A method and device for tracking and predicting the quality of a whole-process furnace charge and a casting blank of a steel mill
By establishing a full-process quality tracking and prediction method in the steelmaking plant, and combining process control parameters and multi-source information, a decision tree is constructed for real-time anomaly judgment, which solves the problem of lagging quality judgment in the steelmaking plant and realizes rapid and accurate quality prediction and production optimization.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- WISDRI ENG & RES INC LTD
- Filing Date
- 2026-03-27
- Publication Date
- 2026-07-21
Smart Images
Figure CN122431265A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of quality control in steelmaking processes, and in particular to a method and apparatus for tracking and predicting the quality of all heats and billets in a steelmaking plant. Background Technology
[0002] The assessment of furnace batch and billet quality in steelmaking plants is a core aspect of production process control, directly impacting product delivery quality, production efficiency, and subsequent process costs. Production control at each stage permeates the entire billet production process, with the level of production control in smelting and casting directly determining the quality of molten steel and billets. Furthermore, rapid and accurate furnace batch and billet quality assessments are crucial for production and quality control in steelmaking. The assessment results not only reflect the level of production control in steelmaking processes, pinpoint weaknesses in the plant's quality control, and assist in optimizing process manufacturing and operational standards, but also directly determine the plant's production efficiency indicators. Many key production, technical, and quality indicators in steelmaking plants are inextricably linked to these assessments.
[0003] The existing technologies have the following main problems: First, the steel determination for each heat relies solely on the composition of representative samples without considering the control parameters of the smelting process, resulting in the failure to detect potential defects caused by internal factors in a timely manner; second, the billet quality tracking models configured in some steel plants only focus on anomalies in the continuous casting process and are not linked to the results of billet surface inspection and low-magnification inspection; third, although the surface inspection system can identify external defects, it cannot detect internal quality problems such as porosity and segregation; fourth, low-magnification experiments require sampling and sample preparation, and the result conveying machine is seriously lagging behind, which cannot meet the requirements for real-time determination; fifth, the hot-rolled direct-rolling process requires rapid and accurate steel determination, and the existing technologies are difficult to support efficient logistics.
[0004] Therefore, how to achieve real-time tracking and prediction of the entire process and significantly improve the accuracy and timeliness of steel judgment is a problem that needs to be solved by those skilled in the art. Summary of the Invention
[0005] This invention provides a method and apparatus for tracking and predicting the quality of heats and billets throughout the entire process of a steelmaking plant. It can solve problems in the prior art such as process control parameters not participating in quality judgment, lack of integration of multi-source information, lack of real-time tracking and prediction and closed-loop correction mechanism.
[0006] The first aspect of this invention provides a method for tracking and predicting the quality of heats and billets throughout the entire process of a steelmaking plant, including:
[0007] According to the steel grade group, separate rules for judging abnormalities in each heat and in each process section are established. The rules for judging abnormalities include the rules for judging abnormalities in quality control points and control parameters, as well as the corresponding severity levels and handling plans. Collect process control parameters for each stage of steelmaking and casting process control parameters for each stage of continuous casting. As the smelting process progresses, the collected control parameters are subjected to real-time anomaly determination based on anomaly determination rules to obtain individual anomaly information in both the furnace and billet dimensions. Based on the pre-generated quality tracking and prediction decision tree, the individual abnormal information is combined and judged in a rule-based manner to determine the comprehensive quality judgment result of the furnace batch and the billet. Based on the comprehensive quality assessment results and the corresponding handling plan for the heat and / or billet, the final handling instructions for the heat and / or billet are output.
[0008] Optionally, rules for judging abnormalities in each heat of each process and rules for judging abnormalities in each process segment of billet casting are established according to steel grade groups, including: Determine the set of tapping marks covered by each steel grade group; For each process in steelmaking, identify the quality control points and their control parameters, and define the exception code, exception content, exception judgment rules, severity level and handling plan for each control parameter. For each process segment of continuous casting, identify the quality control points and their control parameters, and define the exception code, exception content, exception judgment rules, severity level and handling plan for each control parameter.
