Early stage automatic analysis method and device of human error probability analysis system

Abstract

The invention relates to the technical field of data analysis, in particular to an early stage automatic analysis method and device of a human error probability analysis system.The method comprises the steps that an early stage task of the human error probability analysis system is disassembled into a plurality of analysis subtasks; embedding expert knowledge on the basis of a prompt project, establishing an agent for each analysis subtask on a large language model, and realizing information transmission among a plurality of agents through a shared memory or an API (Application Program Interface) calling mechanism; early stage analysis of the human error probability analysis system is achieved through the multiple agents, and the multiple agents analyze the early stage task and the multiple analysis sub-tasks according to the analysis sequence. Therefore, the problems that in the prior art, the early stage of a human error probability analysis system depends on a large amount of manual judgment, and when the human error probability analysis system faces complex and changeable operation scenes, the analysis process is tedious and takes a long time are solved.

Description

Technical Field

[0001] The present application relates to the field of data analysis technology, and in particular to an early-stage automatic analysis method and device for a human error probability analysis system. Background Art

[0002] In the field of Human Reliability Analysis (HRA), the human error probability analysis system, based on event and condition analysis, is one of the most advanced third-generation HRA methods, widely used in high-risk industries such as nuclear power plants. This method uses a structured analytical framework to systematically model and evaluate potential human errors that operators may make in complex situations, thereby improving system safety and reliability.

[0003] The human error probability analysis system based on event and condition analysis achieves a quantitative assessment of the probability of human error through preliminary steps such as scenario analysis, human error pattern recognition, and task modeling. However, this method relies heavily on expert experience and manual judgment. When faced with complex and changing operational scenarios, the analysis process is cumbersome, time-consuming, and susceptible to subjective factors. Furthermore, it struggles to meet the demands of rapid decision-making for emergency response tasks that require high real-time performance. Summary of the Invention

[0004] The present application provides an early-stage automatic analysis method, device, and equipment for a human error probability analysis system to solve the problems that the early stages of the human error probability analysis system in the related art rely on a large amount of manual judgment, and when faced with complex and changeable operating scenarios, there are problems such as cumbersome analysis processes and long time consumption.

[0005] The first aspect of the present application provides an automatic early stage analysis method for a human error probability analysis system, comprising the following steps: decomposing the early stage tasks of the human error probability analysis system into multiple analysis subtasks; embedding expert knowledge based on prompt engineering, building an intelligent agent for each analysis subtask on a large language model, and realizing information transmission between multiple intelligent agents through shared memory or API call mechanism; utilizing multiple intelligent agents to realize early stage analysis of the human error probability analysis system, wherein the multiple intelligent agents analyze the early stage tasks and multiple analysis subtasks in an analysis order.

[0006] Optionally, early-stage tasks include scenario analysis, human error incidents, and mission-critical failure modeling.

[0007] Optionally, scenario analysis includes the subtask of writing an operation process narrative, the subtask of identifying human error events, and the subtask of situational dimension analysis; human error events include the subtask of defining human error events and the subtask of identifying key tasks; and key task failure modeling includes the subtask of describing task characteristics and the subtask of identifying applicable cognitive error patterns.

[0008] Optionally, the task flow of multiple intelligent agents includes: multiple intelligent agents for scenario analysis, which perform the operation process narrative writing subtask, the human error event identification subtask, and the situational dimension analysis subtask; multiple intelligent agents for human error events, which perform the human error event definition subtask and the key task identification subtask based on the analysis results and case description information of the human error event identification subtask; multiple intelligent agents for key task failure modeling, which perform the task feature description subtask and the applicable cognitive error pattern identification subtask based on the analysis results and case description information of the key task identification subtask.

[0009] Optionally, the content of the situational dimension analysis subtask includes environmental content, system content, personnel content, and task content.

[0010] Optionally, the intelligent agent analyzes the environment content, system content, personnel content and task content in sequence, and uses the analysis result of the current content as input for the next content analysis until all content analysis is completed.

[0011] The second aspect of the present application provides an early stage automatic analysis method for a human error probability analysis system, comprising the following steps: obtaining case description information input by a user; identifying early stage tasks of the human error probability analysis system in the case description information; calling multiple intelligent agents for early stage tasks, inputting the case description information into the multiple intelligent agents, and the multiple intelligent agents output analysis results of the early stage tasks, wherein the intelligent agents are built on a large language model based on expert knowledge embedded in prompt engineering, and information transmission is achieved between the multiple intelligent agents through shared memory or API call mechanism.

