Power system simulation file analysis processing method and device based on intelligent agent, computer equipment and storage medium

Abstract

The embodiment of the invention relates to the field of power systems, and provides an agent-based power system simulation file analysis processing method and device, computer equipment and a storage medium, and the method comprises the steps: obtaining natural language analysis processing demand information and a to-be-analyzed simulation form file; analyzing and processing the demand information according to a natural language to obtain standardized task description information and a retrieval result corresponding to the knowledge base; according to the simulation table file and the retrieval result, an overall analysis target corresponding to the natural language analysis processing demand information is disassembled into at least two structured sub-tasks which can be executed in parallel; according to the task type represented by the standardized task description information, executing the structured sub-task to obtain execution result data; and performing result integration according to the execution result data and the template requirement data, and generating a simulation analysis report associated with the simulation form file. By adopting the method, the reusability of simulation report output can be improved, and the report output efficiency is improved.

Description

Technical Field

[0001] This application relates to the field of power system technology, and in particular to a method, apparatus, computer equipment, and storage medium for analyzing and processing power system simulation files based on intelligent agents. Background Technology

[0002] In the field of power system technology, power systems rely heavily on numerical simulations of power flow, short circuits and transient stability in planning, dispatching and operation and maintenance. Mainstream commercial and open source platforms usually export a large number of result tables in batch processing, and engineers then use table editing tools to perform secondary statistics, graphing and writing analysis conclusions.

[0003] Traditional technologies typically employ a chain of "simulation calculation, result export, manual analysis / script compilation, and manual report writing." While this can accomplish routine statistics and graph generation, it has significant limitations when applied to power system simulation tables.

[0004] Therefore, the traditional method of manually processing simulation reports suffers from problems such as non-reusable simulation report output and low output efficiency. Summary of the Invention

[0005] This application provides a method, apparatus, computer equipment, and computer storage medium for analyzing and processing power system simulation files based on intelligent agents, which can improve the reusability of simulation report output and increase report output efficiency.

[0006] In a first aspect, embodiments of this application provide a method for analyzing and processing power system simulation files based on intelligent agents, including:

[0007] Obtain information on natural language processing requirements and the simulation table file to be analyzed;

[0008] Based on the natural language processing requirements, standardized task description information and retrieval results corresponding to the knowledge base are obtained;

[0009] Based on the simulation table file and the search results, the overall analysis objective corresponding to the natural language analysis and processing requirements information is decomposed into at least two structured subtasks that can be executed in parallel.

[0010] Based on the task type represented by the standardized task description information, the structured subtask is executed to obtain execution result data;

[0011] The results are integrated based on the execution result data and template requirement data to generate a simulation analysis report associated with the simulation table file.

[0012] Secondly, embodiments of this application provide an agent-based power system simulation file analysis and processing apparatus, which has the function of implementing the agent-based power system simulation file analysis and processing method provided in the first aspect above. The function can be implemented by hardware or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above function, and the modules can be software and / or hardware.

[0013] In one possible design, the device includes:

[0014] The information acquisition module is used to acquire information on the requirements for natural language analysis and processing, as well as the simulation table file to be analyzed.

[0015] The retrieval module is used to obtain standardized task description information and retrieval results corresponding to the knowledge base based on the natural language processing requirements information.

[0016] The task decomposition module is used to decompose the overall analysis target corresponding to the natural language analysis and processing requirement information into at least two structured subtasks that can be executed in parallel, based on the simulation table file and the search results.

[0017] The task execution module is used to execute the structured subtasks according to the task type represented by the standardized task description information, and obtain execution result data;

[0018] The report generation module is used to integrate the results based on the execution result data and template requirement data to generate a simulation analysis report associated with the simulation table file.

[0019] In another aspect, this application provides a computer device including at least one connected processor and a memory, wherein the memory is used to store program code and a knowledge base, and the processor is used to call the program code in the memory to execute the methods described in the above aspects.

[0020] In another aspect, this application provides a computer storage medium including instructions that, when executed on a computer, cause the computer to perform the methods described in the above aspects.

[0021] In another aspect, this application provides a computer program product containing instructions that, when run on a computer, cause the computer to perform the methods described in the above aspects.

[0022] Compared to traditional simulation report output methods in conventional technologies, the technical solution of this application embodiment achieves end-to-end automation of analysis and report generation. It can maintain system stability under scenarios with field changes, differences in worksheet structure, and mixed use of multiple software, significantly reducing manual intervention and script maintenance costs, improving analysis efficiency, enhancing the reusability of simulation report output, and improving report output efficiency. Attached Figure Description

[0023] Figure 1 This is an application environment diagram from one embodiment;

[0024] Figure 2 This is a flowchart illustrating an agent-based power system simulation file analysis and processing method in one embodiment.

[0025] Figure 3 This is an overall workflow diagram for one embodiment;

[0026] Figure 4 This is a structural block diagram of an agent-based power system simulation file analysis and processing device in one embodiment.