[0009] Optionally, based on a pre-generated quality tracking prediction decision tree, a combined rule-based comprehensive judgment is made on individual anomaly information, including: Construct a multi-level branching structure consisting of sub-item exceptions and set-item exceptions. Set-item exceptions are generated from multiple sub-item exceptions according to preset logical combination rules. Based on the severity level of each individual anomaly, assign corresponding penalty points or weights; The individual abnormal information is combined and judged step by step according to the branch structure to obtain the comprehensive quality judgment result.
[0010] Optionally, the logical combination rules include AND, OR, and NOT logical operations, and nested combinations are allowed between set items.
[0011] Optionally, it also includes a step of feedback correction to the quality tracking prediction decision tree: Obtain at least one of the following: surface inspection results of the billet, low-magnification inspection results, manual judgment of abnormalities, and final judgment results; The obtained results are compared with the comprehensive quality judgment results output by the quality tracking prediction decision tree; Based on the comparison results, adjust the branch structure of the quality control points, anomaly judgment rules, or quality tracking prediction decision tree.
[0012] Optionally, the anomaly determination rules can be adjusted, including modifying the anomaly threshold range or rule expression of the control parameters; the branch structure of the quality tracking prediction decision tree can be adjusted, including adjusting the hierarchical relationship between sub-items and set items, the logical determination rules of set items, or the weight allocation between set items.
[0013] Optionally, the final disposal instructions for the heat and / or the billet can be output, including at least one of the following disposal methods: scrap, downgrade, finishing and repair, block pending judgment, or normal release.
[0014] A second aspect of the present invention provides a device for tracking and predicting the quality of heats and billets throughout the entire process of a steelmaking plant, comprising: The rule building unit is used to establish abnormal judgment rules for each heat of each process and abnormal judgment rules for each process section of billet according to steel grade group. The abnormal judgment rules include abnormal judgment rules for quality control points and control parameters, as well as the corresponding severity level and handling plan. The data acquisition unit is used to collect process control parameters for each steelmaking process and casting process control parameters for each continuous casting process section. The real-time judgment unit is used to perform real-time anomaly judgment on the collected control parameters based on anomaly judgment rules as the smelting process progresses, and obtain single-item anomaly information in the furnace and billet dimensions. The comprehensive judgment unit is used to perform a combined rule-based comprehensive judgment on individual abnormal information based on a pre-generated quality tracking and prediction decision tree, and to determine the comprehensive quality judgment result of the furnace batch and the billet. The instruction output unit is used to output the final disposal instruction for the heat and / or the billet based on the comprehensive quality judgment result and the disposal plan corresponding to the heat and / or the billet.
[0015] A third aspect of this invention provides a device for tracking and predicting the quality of heats and billets throughout the entire process of a steelmaking plant, comprising: One or more processors; A memory on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the steelmaking plant full-process furnace and billet quality tracking and prediction method as described in any of the above.
[0016] The fourth aspect of the present invention provides a computer storage medium for storing a program, which, when executed, is used to implement the method for tracking and predicting the quality of all heats and billets in a steelmaking plant as described in any of the preceding claims.
[0017] Beneficial effects: This invention incorporates process control parameters into the quality judgment system, and combines multi-source information such as surface inspection, low-magnification inspection and human experience to achieve comprehensive prediction of the entire process and multiple dimensions, overcoming the lag and one-sidedness of traditional methods that rely solely on result detection.
[0018] This invention performs real-time anomaly detection during the smelting process, enabling timely warnings and adjustments during production, reducing quality risks, and improving the pass rate of molten steel and billets.
[0019] This invention establishes a quality tracking and prediction decision tree based on a rule engine, which supports multi-level combinations and nested logic of sub-items and set items, and flexibly adapts to different steel types and complex steel judgment logic.
[0020] This invention introduces a feedback correction mechanism, which uses surface detection, low-magnification results, and final judgment results to continuously optimize the decision tree, ensuring that the model evolves with process updates.
