A multi-source data tracing and quality anomaly diagnosis method and system for hot-dip galvanized steel plate pretreatment and corrosion prevention passivation

Abstract

The present application relates to hot galvanizing quality control technical field, especially to a kind of multi-source data tracing and quality abnormal diagnosis method and system for hot galvanizing steel plate pretreatment and anticorrosion passivation, by collecting pretreatment and passivation whole process multi-source data, the standardized processing of protocol unification and time sequence alignment is carried out to heterogeneous data, three-level closed-loop tracing architecture is used to realize whole link traceability, auditability in combination with blockchain technology;Through the fusion mechanism model and data-driven model, real-time diagnosis, prior prediction and root location of quality abnormality are completed, closed-loop control is carried out, and finally the interactive display of data and results is realized through visual interface;The system of the present application includes data acquisition module, data standardization processing module, data tracing module, quality abnormality diagnosis module, closed-loop control module and visual interaction module, effectively solve the problems of data heterogeneity, insufficient tracing accuracy, abnormal diagnosis lag and dependence on artificial experience in prior art.

Description

Technical Field

[0001] This invention relates to the field of hot-dip galvanizing quality control technology, and in particular to a multi-source data traceability and quality anomaly diagnosis method and system for pretreatment and anti-corrosion passivation of hot-dip galvanized steel sheets. Background Technology

[0002] Hot-dip galvanized steel sheets, with their excellent corrosion resistance, formability, and cost advantages, are widely used in various fields such as construction, automobiles, home appliances, and rail transportation. The quality of these products directly determines the safety and durability of downstream applications. Pre-treatment (including alkaline washing and degreasing, pickling and rust removal, ultrasonic cleaning, etc.) and anti-corrosion passivation are the core processes in hot-dip galvanizing production. Controlling the cleanliness of the pre-treatment directly affects the adhesion and bonding strength of the zinc layer, while the uniformity and integrity of the passivation film determine the long-term corrosion resistance of the steel sheet. Together, these two processes constitute the key line of defense for the quality control of hot-dip galvanized steel sheets.

[0003] Traditional quality control models in the pretreatment and passivation processes of hot-dip galvanizing are no longer sufficient to meet the industry's development needs. Current technologies suffer from several significant shortcomings, severely hindering improvements in product quality stability, production efficiency, and the overall quality control level throughout the product lifecycle: Heterogeneous and inconsistent multi-source data leads to insufficient traceability accuracy; existing technologies lack a unified data standardization mechanism, preventing effective fusion and collaborative correlation of multi-source data, resulting in broken data links and only enabling rough batch-level traceability, failing to accurately pinpoint minute quality defects; secondly, quality anomalies rely heavily on manual experience, leading to delayed diagnosis and low accuracy. Currently, quality anomaly diagnosis in the pretreatment and passivation processes of hot-dip galvanizing still primarily relies on the manual experience of on-site operators combined with offline sampling. This approach cannot achieve online real-time identification and diagnosis of latent defects such as uneven passivation film, zinc layer thickness fluctuations, residual oil from pretreatment, and iron ion residue. Therefore, this paper proposes a multi-source data traceability and quality anomaly diagnosis method and system for the pretreatment and passivation of hot-dip galvanized steel sheets to address these issues. Summary of the Invention