[0027] Figure 5 This is an internal structural diagram of a computer device in one embodiment;

[0028] Figure 6 This is a diagram of the internal structure of a computer device in another embodiment. Detailed Implementation

[0029] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or modules is not necessarily limited to those explicitly listed, but may include other steps or modules not explicitly listed or inherent to these processes, methods, products, or devices. The division of modules in the embodiments of this application is merely a logical division; in actual applications, there may be other division methods. For example, multiple modules may be combined into or integrated into another system, or some features may be ignored or not performed. Additionally, the shown or discussed mutual coupling or direct coupling or communication connection may be through some interface, and the indirect coupling or communication connection between modules may be electrical or other similar forms, none of which are limited in the embodiments of this application. Furthermore, the modules or sub-modules described as separate components may or may not be physically separated, may or may not be physical modules, or may be distributed among multiple circuit modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the embodiments of this application.

[0030] Figure 1 As shown in the application environment diagram of one embodiment, this application provides a power system simulation file analysis and processing method based on intelligent agents, which can be applied to, for example... Figure 1 In the application scenario shown, terminal 102 communicates with server 104 via a network.

[0031] The terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, etc. The server 104 can be implemented using a standalone server or a server cluster composed of multiple servers.

[0032] It should be noted that the terminal 102 involved in the embodiments of this application can be a wired terminal or a wireless terminal, and can be a device that provides voice and / or data connectivity to a user, a handheld device with wireless connectivity, or other processing devices connected to a wireless modem. The wireless terminal can communicate with one or more core networks via a wireless access network, and the wireless terminal can be a mobile terminal, such as a mobile phone or a computer with a mobile terminal.

[0033] Figure 2This is a flowchart illustrating an agent-based power system simulation file analysis and processing method in one embodiment, as shown below. Figure 2 As shown in the embodiments of this application, the power system simulation file analysis and processing method based on intelligent agents includes:

[0034] S201, Obtain natural language analysis and processing requirements information and simulation table files to be analyzed.

[0035] Among them, the natural language analysis and processing requirement information is the information that represents the user's requirements for the analysis task, and the simulation table file to be analyzed refers to the table file after simulation to be analyzed.

[0036] S202, based on the natural language processing requirements information, obtain standardized task description information and corresponding retrieval results from the knowledge base.

[0037] Standardized task description information refers to information that has certain standardization in data format and is used to describe the characteristics of a task, while search results refer to the results obtained after searching based on a knowledge base.

[0038] S203, based on the simulation table file and retrieval results, decomposes the overall analysis objective corresponding to the natural language analysis and processing requirements into at least two structured subtasks that can be executed in parallel.

[0039] The overall analysis objective refers to the overall task execution objective described in the natural language analysis and processing requirements information, while the structured subtasks are subtasks with certain compositional norms obtained after decomposition.

[0040] S204. Based on the task type represented by the standardized task description information, execute the structured subtasks to obtain the execution result data.

[0041] Among them, the task type refers to the type name that the planner identifies based on the normalized task description information, used to distinguish the characteristics of different subtasks.

[0042] S205 integrates the results based on the execution result data and the template requirement data to generate a simulation analysis report associated with the simulation table file.

[0043] The simulation analysis report is the final analysis report on the simulation table file generated in this application.

[0044] Compared to traditional simulation report output methods in conventional technologies, this embodiment first obtains natural language processing (NLP) requirements information and the simulation table file to be analyzed. Then, it obtains standardized task description information and the corresponding retrieval results from the knowledge base. Next, it breaks down the overall analysis objective corresponding to the NLP requirements information into at least two parallel executable structured subtasks. These structured subtasks are then executed to obtain execution result data. Finally, the results are integrated based on the execution result data and template requirements to generate a simulation analysis report. This embodiment achieves end-to-end automation of analysis and report generation, maintaining system stability even with field changes, worksheet structure differences, and multi-software usage scenarios. It significantly reduces manual intervention and script maintenance costs, improves analysis efficiency, enhances the reusability of simulation report output, and increases report output efficiency.

[0045] Optionally, in some embodiments of this application, obtaining natural language processing requirements information and the simulation table file to be analyzed includes: using the user interaction terminal as the external interface of the analysis system to receive the natural language processing requirements information and the simulation table file to be analyzed input by the target user.

[0046] The agent-based power system simulation file analysis and processing method provided in this application is applied to an analysis system equipped with a user interaction terminal, a coordinator, a planner, a toolbox, and a reporter, which is used to generate simulation reports.

[0047] Among them, the user interaction terminal serves as the external interface of the analysis system, the natural language analysis and processing requirement information refers to the information that represents the user's natural language analysis and processing requirements, and the simulation table file is the simulation file to be analyzed.

[0048] For example, the power system simulation file analysis and processing method based on intelligent agents provided in this application further includes step S301: the user interaction terminal receives user requests and files to be analyzed, specifically including: the user interaction terminal, as the external interface of the entire analysis system, receives the user's natural language analysis and processing requirements and the uploaded simulation table file to be analyzed and inputs it into the entire analysis system.

[0049] Optionally, in some embodiments of this application, obtaining standardized task description information and retrieval results corresponding to the knowledge base based on natural language processing (NLP) requirement information includes: rewriting the NLP requirement information based on the coordinator to obtain standardized task description information; generating a knowledge base based on the coordinator and retrieval enhancement, and determining the retrieval results based on the NLP requirement information.