[0021] This invention outputs clear disposal instructions (discard, downgrade, rework, blockade, etc.) to support dynamic adjustment of production plans and efficient logistics turnover. Attached Figure Description
[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0023] Figure 1 A flowchart illustrating a method for tracking and predicting the quality of heats and billets throughout the entire process of a steelmaking plant, provided as an embodiment of the present invention; Figure 2 This is a schematic diagram of a device for tracking and predicting the quality of heats and billets throughout the entire process of a steelmaking plant, provided in an embodiment of the present invention. Figure 3 This is a schematic diagram of the structure of a device provided in an embodiment of the present invention. Detailed Implementation
[0024] This invention provides a method and apparatus for tracking and predicting the quality of heats and billets throughout the entire smelting process in a steel plant. Based on the monitoring of abnormalities at quality control points throughout the entire smelting process, it targets the control parameters that affect the quality of heats and billets, closely examines the compliance of the smelting process control parameters, and combines the characterization results of heat quality and billet quality to build a rule structure tree for tracking and predicting the quality of heats and billets.
[0025] To facilitate understanding, the application scenarios of the embodiments of the present invention will be introduced first.
[0026] The quality assessment of steelmaking furnace runs and billets is particularly important. The results of quality assessment lead to measures such as downgrading, finishing, sealing, and verification of the billets to ensure product delivery quality. Simultaneously, it is linked to the closed-loop production plan delivery schedule, allowing for the rescheduling of under-production. Furthermore, the quality assessment results of continuously cast billets significantly impact the finished product quality of subsequent rolling mills. Comprehensive, accurate, and timely billet quality assessment can prevent billets with quality issues from being rolled back, eliminating wasted smelting costs in later processes.
[0027] The existing technology has the following drawbacks: (1) The process control parameters of the smelting process of each heat do not participate in the prediction of the quality of the heat and the billet. Often, only the representative sample composition of the heat participates in the steel judgment of the heat. In the manufacturing management system, the representative sample composition of the heat, the surface quality of the billet (human inspection or machine inspection, human inspection is definitely sampling inspection, and the timeliness of the conveyor cannot be guaranteed), the low magnification inspection results (definitely sampling inspection, and the timeliness of the conveyor cannot be guaranteed), and the expert experience are used to judge the steel manually. This is seriously lagging and unscientific. On the one hand, some billets that may be unqualified have already been hot-sent, which will transfer the quality risk to the downstream process. They may not be delivered according to the order requirements or there will be certain safety risks after use. On the other hand, even if some billets are produced and stacked in time, if there are quality disputes or rejections afterward, it will cause a certain amount of logistics and hoisting work.
[0028] (2) Some steel plants have configured billet quality tracking and prediction models to track the quality of the casting process and guide and optimize cutting. This model only focuses on abnormal events in the continuous casting process and does not link with billet surface inspection results or low-magnification inspection results for feedback correction.
[0029] (3) Some steel plants are equipped with billet surface quality inspection systems. First, the billet surface is photographed using sensors such as cameras and laser rangefinders. Then, production process or quality experts annotate the billet surface photos (defect identification and classification), and the annotated photos are used to train an AI model. Finally, the trained model is applied to online surface defect detection and identification of the billet. However, internal defects such as sparsity and segregation of the billet cannot be detected by images and can only be detected by low-magnification experiments after sampling. (4) Low-magnification experiments are only suitable for sampling inspections because they require sampling, sample preparation, testing, analysis, and then manual observation of the results for defect diagnosis. Finally, the timeliness of manually transmitting the defect diagnosis results to the machine is severely delayed. (5) For the hot-rolled billet process, it is necessary to determine the output furnace number and billet in a timely manner. Without a fast and accurate technology to track and predict the quality of the furnace number and billet, it is impossible to achieve efficient and accurate logistics reversal and ensure product qualification rate.
[0030] In summary, existing technologies cannot meet the needs of steel plants for comprehensive, rapid, and closed-loop quality tracking and prediction. There is an urgent need for a technology to predict the quality of steelmaking furnaces and billets throughout the entire process. The technical solution proposed in this invention uses steel grade groups as keywords, integrating anomaly detection of production process control parameters, billet surface inspection results, low-magnification inspection results, and manual anomaly detection results to establish a quality tracking and prediction rule structure tree, achieving quality tracking and prediction throughout the entire furnace and billet process. As the smelting process progresses, relevant quality control point parameters are collected, and quality results are predicted in real time.