[0012] The third aspect of the present application provides an automatic early stage analysis device for a human error probability analysis system, including: a disassembly module, used to disassemble the early stage tasks of the human error probability analysis system into multiple analysis subtasks; a building module, used to embed expert knowledge based on prompt engineering, build an intelligent agent for each analysis subtask on a large language model, and realize information transmission between multiple intelligent agents through shared memory or API call mechanism; a first analysis module, used to use multiple intelligent agents to realize early stage analysis of the human error probability analysis system, wherein the multiple intelligent agents analyze the early stage tasks and multiple analysis subtasks in an analysis order.

[0013] The fourth aspect of the present application provides an early stage automatic analysis device for a human error probability analysis system, including: an acquisition module for acquiring case description information input by a user; an identification module for identifying early stage tasks of the human error probability analysis system in the case description information; a second analysis module for calling multiple intelligent agents for early stage tasks, inputting case description information into multiple intelligent agents, and multiple intelligent agents output analysis results of the early stage tasks, wherein the intelligent agents are built on a large language model based on expert knowledge embedded in prompt engineering, and information transmission is achieved between multiple intelligent agents through shared memory or API call mechanism.

[0014] The fifth aspect of the present application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and runnable on the processor, wherein the processor executes the program to implement an early-stage automatic analysis method of a human error probability analysis system as in the above-mentioned embodiment.

[0015] Therefore, this application has the following beneficial effects:

[0016] The embodiment of the present application decomposes the early steps in the human error probability analysis system into several analysis subtasks, builds a corresponding intelligent agent for each analysis subtask, embeds expert knowledge through prompt engineering, and guides the large language model to generate structured analysis content. Among them, each subtask intelligent agent realizes information transmission through shared memory or API call mechanism, realizes collaborative work and information flow among multiple intelligent agents, and ensures the full process automation from original input to final analysis result. In this way, it solves the problem that the early stage of the human error probability analysis system in the related technology relies on a lot of manual judgment, and when faced with complex and changeable operation scenarios, there are problems such as cumbersome analysis process and long time consumption.

[0017] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0019] Figure 1 This is a flow chart of an early-stage automatic analysis method of a human error probability analysis system provided according to one embodiment of the present application;

[0020] Figure 2 An example diagram of scenario analysis splitting provided according to an embodiment of the present application

[0021] Figure 3 This is an example diagram of splitting human error events according to one embodiment of the present application;

[0022] Figure 4 This is an example diagram of the critical task failure modeling splitting provided according to one embodiment of the present application;

[0023] Figure 5 This is an example diagram of multi-agent analysis based on a large language model according to one embodiment of the present application;

[0024] Figure 6 A flowchart of an early-stage automatic analysis method of a human error probability analysis system provided according to another embodiment of the present application;

[0025] Figure 7 A block diagram of an early-stage automatic analysis device of a human error probability analysis system according to one embodiment of the present application;

[0026] Figure 8 A block diagram of an early-stage automatic analysis device of a human error probability analysis system according to another embodiment of the present application;

[0027] Figure 9 Schematic diagram of the structure of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0028] The following describes in detail embodiments of the present application. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0029] The following describes the early-stage automatic analysis method, device and equipment of the human error probability analysis system of the embodiment of the present application with reference to the accompanying drawings. In response to the problems mentioned in the above background technology, the present application provides an early-stage automatic analysis method of the human error probability analysis system, in which the early steps in the human error probability analysis system are structured and decomposed into several analysis subtasks. For each analysis subtask, a corresponding intelligent agent is constructed, and expert knowledge is embedded through prompt engineering to guide the large language model to generate structured analysis content. Among them, each subtask intelligent agent realizes information transmission through shared memory or API call mechanism, realizes collaborative work and information flow among multiple intelligent agents, and ensures the full process automation processing from original input to final analysis result.

[0030] Specifically, Figure 1 A flowchart of an early-stage automatic analysis method of a human error probability analysis system provided in an embodiment of the present application.

