A safety early warning method, system, device and medium for a galvanizing process production line

Abstract

The application provides a safety early warning method, system, equipment and medium for a galvanizing process production line, comprising: acquiring multi-source data of the galvanizing process production line; performing multi-modal analysis and vectorization processing on the multi-source data, constructing a galvanizing process production line knowledge graph and storing it in an AI data lake; reading the knowledge graph from the AI data lake to adjust a general large model; horizontally splitting the galvanizing process into multiple process thinking chain nodes according to the process sequence; vertically splitting each process thinking chain node into an analysis node and an auxiliary node executed by a safety hidden danger analysis model, forming a tree structure thinking chain; traversing the tree structure thinking chain based on a preset sequence, analyzing the safety hidden danger of each process; and collecting the safety hidden danger analysis results of all processes to generate a safety hidden danger report and rectification measures for safety early warning. This method helps managers in the galvanizing industry to master the possible safety hidden dangers and rectification measures after the production line is put into operation.

Description

Technical Field

[0001] This invention relates to the field of production safety technology, and in particular to a safety early warning method, system, equipment and medium for a galvanizing production line. Background Technology

[0002] In recent years, the safety situation in steel production has been severe, with various safety accidents causing significant losses to people's lives and companies' property. Traditional safety production management mainly relies on human intervention, depending on the personal experience of enterprise safety supervisors to judge potential production safety hazards. This often leads to incomplete and inadequate analysis of safety hazards in the production process before the production line is put into operation. On-site staff also tend to overlook potential safety risks, becoming complacent and leading to production safety accidents.

[0003] At present, enterprises' safety hazard analysis and management work mainly focuses on engineering project construction, while there are few methods for analyzing safety hazards in the actual production operation of production lines with unique production processes in the steel industry, such as galvanizing. This makes it difficult for managers in the steel industry to fully understand the potential safety hazards after the production line is in operation and the corresponding rectification measures, as well as to do a good job of prevention before accidents occur. As a result, it is impossible to reduce the personal injury accident rate of galvanizing production lines and avoid the occurrence of serious accidents. Summary of the Invention

[0004] In view of this, the purpose of this invention is to provide a safety early warning method, system, equipment and medium for galvanizing production lines, which helps managers in the galvanizing and other steel industries to fully understand the potential safety hazards and corresponding corrective measures after the production line is in operation, and to do a good job of prevention before accidents occur, thereby reducing the personal injury accident rate of galvanizing production lines and preventing the occurrence of serious accidents.

[0005] In a first aspect, embodiments of the present invention provide a safety early warning method for a galvanizing production line, comprising: acquiring multi-source data of the galvanizing production line; performing multimodal analysis and vectorization on the multi-source data to construct a knowledge graph of the galvanizing production line and storing it in an AI data lake; reading the knowledge graph from the AI ​​data lake and adjusting the general large model to obtain a safety hazard analysis model corresponding to the galvanizing process; horizontally dividing the galvanizing process into multiple process thought chain nodes according to the process sequence; vertically dividing each process thought chain node into analysis nodes and auxiliary nodes executed by the safety hazard analysis model to form a tree-structured thought chain; traversing the tree-structured thought chain according to a pre-set order to analyze the safety hazards of each process; summarizing the safety hazard analysis results of all processes to generate a safety hazard report and rectification measures for safety early warning.

[0006] In a preferred embodiment of the present invention, the above-mentioned multimodal parsing and vectorization processing of multi-source data to construct a knowledge graph of the galvanizing process production line and store it in the AI ​​data lake includes: using a multimodal large model to parse the text, image and video content in the multi-source data to obtain the parsed data; and performing vectorization processing on the parsed data and storing it in the AI ​​data lake.

[0007] In a preferred embodiment of the present invention, the above-mentioned method of reading knowledge graphs from the AI ​​data lake and adjusting the general large model to obtain a safety hazard analysis model corresponding to the galvanizing process includes: extracting galvanizing process design, safety production standards and accident information from the AI ​​data lake; training the general large model using the galvanizing process design, safety production standards and accident information to obtain a safety hazard analysis model corresponding to the galvanizing process; the safety hazard analysis model has professional characteristics for safety hazard investigation.

[0008] In a preferred embodiment of the present invention, the above-mentioned galvanizing process is horizontally divided into multiple process thought chain nodes according to the process sequence, including: dividing the galvanizing process into degreasing node, pickling node, water washing node, fluxing node, drying node, preheating node, zinc pot immersion galvanizing node, cooling node and passivation node according to the process sequence; and establishing a corresponding process thought chain node for each node.