[0031] See Figure 1 This diagram illustrates a process flow of a steelmaking plant's full-process heat batch and billet quality tracking and prediction method provided by an embodiment of the present invention. Multiple input sources are used, including production performance data, real-time production data, molten iron composition data, molten steel composition data, billet composition data, billet target data, billet machine inspection data, billet low-magnification data, and manually input data. Heat batch anomaly determination is achieved by performing anomaly determination on the above input data for each process, specifically including: anomaly determination for blast furnace molten iron, outputting blast furnace molten iron anomaly information; anomaly determination for desulfurization process, outputting desulfurization process anomaly information; anomaly determination for converter process, outputting converter process anomaly information; anomaly determination for refining process, outputting refining process anomaly information; anomaly determination for continuous casting process, outputting continuous casting process anomaly information; and anomaly determination for manually input data, outputting manually input anomaly information. The anomaly determination results for each of the above processes are summarized into a heat batch anomaly determination list.
[0032] The billet anomaly detection system performs regional anomaly detection for the continuous casting process, specifically including: anomaly detection for the ladle area, outputting ladle anomaly information; anomaly detection for the tundish area, outputting tundish anomaly information; anomaly detection for the crystallizer area, outputting crystallizer anomaly information; surface inspection anomaly detection based on billet machine inspection data, outputting surface inspection anomaly information; low-magnification inspection anomaly detection based on billet low-magnification data, outputting low-magnification anomaly information; and anomaly detection for manually input data, outputting manually input anomaly information. The anomaly detection results for each of these regions are summarized into a billet anomaly detection list.
[0033] Anomaly information from the furnace anomaly judgment list and the billet anomaly judgment list is input into the anomaly judgment decision tree. Based on pre-generated anomaly judgment rules, the decision tree performs a combined rule-based comprehensive judgment on each anomaly, determines the severity level of the anomaly, and provides a handling plan. Then, according to the handling results output by the decision tree, the corresponding anomaly handling operations are executed, and the handling results are archived. Finally, based on the billet surface inspection results, low-magnification inspection results, manual anomaly judgment results, and the final judgment result, the anomaly judgment rules and the anomaly judgment decision tree are corrected and optimized, forming a closed-loop quality tracking and prediction technology chain.
[0034] The method for tracking and predicting the quality of steelmaking furnaces and billets throughout the entire process provided in this embodiment of the invention can be implemented, for example, through the following steps S101-105.
[0035] S101: Establish abnormal judgment rules for each heat batch of each process and abnormal judgment rules for each process section of billet according to the steel grade group.
[0036] In this embodiment of the invention, the set of tapping marks covered by each steel grade group is determined; for each steelmaking process, quality control points and their control parameters are sorted out, and the abnormal code, abnormal content, abnormal judgment rules, severity level and handling plan of each control parameter are defined; for each continuous casting process section, quality control points and their control parameters are sorted out, and the abnormal code, abnormal content, abnormal judgment rules, severity level and handling plan of each control parameter are defined.
[0037] Specifically, according to steel grade groups, the rules for judging anomalies in each heat of steelmaking processes are compiled, including process, quality control point (control parameter), parameter traceability, anomaly code, anomaly content, anomaly rules, severity level, and handling plan. According to steel grade groups, the rules for judging billet anomalies in each process section of continuous casting are also compiled, including process section, quality control point (control parameter), parameter traceability, anomaly code, anomaly content, anomaly rules, severity level, and handling plan. Table 1 is an exemplary steel grade group configuration table provided by an embodiment of the present invention. Steel grades are grouped according to different characteristics, and each steel grade group corresponds to a set of steel tapping marks.
[0038] Table 1 Steel Grade Group Configuration Table
[0039] S102: Collect process control parameters for each steelmaking process and casting process control parameters for each continuous casting stage.
[0040] In this embodiment of the invention, the quality monitoring of each furnace step and the quality monitoring of each billet in each process segment are combined. First, the control parameter values of each process of desulfurization, converter, refining, and continuous casting, as well as each process segment of continuous casting, are collected.
[0041] S103: As the smelting process progresses, the collected control parameters are subjected to real-time anomaly determination based on anomaly determination rules to obtain individual anomaly information in both the furnace and billet dimensions.
[0042] In this embodiment of the invention, as the smelting process progresses, anomalies are determined for individual process parameters of each process step and individual process parameters of each process segment during the casting process.
[0043] Table 2 is an exemplary casting machine process point configuration table provided by an embodiment of the present invention. According to the process equipment through which the casting flow is generated, the entire casting flow is divided into process sections from the liquid surface of the crystallizer to the cutting car. By monitoring the control parameters of different process sections, quality abnormal events can be obtained.