[0031] like Figure 1As shown, the early stage automatic analysis method of the human error probability analysis system includes the following steps:

[0032] In step S101 , the early stage tasks of the human error probability analysis system are decomposed into multiple analysis subtasks.

[0033] Early-stage tasks include scenario analysis, human error incidents, and critical task failure modeling. Scenario analysis includes the subtasks of writing operational process narratives, identifying human error incidents, and analyzing situational dimensions. Human error incidents include the subtasks of defining human error incidents and identifying critical tasks. Critical task failure modeling includes the subtasks of describing task characteristics and identifying applicable cognitive error patterns.

[0034] It is understandable that the embodiment of the present application can refer to the analysis manual of the human error probability analysis system to structure and decompose the early stage steps into several subtasks, mainly including: scenario analysis, HFE (Human Failure Event), and key task failure modeling. Figure 2 As shown in , scenario analysis includes three subtasks, including operation process narrative writing, HFE identification, and situation dimension analysis (environment, system, personnel, and task); Figure 3 As shown in, HFE analysis includes two subtasks, namely HFE definition and key task identification; Figure 4 As shown in Figure 2, key task failure modeling can be divided into task feature description and applicable cognitive error pattern recognition.

[0035] In step S102, based on the expert knowledge embedded in the prompt engineering, an intelligent agent is built for each analysis subtask on the large language model, and information is transmitted between multiple intelligent agents through shared memory or API call mechanism.

[0036] It is understood that the embodiments of the present application can build multiple intelligent agents on a large language model to implement a human error probability analysis system based on the embedded expert knowledge of the prompt engineering. Specifically, the embodiments of the present application construct a corresponding intelligent agent for each analysis subtask, and guide the large language model to generate structured analysis content by embedding expert knowledge through the prompt engineering. Among them, the large language model can be selected from Claude 3.5, GPT-4, etc., without specific limitation.

[0037] In the actual execution process, each subtask agent realizes information transmission through shared memory or API call mechanism, realizing the automatic flow from operation process description writing, human error event identification to situation dimension analysis. The specific task transmission relationship between them is as follows: Figure 5 shown.

[0038] In step S103, multiple agents are used to implement early stage analysis of the human error probability analysis system, wherein the multiple agents analyze the early stage tasks and multiple analysis subtasks in an analysis order.

[0039] In one embodiment of the present application, the task flow of multiple intelligent agents includes: multiple intelligent agents for scenario analysis, which perform the operation process narrative writing subtask, the human error event identification subtask, and the situational dimension analysis subtask; multiple intelligent agents for human error events, which perform the human error event definition subtask and the key task identification subtask based on the analysis results and case description information of the human error event identification subtask; multiple intelligent agents for key task failure modeling, which perform the task feature description subtask and the applicable cognitive error pattern identification subtask based on the analysis results and case description information of the key task identification subtask.

[0040] Specifically, after the user enters the case information, the multiple scenario analysis agents in this embodiment simultaneously analyze the operation process narrative writing and human error event identification. The analysis results and the user-entered case information are then fed into the scenario dimension analysis subtask, which includes environmental, system, personnel, and task aspects. The agent analyzes the environmental, system, personnel, and task aspects in sequence, using the analysis results of the current content as input for the next content analysis until all content analysis is complete.

[0041] In another embodiment, multiple intelligent agents of human error events can use the analysis results of the human error event subtask and the case information input by the user as input to execute the human error event definition subtask and the key task identification subtask. During the actual execution process, the key task identification subtask can directly analyze the key tasks based on the case information provided by the user.

[0042] In another embodiment, multiple agents modeling critical task failures perform a task feature description subtask and an applicable cognitive failure pattern identification subtask based on the critical tasks analyzed by the critical task identification subtask and user-input case information. The applicable cognitive failure pattern identification subtask inputs user input information and outputs potential cognitive failure patterns applicable to the current situation.

[0043] In summary, the early-stage automatic analysis method of the human error probability analysis system of the embodiment of the present application realizes collaborative work and information flow among multiple agents, ensuring the full-process automated processing from original input to final analysis results.

[0044] Secondly, this application also provides another automatic analysis method for the early stage of the human error probability analysis system, such as Figure 6 As shown, the following steps are included:

[0045] In step S201, the case description information input by the user is obtained.