[0009] In a preferred embodiment of the present invention, the above-mentioned vertical splitting of each process thinking chain node into analysis nodes and auxiliary nodes executed by the safety hazard analysis model to form a tree-structured thinking chain includes: splitting the process thinking chain node into analysis nodes and auxiliary nodes; the analysis nodes are used to analyze the basic information of the process and the four elements of safety hazards; the four elements of safety hazards include: unsafe human behavior, unsafe conditions of objects, management defects and unsafe environmental factors; the auxiliary nodes are used to standardize the output results of the analysis nodes and store short-term memory and long-term memory.

[0010] In a preferred embodiment of the present invention, the above-mentioned traversal of the tree-structured thinking chain based on a pre-set order to analyze the safety hazards of each process includes: performing a depth-first traversal of the tree-structured thinking chain in a left-to-right, top-to-bottom order to analyze the safety hazards of each process; wherein, the safety hazard analysis is performed by the analysis nodes of the tree-structured thinking chain, and at each analysis node, the following are employed: retrieval enhancement generation technology is used to obtain relevant data from the AI ​​data lake; the latest professional information is obtained from an external knowledge base through a model context protocol; and LangChain is used to construct question information for the large model.

[0011] In a preferred embodiment of the present invention, the results of safety hazard analysis of all processes are summarized to generate a safety hazard report and rectification measures for safety early warning. The method further includes: replacing the safety hazard analysis model and rerunning the analysis process to generate a second safety hazard report; comparing the first safety hazard report and the second safety hazard report to obtain the comparison similarity; if the comparison similarity does not meet the preset similarity threshold, the tree structure thinking chain is adjusted.

[0012] Secondly, embodiments of the present invention also provide a safety early warning device for a galvanizing production line, comprising: a multi-source data acquisition module for acquiring multi-source data from the galvanizing production line; a knowledge graph construction module for performing multi-modal parsing and vectorization processing on the multi-source data, constructing a knowledge graph of the galvanizing production line and storing it in an AI data lake; a model adjustment module for reading the knowledge graph from the AI ​​data lake and adjusting a general large model to obtain a safety hazard analysis model corresponding to the galvanizing process; a galvanizing process splitting module for horizontally splitting the galvanizing process into multiple process thinking chain nodes according to the process sequence; a process thinking chain node splitting module for vertically splitting each process thinking chain node into analysis nodes and auxiliary nodes executed by the safety hazard analysis model, forming a tree-structured thinking chain; a safety hazard analysis module for traversing the tree-structured thinking chain based on a pre-set order and analyzing the safety hazards of each process; and a safety early warning module for summarizing the safety hazard analysis results of all processes and generating a safety hazard report and rectification measures for safety early warning.

[0013] Thirdly, embodiments of the present invention also provide an electronic device, including a processor and a memory, wherein the memory stores computer-executable instructions that can be executed by the processor, and the processor executes the computer-executable instructions to implement the safety early warning method for the galvanizing process production line described in the first aspect.

[0014] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing computer-executable instructions. When the computer-executable instructions are invoked and executed by a processor, the computer-executable instructions cause the processor to implement the safety early warning method for the galvanizing process production line described in the first aspect.

[0015] The embodiments of the present invention bring the following beneficial effects:

[0016] This invention provides a safety early warning method, system, equipment, and medium for galvanizing production lines. It acquires multi-source data from the galvanizing production line, performs multimodal analysis and vectorization on this data, constructs a knowledge graph of the galvanizing production line, and stores it in an AI data lake. The knowledge graph is retrieved from the AI ​​data lake, and a general model is adjusted to obtain a safety hazard analysis model corresponding to the galvanizing process. The galvanizing process is horizontally divided into multiple process thought chain nodes according to the process sequence. Each process thought chain node is vertically divided into analysis nodes and auxiliary nodes executed by the safety hazard analysis model, forming a tree-structured thought chain. The tree-structured thought chain is traversed according to a pre-set order to analyze the safety hazards of each process. The analysis results of all process safety hazards are summarized to generate a safety hazard report and rectification measures for safety early warning. This method helps managers in the galvanizing and other steel industries to fully understand the potential safety hazards and corresponding rectification measures after the production line is operational, and to take preventative measures before accidents occur, thereby reducing the personal injury accident rate in galvanizing production lines and preventing serious accidents.

[0017] Other features and advantages of this disclosure will be set forth in the following description, or some features and advantages may be inferred from the description or determined without doubt, or may be learned by practicing the techniques described above.

