized. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM), random access memory (RAM), and software distribution medium, etc. It should be noted that the content included in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction.

[0086] Although the present invention is specifically shown and described in combination with the preferred embodiments, those skilled in the art should understand that various changes can be made to the present invention in form and detail without departing from the spirit and scope of the present invention defined by the appended claims, and all of them are within the protection scope of the present invention.

Claims

1. A steel plant scheduling method, characterized in that: include: Inputting the production data of the current production demand of the steel plant into the pre-trained neural network model; wherein the production data includes: ladle number, current temperature, molten iron composition, starting position, target workstation, number of workers and / or number of transport vehicles; Obtaining a steel plant scheduling plan corresponding to the production data of the current production demand; Wherein, training the neural network model comprises: S1, classify the pre-collected historical production data and historical scheduling plans according to the production stage, equipment type, and timestamp, and process the classified historical production data and historical scheduling plans into a structured data information table according to predetermined rules; S2, inputting the data information table as a training data set into the neural network model for network training to obtain a first model; S3, inputting the selected production data of the steel plant into the first model to obtain a first scheduling plan; S4, using a predetermined active learning strategy and pre-acquired industry expert experience to modify and annotate the first scheduling plan to obtain a second scheduling plan; adding the second scheduling plan to the data information table; S5, calculating a loss function between the first scheduling scheme and the second scheduling scheme, and iteratively optimizing the parameters of the neural network model through a back propagation algorithm according to the loss function; S6, repeating steps S2 to S5 until the first scheduling scheme reaches a predetermined accuracy rate.

2. The method for generating a steel plant scheduling plan according to claim 1, characterized in that: The pre-acquired industry expert experience is obtained through predetermined personnel.

3. The steel plant scheduling method according to claim 1, characterized in that: The neural network model is a neural network model built by combining the TransFormer architecture with the deep separable convolutional architecture.

4. The steel plant scheduling method according to claim 3, characterized in that: The fully connected feedforward neural network of the TransFormer architecture is based on a two-layer deep separable convolutional network, combined with ReLU activation function and regularization technology.

5. The steel plant scheduling method according to claim 3, characterized in that: The multi-layer stacked Transformer modules in the Transformer architecture generate scheduling plans through a multi-head attention mechanism.

6. The steel plant scheduling method according to claim 4, characterized in that: The fully connected feedforward neural network includes: a forward fully connected feedforward neural network module and a reverse fully connected feedforward neural network module; wherein the forward fully connected feedforward neural network module and the reverse fully connected feedforward neural network module perform bidirectional conversion between text information in the data information table and the computer vector representation of the neural network model.

7. A steel plant dispatching system, characterized in that: The steel plant scheduling system is used to implement the steel plant scheduling method according to any one of claims 1 to 6, comprising: The visualization page calls the Python algorithm interface through MATLAB to run the neural network model in claim 1.

8. A steel plant dispatching device, characterized in that: It comprises a memory and a processor, wherein the memory stores at least one program, and the at least one program is executed by the processor to implement the steel plant scheduling method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The storage medium stores at least one program, and the at least one program is executed by a processor to implement the steel plant scheduling method as described in any one of claims 1 to 6.

Images