[0046] It is understandable that the embodiment of the present application can input case description information through the Web interface, such as a human error scenario, where the user inputs "the operator failed to start the injection and cooling operation."

[0047] In step S202, the early stage tasks of the human error probability analysis system in the case description information are identified.

[0048] In step S203, multiple agents of early stage tasks are called, case description information is input into multiple agents, and multiple agents output analysis results of early stage tasks. Among them, the agents are built on a large language model based on the embedded expert knowledge of prompt engineering, and information is transmitted between multiple agents through shared memory or API call mechanism.

[0049] Specifically, after receiving user input, the corresponding agent is called according to the task flow of the human error probability analysis system. Each agent corresponds to multiple subtasks in the human error probability analysis system. Specifically:

[0050] Step 1, scenario analysis, includes three subtasks: Operational Process Narrative Writing Subtask 1.1, Human Error Event Identification Subtask 1.2, and Situational Dimension Analysis Subtask 1.3. Operational Process Narrative Writing Subtask 1.1 presents an operational process narrative for the scenario of "the operator failed to initiate the injection and cooling operation." The output is the context, timeline, operational steps, and relevant environmental information for the operator's failure to initiate the injection and cooling operation. Human Error Event Identification Subtask 1.2 identifies and categorizes the human error events in the case, outputting specific human error events (such as omitted operations and misjudgments) with detailed definitions. Situational Dimension Analysis Subtask 1.3 analyzes the environmental and situational context, system context, personnel context, and task context. The output is information on the physical conditions of the work environment, system status and external influencing factors, system configuration, operating status, and relevant technical parameters, the operator's experience, training, and psychological state, as well as the task's objectives, execution difficulty, and required resources.

[0051] Step 2, Human Error Events, includes two subtasks: Human Error Event Definition Subtask 2.1 and Critical Task Identification Subtask 2.2. The output of Human Error Event Definition Subtask 2.1 is the specific manifestations of the human error event and its scope of impact. Critical Task Identification Subtask 2.2 lists the critical tasks that require special attention in this scenario and assesses their importance.

[0052] Step 3, Critical Task Failure Modeling, includes Task Characterization Subtask 3.1 and Applicable Cognitive Failure Pattern Identification Subtask 3.2. The output of Task Characterization Subtask 3.1 is the detailed characteristics of each critical task, including task type, complexity, and risk points. The output of Applicable Cognitive Failure Pattern Identification Subtask 3.2 is the identification of possible cognitive failure patterns for each critical task, along with corresponding preventive measures or improvement suggestions.

[0053] According to the early-stage automatic analysis method of the human error probability analysis system proposed in the embodiment of the present application, the early steps in the human error probability analysis system are structured and decomposed into several analysis subtasks. For each analysis subtask, a corresponding intelligent agent is constructed, and expert knowledge is embedded through prompt engineering to guide the large language model to generate structured analysis content. Among them, each subtask intelligent agent realizes information transmission through shared memory or API call mechanism, realizes collaborative work and information flow among multiple intelligent agents, and ensures the full process automation processing from original input to final analysis result.

[0054] Next, an early-stage automatic analysis device of a human error probability analysis system proposed in an embodiment of the present application will be described with reference to the accompanying drawings.

[0055] Figure 7 4 is a block diagram of an early-stage automatic analysis device of a human error probability analysis system according to an embodiment of the present application.

[0056] like Figure 7 As shown, the early stage automatic analysis device 10 of the human error probability analysis system includes: a disassembly module 101 , a construction module 102 and a first analysis module 103 .

[0057] Among them, the disassembly module 101 is used to decompose the early stage tasks of the human error probability analysis system into multiple analysis subtasks; the construction module 102 is used to embed expert knowledge based on the prompt engineering, build an intelligent agent for each analysis subtask on the large language model, and realize information transmission between multiple intelligent agents through shared memory or API call mechanism; the first analysis module 103 is used to use multiple intelligent agents to realize the early stage analysis of the human error probability analysis system, wherein the multiple intelligent agents analyze the early stage tasks and multiple analysis subtasks in the analysis order.

[0058] In one embodiment of the present application, early stage tasks include scenario analysis, human error events, and critical mission failure modeling.

