A full-process data tracing system for visualizing a quality map of a galvanizing line and an implementation method thereof

Abstract

The application discloses a full-process data tracing system and implementation method for galvanizing line quality spectrum visualization, belongs to the field of metallurgical industry automation and quality monitoring technology, and high-precision and high-frequency collection of galvanizing line quality parameters, sampling frequency is 100 Hz, data collection period is less than 50 ms, far exceeds the standard of 10 Hz in the industry, can accurately capture the instantaneous quality fluctuation of the strip under the running of the 270 m / min high-speed unit, and the collection accuracy of parameters such as hydrogen dew point and oxygen content meets the industrial high-end detection requirement, at the same time, the invalid area data of the loop is excluded, the effectiveness of the collected data is ensured, high-precision and high-frequency collection of the full-process quality parameters of the galvanizing line production is realized, the time domain data is accurately mapped to the length dimension of the strip, real-time visual monitoring is realized through the quality spectrum, a standardized report generation and double-dimension historical data tracing mechanism is constructed, and the intelligent and accurate level of the galvanizing line quality control is improved.

Description

Technical Field

[0001] This invention relates to a full-process data traceability system and implementation method for visualizing the quality spectrum of a galvanizing line. The invention belongs to the field of metallurgical industrial automation and quality monitoring technology, specifically to the field of a full-process data traceability system and implementation method for visualizing the quality spectrum of a galvanizing line. Background Technology

[0002] Galvanizing is a crucial process in the metallurgical industry. Process parameters such as strip thickness, zinc coating thickness, annealing temperature, and rolling force directly determine product quality. Full-process quality monitoring and data traceability are core elements for ensuring the stability of galvanized strip production and improving product qualification rates. Currently, quality monitoring systems in galvanizing production suffer from numerous technical deficiencies: The data acquisition accuracy is low and the sampling frequency is insufficient. The industry standard sampling frequency is only 10Hz, which cannot capture the instantaneous fluctuations in strip quality under the operation of a high-speed unit of 270m / min. Moreover, most of the data acquisition is time-dimensional, which is disconnected from the physical length dimension of the strip. The quality parameters cannot be accurately mapped to the specific location of the product, making it difficult to accurately locate quality problems. The measurement point layout lacks a systematic approach, fails to exclude invalid areas such as loopholes, results in low data validity, and multiple measurement point data are stored independently without a unified length axis aggregation mechanism, leading to fragmented quality data throughout the entire process and an inability to form a complete product quality archive. The visualization of quality data is poor, mostly displayed as a single value or simple curve, which cannot intuitively present the quality fluctuation pattern of the whole process. Furthermore, the response to anomaly identification is slow, and alarms cannot be triggered in time when process parameters exceed the tolerance, leading to the expansion of quality problems. The historical data traceability capability is insufficient, the data storage cycle is short and the format is inconsistent. There is a lack of a traceability mechanism that combines time and length, making it difficult to quickly review and trace the source after quality problems occur. At the same time, there is no standardized quality report generation function, and manual statistics are inefficient and prone to errors. The system has poor compatibility and scalability, complex integration with existing PLC control systems and material tracking systems on the production line, and cannot reserve interfaces for AI algorithms, making it difficult to achieve intelligent upgrades for quality prediction. Furthermore, the hardware is not capable of adapting to the high temperature and high interference environment of industrial sites, resulting in poor stability during continuous operation.

[0003] To address the aforementioned technical challenges, there is an urgent need to develop a quality monitoring system and implementation method that is compatible with the high-speed production process of galvanizing lines, enabling high-precision data acquisition, accurate length dimension mapping, real-time quality spectrum visualization, and full-process data traceability. This system would solve problems such as data disconnect, poor visualization, and difficulty in traceability in traditional systems, thereby improving the quality control level of galvanizing line production. Summary of the Invention

[0004] The technical problem to be solved by this invention is to overcome the existing defects and provide a full-process data traceability system and implementation method for visualizing the quality spectrum of galvanized lines. This system enables high-precision and high-frequency acquisition of quality parameters throughout the entire galvanized line production process, accurately maps time-domain data to the strip length dimension, achieves real-time visual monitoring through quality spectrum, and simultaneously constructs a standardized report generation and dual-dimensional historical data traceability mechanism to improve the intelligence and precision of galvanized line quality control.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A full-process data traceability system for visualizing the quality spectrum of galvanized wire includes a hardware layer and a software layer. The hardware layer includes a bus monitoring module, an optical fiber communication module, an XGN customized server, a display, a Siemens PLC active addressing communication interface, multimode optical fiber, and auxiliary materials. The software layer includes a length reference data mapping mechanism, a multi-measurement point distributed acquisition architecture, a data aggregation mechanism storage algorithm, a quality spectrum visualization system, a visualization report client, a historical data traceability client, and client software. The bus monitoring module is a BM-DP type PROFIBUS DIP network bus monitor, which establishes a millisecond-level communication link with the tracking PLC to achieve data fusion of heterogeneous devices. It supports two independent PROFIBUS lines with transmission rates ranging from 9.6 kbit / s to 12 Mbit / s. The fiber optic communication module supports the 32 Mbit Flex protocol and DMA direct memory access technology, allowing data to be directly written to the host memory, reducing CPU load. The XGN customized server is equipped with a W-2265 processor, 32 GB of memory, a 24 TB hard drive, and an Ubuntu 20.04 operating system. It also reserves an AI algorithm API interface for system operation support and full-process quality data storage. The system is compatible with the technical parameters of a galvanizing production line with a strip inspection width of 2030mm, a roll width of 2500mm, a strip thickness range of 0.4mm-2.5mm, a maximum unit speed of 270m / min, and a deviation of ±100mm. It has an installed capacity of 425W and uses a separate power supply according to the AC 220V standard for computer rooms. All components of the hardware layer are adapted to industrial field working environments ranging from 0°C to 50°C.