[0018] To make the above-mentioned objects, features and advantages of this disclosure more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0019] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 A flowchart of a safety early warning method for a galvanizing production line provided in an embodiment of the present invention;

[0021] Figure 2 A flowchart of another safety early warning method for a galvanizing production line provided in an embodiment of the present invention;

[0022] Figure 3 This is a schematic diagram illustrating a "horizontal splitting" method provided in an embodiment of this application;

[0023] Figure 4This is a schematic diagram illustrating an executable intelligent agent tree-shaped thought chain formed through "vertical splitting" and "horizontal splitting" in an embodiment of this application.

[0024] Figure 5 This is a schematic diagram of the structure of a safety early warning device for a galvanizing production line provided in an embodiment of the present invention;

[0025] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions 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, 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.

[0027] In recent years, the safety situation in steel production has been severe, with various safety accidents causing significant losses to people's lives and companies' property. Traditional safety production management mainly relies on human intervention, depending on the personal experience of enterprise safety supervisors to judge potential production safety hazards. This often leads to incomplete and inadequate analysis of safety hazards in the production process before the production line is put into operation. On-site staff also tend to overlook potential safety risks, becoming complacent and leading to production safety accidents.

[0028] At present, enterprises' safety hazard analysis and management work mainly focuses on engineering project construction, while there are few methods for analyzing safety hazards in the actual production operation of production lines with unique production processes in the steel industry, such as galvanizing. This makes it difficult for managers in the steel industry to fully understand the potential safety hazards after the production line is in operation and the corresponding rectification measures, as well as to do a good job of prevention before accidents occur. As a result, it is impossible to reduce the personal injury accident rate of galvanizing production lines and avoid the occurrence of serious accidents.

[0029] Based on this, the present invention provides a safety early warning method, system, equipment, and medium for a galvanizing production line. This method acquires multi-source data from the galvanizing production line, performs multi-modal analysis and vectorization on the data, constructs a knowledge graph of the galvanizing production line, and stores it in an AI data lake. The knowledge graph is then read from the AI ​​data lake to adjust a general model, resulting in a safety hazard analysis model corresponding to the galvanizing process. The galvanizing process is horizontally divided into multiple process thought chain nodes according to the process sequence. Each process thought chain node is vertically divided into analysis nodes and auxiliary nodes executed by the safety hazard analysis model, forming a tree-structured thought chain. The tree-structured thought chain is traversed according to a pre-set order to analyze the safety hazards of each process. The analysis results of all process safety hazards are summarized to generate a safety hazard report and rectification measures for safety early warning. This approach helps managers in the galvanizing and other steel industries fully understand the potential safety hazards and corresponding rectification measures after the production line is operational, and to take preventative measures before accidents occur, thereby reducing the personal injury accident rate in the galvanizing production line and preventing serious accidents.

[0030] To facilitate understanding of this embodiment, a safety early warning method for a galvanizing production line disclosed in this embodiment of the invention will first be described in detail.

[0031] Example 1

[0032] This invention provides a safety early warning method for a galvanizing production line. Figure 1 A flowchart illustrating a safety early warning method for a galvanizing production line provided in an embodiment of the present invention. Figure 1 As shown, the safety early warning method for this galvanizing production line may include the following steps:

[0033] Step S101: Obtain multi-source data from the galvanizing process production line.

[0034] Among them, multi-source data can include collected process flow designs, on-site video photos, accident cases, etc.

[0035] Step S102: Perform multimodal parsing and vectorization processing on the multi-source data, construct a knowledge graph of the galvanizing process production line, and store it in the AI ​​data lake.

[0036] Among them, the collected multi-source data is analyzed using multimodal large model technology to separate text, image and video content, and then vectorized and stored in the AI ​​data lake.

[0037] Step S103: Read the knowledge graph from the AI ​​data lake and adjust the general large model to obtain the safety hazard analysis model corresponding to the galvanizing process.

[0038] Among these, data such as galvanizing process design, safety production, and accident information in the AI ​​data lake can be used for training or fine-tuning to compensate for the lack of professional knowledge in the large model, enabling the large model to learn specific characteristics and patterns related to safety hazard investigation in order to adapt to the current task.

[0039] Step S104: The galvanizing process is horizontally divided into multiple process thought chain nodes according to the process sequence.

[0040] Among these, the galvanizing process can be horizontally broken down into multiple process thought chain nodes according to the process sequence using workflow tools.

[0041] The workflow tool can be Dify or Coze, etc.

[0042] Step S105: Vertically split each process thinking chain node into analysis nodes and auxiliary nodes executed by the safety hazard analysis model, forming a tree-structured thinking chain.

[0043] Among them, the analysis nodes are mainly used to analyze the basic information of the process and the safety hazard analysis dimensions of related processes.