[0004] The purpose of this invention is to solve the problems in the background art, and to propose a multi-source data traceability and quality anomaly diagnosis method and system for pretreatment and anti-corrosion passivation of hot-dip galvanized steel sheets.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: A method for multi-source data traceability and quality anomaly diagnosis for pretreatment and corrosion passivation of hot-dip galvanized steel sheets includes the following steps: Step 1: Collect multi-source data for the entire process of pretreatment and anti-corrosion passivation of hot-dip galvanized steel sheets; the multi-source data includes process data, equipment operation status data, quality inspection data, material batch data and environmental condition data for the entire process of pretreatment, hot-dip galvanizing, passivation and finished product testing of hot-dip galvanized steel sheets, breaking through the limitations of traditional scattered multi-source data collection and achieving comprehensive coverage of the entire process data; Step 2: Standardize the collected multi-source data to eliminate data heterogeneity and temporal misalignment; the standardization process includes unifying protocols and aligning time sequences of multi-source heterogeneous data, establishing a unified multi-source data metadata dictionary, realizing the collaborative fusion of multi-source data, and solving the problems of ineffective association of multi-source data and broken data links in the existing technology. Step 3: Perform full-chain traceability on the standardized multi-source data to achieve queryable, traceable, and auditable multi-dimensional data; the full-chain traceability adopts a three-level closed-loop traceability architecture to achieve full-process traceability at the batch level, process level, and data level, and the data-level traceability uses blockchain technology to ensure the immutability and auditability of the data, improve the accuracy and reliability of data traceability, and solve the problem of inaccurate location of quality defects; Step 4: Analyze the standardized multi-source data to complete real-time diagnosis, pre-prediction, and root cause localization of quality anomalies; the diagnosis and prediction of quality anomalies are achieved by integrating the mechanism model and the data-driven model, mining the hidden correlations between multi-source data, and completing the accurate identification of latent anomalies and new types of anomalies, eliminating the dependence on human experience and solving the defects of lagging and low accuracy in quality anomaly diagnosis. Step 5: Based on the anomaly diagnosis results, automatically adjust the production process parameters to achieve closed-loop control of quality anomalies; the automatic adjustment of process parameters adopts a combination of model predictive control and reinforcement learning to achieve personalized process control and continuous optimization of the process window, avoiding the limitations of traditional post-event remediation and reducing quality losses. Step Six: Visualize and interactively display traceability data, anomaly diagnosis results, and process adjustment parameters to improve the convenience and intuitiveness of production control, enabling operators to monitor production status and quality in real time.

[0006] A multi-source data traceability and quality anomaly diagnosis system for pretreatment and corrosion passivation of hot-dip galvanized steel sheets is disclosed. This system implements the aforementioned multi-source data traceability and quality anomaly diagnosis method. It includes a data acquisition module, a data standardization processing module, a data traceability module, a quality anomaly diagnosis module, a closed-loop control module, and a visualization interaction module. These modules work collaboratively to form a complete quality control closed loop. Specific module functions are as follows: Data acquisition module: It is used to overcome the shortcomings of scattered multi-source data acquisition, realize the comprehensive acquisition of multi-type data in the entire process of hot-dip galvanized steel sheet pretreatment and anti-corrosion passivation, and provide complete data support for subsequent data processing, traceability and diagnosis; Data standardization processing module: It is used to solve the problems of heterogeneous and inconsistent time series of multi-source data. It performs protocol unification, time series alignment and standardization processing on the collected multi-source data, establishes a unified multi-source data metadata dictionary, realizes the collaborative integration of multi-source data, and eliminates the risk of data link breakage. Data traceability module: It is used to solve the problems of insufficient accuracy and poor reliability of data traceability. It adopts a three-level closed-loop traceability architecture to achieve full-link traceability and auditability of multi-source data. It also integrates blockchain technology to ensure the immutability and auditability of traceability data, and can accurately locate the process segment and parameter range corresponding to quality defects. Quality Anomaly Diagnosis Module: This module addresses the issues of delayed and low-accuracy quality anomaly diagnosis. It integrates mechanistic and data-driven models to uncover hidden correlations between multi-source data, enabling real-time anomaly diagnosis, pre-emptive prediction, and root cause localization. It can accurately identify latent and new types of anomalies, eliminating reliance on human experience. Closed-loop control module: Used to realize closed-loop management of quality anomalies, avoiding the limitations of post-event remediation. Based on the diagnostic results output by the quality anomaly diagnosis module, it automatically adjusts the production process parameters and adopts a combination of model predictive control and reinforcement learning to achieve personalized process control and continuous optimization of the process window. Visual interaction module: Used to intuitively display and interact with various data and diagnostic and control results, improve the convenience of management and control, and make it easy for operators to view traceability data, abnormal diagnosis results and process adjustment parameters in real time, so as to achieve efficient management and control of the production process.