[0059] In one embodiment of the present application, scenario analysis includes an operation process narrative writing subtask, a human error event identification subtask, and a situational dimension analysis subtask; human error events include a human error event definition subtask and a key task identification subtask; and key task failure modeling includes a task feature description subtask and an applicable cognitive error pattern identification subtask.

[0060] In one embodiment of the present application, the task flow of multiple intelligent agents includes: multiple intelligent agents for scenario analysis, which perform the operation process narrative writing subtask, the human error event identification subtask, and the situational dimension analysis subtask; multiple intelligent agents for human error events, which perform the human error event definition subtask and the key task identification subtask based on the analysis results and case description information of the human error event identification subtask; multiple intelligent agents for key task failure modeling, which perform the task feature description subtask and the applicable cognitive error pattern identification subtask based on the analysis results and case description information of the key task identification subtask.

[0061] In one embodiment of the present application, the content of the situational dimension analysis subtask includes environmental content, system content, personnel content, and task content.

[0062] In one embodiment of the present application, the intelligent agent analyzes the environmental content, system content, personnel content and task content in sequence, and uses the analysis results of the current content as input for the next content analysis until all content analysis is completed.

[0063] Secondly, the present application also provides another automatic analysis device 20 for the early stage of the human error probability analysis system, such as Figure 8 As shown, it includes: an acquisition module 201, an identification module 202 and a second analysis module 203

[0064] Among them, the acquisition module 201 is used to obtain the case description information input by the user; the identification module 202 is used to identify the early stage tasks of the human error probability analysis system in the case description information; the second analysis module 203 is used to call multiple intelligent agents of the early stage tasks, input the case description information into multiple intelligent agents, and multiple intelligent agents output the analysis results of the early stage tasks, wherein the intelligent agents are based on the embedded expert knowledge of the prompt engineering and are built on the large language model, and information transmission is realized between multiple intelligent agents through shared memory or API call mechanism.

[0065] It should be noted that the above explanation of the embodiment of the early stage automatic analysis method of the human error probability analysis system is also applicable to the early stage automatic analysis device of the human error probability analysis system of this embodiment, and will not be repeated here.

[0066] According to the early stage automatic analysis device of the human error probability analysis system proposed in the embodiment of the present application, the early steps in the human error probability analysis system are structured and decomposed into several analysis subtasks. For each analysis subtask, a corresponding intelligent agent is constructed, and expert knowledge is embedded through prompt engineering to guide the large language model to generate structured analysis content. Among them, each subtask intelligent agent realizes information transmission through shared memory or API call mechanism, realizes collaborative work and information flow among multiple intelligent agents, and ensures the full process automation processing from original input to final analysis result.

[0067] Figure 9 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device may include:

[0068] A memory 901 , a processor 902 , and a computer program stored in the memory 901 and executable on the processor 902 .

[0069] When the processor 902 executes the program, the early-stage automatic analysis method of the human error probability analysis system provided in the above embodiment is implemented.

[0070] Furthermore, the electronic device further includes:

[0071] The communication interface 903 is used for communication between the memory 901 and the processor 902 .

[0072] The memory 901 is used to store computer programs that can be run on the processor 902 .

[0073] The memory 901 may include a high-speed RAM (Random Access Memory) memory, and may also include a non-volatile memory, such as at least one disk memory.

[0074] If the memory 901, the processor 902, and the communication interface 903 are implemented independently, the communication interface 903, the memory 901, and the processor 902 can be connected to each other via a bus and communicate with each other. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 9 Only one thick line is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0075] Optionally, in a specific implementation, if the memory 901, the processor 902 and the communication interface 903 are integrated on a chip, the memory 901, the processor 902 and the communication interface 903 can communicate with each other through an internal interface.

[0076] The processor 902 may be a CPU (Central Processing Unit), or an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present application.

[0077] An embodiment of the present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned early-stage automatic analysis method of the human error probability analysis system.

[0078] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and features of different embodiments or examples without contradiction.

[0079] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of technical features indicated. Thus, a feature specified as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of this application, "N" means at least two, for example, two, three, etc., unless otherwise specifically defined.

[0080] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or N executable instructions for implementing a custom logical function or process step, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed in a different order than shown or discussed, including performing functions in a substantially simultaneous manner or in a reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application pertain.