[0044] In addition to standardizing the output of the security hazard analysis model, the auxiliary node is also responsible for the agent's memory function, which includes short-term memory and long-term memory.

[0045] Step S106: Traverse the tree-structured thought chain based on a pre-set order and analyze the safety hazards of each process.

[0046] The preset order is from left to right and from top to bottom.

[0047] The intelligent agent performs a depth-first traversal of the tree-structured thought chain from left to right and from top to bottom, and autonomously performs analysis and processing.

[0048] Step S107: Summarize the safety hazard analysis results of all processes, generate a safety hazard report and rectification measures for safety early warning.

[0049] This system can summarize the safety hazard analysis results of all processes, mark the data containing safety hazards in the analysis results, compile a safety hazard report, generate corresponding rectification measures for the existing safety hazards, and issue early warnings so that staff can eliminate safety hazards according to the rectification measures in the alarm after seeing the alarm.

[0050] The safety early warning method for galvanizing production lines provided in this invention can acquire multi-source data from the galvanizing production line, perform multi-modal analysis and vectorization processing on the multi-source data, construct a knowledge graph of the galvanizing production line and store it in an AI data lake, read the knowledge graph from the AI ​​data lake to adjust the general large model, and obtain a safety hazard analysis model corresponding to the galvanizing process. The galvanizing process is horizontally divided into multiple process thought chain nodes according to the process sequence, and each process thought chain node is vertically divided into analysis nodes and auxiliary nodes executed by the safety hazard analysis model, forming a tree-structured thought chain. The tree-structured thought chain is traversed according to a pre-set order to analyze the safety hazards of each process. The results of the safety hazard analysis of all processes are summarized to generate a safety hazard report and rectification measures for safety early warning. This method helps managers in the galvanizing and other steel industries to fully understand the potential safety hazards after the production line is operational and the corresponding rectification measures, as well as to carry out preventative work before accidents occur, eliminate safety hazards in advance, reduce the personal injury accident rate of the galvanizing production line, and prevent serious accidents.

[0051] Example 2

[0052] This invention also provides another safety early warning method for a galvanizing production line; this method is implemented based on the method described in the above embodiments.

[0053] Figure 2 A flowchart of another safety early warning method for a galvanizing production line provided in an embodiment of the present invention is shown below. Figure 2 As shown, the safety early warning method for this galvanizing production line may include the following steps:

[0054] Step S201: Obtain multi-source data from the galvanizing process production line.

[0055] The multi-source data may further include: galvanizing production line process design, production line design scheme, construction blueprints, equipment drawings, safety operation specifications, historical safety accident information database, relevant national safety management policies and regulations, enterprise safety management requirements, national standards, industry specifications, and videos and photos of the production line after its completion.

[0056] Step S202: Perform multimodal parsing and vectorization processing on the multi-source data, construct a knowledge graph of the galvanizing process production line, and store it in the AI ​​data lake.

[0057] Specifically, performing multimodal parsing and vectorization on multi-source data to construct a knowledge graph of the galvanizing process production line and store it in the AI ​​data lake can include: using a multimodal large model to parse the text, image, and video content in the multi-source data to obtain the parsed data; and then vectorizing the parsed data and storing it in the AI ​​data lake.

[0058] The multimodal large models used may include: Generalized Questions QVQ-Max, VL-Max, etc.

[0059] Step S203: Read the knowledge graph from the AI ​​data lake and adjust the general large model to obtain the safety hazard analysis model corresponding to the galvanizing process.

[0060] Specifically, the knowledge graph read from the AI ​​data lake is used to adjust the general model to obtain a safety hazard analysis model corresponding to the galvanizing process. This can include: extracting galvanizing process design, safety production standards and accident information from the AI ​​data lake; using the galvanizing process design, safety production standards and accident information to train the general model to obtain a safety hazard analysis model corresponding to the galvanizing process; the safety hazard analysis model has professional characteristics for safety hazard investigation.

[0061] For general large models like Deepseek and Qwen3, fine-tuning can be done using data from the AI ​​data lake, such as galvanizing process design, safety production, and accident information, to compensate for the lack of professional knowledge in the large models. This allows the large models to learn specific characteristics and patterns related to safety hazard investigation in order to adapt to the current task.

[0062] Step S204: The galvanizing process is horizontally divided into multiple process thought chain nodes according to the process sequence.

[0063] One approach is to use workflow to build an intelligent agent thinking chain. First, the entire galvanizing process is "horizontally broken down" into steps according to their sequence, thus establishing process thinking chain nodes.