[0044] Table 2. Casting Machine Process Point Configuration Table
[0045] S104: Based on the pre-generated quality tracking and prediction decision tree, perform a combined rule-based comprehensive judgment on individual abnormal information to determine the comprehensive quality judgment result of the furnace batch and the billet.
[0046] In this embodiment of the invention, a multi-level branching structure consisting of sub-item anomalies and aggregate anomalies is constructed. Aggregate anomalies are generated from multiple sub-item anomalies according to preset logical combination rules. Based on the severity level of each individual anomaly, corresponding penalty points or weights are assigned. Individual anomaly information is combined and judged level by level according to the branching structure to obtain a comprehensive quality judgment result. The logical combination rules include AND, OR, and NOT logical operations, and nested combinations between aggregate anomalies are allowed.
[0047] Specifically, a quality tracking prediction decision tree is established according to steel grade groups. The branch structure of the prediction decision tree is built based on a rule engine, mainly including the relationship between sub-items and set items. Set items can be understood as derived virtual anomalies. Finally, there is a comprehensive judgment based on multiple combination rules for single process anomaly determination. This combination method is based on the correlation of metallurgical mechanisms. The judgment logic for sub-items within a set item (AND, OR, NOT) allows for nesting of sets. Penalties and weights exist between sets, with smaller severity level values (1-10, 1 being the most severe, decreasing sequentially) resulting in larger penalties. See Tables 3, 4, and 5. Table 3 is an exemplary furnace anomaly code definition table provided in this embodiment of the invention; Table 4 is an exemplary billet anomaly code definition table provided in this embodiment of the invention; and Table 5 is an exemplary severity level and handling rule definition table provided in this embodiment of the invention.
[0048] Table 3 Furnace Exception Code Definition Table
[0049] Table 4 Definition of Cast Billet Abnormal Codes
[0050] Table 5. Definition of Severity Levels and Handling Rules
[0051] S105: Based on the comprehensive quality assessment results and the corresponding handling plan for the heat and / or billet, output the final handling instruction for the heat and / or billet.
[0052] In this embodiment of the invention, the final disposal instruction for the output furnace batch and / or billet includes at least one of the following disposal methods: scrapping, downgrading, finishing and repair, sealing for further judgment, or normal release.
[0053] In one implementation of this invention, at least one of the following is obtained: surface inspection results of the cast billet, low-magnification inspection results, manually determined anomaly results, and final judgment results. The obtained results are compared with the comprehensive quality judgment result output by the quality tracking and prediction decision tree. Based on the comparison results, the quality control points, anomaly judgment rules, or the branch structure of the quality tracking and prediction decision tree are adjusted. Adjusting the anomaly judgment rules includes correcting the anomaly threshold range or rule expression of the control parameters. Adjusting the branch structure of the quality tracking and prediction decision tree includes adjusting the hierarchical relationship between sub-items and set items, the logical judgment rules of set items, or the weight allocation between set items.
[0054] Specifically, the prediction technique involves comparing and correcting the results of surface inspection of the cast billet, low-magnification inspection, manual anomaly judgment, and final judgment with the results of the quality tracking and prediction decision tree. This involves adjustments to quality control points, anomaly rules, and the prediction decision tree. Quality control point adjustment refers to adding or removing quality control points, or modifying the traceability of quality control point parameters; anomaly rule adjustment refers to adjusting the anomaly threshold range or rule expression based on steel grade manufacturing standards and operational standards, combined with statistical analysis of historical furnace data; and prediction decision tree adjustment refers to adjusting the branch structure built on the rule engine, mainly including the relationship between sub-items and set items, set item judgment logic (AND, OR, NOT), and penalties and weights between set items.
[0055] Beneficial Effects: This invention proposes a method for tracking and predicting the quality of heats and billets throughout the entire process of a steelmaking plant. First, it establishes a process quality and output quality assessment system, including quality control points for heats and billets, quality control standards, anomaly judgment rules, and product disposal rules. Second, based on heat and billet production process data, it performs anomaly judgments on multiple quality control points, including anomaly codes, anomaly content, severity, and disposal methods. Finally, it comprehensively assesses multiple dimensions, including composition, process, and surface analysis, ultimately determining the disposal plan for the entire heat, a single billet, or multiple billets. This method is applicable to the entire process and product quality control scenarios of desulfurization → converter → refining → continuous casting, covering the quality control needs of different types of continuous casting billets and possessing broad industrial applicability.