[0007] Compared with existing technologies, the advantages of this invention are: 1. This invention effectively solves the technical defects of heterogeneous multi-source data, inconsistent time sequence, insufficient traceability accuracy, and poor data reliability. By standardizing the multi-source data throughout the entire process, a unified data fusion system is established. Combined with a three-level closed-loop traceability architecture and blockchain technology, it achieves accurate traceability across the entire chain from the batch level to the data level, ensuring the immutability and auditability of traceability data. It can quickly locate the process segment and parameter range corresponding to quality defects, solving the problems of broken data links, inaccurate positioning, and unreliable data in traditional traceability models, and improving the quality control level of hot-dip galvanized steel sheets throughout their entire life cycle.

[0008] 2. This invention addresses the pain points of relying on human experience for quality anomalies, resulting in delayed diagnosis and low accuracy. By integrating mechanistic models and data-driven models, it uncovers hidden correlations between multi-source data, enabling online real-time diagnosis, prediction, and root cause localization of quality anomalies during the pretreatment and passivation processes of hot-dip galvanizing. It can accurately identify latent anomalies such as uneven passivation film and pretreatment residues, as well as new types of anomalies. This eliminates reliance on human experience and offline sampling inspections, avoids the limitations of post-event remediation of quality anomalies, effectively reduces quality losses, and improves product quality stability.

[0009] 3. This invention addresses the problem that traditional control models cannot achieve closed-loop management of quality anomalies. By coordinating the closed-loop control module with other modules and combining model predictive control with reinforcement learning for process adjustment, it achieves closed-loop management of quality anomaly diagnosis, adjustment, feedback, and optimization. It can automatically optimize process parameters and improve process windows based on anomaly diagnosis results. At the same time, it enhances the convenience of management through a visual interaction module, thereby improving the overall efficiency of hot-dip galvanizing production. It is suitable for the product quality requirements and environmental compliance requirements of high-end manufacturing industries and has strong practicality and promotional value. Attached Figure Description

[0010] Fig. 1 This is a schematic diagram of the system module structure proposed in this invention; Fig. 2 This is a schematic diagram of the method steps proposed in this invention. Detailed Implementation

[0011] 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.