[0064] Specifically, the galvanizing process can be horizontally divided into multiple process thought chain nodes according to the process sequence. This can include: dividing the galvanizing process into degreasing node, pickling node, water washing node, fluxing node, drying node, preheating node, zinc pot immersion galvanizing node, cooling node, and passivation node; and establishing a corresponding process thought chain node for each node.

[0065] Among them, the analytical thinking chain of constructing intelligent agents can be compiled using tools such as Dify or Coze. First, the entire galvanizing process is "horizontally broken down" in a chain, and each node on the chain is a process that constitutes the galvanizing process.

[0066] For ease of understanding, Figure 3 This is a schematic diagram of a "horizontal splitting" method provided in an embodiment of this application.

[0067] like Figure 3As shown, taking the hot-dip galvanizing production process as an example, it mainly includes the following processes: degreasing, pickling, water washing, fluxing, drying, preheating, zinc pot immersion galvanizing, cooling, passivation, etc. First, according to the order of each process in the process, it is decomposed into multiple process chain nodes horizontally, and each node corresponds to a galvanizing process.

[0068] Step S205: Vertically split each process thinking chain node into analysis nodes and auxiliary nodes executed by the safety hazard analysis model, forming a tree-structured thinking chain.

[0069] In this process, based on the "horizontal splitting," a "vertical splitting" is carried out. Based on the nodes of the process thinking chain, it is further refined and split into analysis nodes and auxiliary nodes that are executed by the safety hazard analysis model. The analysis is carried out according to the four dimensions of safety hazard analysis, and finally an executable tree structure thinking chain is formed.

[0070] For ease of understanding, Figure 4 This application provides a schematic diagram of an executable intelligent agent tree-shaped thought chain formed through "horizontal splitting" and "vertical splitting" in an embodiment of the present application, as shown below. Figure 4 As shown, the "Analysis of Safety Hazards and Rectification Measures in Galvanizing" instruction is used as input for analysis. The process includes degreasing, pickling, washing, fluxing, drying, preheating, zinc pot immersion plating, cooling, and passivation. First, the process is broken down horizontally into multiple process-related thought chain nodes, each corresponding to a specific galvanizing process. Taking the preheating process as an example, it is vertically broken down according to the analysis of basic information and the sequence of safety hazard factors, forming analysis nodes and auxiliary nodes that can be executed by the analysis model, creating a tree-structured thought chain. The goal is to integrate the safety hazards and rectification suggestions from each process, forming a summary report, and finally outputting a safety hazard and rectification measure analysis report.

[0071] Specifically, each process thinking chain node is vertically split into analysis nodes and auxiliary nodes executed by the safety hazard analysis model, forming a tree-structured thinking chain. This can include: splitting the process thinking chain node into analysis nodes and auxiliary nodes; analysis nodes are used to analyze basic process information and the four elements of safety hazards; auxiliary nodes are used to standardize the output results of analysis nodes and store short-term and long-term memory.

[0072] The four elements of safety hazards include: unsafe acts of people, unsafe conditions of objects, management deficiencies, and unsafe environmental factors.

[0073] Among them, the analysis nodes are mainly used to analyze the basic information of the process and the safety hazard analysis dimensions of related processes. The basic information of the process may include: process design, equipment data, construction drawings, actual site conditions, etc. The safety hazard analysis dimensions may include: unsafe human behavior, unsafe conditions of objects, management defects, and unsafe environmental factors.

[0074] In addition to standardizing the output of large models, auxiliary nodes are also responsible for the agent's memory function, which includes short-term memory and long-term memory.

[0075] Short-term memory is used to store intermediate results and session context of thought chain node processing in order to support the loop sessions and error analysis and troubleshooting required in thought chain analysis.

[0076] Long-term memory is used to record the final output of the agent and the intermediate thinking process, and stores them in a vector database for use in the agent's continuous learning function.

[0077] Step S206: Traverse the tree-structured thought chain based on a pre-set order and analyze the safety hazards of each process.

[0078] Specifically, traversing the tree-structured thinking chain based on a pre-set order and analyzing the safety hazards of each process can include: performing a depth-first traversal of the tree-structured thinking chain in a left-to-right and top-to-bottom order to analyze the safety hazards of each process.

[0079] The analysis process utilizes a tree-structured thought chain to identify security vulnerabilities. At each node, the following techniques are employed: retrieval enhancement generation technology is used to acquire relevant data from the AI ​​data lake; the latest professional information is obtained from external knowledge bases through model context protocols; and LangChain is used to construct query information for the large model. This effectively reduces the illusion of a large model and improves the accuracy of the analysis results.

