Production method of high-strength boron-containing steel for ocean engineering

Abstract

The invention relates to the technical field of boron-containing steel production, in particular to a production method of high-strength boron-containing steel for ocean engineering, which comprises the following steps: acquiring equipment parameters in a production line for actually producing the boron-containing steel, and constructing a three-dimensional model based on the equipment parameters; constructing the three-dimensional model in a cloud server, arranging a plurality of monitoring points on entity equipment of the boron-containing steel production line, connecting the plurality of monitoring points to form a monitoring network, and constructing a transmission chain between the monitoring network and the cloud server; the production parameters of the current working procedure on the production line are collected in real time based on the monitoring points, and the production parameters of the current working procedure are transmitted to the three-dimensional model based on the transmission chain; and the production parameters of the next procedure are determined based on the three-dimensional model, the production parameters of the next procedure are issued to the corresponding production procedure in the production line for production control, the accuracy of technological parameter prediction can be improved, and then the production quality of the boron-containing steel is improved.

Description

Technical Field

[0001] This invention relates to the field of boron-containing steel production technology, specifically a method for producing high-strength boron-containing steel for marine engineering. Background Technology

[0002] The performance of boron-containing steel depends on parameters such as rolling temperature, cooling rate, and final cooling temperature. Even slight deviations in these parameters can lead to the final microstructure not being the expected bainitic / martensite dual phase structure, but instead exhibiting ferrite, pearlite, or coarse microstructure, resulting in substandard strength and toughness. In existing production methods, all similar products are processed using the same set of fixed process parameters. These fixed process parameters cannot flexibly respond to dynamic adjustments in the process path of production batches, affecting the production quality of boron-containing steel.

[0003] To address these issues, we propose a production method for high-strength boron-containing steel for marine engineering. Summary of the Invention

[0004] The purpose of this invention is to provide a method for producing high-strength boron-containing steel for marine engineering, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for producing high-strength boron-containing steel for marine engineering, the method comprising the following steps: Obtain the equipment parameters in the actual production line for producing boron-containing steel, and construct a 3D model based on the equipment parameters; The three-dimensional model is built in the cloud server, multiple monitoring points are set up on the physical equipment of the boron-containing steel production line, the multiple monitoring points are connected to form a monitoring network, and a transmission chain is built between the monitoring network and the cloud server. The production parameters of the current process on the production line are collected in real time based on monitoring points, and the production parameters of the current process are transmitted to the three-dimensional model based on the transmission chain. Based on the 3D model, the production parameters for the next process are determined, and these parameters are then sent to the corresponding production process on the production line for production control.

[0006] Preferably, the step of obtaining equipment parameters in the actual production line for producing boron-containing steel and constructing a three-dimensional model based on the equipment parameters includes: Obtain the equipment parameters of the boron-containing steel production line, including equipment layout information, equipment shape information, and equipment connection information; A three-dimensional model is obtained by performing three-dimensional modeling based on the equipment parameters, and the process equipment models corresponding to different production processes are identified in the three-dimensional model. For each pair of adjacent processes with a logical sequence, a pre-trained prediction model is integrated between their corresponding process equipment models. The prediction model is used to predict the process operation information of the downstream process based on initial information, which is the boron-containing steel process parameters at the start of the production line or the boron-containing steel process parameters output by the upstream process.

[0007] Preferably, the step of setting up multiple monitoring points on the physical equipment of the boron-containing steel production line and connecting the multiple monitoring points to form a monitoring network includes: Multiple monitoring points are set up on the physical equipment of the boron-containing steel production line. A moving window is configured for each monitoring point. The moving window is used to collect and cache the historical data of the corresponding monitoring point stored in time series. Configure a unified clock synchronization source for the moving windows to ensure that the time base of all moving windows is consistent; establish communication connections between multiple moving windows to connect multiple monitoring points to form a monitoring network.