[0056] Compared with the prior art, the present invention has the following advantages and positive effects: This invention establishes a technical framework for judging the quality of steelmaking furnace batches and cast billets, with quality control throughout the entire billet production process. Simultaneously, it incorporates billet surface quality, low-magnification inspection results, and expert experience as influencing factors in billet quality judgment.
[0057] This invention allows for real-time prediction of furnace or billet quality as the smelting process progresses, which helps to optimize and adjust control parameters during production, reduce and correct the risk of quality anomalies, and improve the pass rate of molten steel and billet.
[0058] This invention identifies the quality control points for each process, clarifies parameter traceability, exception codes, exception content, exception rules, severity levels, and handling solutions, enabling a very clear understanding of the quality control status of each process.
[0059] This invention establishes a quality tracking and prediction decision tree, defining the branch structure of the prediction decision tree according to steel type. It can arbitrarily combine single anomalies of different quality control points, billet surface quality, low-magnification inspection results, and expert experience to adapt to the complex steel judgment logic of different steel plants and different steel types.
[0060] This invention compares and corrects the results of billet surface inspection, low-magnification inspection, manual anomaly judgment, final judgment, and rolled material performance testing and analysis with the results of the quality tracking and prediction decision tree, and then adjusts the heat batch and billet quality judgment technology framework.
[0061] Based on the methods provided in the above embodiments, this invention also provides a device for tracking and predicting the quality of heats and billets throughout the entire process of a steelmaking plant. The following description, in conjunction with the accompanying drawings, introduces this device for tracking and predicting the quality of heats and billets throughout the entire process of a steelmaking plant.
[0062] See Figure 2 The figure is a schematic diagram of a steelmaking plant's full-process furnace batch and billet quality tracking and prediction device provided in an embodiment of the present invention.
[0063] The steelmaking plant full-process furnace batch and billet quality tracking and prediction device 200 provided in this embodiment of the invention includes: a rule construction unit 201, a data acquisition unit 202, a real-time judgment unit 203, a comprehensive judgment unit 204, and an instruction output unit 205.
[0064] The rule construction unit 201 is used to establish the abnormal judgment rules for each heat of each process and the abnormal judgment rules for each process section of billet according to the steel grade group. The abnormal judgment rules include the abnormal judgment rules for quality control points and control parameters, as well as the corresponding severity level and handling plan. The data acquisition unit 202 is used to collect process control parameters of each steelmaking process and casting process control parameters of each continuous casting process section; The real-time judgment unit 203 is used to perform real-time anomaly judgment on the collected control parameters based on anomaly judgment rules as the smelting process progresses, and obtain single-item anomaly information in the furnace and billet dimensions. The comprehensive judgment unit 204 is used to perform a combined rule-based comprehensive judgment on individual abnormal information based on a pre-generated quality tracking and prediction decision tree, and to determine the comprehensive quality judgment result of the furnace batch and the billet. The instruction output unit 205 is used to output the final disposal instruction for the heat and / or the billet based on the comprehensive quality judgment result and the disposal plan corresponding to the heat and / or the billet.
[0065] In one possible implementation, the rule construction unit 201 is specifically used for: Determine the set of tapping marks covered by each steel grade group; For each process in steelmaking, identify the quality control points and their control parameters, and define the exception code, exception content, exception judgment rules, severity level and handling plan for each control parameter. For each process segment of continuous casting, identify the quality control points and their control parameters, and define the exception code, exception content, exception judgment rules, severity level and handling plan for each control parameter.
[0066] In one possible implementation, the synthesis determination unit 204 is specifically used for: Construct a multi-level branching structure consisting of sub-item exceptions and set-item exceptions. Set-item exceptions are generated from multiple sub-item exceptions according to preset logical combination rules. Based on the severity level of each individual anomaly, assign corresponding penalty points or weights; The individual abnormal information is combined and judged step by step according to the branch structure to obtain the comprehensive quality judgment result.
[0067] In one possible implementation, the logical combination rules include AND, OR, and NOT logical operations, and nested combination is allowed between set item exceptions.