[0006] Furthermore, the multi-measurement point distributed acquisition architecture deploys measurement points at key locations such as the galvanizing line entrance section, furnace area, process section, and exit section. Each production line can define up to 28 measurement points, excluding invalid measurement points in looped areas. Each measurement point supports dual-position recording at the entrance / exit to analyze material deformation. This architecture collects over 200 process parameters in real time, including strip speed, weld signal, hydrogen dew point, oxygen content, exit thickness, rolling force, annealing temperature, and zinc layer thickness. The hydrogen dew point acquisition accuracy is ±0.5℃, the oxygen content acquisition resolution is 0.1%, the exit thickness sampling frequency is 100Hz, the overall data acquisition cycle is less than 50ms, and the sampling frequency can be set to a minimum of 100Hz, which meets the requirement of accurately capturing instantaneous fluctuations in high-speed strip production at the highest unit speed of 270m / min.

[0007] Furthermore, the data mapping mechanism of the length reference integrates the real-time collected time domain data and the location information provided by the material tracking system, accurately mapping the quality parameters scattered at different measurement points to the physical length coordinates of the strip steel. The standard length accuracy is 1m and the resolution can be adjusted according to process requirements. This mechanism is based on the tension roller speed, zone speed, strip tracking color number in the tracking PLC, hole finding instrument signal, inlet shearing signal, and outlet shearing signal for debugging. It is interfaced with the S7-400 PLC under the PCS7 system through DP substation hardware configuration and communication program. The supplier provides standard GSD files, communication program blocks and sample programs, and the client completes the hardware configuration and program writing.

[0008] Furthermore, the data aggregation mechanism storage algorithm controls each measurement point to independently record and collect data based on the time axis. After the steel coil completes the exit shearing, it automatically extracts related data from the temporary independent file according to the material ID and length position, and generates a *.dat data file bound to the final exit steel coil. All measurement values ​​belonging to this strip are stored as the length of the final product. The algorithm also supports parallel storage of incoming steel coil data. The coil number is obtained from the L1 system, and the file name contains information such as the coil ID and production date for easy traceability. When a steel coil has been cut multiple times, a suffix of ≤99 is automatically added to avoid file name conflicts and ensure the uniqueness of data storage.

[0009] Furthermore, the quality spectrum visualization system is the core visualization module at the front end of the system. With strip length as the horizontal axis and process parameters as the vertical axis, it renders the quality fluctuation curve of the entire process in real time and supports zooming to 1m accuracy to view local quality data, including strip shape, zinc layer thickness deviation, thickness fluctuation, etc. The system uses a layout manager to centrally configure interfaces for multiple scenarios. It can assign exclusive display schemes according to user roles and define the default startup layout. The layout elements adopt a dockable window design, which supports tab arrangement or free combination to adapt to the different monitoring needs of operators, technicians and managers. It also supports the synchronous integrated display of process data, quality indicators and event information, as well as offline trend comparison and historical event backtracking.

[0010] Furthermore, the visualization report client integrates multi-source data from the entire galvanizing line production process to form a structured dataset with in-depth analytical value. The structured dataset includes a basic information layer, a quality parameter layer, a process event layer, and a spatiotemporal coordinate layer. This client integrates a custom reporting engine that can automatically generate standardized PDF quality reports containing trend analysis, anomaly tracing, and quality evaluation without manual intervention. The reports can be printed directly or stored electronically as proof of product quality testing and traceability.

[0011] Furthermore, the historical data traceability client enables uninterrupted recording and long-term retention of quality data, breaking through the capacity limitations of traditional file storage. It supports data retention ranging from high-frequency collection at the millisecond level to long-term data retention over several years, and can retrieve details of all steel coils produced within the past three years. All recorded data is synchronized across the entire domain through a central timestamp, ensuring strict consistency of heterogeneous data across devices and production lines in the time dimension. It supports bidirectional traceability from both time and length dimensions, and allows for rapid zooming from an overview view to millisecond-level details. The visual timeline control enables hierarchical jumps from macro to micro perspectives.

[0012] Furthermore, the bus monitoring module is installed in the PLC cabinet, and the XGN customized server is located in the electrical room of the hot-dip galvanizing production line. The two are connected by FO / p2-30 multimode optical cable. The multimode optical cable is 62.5 / 125µm in size and uses ST type plug. The maximum transmission distance can reach 2000 meters without a repeater. Both the bus monitoring module and the fiber optic communication module adopt a passive cooling design. The bus monitoring module has an IP20 protection rating, supports DIN rail mounting, and can be stored at temperatures ranging from -25°C to 70°C, making it suitable for the complex environment of the galvanizing workshop industrial site.