[0080] Step S207: Summarize the safety hazard analysis results of all processes, generate a safety hazard report and rectification measures for safety early warning.

[0081] Step S208: Replace the safety hazard analysis model and rerun the analysis process to generate a second safety hazard report.

[0082] One approach is to use another safety hazard analysis model to re-analyze the process and generate a new safety hazard report to verify the results of the previous analysis.

[0083] Step S209: Compare the first generated safety hazard report and the second generated safety hazard report to obtain the similarity score.

[0084] Among them, the similarity ratio characterizes the applicability of the tree-structured thinking chain. If the similarity ratio is too low, it can be considered that there is a major conflict between the two results, and the applicability of the tree-structured thinking chain is not strong.

[0085] Step S210: If the comparison similarity does not meet the preset similarity threshold, adjust the tree structure thinking chain.

[0086] This allows for the examination of the original thought chain structure, and the readjustment of the tree-structured thought chain based on the agent's processing steps, the thinking process of the large model, and the intermediate processing results to ensure its applicability.

[0087] The safety early warning method for galvanizing production lines provided in this invention is not only highly professional, capable of analyzing potential safety hazards and corresponding rectification measures based on the actual layout of galvanizing production lines in the steel industry after construction, but also comprehensive. It can investigate and analyze the actual safety hazards existing in each process from four dimensions: unsafe human behavior, unsafe conditions of equipment, management deficiencies, and unsafe environmental factors. Furthermore, it is advanced, employing the latest large-scale intelligent model technology to help safety supervisors discover potential safety hazards and help managers fully understand the potential safety hazards and corresponding rectification measures after the production line is operational. This allows for preventative work before accidents occur, early elimination of safety hazards, reduction of the personal injury accident rate in galvanizing production lines, and prevention of serious accidents.

[0088] Example 3

[0089] Corresponding to the above method embodiments, this invention provides a safety early warning device for a galvanizing production line. Figure 5 This is a schematic diagram of the structure of a safety early warning device for a galvanizing production line provided in an embodiment of the present invention, as shown below. Figure 5 As shown, the safety early warning device for this galvanizing production line may include:

[0090] The multi-source data acquisition module 301 is used to acquire multi-source data from the galvanizing process production line.

[0091] The knowledge graph construction module 302 is used to perform multimodal parsing and vectorization processing on multi-source data, construct a knowledge graph of the galvanizing process production line, and store it in the AI ​​data lake.

[0092] The model adjustment module 303 is used to read the knowledge graph from the AI ​​data lake to adjust the general large model and obtain the safety hazard analysis model corresponding to the galvanizing process.

[0093] The galvanizing process splitting module 304 is used to horizontally split the galvanizing process into multiple process thought chain nodes according to the process sequence.

[0094] The process thinking chain node splitting module 305 is used to vertically split each process thinking chain node into analysis nodes and auxiliary nodes that are executed by the safety hazard analysis model, forming a tree-structured thinking chain.

[0095] The safety hazard analysis module 306 is used to traverse the tree-structured thinking chain based on a pre-set order and analyze the safety hazards of each process.

[0096] The safety early warning module 307 is used to summarize the safety hazard analysis results of all processes, generate safety hazard reports and rectification measures for safety early warning.

[0097] The safety early warning device for a galvanizing production line provided in this invention can acquire multi-source data from the galvanizing production line, perform multi-modal analysis and vectorization processing on the multi-source data, construct a knowledge graph of the galvanizing production line and store it in an AI data lake, read the knowledge graph from the AI ​​data lake to adjust the general large model, and obtain a safety hazard analysis model corresponding to the galvanizing process. The galvanizing process is horizontally divided into multiple process thought chain nodes according to the process sequence, and each process thought chain node is vertically divided into analysis nodes and auxiliary nodes executed by the safety hazard analysis model, forming a tree-structured thought chain. Based on a pre-set order, the tree-structured thought chain is traversed to analyze the safety hazards of each process, and the analysis results of all process safety hazards are summarized to generate a safety hazard report and rectification measures for safety early warning. This method helps managers in the galvanizing and other steel industries to fully understand the potential safety hazards after the production line is operational and the corresponding rectification measures, as well as to carry out preventative work before accidents occur, eliminate safety hazards in advance, reduce the personal injury accident rate of the galvanizing production line, and prevent serious accidents.

[0098] In some embodiments, the knowledge graph construction module is also used to parse the text, image and video content in multi-source data using a multimodal large model to obtain parsed data; and to vectorize the parsed data and store it in the AI ​​data lake.