[0068] In one possible implementation, a correction unit is also included, having the ability to: Obtain at least one of the following: surface inspection results of the billet, low-magnification inspection results, manual judgment of abnormalities, and final judgment results; The obtained results are compared with the comprehensive quality judgment results output by the quality tracking prediction decision tree; Based on the comparison results, adjust the branch structure of the quality control points, anomaly judgment rules, or quality tracking prediction decision tree.
[0069] In one possible implementation, the anomaly determination rules are adjusted, including modifying the anomaly threshold range or rule expression of the control parameters; the branch structure of the quality tracking prediction decision tree is adjusted, including adjusting the hierarchical relationship between sub-items and set items, the logical determination rules of set items, or the weight allocation between set items.
[0070] In one possible implementation, the final disposal instructions for the heat and / or billet are output, including at least one of the following disposal methods: scrapping, downgrading, finishing and repair, blockade pending judgment, or normal release.
[0071] Since the steelmaking plant's full-process heat number and billet quality tracking and prediction device 200 is the same device as the steelmaking plant's full-process heat number and billet quality tracking and prediction method provided in the above method embodiments, the specific implementation of each unit of the steelmaking plant's full-process heat number and billet quality tracking and prediction device 200 is based on the same concept as the above method embodiments. Therefore, for the specific implementation of each unit of the device 200, please refer to the description of the steelmaking plant's full-process heat number and billet quality tracking and prediction method in the above method embodiments, and it will not be repeated here.
[0072] This invention also provides a device for tracking and predicting the quality of heats and billets throughout the entire process of a steelmaking plant, the device comprising: a processor and a memory; The memory is used to store instructions; The processor is used to execute the instructions in the memory to perform the steelmaking plant's full-process furnace batch and billet quality tracking and prediction method mentioned in the above embodiments.
[0073] It should be noted that the hardware structure of the steelmaking plant's full-process heat cycle and billet quality tracking and prediction equipment provided in the embodiments of the present invention can be as follows: Figure 3 The structure shown, Figure 3 This is a schematic diagram of the structure of a device provided in an embodiment of the present invention.
[0074] Please see Figure 3 As shown, device 300 includes: a processor 310, a communication interface 320, and a memory 330. The number of processors 310 in device 300 can be one or more. Figure 3 Taking a processor as an example, in this embodiment of the invention, the processor 310, communication interface 320, and memory 330 can be connected via a bus system or other means. Figure 3 Taking the connection between China and Israel via bus system 340 as an example.
[0075] Processor 310 may be a central processing unit (CPU), a network processor (NP), or a combination of a CPU and an NP. Processor 310 may further include hardware chips. These hardware chips may be application-specific integrated circuits (ASICs), programmable logic devices (PLDs), or combinations thereof. The PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), generic array logic (GAL), or any combination thereof.
[0076] The memory 330 may include volatile memory, such as random-access memory (RAM); the memory 330 may also include non-volatile memory, such as flash memory, hard disk drive (HDD) or solid-state drive (SSD); the memory 330 may also include a combination of the above types of memory.
[0077] Optionally, the memory 330 stores an operating system and programs, executable modules, or data structures, or subsets thereof, or extended sets thereof. The programs may include various operation instructions for implementing various operations. The operating system may include various system programs for implementing various basic business processes and handling hardware-based tasks. The processor 310 can read the programs in the memory 330 to implement the steelmaking plant's full-process furnace batch and billet quality tracking and prediction method provided in this embodiment of the invention.
[0078] The bus system 340 can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The bus system 340 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 3 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.
[0079] This invention also provides a computer-readable storage medium including instructions that, when run on a computer, cause the computer to execute the steelmaking plant's full-process furnace batch and billet quality tracking and prediction method mentioned in the above embodiments.
[0080] This invention also provides a computer program product containing instructions that, when run on a computer, causes the computer to execute the steelmaking plant's full-process furnace batch and billet quality tracking and prediction method mentioned in the above embodiments.