[0008] Preferably, the step of establishing a transmission link between the monitoring network and the cloud server includes: Using the target production batch on the production line as the fixed point, a selection group and a management terminal are configured for the fixed point. The selection group includes multiple interconnected random points. A temporary communication channel is established between the random points and the monitoring point. Communication connections are established between the fixed point and the multiple random points. A transmission channel is established between the selected group and the cloud server based on the management terminal, and the transmission channel is used as the transmission link between the monitoring network and the cloud server.

[0009] Preferably, the step of collecting production parameters of the current process on the production line in real time based on monitoring points, and transmitting the production parameters of the current process to the three-dimensional model based on the transmission chain includes: Real-time collection of production parameters for the current process on the production line; information chain generated by acquiring production parameters at the same time point from various monitoring points based on a moving window. Obtain the current process corresponding to the current fixed position, take the next process corresponding to the current process as the target process, determine the target prediction model corresponding to the target process from the 3D model, obtain the initial demand information of the target prediction model, and determine the target monitoring group corresponding to the target prediction model based on the initial demand information. Multiple random points are assigned to the target monitoring points of the corresponding target monitoring group, and a temporary communication channel is established between the random points and the target monitoring points. The random points obtain parameter information located on the information chain from the target monitoring points based on the temporary communication channel and transmit it to the management terminal. The management terminal packages and transmits multiple parameter information transmitted from various points to the target prediction model.

[0010] Preferably, the step of determining the monitoring group corresponding to the target prediction model based on the initial demand information includes: Obtain the demand information category corresponding to the initial demand information of the target prediction model, and send the demand information category to different random points through the management terminal; Based on the monitoring network, the monitoring point will be moved from the current monitoring point to the target monitoring point corresponding to the demand information category; multiple target monitoring points will be combined to obtain the monitoring group corresponding to the target prediction model.

[0011] Preferably, the specific steps of assigning multiple random points to the target monitoring points of the corresponding target monitoring group include: Based on the management terminal assigning target monitoring points to each random point, each random point sends a transfer request to its currently bound target monitoring point. The transfer request contains the identity identifier of the target monitoring point assigned to it. The current monitoring point establishes sensor communication with the target monitoring point based on the identity identifier, and then unbinds the random point; The target monitoring verifies the access credentials of the monitoring point, establishes a binding relationship with the monitoring point, and assigns the monitoring point to the corresponding target monitoring point.

[0012] Preferably, the step of determining the production parameters for the next process based on the 3D model and then distributing these parameters to the corresponding production process on the production line for production control includes: The production parameters of the current process on the production line are collected in real time and input into the prediction model of the next process in the 3D model to obtain the production parameters of the next process; The production parameters for the next process are sent to the corresponding physical actuator in the production line. The physical actuator adjusts its operating state according to the received production parameters to control the actual production of the next process.

[0013] Compared with the prior art, the beneficial effects of the present invention are: By selecting different information based on changes in fixed points, the operation information for the next process is optimized. This allows for better formulation of corresponding process operation information based on the production parameters of the previous process, improving the accuracy of process parameter prediction and thus improving the production quality of boron-containing steel. When a production batch (fixed point) moves between different processes, the dynamic binding of the following point, monitoring group, and fixed point ensures that the required data flow always follows the target production batch. Data irrelevant to the current batch is automatically removed, and the following point is directly transferred between monitoring points. This reduces the end-to-end delay from event occurrence to data availability, improves the efficiency of obtaining the fixed point's production parameters in the current process, improves the efficiency of optimizing the production parameters for the next process, enables real-time monitoring of the production process, allows for timely intervention in production anomalies, and improves the production quality of boron-containing steel. Attached Figure Description

[0014] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a schematic diagram of the method flow of the present invention; Figure 2 This is a schematic diagram of the transmission chain of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Example