[0013] A method for full-process data traceability based on the visualization of galvanizing line quality maps using the above system includes the following steps: S1. System Deployment and Configuration: Complete the physical deployment and electrical connection of each component in the hardware layer, complete the hardware configuration and communication program writing of the DP substation of S7-400PLC under PCS7 system, and complete the configuration settings of measurement points, quality parameters, acquisition thresholds and alarm thresholds in the system configuration dialog box. S2. Multi-source data distributed acquisition: A millisecond-level communication link is established between the bus monitoring module and the tracking PLC. The industrial protocol is parsed to realize the data fusion of heterogeneous equipment. Quality parameters are collected at each measurement point according to a preset cycle. Invalid data in the loop area is excluded. The collected data is temporarily and independently stored based on the time axis and bound to the steel coil ID. S3. Length dimension data mapping and aggregation: Through the length benchmark data mapping mechanism, the time domain acquisition data and material tracking location information are integrated to accurately map the quality parameters to the physical length coordinates of the strip steel. After the steel coil is sheared, the data aggregation mechanism storage algorithm extracts the associated data and generates a *.dat data file bound to the export steel coil, while storing the inbound steel coil data in parallel. S4. Real-time visualization and intelligent monitoring of quality data: The quality map visualization system renders the mapped length domain quality data in the form of a map in real time. The system performs real-time analysis of the collected data, completes data analysis in ≤1 second, identifies abnormal process parameters within 5 seconds and automatically triggers the alarm mechanism. Technicians intervene in a timely manner based on the map and alarm information. S5. Standardized report generation and dual-dimensional historical data traceability: After the steel coil is produced, the visual report client automatically generates a PDF quality report. The historical data traceability client enables long-term retention of quality data and supports millisecond-level detailed traceability of steel coil quality data within 3 years from both time and length dimensions, providing data support for quality review and process optimization.

[0014] Furthermore, in step S1, configure the two actual length values ​​for the inlet and outlet for each measurement point in the configuration dialog box, select the corresponding quality measurement value for each measurement point, determine the collection items of 200+ process parameters according to the requirements of the production department and the galvanizing workshop, set the system sampling frequency to 100Hz, and the data acquisition cycle to less than 50ms. At the same time, complete the display scheme allocation for user roles and the setting of abnormal alarm thresholds.

[0015] Furthermore, in step S4, the system achieves a 100% tracking rate for key process parameters such as rolling force and annealing temperature, and a problem resolution rate of over 95%. Through the AI ​​algorithm API interface reserved on the server, machine learning models can be accessed, historical quality data can be used as training samples to optimize the accuracy of quality prediction and achieve early warning of quality problems.

[0016] Compared with the prior art, the beneficial effects of the present invention are: This invention achieves high-precision, high-frequency acquisition of galvanizing line quality parameters, with a sampling frequency of 100Hz and a data acquisition cycle of less than 50ms, far exceeding the industry standard of 10Hz. It can accurately capture the instantaneous quality fluctuations of strip steel under the operation of a high-speed unit of 270m / min, and the acquisition accuracy of parameters such as hydrogen dew point and oxygen content meets the needs of high-end industrial testing. At the same time, it excludes invalid data from looper areas to ensure the validity of the acquired data. An innovative data mapping mechanism based on length benchmarks was designed, which for the first time accurately converts the quality parameters collected in the time domain to the strip length dimension. The standard length accuracy is 1m and can be adjusted as needed. The material elongation rate is taken into account during the mapping process to ensure that the quality data corresponds one-to-one with the physical location of the strip, thus solving the core technical problem of data disconnect between the traditional system and the product location. A distributed acquisition and length axis data aggregation architecture with multiple measurement points was constructed, supporting the deployment of 28 measurement points and dual entry / exit position recording. After the steel coil is cut, a *.dat data file bound to the product is automatically generated, with the file name bound to the steel coil ID, realizing centralized management and unique identification of quality data throughout the entire process, and providing a standardized data foundation for traceability. A length domain quality spectrum visualization system was developed, which intuitively presents the quality fluctuation of the entire process with the strip length as the horizontal axis. It supports local zoom viewing with 1m accuracy and realizes the synchronous display of process data, quality indicators and event information. At the same time, the system completes data analysis in ≤1 second and identifies anomalies and alarms within 5 seconds, which greatly improves the response speed of quality problems and the anomaly problem resolution rate is over 95%. A standardized report generation and dual-dimensional traceability mechanism has been established. The visual report client automatically generates PDF quality reports, and the historical data traceability client achieves full-domain data synchronization through a central timestamp. It supports dual-dimensional traceability of steel coil data within 3 years in terms of time and length, and can zoom from an overview to millisecond-level details, enabling rapid review and tracing of quality issues. The system's hardware and software are highly adapted to the galvanizing line production process. The hardware adopts a passive cooling design, adapting to industrial environments ranging from 0°C to 50°C. The bus monitoring module and server achieve 2000-meter repeater-free transmission via fiber optic cable, exhibiting strong anti-interference capabilities and enabling continuous trouble-free operation for 3 months. Simultaneously, the system reserves an AI algorithm API interface, allowing access to machine learning models for quality prediction. It seamlessly integrates with existing PCS7 and L1 systems, and the hardware includes expansion interfaces while the software supports configuration adjustments, demonstrating excellent compatibility and scalability. This invention constructs a complete quality control system for galvanizing lines, encompassing data acquisition, length mapping, real-time monitoring, report generation, and historical traceability. It replaces the traditional manual statistics and single-dimensional monitoring model, improving the intelligence and precision of quality control in galvanizing line production, effectively reducing the defect rate, increasing the product qualification rate, and possessing significant industrial application value and economic benefits. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall architecture of the system of the present invention; Figure 2 This is a schematic diagram illustrating the deployment of the multi-measurement point distributed acquisition architecture of the present invention; Figure 3 This is a schematic diagram illustrating the workflow of the data mapping mechanism for the length reference of this invention. Figure 4 This is a flowchart of the method for implementing the present invention. Detailed Implementation