Claims
1. A method for tracking and predicting the quality of heats and billets throughout the entire process of a steelmaking plant, characterized in that, include: According to the steel grade group, separate rules for judging abnormalities in each heat of each process and rules for judging abnormalities in each process section of billet are established. The rules for judging abnormalities include the rules for judging abnormalities of quality control points and control parameters, as well as the corresponding severity levels and handling plans. Collect process control parameters for each stage of steelmaking and casting process control parameters for each stage of continuous casting. As the smelting process progresses, the collected control parameters are subjected to real-time anomaly determination based on the anomaly determination rules to obtain individual anomaly information in both the furnace batch and billet dimensions. Based on the pre-generated quality tracking and prediction decision tree, the individual abnormal information is combined and judged in a rule-based manner to determine the comprehensive quality judgment result of the furnace batch and the billet. Based on the comprehensive quality assessment results and the corresponding handling plan for the heat and / or the billet, the final handling instruction for the heat and / or the billet is output.
2. The method according to claim 1, characterized in that, The aforementioned rules for determining anomalies in each heat of each process and for determining anomalies in each process segment of billet casting are established according to steel grade groups, including: Determine the set of tapping marks covered by each steel grade group; For each process in steelmaking, identify the quality control points and their control parameters, and define the exception code, exception content, exception judgment rules, severity level and handling plan for each control parameter. For each process segment of continuous casting, identify the quality control points and their control parameters, and define the exception code, exception content, exception judgment rules, severity level and handling plan for each control parameter.
3. The method according to claim 1, characterized in that, The method of combining and comprehensively judging the individual anomaly information based on the pre-generated quality tracking prediction decision tree includes: Construct a multi-level branch structure consisting of sub-item exceptions and set-item exceptions, wherein the set-item exceptions are generated by multiple sub-item exceptions according to a preset logical combination rule; Based on the severity level of each individual anomaly, assign corresponding penalty points or weights; The individual abnormal information is combined and judged step by step according to the branch structure to obtain the comprehensive quality judgment result.
4. The method according to claim 3, characterized in that, The logical combination rules include AND, OR, and NOT logical operations, and nested combinations are allowed between the set items.
5. The method according to claim 1, characterized in that, It also includes a step of feedback correction to the quality tracking prediction decision tree: Obtain at least one of the following: surface inspection results of the billet, low-magnification inspection results, manual judgment of abnormalities, and final judgment results; The obtained results are compared with the comprehensive quality judgment results output by the quality tracking prediction decision tree; Based on the comparison results, adjust the branch structure of the quality control points, the anomaly judgment rules, or the quality tracking prediction decision tree.
6. The method according to claim 5, characterized in that, The adjustment of the anomaly determination rule includes modifying the anomaly threshold range or rule expression of the control parameters; the adjustment of the branch structure of the quality tracking prediction decision tree includes adjusting the hierarchical relationship between sub-items and set items, the logical determination rule of set items, or the weight allocation between set items.
7. The method according to claim 1, characterized in that, The final disposal instructions for the output furnace batch and / or billet include at least one of the following disposal methods: scrapping, downgrading, finishing and repair, sealing for further judgment, or normal release.
8. A device for tracking and predicting the quality of heats and billets throughout the entire process of a steelmaking plant, characterized in that, include: The rule construction unit is used to establish abnormal judgment rules for each heat of each process and abnormal judgment rules for each process section of billet according to the steel grade group. The abnormal judgment rules include abnormal judgment rules for quality control points and control parameters, as well as the corresponding severity level and handling plan. The data acquisition unit is used to collect process control parameters for each steelmaking process and casting process control parameters for each continuous casting process section. The real-time determination unit is used to perform real-time anomaly determination on the collected control parameters based on the anomaly determination rules as the smelting process progresses, and to obtain single-item anomaly information in the furnace and billet dimensions. The comprehensive judgment unit is used to perform a combined rule-based comprehensive judgment on the individual abnormal information based on the pre-generated quality tracking and prediction decision tree, and to determine the comprehensive quality judgment result of the furnace batch and the billet. The instruction output unit is used to output the final disposal instruction for the heat and / or the billet based on the comprehensive quality judgment result and the disposal plan corresponding to the heat and / or the billet.
9. An electronic device, characterized in that, The device includes: a processor and a memory; The memory is used to store instructions; The processor is configured to execute the instructions in the memory to perform the method according to any one of claims 1-7.
10. A computer-readable storage medium, characterized in that, Including instructions that, when run on a computer, cause the computer to perform the method described in any one of claims 1-7 above.