[0018] Please see Figures 1 to 2 This invention provides a technical solution for the production method of high-strength boron-containing steel for marine engineering: a production method of high-strength boron-containing steel for marine engineering includes the following steps: S1: Obtain the equipment parameters in the actual production line for producing boron-containing steel, and build a three-dimensional model based on the equipment parameters; The steps for obtaining equipment parameters in the actual production line for boron-containing steel and constructing a 3D model based on these parameters include: obtaining equipment parameters for the boron-containing steel production line, including equipment layout information, equipment shape information, and equipment connection information; performing 3D modeling based on the equipment parameters to obtain a 3D model, and identifying the process equipment models corresponding to different production processes in the 3D model; for every two adjacent processes with a logical sequence relationship, integrating a pre-trained prediction model between their corresponding process equipment models, the prediction model is used to predict the process operation information of the downstream process based on initial information, which is the boron-containing steel process parameters at the beginning of the production line or the boron-containing steel process parameters output by the upstream process; Specifically, by analyzing the engineering design drawings of the boron-containing steel production line, basic equipment layout information, equipment shape information, and equipment connection information are obtained. A 3D laser scanner is used to scan the actual production line to obtain point cloud data. The point cloud data is then fitted and calibrated with a 3D model constructed based on the engineering design drawings to generate a 3D equipment model consistent with the physical production line. The equipment in the actual boron-containing steel production line includes at least a continuous casting machine, a heating furnace, a hot rolling mill, and a cooling device. The 3D model is used to simulate the phase transformation process and boron precipitation behavior of the boron-containing steel in the production line when corresponding process parameters are input later. The prediction model is a machine learning model, a pre-trained model used to predict the process operation information of the current process based on initial information. The specific construction process involves collecting historical production data of the boron-containing steel production line, including initial information for multiple production batches, as well as the actual process operation information and corresponding product quality indicators. Using the initial information as input features, and the process operation information that can produce qualified boron-containing steel products as the input features, the model is then used to predict the process operation information of the current process. The information is used as output labels to train the machine learning algorithm and obtain a predictive model. Each predictive model is used as attribute data and associated with the corresponding downstream process equipment model in the initial 3D model to form a predictive 3D model. In the predictive 3D model, the predictive model can be driven to run sequentially based on the input initial process parameters, and the operation status of the entire production line can be visualized and predicted. The initial information of the current production batch is accessed in real time. In the predictive 3D model, the predictive model corresponding to the current production process is run to obtain predictive process operation information for subsequent processes. The predictive process operation information is sent to the production execution system or displayed to the operator as a decision reference for production settings. The initial information is equivalent to production parameter information, including any one or a combination of the following: the initial boron content of the slab, the furnace outlet temperature, and the finishing mill outlet temperature. The process operation information includes any one or a combination of the following: the set temperature of each section of the furnace, the reduction amount distribution of the rolling mill stand, and the manifold opening mode of the cooling system.

[0019] S2: Build a 3D model in a cloud server, set up multiple monitoring points on the physical equipment of the boron-containing steel production line, connect the multiple monitoring points to form a monitoring network, and build a transmission chain between the monitoring network and the cloud server; The steps of deploying multiple monitoring points on the physical equipment of a boron-containing steel production line and connecting these monitoring points to form a monitoring network include: deploying multiple monitoring points on the physical equipment of the boron-containing steel production line; configuring a moving window for each monitoring point, wherein the moving window is used to collect and cache historical data stored in time series for the corresponding monitoring point; configuring a unified clock synchronization source for the moving windows to ensure that the time base of all moving windows is consistent; establishing communication connections between multiple moving windows to connect the multiple monitoring points and obtain the monitoring network. Specifically, monitoring points are equivalent to multiple sensors deployed on the actual production line. When data is collected at each monitoring point, a high-precision timestamp based on a unified clock synchronization source is added to each data point. The original data with timestamps is stored in the local data sequence of the monitoring point for collection by the moving window. Multiple moving window modules are driven to collect data at the same time point concurrently to generate an information chain sequence. This information is then used to extract the corresponding monitoring point's information from the information chain sequence at subsequent points. All data collected at these points is packaged into a data packet at the management end and transmitted to the corresponding target prediction model. This maintains the consistency of data collected from each monitoring point and ensures that consistent data is transmitted to the corresponding target prediction model at the same time, improving the accuracy of the prediction model in predicting the operation information of the next process, thereby improving the production quality of boron-containing steel.