[0099] In some embodiments, the model adjustment module is also used to extract galvanizing process design, safety production standards and accident information from the AI ​​data lake; to train a general large model using the galvanizing process design, safety production standards and accident information to obtain a safety hazard analysis model corresponding to the galvanizing process; the safety hazard analysis model has professional characteristics for safety hazard investigation.

[0100] In some embodiments, the galvanizing process splitting module is also used to split the galvanizing process into degreasing nodes, pickling nodes, water washing nodes, fluxing nodes, drying nodes, preheating nodes, zinc pot immersion galvanizing nodes, cooling nodes, and passivation nodes according to the process sequence; and to establish a corresponding process thinking chain node for each node.

[0101] In some embodiments, the process thinking chain node splitting module is further used to split the process thinking chain node into analysis nodes and auxiliary nodes; the analysis node is used to analyze the basic information of the process and the four elements of safety hazards; the four elements of safety hazards include: unsafe human behavior, unsafe conditions of objects, management deficiencies and unsafe environmental factors; the auxiliary node is used to standardize the output results of the analysis node, and to store short-term memory and long-term memory.

[0102] In some embodiments, the safety hazard analysis module is further configured to perform a depth-first traversal of the tree-structured thinking chain in a left-to-right, top-to-bottom order to analyze safety hazards in each process; wherein, the safety hazard analysis is performed by the nodes of the tree-structured thinking chain, and at each node, the following are employed: retrieval enhancement generation technology is used to obtain relevant data from the AI ​​data lake; the latest professional information is obtained from an external knowledge base through a model context protocol; and LangChain is used to construct question information for the large model.

[0103] In some embodiments, the safety warning module is also used to replace the safety hazard analysis model and rerun the analysis process to generate a second safety hazard report; compare the first generated safety hazard report and the second safety hazard report to obtain the comparison similarity; if the comparison similarity does not meet the preset similarity threshold, adjust the tree structure thinking chain.

[0104] The device provided in this embodiment of the invention has the same implementation principle and technical effect as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0105] Example 4

[0106] This invention also provides an electronic device for a safety early warning method for operating the above-mentioned galvanizing production line; see [link to related documentation]. Figure 6 The diagram shows the structure of an electronic device, which includes a memory 400 and a processor 401. The memory 400 stores one or more computer instructions, which are executed by the processor 401 to implement the aforementioned safety warning method for the galvanizing production line.

[0107] Furthermore, Figure 6 The electronic device shown also includes a bus 402 and a communication interface 403. The processor 401, the communication interface 403 and the memory 400 are connected via the bus 402.

[0108] The memory 400 may include high-speed random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 403 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc. The bus 402 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 6 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0109] Processor 401 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 401 or by instructions in software form. Processor 401 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this invention. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this invention can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software module can reside in a readily available storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory 400, and processor 401 reads information from memory 400 and, in conjunction with its hardware, completes the steps of the method described in the foregoing embodiments.

[0110] This invention also provides a computer-readable storage medium storing computer-executable instructions. When these computer-executable instructions are called and executed by a processor, they cause the processor to implement the aforementioned safety warning method for the galvanizing production line. For specific implementation details, please refer to the method embodiments, which will not be repeated here.

[0111] The computer program product for the safety early warning method for a galvanizing production line provided in this embodiment of the invention includes a computer-readable storage medium storing non-volatile program code executable by a processor. The instructions included in the program code can be used to execute the methods described in the preceding method embodiments. For specific implementation details, please refer to the method embodiments, which will not be repeated here.