[0018] 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.

[0019] Please see Figures 1-4 The present invention provides a technical solution: A full-process data traceability system and implementation method for visualizing the quality spectrum of a galvanizing line are applied to this production line, and the specific implementation is as follows: I. System Overall Architecture Setup The system of this invention includes a hardware layer and a software layer. The components are deployed collaboratively according to their functions. The specific setup requirements are as follows: 1. Hardware layer deployment (1) Bus monitoring module: Two BM-DP type PROFIBUS DAP network bus monitors are selected, equipped with two independent PROFIBUS interfaces (9-pin D-Sub plugs), with a transmission rate covering 9.6kbit / s to 12Mbit / s, automatic rate adaptation, protection level IP20, and DIN rail mounting in the PLC cabinet in the electrical room of the galvanizing workshop. It has a passive cooling design, an operating temperature of 0°C to 50°C, and a storage temperature of -25°C to 70°C. (2) Fiber optic communication module: One 2I / O fiber optic communication card is selected, which has two fiber optic input / output ports, supports 32MbitFlex bidirectional data transmission protocol and DMA direct memory access technology, and writes data directly into the host memory to reduce CPU load. It has a passive cooling design and an operating temperature of 0°C to 50°C. (3) XGN customized server: 1 unit, located in the electrical room of the hot-dip galvanizing production line, with hardware configuration of W-2265 processor, 32G memory, 24TB hard disk, 2 gigabit network ports, pre-installed Ubuntu20.04 operating system, and reserved AI algorithm API interface for system operation support and full-process quality data storage; (4) Display: One 27" LCD display, which serves as the system's human-computer interaction terminal and is connected to the XGN customized server; (5) Siemens PLC active addressing communication interface: 1 customized model to achieve seamless connection with the S7-400 PLC under the L1 system and PCS7 system of the production line; (6) Multimode optical cable: 2 FO / p2-30 multimode optical cables, 62.5 / 125µm specification, using ST type plug, connecting the bus monitoring module and the XGN customized server, with a maximum transmission distance of 2000 meters without repeaters; (7) Auxiliary materials: including power supply cables, fixing brackets, connecting connectors, etc., to complete the electrical connection and physical fixation of each component in the hardware layer; (8) Power supply: The total installed capacity of the system is 425W. It adopts a separate power supply according to the AC 220V computer room standard, which is isolated from the power supply of other equipment on the production line to improve the stability of power supply.

[0020] 2. Software layer configuration Each module in the software layer is configured and deployed based on the XGN customized server, including the length benchmark data mapping mechanism, multi-measurement point distributed acquisition architecture, data aggregation mechanism storage algorithm, quality spectrum visualization system, visualization report client, historical data traceability client and client software. The configuration process is as follows: (1) The I / O manager is used to centrally configure the measurement points, signals, and signal groups. The automatic detection function is used to identify the connected hardware and insert it into the configuration table. Loose loop areas are excluded. A total of 16 measurement points are deployed in the inlet section, furnace area, process section, and outlet section. Figure 2 At position m (less than the maximum of 28 that can be defined), reserve ≥10% of backup data access points; (2) Configure two actual length values ​​for the inlet and outlet for each measurement point to analyze the material deformation. According to the requirements of the production department and the galvanizing workshop, determine the collection items of 200+ process parameters, including strip speed, weld signal, hydrogen dew point (accuracy ±0.5℃), oxygen content (resolution 0.1%), outlet thickness (sampling frequency 100Hz), rolling force, annealing temperature, zinc layer thickness, etc. (3) Set the system sampling frequency to 100Hz, the data acquisition period to less than 50ms, and configure abnormal alarm thresholds, such as annealing temperature > set value ±5℃, zinc layer thickness deviation ±0.1mm, strip thickness deviation ±0.05mm, etc.; (4) Complete the hardware configuration and communication program writing of the DP substation of S7-400PLC under PCS7 system, import the standard GSD file, communication program block and sample program provided by Party B, realize the docking with the material tracking system and L1 system, and ensure the automatic acquisition of steel coil ID and real-time synchronization of material location information. (5) Assign display schemes for the quality map visualization system according to user roles (operators, technicians, managers), define the default startup layout for each role, set layout elements as dockable windows, and support tab arrangement and free combination.