[0081] It should be understood that various parts of the present application can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, the steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement the method: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array, a field programmable gate array, etc.

[0082] A person skilled in the art may understand that all or part of the steps carried out in the method for implementing the above-mentioned embodiment may be completed by instructing the relevant hardware through a program, and the above-mentioned program may be stored in a computer-readable storage medium, which, when executed, includes one of the steps of the method embodiment or a combination thereof.

[0083] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. An early-stage automatic analysis method for a human error probability analysis system, characterized in that: The following steps are involved: Decompose the early stage tasks of the human error probability analysis system into multiple analysis subtasks; Based on the embedding of expert knowledge in prompt engineering, an intelligent agent is built for each analysis subtask on the large language model, and information is transferred between the multiple intelligent agents through shared memory or API call mechanism; The multiple intelligent agents are used to implement the early stage analysis of the human error probability analysis system, wherein the multiple intelligent agents analyze the early stage tasks and multiple analysis subtasks in an analysis order.

2. The early stage automatic analysis method of the human error probability analysis system according to claim 1, characterized in that: The early phase tasks include scenario analysis, human error incidents, and mission-critical failure modeling.

3. The early stage automatic analysis method of the human error probability analysis system according to claim 2, characterized in that: The scenario analysis includes the subtask of writing an operation process narrative, the subtask of identifying human error events, and the subtask of situational dimension analysis. The human error events include the subtask of defining human error events and the subtask of identifying key tasks. The key task failure modeling includes the subtask of describing task characteristics and the subtask of identifying applicable cognitive error patterns.

4. The early stage automatic analysis method of the human error probability analysis system according to claim 3, characterized in that: The task flow of the multiple agents includes: The multiple intelligent agents of the scenario analysis perform the subtasks of writing the operation process description, identifying the human error event, and analyzing the situational dimensions; The multiple intelligent agents of the human error event execute the human error event definition subtask and the key task identification subtask according to the analysis results and case description information of the human error event identification subtask; The multiple intelligent agents of the critical task failure modeling perform the task feature description subtask and the applicable cognitive error pattern recognition subtask based on the analysis results and case description information of the critical task identification subtask.

5. The early stage automatic analysis method of the human error probability analysis system according to claim 4, characterized in that: The content of the situational dimension analysis subtask includes environmental content, system content, personnel content and task content.

6. The early stage automatic analysis method of the human error probability analysis system according to claim 5, characterized in that: The intelligent agent analyzes the environmental content, the system content, the personnel content and the task content in sequence, and uses the analysis result of the current content as input for the next content analysis until all content analysis is completed.

7. An early-stage automatic analysis method for a human error probability analysis system, characterized in that: The following steps are involved: Get the case description information entered by the user; Identify the early stage tasks of the human error probability analysis system in the case description information; Call multiple agents for the early stage task, input the case description information into the multiple agents, and the multiple agents output analysis results of the early stage task, wherein the agents are built on a large language model based on expert knowledge embedded in prompt engineering, and information is transmitted between the multiple agents through shared memory or API call mechanism.

8. An early-stage automatic analysis device for a human error probability analysis system, characterized in that: include: A disassembly module is used to decompose the early stage tasks of the human error probability analysis system into multiple analysis subtasks; A building module is used to embed expert knowledge based on prompt engineering and build an intelligent agent for each analysis subtask on the large language model. The multiple intelligent agents can realize information transmission through shared memory or API call mechanism; The first analysis module is used to use the multiple intelligent agents to implement the early stage analysis of the human error probability analysis system, wherein the multiple intelligent agents analyze the early stage tasks and multiple analysis subtasks in an analysis order.

9. An early-stage automatic analysis device for a human error probability analysis system, characterized in that: include: The acquisition module is used to obtain the case description information input by the user; An identification module, configured to identify an early stage task of a human error probability analysis system in the case description information; The second analysis module is used to call multiple agents of the early stage tasks, input the case description information into the multiple agents, and the multiple agents output the analysis results of the early stage tasks, wherein the agents are built on a large language model based on the embedded expert knowledge of the prompt engineering, and information is transmitted between the multiple agents through shared memory or API call mechanism.

10. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the early-stage automatic analysis method of the human error probability analysis system according to any one of claims 1 to 9.

Images