[0020] The steps for building a transmission link between the monitoring network and the cloud server include: taking the target production batch on the production line as the fixed point, configuring a selection group and a management terminal for the fixed point, wherein the selection group includes multiple inter-communicating random points, temporarily establishing a communication channel between the random points and the monitoring points, and establishing communication connections between the fixed point and multiple random points; establishing a transmission channel between the selection group and the cloud server based on the management terminal, and using the transmission channel as the transmission link between the monitoring network and the cloud server. Specifically, the monitoring network consists of multiple monitoring points with mutual sensing and communication capabilities, used to perform point-to-point transmission, unbinding, and binding operations. A fixed point refers to a target production batch (such as a steel slab with a specific number or a workpiece from a specific order) that is uniquely identified and tracked within the 3D model and physical production line. It serves as a carrier of virtual identity information, used to carry the identity, status, and location information of the produced product, and as an anchor point for position synchronization between the 3D model and the physical world. The movement of fixed points drives the dynamic reorganization of the corresponding monitoring group. A point-to-point is equivalent to a functional execution unit; multiple points-to-points form a selection group, configured to migrate between monitoring points. It represents the management end in executing specific data acquisition tasks, including task logic, data caching, and communication interfaces. It can change the location of the monitoring point based on the sensor communication between monitoring points, establish temporary communication channels with the monitoring points, acquire corresponding data from the bound monitoring points, and collaborate with other monitoring points and the management end to exchange relevant information and status. A selection group is a dynamic set of multiple communicating monitoring points, managed by a single management end. It belongs to the fixed-point data acquisition task execution team. The monitoring points in the selection group are not fixed and can be dynamically added, removed, or replaced according to the different processes in which the fixed points are located. Multiple monitoring points achieve data collaboration through communication connections (such as data...). The fusion and packaging (or integration) is a functional collaborative entity; the monitoring group is a collection of all physical monitoring points currently bound to all random points within a selected group at a specific moment. It dynamically changes as the selected group moves and reorganizes. For example, when the fixed point is in the heating process, the monitoring group mainly consists of temperature sensors in the heating furnace area; when the fixed point moves to the rolling process, the monitoring group becomes composed of pressure and speed sensors in the rolling mill area. The management end is a central coordinator associated with the fixed point, logically moving synchronously with it. It is responsible for the global management and scheduling of its dedicated selected group. The management end is a control and decision-making unit used to receive predictive model requirements from the cloud server and decompose them. For specific point-following tasks, the system determines the number and types of monitoring points needed, directs their movement, receives data from selected groups, and uploads it to the cloud server. During point-to-point movement, dynamic binding between points, monitoring groups, and fixed points enables "data accompaniment," ensuring the data stream always follows the target production batch. Data irrelevant to the current batch is automatically removed, improving the efficiency of acquiring data corresponding to fixed points and thus enhancing the efficiency of production parameter optimization. By directly transferring points between monitoring points, the system reduces end-to-end latency from event occurrence to data availability, enabling real-time monitoring of the production process and timely intervention in production anomalies, thereby improving the production quality of boron-containing steel. It should be noted that the process of moving a random point from its current monitoring point to the target monitoring point corresponding to the required information category based on the monitoring network is executed by inductive communication between multiple monitoring points in the monitoring network. Specifically, the random point sends a transfer request containing the identity information of the target monitoring point to its currently bound monitoring point; the current monitoring point establishes inductive communication with the target monitoring point based on the identity information and unbinds the random point; after verifying the access credentials of the random point, the target monitoring point establishes a binding relationship with it, completing the movement of the random point; when the number of random points does not match the number of monitoring points, if the number of random points in the current selection group is less than the total number of monitoring points in the monitoring group of the next process, one random point can be arbitrarily selected to send the identity information of multiple monitoring points corresponding to the next process, and the random point can be differentiated into multiple random points, and the total number of the differentiated random points (including itself) is sent to the management terminal. The number of identity information points for the next process at the current monitoring point is the same. The monitoring point corresponding to this point acquires the identity information carried by the current point and other points derived from it, and transfers the corresponding current point to the monitoring point corresponding to its identity information, so that it becomes a member of the monitoring group corresponding to the next process. If the number of current points in the current selection group is greater than the total number of monitoring points in the monitoring group of the next process, the excess current points (that is, current points whose identity information has not been assigned to the monitoring points required by the next process) are directly merged into other current points whose identity information has been assigned to the monitoring points required by the next process. Here, merging means combining two into one current point and transferring it to the monitoring point corresponding to the identity information. If the number of current points in the current selection group is equal to the total number of monitoring points in the monitoring group of the next process, the current points are directly transferred to the corresponding monitoring point. By transferring data from point to point on the monitoring network, the efficiency of data collection at the monitoring points is improved. Different information can be selected based on changes in fixed points, thereby optimizing the operation information of the next process. This allows the corresponding process operation information to be better formulated based on the production parameters of the previous process, thereby improving the production quality of boron-containing steel. By setting "access credentials" and "temporary binding", the security of data access can be guaranteed, and the on-demand allocation and release of resources can be realized, thereby improving resource utilization. The following scheme enables inductive communication between multiple monitoring points in a monitoring network: Multiple monitoring points on the production line are organized into a distributed hash table network, where each monitoring point acts as a node. Each monitoring point publishes its own identity and the data category information it can provide as key-value pairs and stores them in the distributed hash table network. When the first monitoring point needs to transmit a random data point, it uses the target data category carried by the random data point as the query key to route and query within the distributed hash table network, directly locating the second monitoring point responsible for that target data category. The first and second monitoring points establish a direct communication connection, performing unbinding and binding operations to complete the transmission from the first to the second monitoring point. In the query step, the target data category is hashed into a second hash value, and the distributed hash table network locates the network address of the second monitoring point by matching the second hash value. The distributed hash table network uses any one of the Chord, Kademlia, or Pastry protocols to organize nodes and implement key-value pair storage and retrieval. The query step is initiated independently by the first monitoring point and completed collaboratively across multiple hops in the distributed hash table network, without relying on the coordination of a central server. Specifically, the system acquires the categories of information collected from each monitoring point and the categories of initial information required by the target prediction model. These requirement categories are then sent to the management terminal, which in turn sends them to different random points. These random points move to their corresponding monitoring points, and the monitoring points sense each other, directly transmitting each random point from its current location to the target sensing point (unbinding the random point from itself and transmitting it to the next monitoring point, binding it to the next monitoring point based on the access credentials). The system then obtains information from the next monitoring point for subsequent process evaluation. The target monitoring points where multiple random points are moved are then identified as a monitoring group. Once a random point moves to the monitoring group, it directly acquires data. The data is transmitted to the management end. The corresponding monitoring group is found through the random point, and the random point is connected to the management end. The monitoring group corresponding to the random point is different depending on the process where the random point is located. Therefore, the location of the random point is different, which in turn leads to the monitoring point connected to the random point. The position of each random point on the monitoring network is determined according to the demand information of the target prediction model. The monitoring group is obtained according to the location. The data of the monitoring group is aligned and uniformly transmitted to the management end. After the corresponding monitoring point is determined, a temporary connection channel between the random point and the monitoring point is temporarily set up. When the target production batch moves to the next process, the current temporary channel is automatically released, and a new temporary channel is set up for the random point and the monitoring point corresponding to the next process.