[0012] Reference Figs. 1-2 The present invention will be further described in detail below with reference to specific embodiments. These embodiments are used to explain the present invention and are not intended to limit the scope of protection of the present invention. In this embodiment, a multi-source data traceability and quality anomaly diagnosis system for pretreatment and corrosion passivation of hot-dip galvanized steel sheets includes a data acquisition module, a data standardization processing module, a data traceability module, a quality anomaly diagnosis module, a closed-loop control module, and a visualization interaction module. These modules work collaboratively to form a complete end-to-end quality control closed loop. The specific implementation methods of each module are as follows: Data Acquisition Module: The data acquisition module is used to overcome the shortcomings of scattered multi-source data acquisition, and realize the comprehensive acquisition of multi-type data of the entire process of hot-dip galvanized steel sheet pretreatment and anti-corrosion passivation. The acquisition scope covers the entire process of hot-dip galvanized steel sheet pretreatment, hot-dip galvanizing, passivation and finished product inspection. The acquired multi-source data includes process data, equipment operating status data, quality inspection data, material batch data and environmental condition data. In practice, the data acquisition module collects various types of data synchronously through acquisition terminals deployed in each production process. These include process data corresponding to relevant operating parameters for pretreatment processes such as alkaline washing, acid washing, ultrasonic cleaning, and passivation; equipment operating status data corresponding to the operating status information of production equipment such as annealing furnaces, zinc pots, and circulating pumps; quality inspection data corresponding to relevant inspection information such as the surface quality and coating adhesion of finished steel plates; material batch data corresponding to batch information of raw materials such as substrates, zinc ingots, and passivating agents; and environmental condition data corresponding to environmental information such as temperature, humidity, and acid mist concentration in the production workshop. The acquisition terminals establish communication connections with each production and testing device to ensure the comprehensiveness and real-time nature of data collection, providing complete data support for subsequent data processing, traceability, and anomaly diagnosis. Data standardization processing module: The data standardization processing module is used to solve the problems of heterogeneity and time sequence inconsistency of multi-source data, and realize the unified fusion of multi-source data. Specifically, it is used to standardize the multi-source data collected by the data acquisition module to eliminate data heterogeneity and time sequence misalignment. In practical implementation, the data standardization processing module first performs protocol unification processing on the collected multi-source heterogeneous data, converting heterogeneous data output from different devices and systems into a unified communication protocol to solve the problem of incompatibility between multi-source data protocols. Subsequently, it performs time-series alignment processing on various types of data, adjusting multi-source data with different sampling frequencies to a unified time sequence based on the data collection timestamp, eliminating the risk of time sequence misalignment. At the same time, it establishes a unified multi-source data metadata dictionary, uniformly defining the names, types, and meanings of various types of data, realizing the collaborative fusion of multi-source data, eliminating the problems of ineffective association of multi-source data and broken data links in existing technologies, and providing a standardized data foundation for subsequent data traceability and anomaly diagnosis. Data traceability module: The data traceability module is used to solve the problems of insufficient accuracy and poor reliability of data traceability, realize full-link traceability and auditability, and integrate blockchain technology to realize full-link traceability of multi-source data; In practical implementation, the data traceability module adopts a three-level closed-loop traceability architecture to achieve full-process traceability at the batch, process, and data levels: Batch-level traceability uses raw material batches and finished product batches as core indexes, linking batch information of raw materials such as substrates and passivating agents with batch information of finished steel plates to achieve overall traceability at the batch level; Process-level traceability links the operation data of each production process, recording the processing trajectory and corresponding process parameters of each batch of steel plates in each process to achieve precise traceability at the process level; Data-level traceability integrates blockchain technology to upload standardized key process parameters, quality inspection data, equipment operation data, and other core data to the blockchain in real time. Utilizing the immutable and auditable characteristics of blockchain, the credibility of traceability data is guaranteed, while also supporting reverse traceability of data. By quickly locating the corresponding process and parameter range through finished product quality anomalies, it solves the problems