[0021] II. Working Principle of the System Core Module 1. Multi-measurement point distributed acquisition architecture This architecture establishes a millisecond-level communication link between the DP bus and the tracking PLC (MTRPLC). The bus monitoring module parses the industrial protocol of the control system to achieve data fusion of heterogeneous equipment. Eighteen measurement points independently collect quality parameters according to a preset acquisition cycle. The measurement points in the inlet section collect strip speed, weld signal, and hole finding instrument signal. The measurement points in the furnace area collect annealing temperature, hydrogen dew point, and oxygen content. The measurement points in the process section collect rolling force and tension. The measurement points in the outlet section collect strip thickness, zinc layer thickness, and deviation. During the data acquisition process, invalid measurement data from the looper area is automatically excluded, and valid data is temporarily and independently stored on the XGN customized server based on the time axis. At the same time, the steel coil number is obtained from the L1 system and bound to the acquired data to provide a unique product identifier for subsequent data aggregation and traceability. All acquired process parameters meet the accuracy requirements, and the export thickness sampling frequency is 100Hz, which can accurately capture thickness fluctuations under high-speed production.

[0022] 2. Data mapping mechanism for length reference This mechanism is the core innovative module of the system. During the debugging phase, it connects to the tension roller speed, area speed, the strip tracking color number (ID jump signal) in the tracking PLC, the hole finding instrument signal, the inlet shear signal, and the outlet shear signal to complete the calibration. During operation, the system integrates time-domain acquired data with location information provided by the material tracking system in real time, accurately mapping the quality parameters (thickness, temperature, rolling force, zinc layer thickness, etc.) scattered at 18 measurement points to the physical length coordinates of the strip steel. The standard length accuracy is 1m. The material elongation rate is taken into account during the mapping process to ensure that the quality data corresponds one-to-one with the position of the strip steel. This mechanism seamlessly integrates with the S7-400 PLC under the PCS7 system, enabling precise conversion of time-domain data to length-domain data and solving the problem of data disconnect between traditional systems and the physical location of products.

[0023] 3. Data aggregation mechanism and storage algorithm Each measurement point independently records and collects data based on the time axis. When the steel coil is completed and exited after shearing, the algorithm is triggered. The algorithm automatically extracts all the quality data of the steel coil from the temporary storage file of each measurement point according to the material ID and length position, writes it into a new *.dat data file, and binds all quality values ​​to the length of the final exported steel coil. It also supports parallel storage of incoming steel coil data to meet the needs of full-process quality analysis. The file name is named according to "steel coil ID + production date", such as "HG20250704001+20250704". When the steel coil has been cut multiple times, a suffix (such as 01, 02) is automatically added to avoid file name conflicts. The generated *.dat data files are uniformly stored in the 24TB hard drive of the XGN customized server, providing a standardized data foundation for historical data traceability.

[0024] 4. Quality Spectrum Visualization System This system is the core front-end visualization module. With strip length as the horizontal axis and process parameters as the vertical axis, it renders the quality fluctuation curve of the entire process in real time. Operators can view the real-time graph through a 27" LCD monitor and zoom to 1m accuracy to view local quality data, such as zinc layer thickness deviation and plate shape defects in a certain 1m section. The system uses a layout manager to centrally configure interfaces for multiple scenarios. The operator view focuses on real-time quality fluctuations and abnormal alarms, the technician view focuses on process parameter trends and offline comparisons, and the management view focuses on quality pass rates and report summaries. The layout elements adopt a dockable window design, supporting tab arrangement or free combination. It also supports the synchronous integrated display of process data, quality indicators, and event information, as well as offline trend comparisons and historical event backtracking.

[0025] 5. Visual Reporting Client This client integrates multi-source data from the entire production process to form a structured dataset with a four-layer structure: (1) Basic information layer: includes metadata such as steel coil ID, production batch, strip steel specifications, and production team, to achieve accurate association between products and production processes; (2) Quality parameter layer: covers key quality indicators of each measurement point, supports multi-dimensional presentation of raw values, average values, extreme values, etc., and realizes logical aggregation of data through signal group classification; (3) Process event layer: embed key event records such as equipment start-up and shutdown, parameter adjustment, and alarm triggering, and link them with the quality data of the corresponding time / length nodes to build an "event-quality" correlation analysis model; (4) Spatiotemporal coordinate layer: Synchronously records the timestamp of data collection and the product length coordinates, supporting traceability in both time and length dimensions; The client integrates a custom reporting engine, which automatically generates a standardized PDF quality report containing trend analysis, anomaly tracing, and quality evaluation after the steel coil comes off the production line. No manual intervention is required, and the report can be printed directly or stored electronically.

[0026] 6. Historical Data Tracing Client This client enables uninterrupted recording and long-term retention of quality data, breaking through the capacity limitations of traditional file storage. It supports data retention ranging from high-frequency collection at the millisecond level to long-term data retention over several years, and can review the details of all steel coils produced within the past 3 years. All recorded data is synchronized across the entire domain through a central timestamp, ensuring strict consistency of heterogeneous data across devices and production lines in the time dimension. It supports bidirectional traceability from the time dimension (such as production quality fluctuations from 8:00 to 10:00 on July 4, 2025) and the length dimension (such as parameter deviations in the 100-150m segment of steel coil HG20250704001). It can quickly zoom from the overview view to millisecond-level details, and realize hierarchical jumps from "macro" to "micro" through a visual timeline control. After a quality problem occurs, technicians can quickly review and trace the source to locate the cause of the problem.