[0021] S3: Real-time collection of production parameters of the current process on the production line based on monitoring points, and transmission of the production parameters of the current process to the three-dimensional model based on the transmission chain; The steps of collecting production parameters of the current process on the production line in real time based on monitoring points and transmitting these parameters to the 3D model based on the transmission chain include: collecting production parameters of the current process on the production line in real time; generating an information chain by acquiring production parameters of the same time node from various monitoring points based on a moving window; obtaining the current process corresponding to the current fixed position; taking the next process corresponding to the current process as the target process; determining the target prediction model corresponding to the target process from the 3D model; obtaining the initial demand information of the target prediction model; determining the target monitoring group corresponding to the target prediction model based on the initial demand information; assigning multiple random points to the target monitoring points of the corresponding target monitoring groups; establishing a temporary communication channel between the random points and the target monitoring points; the random points acquiring parameter information located on the information chain from the target monitoring points based on the temporary communication channel and transmitting it to the management end; and packaging the parameter information transmitted by multiple random points and transmitting it to the target prediction model based on the management end. The steps for determining the monitoring group corresponding to the target prediction model based on the initial demand information include: obtaining the demand information category corresponding to the initial demand information of the target prediction model, and sending the demand information category to different random points through the management terminal; moving the random points from the current monitoring points to the target monitoring points corresponding to the demand information categories based on the monitoring network; and combining multiple target monitoring points to obtain the monitoring group corresponding to the target prediction model. The specific steps for assigning multiple random points to target monitoring points of corresponding target monitoring groups include: assigning target monitoring points to each random point based on the management terminal; each random point sending a transfer request to its currently bound target monitoring point, wherein the transfer request contains the identity identifier of the target monitoring point to which it is assigned; the current monitoring point establishing sensor communication with the target monitoring point based on the identity identifier and unbinding the random point; the target monitoring verifying the access credentials of the random point, establishing a binding relationship with the random point, and assigning the random point to the corresponding target monitoring point.