of insufficient accuracy and unreliable data in traditional traceability. Quality Anomaly Diagnosis Module: The quality anomaly diagnosis module is used to solve the problems of lagging quality anomaly diagnosis and low accuracy. It realizes real-time diagnosis, pre-prediction and root cause location of anomalies. Specifically, it is used to analyze standardized multi-source data to complete real-time diagnosis, pre-prediction and root cause location of quality anomalies. In practical implementation, the quality anomaly diagnosis module integrates a mechanistic model and a data-driven model to uncover hidden correlations between multi-source data, enabling accurate identification of latent and novel anomalies. The mechanistic model, based on the production mechanism of hot-dip galvanizing pretreatment and passivation, describes the intrinsic relationship between process parameters, equipment status, and product quality. The data-driven model uncovers hidden patterns in multi-source data, compensating for the limitations of the mechanistic model. After integration, standardized multi-source data can be received in real time to perform online real-time diagnosis of quality anomalies in the production process. Simultaneously, it can predict production quality trends in advance. When an anomaly is detected or a potential anomaly is predicted, the root cause of the anomaly can be quickly located, and the corresponding process, parameters, and equipment can be identified. This eliminates reliance on human experience and solves the shortcomings of traditional diagnosis, such as lag and low accuracy. Closed-loop control module: The closed-loop control module is used to realize closed-loop management of quality anomalies, avoiding the limitations of post-event remediation. Specifically, it is used to automatically adjust production process parameters based on the anomaly diagnosis results. In practice, the closed-loop control module establishes a communication connection with the quality anomaly diagnosis module and the production control system, and receives the anomaly diagnosis results and root cause location information output by the quality anomaly diagnosis module in real time. It adopts a combination of model predictive control and reinforcement learning to automatically adjust the production process parameters. Among them, model predictive control is used to predict the quality effect after the process parameter adjustment based on the root cause of the anomaly, and reinforcement learning is used to continuously optimize the adjustment strategy based on historical adjustment data, so as to realize personalized process control and continuous optimization of the process window, avoid the limitations of traditional post-event remediation, effectively reduce quality loss, and improve product quality stability. Visualization and Interaction Module: The visualization and interaction module is used to realize the intuitive display and interaction of various data and diagnostic and control results, improve the convenience of management and control, and is specifically used to visualize and interact with traceability data, abnormal diagnosis results and process adjustment parameters. In practical implementation, the visualization interaction module adopts a graphical interface design, which displays the full-link traceability data output by the data traceability module, the abnormal diagnosis results and root cause information output by the quality abnormality diagnosis module, and the process adjustment parameters output by the closed-loop control module in an intuitive form such as charts, curves, and text. At the same time, it supports interactive operation by operators. Operators can query traceability data of specific batches and specific processes through the visualization interface, view abnormal diagnosis details, and manually intervene in the process parameter adjustment process, thereby improving the convenience and intuitiveness of production control and making it easier for operators to grasp the production status and quality situation in real time. In this embodiment, a multi-source data traceability and quality anomaly diagnosis method for pretreatment and anti-corrosion passivation of hot-dip galvanized steel sheets is applied to the above-mentioned system, and specifically includes the following steps: Step 1: Collect multi-source data on the entire process of pretreatment and anti-corrosion passivation of hot-dip galvanized steel sheets: Through the system's data acquisition module, acquisition terminals deployed in each production process synchronously collect multi-source data from the entire process of hot-dip galvanized steel sheet pretreatment and anti-corrosion passivation. The multi-source data includes process data, equipment operating status data, quality inspection data, material batch data, and environmental condition data for the entire process of hot-dip galvanized steel sheet pretreatment, hot-dip galvanizing, passivation, and finished product testing. During the acquisition process, stable communication between the acquisition terminals and each production and testing equipment is ensured to achieve comprehensive and real-time acquisition of data throughout the entire process, avoid data omissions, and provide complete data support for subsequent data processing. Step 2: Standardize the collected multi-source data to eliminate data heterogeneity and temporal misalignment: The system's data standardization module standardizes the multi-source data collected in step one: First, it unifies the protocol of the heterogeneous