[0027] III. Full-process data traceability implementation method for visualizing the quality spectrum of galvanizing lines The implementation method of this invention is based on the above system architecture and follows the full-process logic of "data acquisition - mapping aggregation - real-time monitoring - report generation - historical tracing". The specific steps are as follows: S1. System Deployment and Configuration First, complete the physical deployment and electrical connection of each component in the hardware layer. Place the XGN customized server in the electrical room of the hot-dip galvanizing production line, install the bus monitoring module in the PLC cabinet, realize fiber optic communication between the two through multimode optical cable, connect the display and the Siemens PLC active addressing communication interface, and complete the separate AC 220V power supply for the system. Subsequently, the software configuration was completed. Under the PCS7 system, the hardware configuration and communication program writing of the DP substation of the S7-400 PLC were completed, and the standard GSD file, communication program block and sample program were imported. In the system configuration dialog box, the names of 18 measurement points, material tracking control signals, material IDs and actual length positions were entered. For each measurement point, two actual length values ​​were configured for the inlet and outlet. The corresponding mass measurement values ​​for each measurement point were selected. The sampling frequency was set to 100Hz and the data acquisition period was less than 50ms. The abnormal alarm threshold was configured. The display scheme allocation of user roles and the classification of signal groups were completed.

[0028] S2, Distributed Acquisition of Multi-Source Data After the system starts up, the bus monitoring module establishes a millisecond-level communication link with the tracking PLC, parses the industrial protocol to realize the data fusion of heterogeneous equipment; 18 measurement points collect corresponding process parameters according to the preset cycle, the hydrogen dew point collection accuracy is controlled within ±0.5℃, the oxygen content collection resolution is 0.1%, and the outlet thickness sampling frequency is 100Hz. During the data collection process, invalid data in the loop area is automatically excluded, and valid data is temporarily and independently stored on the XGN customized server based on the timeline. At the same time, the steel coil number is obtained from the L1 system and bound to the collected data in real time to ensure that each data has a unique product identifier.

[0029] S3, Length Dimension Data Mapping and Aggregation The data mapping mechanism of the length benchmark integrates the time-domain acquired data and the location information of the material tracking system in real time, accurately mapping the quality parameters of each measurement point to the physical length coordinates of the strip steel. The length accuracy is controlled to 1m. The material elongation rate is considered in the mapping process to ensure that the data corresponds one-to-one with the position of the strip steel. Once the steel coil is produced and sheared at the exit, a data aggregation mechanism and storage algorithm are triggered. Based on the material ID and length position, the algorithm extracts all quality data of the steel coil from the temporary storage files of each measurement point, generating a *.dat data file bound to the exit steel coil. All measurement values ​​belonging to this strip are stored as the length of the final product. At the same time, the entry steel coil data is stored in parallel. The file name is bound to the steel coil ID and production date. A suffix is ​​automatically added during multiple shearing operations to ensure the uniqueness of data storage.

[0030] S4. Real-time visualization and intelligent monitoring of quality data The quality map visualization system uses the mapped length domain quality data, with strip length as the horizontal axis and process parameters as the vertical axis, to render the quality fluctuation curve of the entire process in real time. Operators can view the real-time map on the monitor and zoom in to 1m to view local quality data, quickly identifying problems such as strip shape and zinc layer thickness deviation. The system performs real-time analysis on the collected quality data, with an analysis response time of ≤1 second. When process parameters are detected to be out of tolerance (such as excessive annealing temperature or zinc layer thickness deviation), the system identifies the anomaly within 5 seconds and automatically triggers an audible and visual alarm mechanism. At the same time, it records the timestamp, length coordinates, and parameter deviation values ​​of the abnormal event. Technicians intervene and adjust process parameters in a timely manner based on real-time graphs and alarm information. The system achieves a 100% tracking rate for key process parameters such as rolling force and annealing temperature, and a problem resolution rate of over 95%. At the same time, through the AI ​​algorithm API interface reserved on the XGN customized server, machine learning models are connected, and historical quality data is used as training samples to optimize the accuracy of quality prediction and achieve early warning of quality problems.

[0031] S5, Standardized Report Generation and Two-Dimensional Historical Data Tracing After the steel coil comes off the production line, the visual reporting client automatically generates a PDF quality report based on the structured dataset. The report includes basic information about the steel coil, trend analysis of quality parameters at each measurement point, source tracing of abnormal events, quality evaluation, etc. It can be printed directly or stored electronically as a product quality inspection certificate. The historical data traceability client stores *.dat data files long-term on XGN's customized server's 24TB hard drive, enabling detailed review of all steel coils produced within the past three years. It supports traceability from both time and length dimensions, and achieves full-domain data synchronization through a central timestamp. The system allows for rapid zooming from an overview view to millisecond-level details, providing comprehensive data support for quality issue review, process optimization, and customer quality traceability. Simultaneously, the system supports offline trend comparison, allowing for comparative analysis of quality data from different batches of steel coils to uncover production patterns and continuously improve production processes.