[0022] Specifically, before a transfer request is issued by a monitoring point, the management terminal generates a temporary access credential for that monitoring point and sends it to the monitoring point along with the target monitoring point's identity information. The management terminal dynamically generates new demand categories based on the evaluation results of the target prediction model or the movement of production batches. This triggers a new round of sensing and transmission by the monitoring point to dynamically adjust the composition of the monitoring group. By establishing a data transmission chain between the monitoring points in the downstream process and the 3D model, the output status data is synchronized to the 3D model in real time and automatically, serving as the trigger input for the prediction model.

[0023] Specifically, the management end dynamically generates new demand categories based on the evaluation results of the target prediction model or the movement of production batches; triggers a new round of sensing and transmission at the monitoring points to dynamically adjust the composition of the monitoring group. The binding relationship between the monitoring points and the monitoring points is temporary, and at any given time, a monitoring point is bound to only one monitoring point. The monitoring points can communicate directly with each other and autonomously complete the handover of monitoring points, reducing transmission latency and alleviating the workload of the management end. The monitoring group can be dynamically generated and adjusted with the movement of demand and production batches, improving the adaptability of the production method and process. By directly sending the demand information category to the corresponding monitoring point, and then the monitoring point directly transmits the location of the monitoring point to the next corresponding monitoring point, the monitoring point can directly collect data information from the corresponding monitoring point, saving the time of finding the target monitoring point from multiple monitoring points. This facilitates the acquisition of information corresponding to the monitoring groups between different processes, improving the efficiency of acquiring monitoring group information. The information of the monitoring group changes according to the change of the product's location, thereby optimizing the operating parameters of the next process of the product and improving the efficiency of product processing.