multi-source data, converting heterogeneous data output from different devices and systems into a unified communication protocol; then, it aligns the time sequence of each type of data according to the data collection timestamp to eliminate potential time sequence misalignment; at the same time, it establishes a unified multi-source data metadata dictionary to uniformly define the names, types, and meanings of various types of data, realizing the collaborative fusion of multi-source data, eliminating data heterogeneity and time sequence misalignment, and solving the problems of ineffective association of multi-source data and broken data links in existing technologies. Step 3: Perform end-to-end traceability on the standardized multi-source data to achieve queryable, traceable, and auditable multi-dimensional data: Through the system's data traceability module, the standardized multi-source data from step two is traced across the entire chain. A three-level closed-loop traceability architecture is adopted to achieve full-process traceability at the batch, process, and data levels: Batch-level traceability uses raw material batches and finished product batches as core indexes, linking batch information of raw materials and finished products to achieve overall traceability at the batch level; Process-level traceability links the operational data of each production process, recording the processing trajectory of the steel plate in each process and the corresponding process parameters, achieving precise traceability at the process level; Data-level traceability uses blockchain technology to upload key core data to the chain in real time, ensuring the immutability and auditability of the data, while also supporting reverse traceability, quickly locating the corresponding process and parameter range through finished product quality anomalies, and achieving multi-dimensional data queryability, traceability, and auditability. Step 4: Analyze the standardized multi-source data to complete real-time diagnosis, pre-emptive prediction, and root cause localization of quality anomalies. The system's quality anomaly diagnosis module analyzes the standardized multi-source data from step two, integrates mechanistic and data-driven models, uncovers hidden correlations between the multi-source data, and performs online real-time diagnosis of quality anomalies in the production process, identifying latent and new types of anomalies. Simultaneously, it predicts production quality trends over a future period, anticipating potential quality anomalies. When an anomaly is detected or a potential anomaly is predicted, the system quickly locates the root cause, identifies the corresponding process, parameters, and equipment, providing a basis for subsequent process adjustments. This eliminates reliance on manual experience and solves the problems of delayed and low-accuracy quality anomaly diagnosis. Step 5: Based on the anomaly diagnosis results, automatically adjust the production process parameters to achieve closed-loop control of quality anomalies. The system's closed-loop control module receives the anomaly diagnosis results and root cause location information output from step four. It then uses a combination of model predictive control and reinforcement learning to automatically adjust the production process parameters. Model predictive control predicts the quality effect after the process parameter adjustment based on the anomaly root cause, while reinforcement learning continuously optimizes the adjustment strategy based on historical adjustment data. This achieves personalized process control and continuous optimization of the process window, forming a closed-loop management system of diagnosis-adjustment-feedback-optimization. This avoids the limitations of post-event remediation and effectively reduces quality losses. Step Six: Visualize and interactively display traceability data, anomaly diagnosis results, and process adjustment parameters: Through the system's visual interaction module, the traceability data of step three, the anomaly diagnosis results of step four, and the process adjustment parameters of step five are displayed in an intuitive form such as charts, curves, and text. Operators can use the visual interface to query traceability data for specific batches and specific processes, view anomaly diagnosis details, and manually intervene in the process parameter adjustment process. This enables intuitive display and interactive operation of various data and diagnosis and control results, improving the convenience and intuitiveness of production management. The method and system of this embodiment, through the collaborative work of each module and the orderly execution of each step, effectively solves the technical defects in the existing technology, such as heterogeneous multi-source data, insufficient traceability accuracy, lagging and low accuracy of quality anomaly diagnosis, and poor data reliability in the pretreatment and anti-corrosion passivation process of hot-dip galvanized steel sheets. It realizes the unified integration of multi-source data throughout the entire process and precise traceability across the entire chain, enabling real-time diagnosis, pre-prediction, and closed-loop control of quality anomalies. It eliminates the reliance on human experience, improves product quality stability and production control efficiency, adapts to the entire production process of pretreatment and anti-corrosion passivation of hot-dip galvanized steel sheets, and meets environmental compliance audit requirements, thus having broad application prospects.