[0032] IV. System Acceptance and Performance Assurance The system of this invention will undergo acceptance testing within 3 months of its commissioning. The testing follows the principle of confirmation of startup by representatives from both parties on-site. If the initial testing fails to meet the standards, a second testing can be conducted within a specified time. The testing content includes the accuracy of the correlation between quality data and length, sampling frequency and high-speed capture capability, full production line data coverage and parameter tracking rate, real-time analysis and anomaly handling capability, system continuous operation stability, and quality graph and report output capability. In this embodiment, all testing indicators have met the preset requirements. Accuracy of correlation between quality data and length: average deviation ≤ 1m, meeting the standard; Sampling frequency and high-speed capture capability: When the production line is running at 270m / min, the sampling frequency is stable at 100Hz, and the artificially simulated instantaneous thickness deviation of ±0.1mm is completely captured. The timestamp accuracy is 8ms≤10ms, which meets the standard. Full production line data coverage and parameter tracking rate: Full coverage of 18 measurement points throughout the entire process; 10 sets of process parameters were randomly selected, with a tracking rate of 100%; the proportion of backup data access points was ≥10% (15%), meeting the standard. Real-time analysis and anomaly handling capabilities: The system identifies 5 typical process anomalies in advance, with an identification time of ≤4 seconds for each anomaly. The anomaly resolution rate is ≥95% within 10 consecutive days, and the anomaly feature value extraction error is ≤3% to 5%, meeting the standards. System continuous operation stability: It has been running without failure for 90 consecutive days, without any crashes, data interruptions or other anomalies. All data collected within 3 months has been stored completely without any file corruption or loss, which meets the standard. Quality graph and report output capabilities: The quality graph supports 1m precision scaling, the length axis and parameter curves correspond accurately, and PDF reports are automatically generated, with complete content, accurate data, and compliance with standards.

[0033] After the system passes the assessment and acceptance, the quality guarantee period is 12 months. During the warranty period, the seller will dispatch professional technicians to provide technical support, including equipment maintenance, troubleshooting, and software upgrades. At the same time, the system has good scalability and can increase the number of measurement points, expand the process parameter collection items, and connect new AI algorithm models according to the process upgrade needs of the production line, so as to adapt to the long-term development needs of galvanizing line production.

[0034] V. Application Results The full-process data traceability system and implementation method for visualizing the quality spectrum of galvanizing lines, as described in this invention, has achieved significant application results after being put into use at the galvanizing line of Handan Iron and Steel Group Co., Ltd. The accuracy and efficiency of quality parameter acquisition have been greatly improved, enabling precise capture of strip steel quality fluctuations under high-speed production of 270m / min, and improving the effectiveness of the acquired data by 30%. It achieves precise binding between quality data and strip length dimension, reducing the time for locating quality problems from several hours to several minutes, with a positioning accuracy of 1m; The quality visualization system makes quality fluctuations throughout the entire process readily visible, significantly improving anomaly identification and response speed. The alarm time for out-of-tolerance process parameters is ≤5 seconds, and the anomaly resolution rate has increased from 80% to 98%. Standardized PDF quality reports are automatically generated, replacing the traditional manual statistical method, improving report generation efficiency by 90% and reducing the error rate to 0. The dual-dimensional historical data traceability mechanism enables millisecond-level traceability of steel coil data within 3 years, improving the efficiency of quality problem review and source tracing by 80% and providing complete data support for process optimization. The system has strong continuous operation stability and can run without failure for 3 months. It can adapt to the complex environment of industrial sites, seamlessly connect with existing production line systems, and the reserved AI algorithm interface lays the foundation for subsequent intelligent upgrades. After being put into use, the strip steel defect rate of the galvanizing line decreased by 25%, the product qualification rate increased by 3%, and it can create significant economic benefits every year. At the same time, it improves the intelligent level of enterprise quality control and has broad industrial promotion value.

[0035] 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 full-process data traceability system for visualizing the quality spectrum of a galvanizing line, characterized in that, It includes a hardware layer and a software layer. The hardware layer includes a bus monitoring module, an optical fiber communication module, an XGN customized server, a display, a Siemens PLC active addressing communication interface, multimode optical fiber and auxiliary materials. The software layer includes a length reference data mapping mechanism, a multi-measurement point distributed acquisition architecture, a data aggregation mechanism storage algorithm, a quality spectrum visualization system, a visualization report client, a historical data traceability client and client software. The bus monitoring module is a BM-DP type PROFIBUSDP network bus monitor, which establishes a millisecond-level communication link with the tracking PLC to realize data fusion of heterogeneous devices; the fiber optic communication module supports 32Mbit Flex protocol and DMA direct memory access technology; the XGN customized server is equipped with a W-2265 processor, 32G memory, 24TB hard drive and Ubuntu 20.04 operating system, and reserves an AI algorithm API interface. The system is compatible with the technical parameters of a galvanizing production line with a strip detection width of 2030mm, a roll width of 2500mm, a strip thickness range of 0.4mm-2.5mm, a maximum unit speed of 270m / min, and a deviation of ±100mm. The installed capacity is 425W, and it uses a separate power supply according to the AC 220V standard for the computer room.