[0024] S4: Determine the production parameters for the next process based on the 3D model, and send the production parameters of the next process to the corresponding production process in the production line for production control; The steps of determining the production parameters for the next process based on a 3D model and then sending these parameters to the corresponding production process on the production line for production control include: inputting the production parameters of the current process on the production line into the prediction model corresponding to the next process in the 3D model to obtain the production parameters for the next process; sending the production parameters for the next process to the physical actuators corresponding to the next process on the production line, and adjusting the operating status of the physical actuators according to the received production parameters to control the actual production of the next process. Specifically, based on preset product quality targets, preliminary production parameters are simulated, verified, and iterated in a 3D model to generate final optimized production parameters. The system receives adjustment instructions from users via an interactive interface for at least one parameter in the production process parameter set. In response to the adjustment instructions, the 3D model is rerun, and the microstructure evolution and mechanical property distribution displayed in the 3D model are updated in real time. This can be used for production parameter optimization, prediction through the simulation model, and the predicted process parameters are then distributed to the corresponding production processes for steel production. Specifically, multiple monitoring points are deployed on the production line to collect production parameters. A 3D model is constructed based on these parameters, and a prediction model is inserted between adjacent processes in the 3D model. A transmission chain is established between the processes on the production line and their corresponding prediction models. A selection cluster is set up along the transmission chain with the target production batch as the fixed point. Each selection cluster includes multiple random points that change with the fixed point. A corresponding monitoring cluster is selected based on the location of the fixed point, and the random points are placed on the corresponding monitoring clusters. Data from the monitoring clusters is collected simultaneously. Each monitoring point in the monitoring cluster corresponds to a memory used to store the data sequence collected by the monitoring point in chronological order. The window moves along the corresponding data sequence, collecting data at the same time point as an information chain sequence. At this time, the monitoring point collects parameter information from the corresponding monitoring point and is located on the information chain sequence, transmitting it to the management terminal. The parameter information collected from multiple monitoring points is packaged and uniformly transmitted to the corresponding prediction model in the 3D model. Based on the data package, the data for the next process is predicted. The data for the next process is predicted in real time based on the steel production data of the current process, and the prediction results are sent to the corresponding process to complete production. This allows the production data of each process to change in real time based on the data of the previous process and the current environmental data, improving the quality of steel production. Through real-time intervention, defects can be corrected before they occur (such as adjusting the reduction amount to close micro-cracks), reducing the flow of defects into the next process and their amplification, thereby reducing scrap rate and subsequent processing costs. Based on the actual data provided by the previous process (such as the actual temperature and slight compositional differences of the billet), corresponding process parameters are formulated for the current steel plate, effectively reducing the impact of upstream fluctuations on the final product quality and achieving precise and high-speed adaptive control.

[0025] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0026] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for producing high-strength boron-containing steel for marine engineering, characterized in that, Includes the following steps: Obtain the equipment parameters in the actual production line for producing boron-containing steel, and construct a 3D model based on the equipment parameters; The three-dimensional model is built in the cloud server, multiple monitoring points are set up on the physical equipment of the boron-containing steel production line, the multiple monitoring points are connected to form a monitoring network, and a transmission chain is built between the monitoring network and the cloud server. The production parameters of the current process on the production line are collected in real time based on monitoring points, and the production parameters of the current process are transmitted to the three-dimensional model based on the transmission chain. Based on the 3D model, the production parameters for the next process are determined, and these parameters are then sent to the corresponding production process on the production line for production control.

2. The method for producing high-strength boron-containing steel for marine engineering according to claim 1, characterized in that: The steps of obtaining equipment parameters in the actual production line for producing boron-containing steel and constructing a three-dimensional model based on the equipment parameters include: Obtain the equipment parameters of the boron-containing steel production line, including equipment layout information, equipment shape information, and equipment connection information; A three-dimensional model is obtained by performing three-dimensional modeling based on the equipment parameters, and the process equipment models corresponding to different production processes are identified in the three-dimensional model. For each pair of adjacent processes with a logical sequence, a pre-trained prediction model is integrated between their corresponding process equipment models. The prediction model is used to predict the process operation information of the downstream process based on initial information, which is the boron-containing steel process parameters at the start of the production line or the boron-containing steel process parameters output by the upstream process.