[0112] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0113] In the several embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the coupling or direct coupling or communication connection shown or discussed may be through some communication interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0114] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0115] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0116] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0117] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A safety early warning method for a galvanizing production line, characterized in that, The method includes: Obtain multi-source data from the galvanizing process production line; The multi-source data is subjected to multimodal parsing and vectorization processing to construct a knowledge graph of the galvanizing process production line and store it in the AI ​​data lake; The knowledge graph is read from the AI ​​data lake and the general model is adjusted to obtain a safety hazard analysis model corresponding to the galvanizing process; The galvanizing process is horizontally broken down into multiple process nodes according to the sequence of operations; Each of the aforementioned process thinking chain nodes is vertically divided into analysis nodes and auxiliary nodes that are executed by the safety hazard analysis model, forming a tree-structured thinking chain; The tree-structured thought chain is traversed according to a pre-set order to analyze the safety hazards of each process. The results of safety hazard analysis for all processes are summarized, and a safety hazard report and rectification measures are generated for safety early warning. The auxiliary node is used to normalize the output of the analysis node and to store short-term and long-term memory. Short-term memory is used to store intermediate results and session context of thought chain node processing in order to support the loop sessions and error analysis and troubleshooting required in thought chain analysis. Among them, long-term memory is used to record the final result of the agent's output and the intermediate thinking process, and store them in a vector database for use in the agent's continuous learning function. The method of horizontally dividing the galvanizing process into multiple process nodes according to the process sequence includes: The galvanizing process is broken down into the following steps in sequence: degreasing, pickling, washing, fluxing, drying, preheating, zinc pot immersion, cooling, and passivation. Establish a corresponding process thinking chain node for each node; Among them, the analytical thinking chain of constructing intelligent agents by using Dify or Coze tools to compile workflows is used to break down the entire galvanizing process into a chain-like horizontal breakdown. Each node in the chain is a process that constitutes the galvanizing process. The step of vertically splitting each of the process thought chain nodes into analysis nodes and auxiliary nodes executed by the safety hazard analysis model to form a tree-structured thought chain includes: The process thinking chain nodes are broken down into analysis nodes and auxiliary nodes; The analysis node is used to analyze basic process information and the four elements of safety hazards; the four elements of safety hazards include: unsafe acts of people, unsafe conditions of equipment, management deficiencies, and unsafe environmental factors; The auxiliary node is used to normalize the output of the analysis node and to store short-term and long-term memory.

2. The method according to claim 1, characterized in that, The process of performing multimodal parsing and vectorization on the multi-source data, constructing a knowledge graph of the galvanizing process production line, and storing it in the AI ​​data lake includes: A multimodal large model is used to parse the text, image, and video content in the multi-source data to obtain the parsed data; The parsed data is vectorized and stored in the AI ​​data lake.

3. The method according to claim 1, characterized in that, The step of reading the knowledge graph from the AI ​​data lake and adjusting the general model to obtain the safety hazard analysis model corresponding to the galvanizing process includes: Extract galvanizing process design, safety production standards, and accident information from the AI ​​data lake; The general model is trained using the galvanizing process design, the safety production standards, and the accident information to obtain a safety hazard analysis model corresponding to the galvanizing process; the safety hazard analysis model has professional characteristics for safety hazard investigation.

4. The method according to claim 1, characterized in that, The process of traversing the tree-structured thought chain according to a pre-set order and analyzing the safety hazards of each process includes: Perform a depth-first traversal of the tree-structured thinking chain in order from left to right and from top to bottom to analyze the safety hazards of each process. The analysis of security risks is conducted through the analysis nodes of the tree-structured thinking chain, and the following is adopted at each analysis node: The retrieval enhancement generation technology obtains relevant data from the AI ​​data lake; Retrieve the latest professional information from external knowledge bases through model context protocols; Use LangChain to construct query information for large models.

5. The method according to claim 1, characterized in that, The method further includes summarizing the safety hazard analysis results of all processes to generate a safety hazard report and rectification measures for safety early warning. Replace the safety hazard analysis model and rerun the analysis process to generate a second safety hazard report; The first safety hazard report and the second safety hazard report are compared to obtain the similarity score. If the comparison similarity does not meet the preset similarity threshold, the tree-structured thought chain will be adjusted.

6. A safety early warning device for a galvanizing production line, characterized in that, The apparatus for implementing a safety early warning method for a galvanizing process production line according to any one of 1 to 5, the apparatus comprising: The multi-source data acquisition module is used to acquire multi-source data from the galvanizing process production line. The knowledge graph construction module is used to perform multimodal parsing and vectorization processing on the multi-source data, construct a knowledge graph of the galvanizing process production line, and store it in the AI ​​data lake; The model adjustment module is used to read the knowledge graph from the AI ​​data lake and adjust the general large model to obtain the safety hazard analysis model corresponding to the galvanizing process. The galvanizing process breakdown module is used to horizontally break down the galvanizing process into multiple process thought chain nodes according to the process sequence; The process thinking chain node splitting module is used to vertically split each process thinking chain node into analysis nodes and auxiliary nodes that are executed by the safety hazard analysis model, forming a tree-structured thinking chain; The safety hazard analysis module is used to traverse the tree-structured thought chain based on a pre-set order and analyze the safety hazards of each process. The safety early warning module is used to summarize the safety hazard analysis results of all processes, generate safety hazard reports and rectification measures for safety early warning.

7. An electronic device, characterized in that, The device includes a processor and a memory, the memory storing computer-executable instructions that can be executed by the processor, the processor executing the computer-executable instructions to implement the safety early warning method for a galvanizing production line according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when invoked and executed by a processor, cause the processor to implement the safety early warning method for the galvanizing production line as described in any one of claims 1 to 5.

Images