2. The system according to claim 1, characterized in that, The multi-measurement point distributed acquisition architecture deploys measurement points in the galvanizing line entrance section, furnace area, process section, and exit section. Each production line can define up to 28 measurement points, excluding invalid measurement points in loop areas. Each measurement point supports dual-position recording at the entrance and exit. The architecture collects more than 200 process parameters in real time, including hydrogen dew point with an accuracy of ±0.5℃, oxygen content with a resolution of 0.1%, outlet thickness with a sampling frequency of 100Hz, and an overall data acquisition cycle of less than 50ms. The minimum sampling frequency can be set to 100Hz.

3. The system according to claim 1, characterized in that, The data mapping mechanism of the length reference integrates time domain data and material tracking location information to accurately map quality parameters to the physical length coordinates of the strip steel. The standard length accuracy is 1m and resolution adjustment is supported. This mechanism is based on tension roller speed, zone speed, tracking PLC strip color number tracking, hole finding instrument signal, and shearing signal to complete the debugging. It is connected to the S7-400 PLC under the PCS7 system through DP substation hardware configuration and communication program.

4. The system according to claim 1, characterized in that, The data aggregation mechanism storage algorithm controls each measurement point to independently record data based on the time axis. After the steel coil is cut at the exit, the data is automatically extracted according to the material ID and length position to generate a *.dat data file bound to the exit steel coil. At the same time, it supports parallel storage of the inlet steel coil data. The file name contains the coil ID and production date information. When cutting multiple times, a suffix of ≤99 is automatically added to avoid conflicts. The coil number is obtained from the L1 system.

5. The system according to claim 1, characterized in that, The quality graph visualization system uses strip length as the horizontal axis and process parameters as the vertical axis to render the quality fluctuation curve of the entire process in real time, and supports zooming to 1m accuracy to view local quality data. The system enables centralized configuration of multi-scenario interfaces through a layout manager, assigns display schemes according to user roles, and adopts a dockable window design for layout elements, supporting tab arrangement or free combination. It also supports the synchronous integrated display of process data, quality indicators, and event information, as well as offline trend comparison.

6. The system according to claim 1, characterized in that, The visualization report client integrates multi-source data to form a structured dataset, which includes a basic information layer, a quality parameter layer, a process event layer, and a spatiotemporal coordinate layer. This client integrates a custom reporting engine that can automatically generate standardized, high-quality PDF reports that include trend analysis and anomaly tracing.

7. The system according to claim 1, characterized in that, The historical data traceability client enables uninterrupted recording and long-term retention of quality data, and allows for review of details of all steel coils produced within the past 3 years. All recorded data is synchronized across the entire domain through a central timestamp, supporting traceability in both time and length dimensions. It allows for rapid zooming from an overview view to millisecond-level details, enabling "macro-micro" hierarchical jumps.

8. The system according to claim 1, characterized in that, The bus monitoring module is installed in the PLC cabinet, and the XGN customized server is located in the electrical room of the hot-dip galvanizing production line. The two are connected by a 62.5 / 125µm FO / p2-30 multimode optical cable with an ST type plug. The maximum transmission distance can reach 2000 meters without a repeater. Both the bus monitoring module and the fiber optic communication module adopt a passive cooling design, with an operating temperature of 0°C to 50°C. The bus monitoring module has an IP20 protection rating, supports DIN rail mounting, and has a storage temperature of -25°C to 70°C.

9. A method for full-process data traceability based on visualization of galvanizing line quality maps using the system described in any one of claims 1-8, characterized in that, Includes the following steps: S1. System Deployment and Configuration: Complete the physical deployment and electrical connection of the hardware layer, complete the hardware configuration and communication program writing of the DP substation with S7-400PLC, complete the configuration settings of measurement points, quality parameters, acquisition thresholds and alarm thresholds in the configuration dialog box, and configure two actual length values ​​for the inlet and outlet for each measurement point. S2. Multi-source data distributed acquisition: A millisecond-level communication link is established between the bus monitoring module and the tracking PLC. Quality parameters are collected at each measurement point according to a preset cycle. Invalid data in the loop area is excluded. The collected data is temporarily and independently stored based on the time axis and bound to the steel coil ID. S3. Length dimension data mapping and aggregation: The time domain data is mapped to the strip length coordinate through the length benchmark data mapping mechanism. After the steel coil is cut, the data aggregation mechanism stores the related data through the storage algorithm to generate a *.dat data file bound to the exported steel coil, while storing the imported steel coil data in parallel. S4. Real-time visualization and intelligent monitoring of quality data: The quality map visualization system renders the length domain quality map in real time. The system completes data analysis in ≤1 second, identifies abnormal process parameters within 5 seconds and automatically triggers alarms. The tracking rate of key process parameters is 100%, and the resolution rate of abnormal problems is ≥95%. The system can be connected to machine learning models through the AI ​​algorithm API interface to achieve quality prediction. S5. Standardized report generation and dual-dimensional historical data traceability: After the steel coil is produced, the visual report client automatically generates a PDF quality report, and the historical data traceability client enables millisecond-level traceability of the time and length of the steel coil data within 3 years.

Images