[0013] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for multi-source data traceability and quality anomaly diagnosis for pretreatment and corrosion passivation of hot-dip galvanized steel sheets, characterized in that, Includes the following steps: Step 1: Collect multi-source data on the entire process of pretreatment and anti-corrosion passivation of hot-dip galvanized steel sheets; Step 2: Standardize the collected multi-source data to eliminate data heterogeneity and temporal misalignment; Step 3: Perform full-chain traceability on the standardized multi-source data to achieve queryable, traceable, and auditable multi-dimensional data; Step 4: Analyze the standardized multi-source data to complete real-time diagnosis, prediction and root cause location of quality anomalies; Step 5: Based on the anomaly diagnosis results, automatically adjust the production process parameters to achieve closed-loop control of quality anomalies; Step Six: Visualize and interactively display traceability data, anomaly diagnosis results, and process adjustment parameters.

2. The method for multi-source data traceability and quality anomaly diagnosis for pretreatment and anti-corrosion passivation of hot-dip galvanized steel sheets according to claim 1, characterized in that, The multi-source data includes process data for the entire process of hot-dip galvanized steel sheet pretreatment, hot-dip galvanizing, passivation and finished product inspection, equipment operating status data, quality inspection data, material batch data and environmental condition data.

3. The method for multi-source data traceability and quality anomaly diagnosis for pretreatment and anti-corrosion passivation of hot-dip galvanized steel sheets according to claim 1, characterized in that, The standardization process includes unifying protocols and aligning time sequences for multi-source heterogeneous data, establishing a unified multi-source data metadata dictionary, and realizing the collaborative fusion of multi-source data.

4. The method for multi-source data traceability and quality anomaly diagnosis for pretreatment and anti-corrosion passivation of hot-dip galvanized steel sheets according to claim 1, characterized in that, The full-chain traceability adopts a three-level closed-loop traceability architecture to achieve full-process traceability at the batch, process, and data levels. Furthermore, data-level traceability uses blockchain technology to ensure the immutability and auditability of data.

5. The method for multi-source data traceability and quality anomaly diagnosis for pretreatment and anti-corrosion passivation of hot-dip galvanized steel sheets according to claim 1, characterized in that, The diagnosis and prediction of quality anomalies are achieved by integrating mechanistic models and data-driven models, uncovering hidden correlations between multi-source data, and completing the accurate identification of latent anomalies and new types of anomalies.

6. The method for multi-source data traceability and quality anomaly diagnosis for pretreatment and anti-corrosion passivation of hot-dip galvanized steel sheets according to claim 1, characterized in that, The automatic adjustment of process parameters adopts a combination of model predictive control and reinforcement learning to achieve personalized process control and continuous optimization of the process window.

7. A multi-source data traceability and quality anomaly diagnosis system for pretreatment and corrosion passivation of hot-dip galvanized steel sheets, characterized in that, It includes a data acquisition module, a data standardization and processing module, a data traceability module, a quality anomaly diagnosis module, a closed-loop control module, and a visualization and interaction module, and all modules work together. The data acquisition module is used to overcome the shortcomings of scattered multi-source data acquisition and realize the comprehensive acquisition of multiple types of data throughout the entire process of hot-dip galvanized steel sheet pretreatment and anti-corrosion passivation. The data standardization processing module is used to solve the problems of heterogeneity and time inconsistency of multi-source data, and to achieve unified integration of multi-source data; The data traceability module is used to solve the problems of insufficient accuracy and poor reliability of data traceability, and to achieve full-chain traceability and auditability; The quality anomaly diagnosis module is used to solve the problems of lagging quality anomaly diagnosis and low accuracy, and to realize real-time diagnosis, pre-prediction and root cause location of anomalies. The closed-loop control module is used to achieve closed-loop management of quality anomalies, avoiding the limitations of post-event remediation; The visualization and interaction module is used to intuitively display and interact with various data and diagnostic and control results, thereby improving the convenience of management and control.

8. A multi-source data traceability and quality anomaly diagnosis system for pretreatment and corrosion passivation of hot-dip galvanized steel sheets according to claim 7, characterized in that, The data acquisition module is used to collect multi-source data from the entire process of pretreatment and anti-corrosion passivation of hot-dip galvanized steel sheets, and the data standardization processing module is used to standardize the collected multi-source data.

9. A multi-source data traceability and quality anomaly diagnosis system for pretreatment and corrosion passivation of hot-dip galvanized steel sheets according to claim 7, characterized in that, The data traceability module is used to realize full-link traceability of multi-source data and integrates blockchain technology; the quality anomaly diagnosis module is used to complete real-time diagnosis, pre-prediction and root cause location of quality anomalies.

10. A multi-source data traceability and quality anomaly diagnosis system for pretreatment and corrosion passivation of hot-dip galvanized steel sheets according to claim 7, characterized in that, The closed-loop control module is used to automatically adjust the production process parameters based on the anomaly diagnosis results, and the visualization interaction module is used to realize the visualization display and interactive operation of traceability data, anomaly diagnosis results and process adjustment parameters.

Images