3. The method for producing high-strength boron-containing steel for marine engineering according to claim 1, characterized in that: The step of setting up multiple monitoring points on the physical equipment of the boron-containing steel production line and connecting the multiple monitoring points to form a monitoring network includes: Multiple monitoring points are set up on the physical equipment of the boron-containing steel production line. A moving window is configured for each monitoring point. The moving window is used to collect and cache the historical data of the corresponding monitoring point stored in time series. Configure a unified clock synchronization source for the moving windows to ensure that the time base of all moving windows is consistent; establish communication connections between multiple moving windows to connect multiple monitoring points to form a monitoring network.

4. The method for producing high-strength boron-containing steel for marine engineering according to claim 1, characterized in that: The steps for establishing a transmission link between the monitoring network and the cloud server include: Using the target production batch on the production line as the fixed point, a selection group and a management terminal are configured for the fixed point. The selection group includes multiple interconnected random points. A temporary communication channel is established between the random points and the monitoring point. Communication connections are established between the fixed point and the multiple random points. A transmission channel is established between the selected group and the cloud server based on the management terminal, and the transmission channel is used as the transmission link between the monitoring network and the cloud server.

5. The method for producing high-strength boron-containing steel for marine engineering according to claim 1, characterized in that: The step of collecting production parameters of the current process on the production line in real time based on monitoring points and transmitting the production parameters of the current process onto the three-dimensional model based on the transmission chain includes: Real-time collection of production parameters for the current process on the production line; information chain generated by acquiring production parameters at the same time point from various monitoring points based on a moving window. Obtain the current process corresponding to the current fixed position, take the next process corresponding to the current process as the target process, determine the target prediction model corresponding to the target process from the 3D model, obtain the initial demand information of the target prediction model, and determine the target monitoring group corresponding to the target prediction model based on the initial demand information. Multiple random points are assigned to the target monitoring points of the corresponding target monitoring group, and a temporary communication channel is established between the random points and the target monitoring points. The random points obtain parameter information located on the information chain from the target monitoring points based on the temporary communication channel and transmit it to the management terminal. The management terminal packages and transmits multiple parameter information transmitted from various points to the target prediction model.

6. The method for producing high-strength boron-containing steel for marine engineering according to claim 5, characterized in that: The steps for determining the monitoring group corresponding to the target prediction model based on initial demand information include: Obtain the demand information category corresponding to the initial demand information of the target prediction model, and send the demand information category to different random points through the management terminal; Based on the monitoring network, the monitoring point will be moved from the current monitoring point to the target monitoring point corresponding to the demand information category; multiple target monitoring points will be combined to obtain the monitoring group corresponding to the target prediction model.

7. A method for producing high-strength boron-containing steel for marine engineering according to claim 5, characterized in that: The specific steps for assigning multiple random points to the target monitoring points of the corresponding target monitoring group include: Based on the management terminal assigning target monitoring points to each random point, each random point sends a transfer request to its currently bound target monitoring point. The transfer request contains the identity identifier of the target monitoring point assigned to it. The current monitoring point establishes sensor communication with the target monitoring point based on the identity identifier, and then unbinds the random point; The target monitoring verifies the access credentials of the monitoring point, establishes a binding relationship with the monitoring point, and assigns the monitoring point to the corresponding target monitoring point.

8. The method for producing high-strength boron-containing steel for marine engineering according to claim 1, characterized in that: The step of determining the production parameters for the next process based on a 3D model and then distributing these parameters to the corresponding production process on the production line for production control includes: The production parameters of the current process on the production line are collected in real time and input into the prediction model of the next process in the 3D model to obtain the production parameters of the next process; The production parameters for the next process are sent to the corresponding physical actuator in the production line. The physical actuator adjusts its operating state according to the received production parameters to control the actual production of the next process.

Images