[0050] Among them, the standardized task description information is also known as the standardized task description.

[0051] In this system, the coordinator is an intelligent agent, and the rewriting process refers to rewriting the natural language request into a standardized task description, mapping it to a "task objective—constraints—output delivery" triple. A standardized task description is a task description with certain specifications.

[0052] Among them, the retrieval-augmented generation knowledge base is also known as the RAG (Retrieval-Augmented Generation) knowledge base.

[0053] Optionally, in some embodiments of this application, based on the coordinator and the retrieval enhancement generation knowledge base, the retrieval results are determined according to the natural language processing (NLP) requirement information, including: extracting structured elements from the NLP requirement information; inputting the structured elements as retrieval vectors into the retrieval enhancement generation knowledge base for retrieval, and obtaining the retrieval results.

[0054] The extraction of structured elements refers to the coordinator's use of large language models to extract structured elements from user analysis and processing needs. These structured elements include region, time, target indicators, and statistical definitions.

[0055] The phrase "using structured elements as retrieval vectors to input into the retrieval enhancement knowledge base for retrieval" refers to calling the retrieval enhancement knowledge base to retrieve relevant industry knowledge and terminology, and then inputting the extracted structured elements as retrieval vectors into the industry knowledge base.

[0056] Optionally, in some embodiments of this application, structured elements are input as retrieval vectors to a retrieval enhancement knowledge base for retrieval to obtain retrieval results, including: determining industry standards, terminology definitions, statistical standards, and calculation formulas related to the current task through vector similarity matching and semantic expansion retrieval based on a hybrid architecture of vector index and keyword index; and using industry standards, terminology definitions, statistical standards, and calculation formulas as retrieval results.

[0057] Furthermore, the search results can be further processed to obtain processed search results, which can then be used to continue subsequent processing steps to generate the simulation report. For example, the search results can be fused using a context reordering algorithm to output a set of optimal knowledge fragments, which can guide the unified approach to subsequent data retrieval, calculation, and expression.

[0058] For example, the agent-based power system simulation file analysis and processing method provided in this application further includes step S302: the coordinator performs intent parsing and knowledge retrieval, specifically including:

[0059] The coordinator, leveraging the capabilities of a large language model, extracts elements such as region, time, target indicators, and statistical definitions from user analysis and processing needs. It then rewrites the natural language requests into standardized task descriptions, mapping them to a "task objective—constraints—output delivery" triple. Next, it invokes retrieval enhancement to generate a knowledge base that retrieves relevant industry knowledge and terminology, inputting the extracted structured elements as retrieval vectors into the industry knowledge base.

[0060] The knowledge base adopts a hybrid architecture of vector index and keyword index. Through vector similarity matching and semantic expansion retrieval, it finds industry standards, terminology definitions, statistical methods and calculation formulas related to the current task.

[0061] The retrieval results are then fused using a context reordering algorithm to output a set of optimal knowledge fragments, which are used to guide the unified approach in subsequent data retrieval, calculation, and expression.

[0062] Optionally, in some embodiments of this application, based on the simulated table file and the search results, the overall analysis objective corresponding to the natural language analysis and processing requirements is decomposed into at least two structured subtasks that can be executed in parallel, including: calling a table parsing tool to perform structured reading and data type identification on the simulated table file to obtain table reading and identification information; and refining the overall analysis objective into atomic-level operator operations based on the table reading and identification information and the search results to determine the structured subtasks.

[0063] The search results are also known as industry knowledge search results, and the simulated table file is also known as the table to be analyzed.

[0064] It should be noted that the execution of the steps in this embodiment depends on the planner in the intelligent agent, also known as the planner module.

[0065] The table parsing tool refers to the MCP (Model Context Protocol) interface calling the table parsing tool.

[0066] The information read and identified from the table includes, but is not limited to, table header fields, unit dimensions, and time series information.

[0067] For example, the agent-based power system simulation file analysis and processing method provided in this application further includes step S303: the planner performs task decomposition, specifically including:

[0068] After receiving the standardized task description, industry knowledge retrieval results, and table to be analyzed from the coordinator, the planner module breaks down the overall analysis objective into a series of structured subtasks that can be executed in parallel.

[0069] Specifically, the planner first calls a table parsing tool through the MCP interface to perform structured reading and data type identification of the input file. It then uses a Pandas-based parsing engine with regular expressions to automatically identify header fields, unit dimensions, and time series information. Next, based on predefined statistical definitions and indicator calculation formulas from industry knowledge, it further refines the task objectives into atomic-level operator operations.

[0070] Specifically, this includes using the MCP tool to read tables, using the MCP tool to locate relevant indicator columns, using the MCP tool to perform aggregation operations, writing code for large models and using the code interpreter MCP to generate bar charts, and having the reporter integrate and analyze reports based on specific templates.

[0071] Optionally, in some embodiments of this application, according to the task type represented by the standardized task description information, a structured subtask is executed to obtain execution result data, including: when the task type is one that can be completed by simple table statistical operations, a table processing tool is called to execute the structured subtask to obtain execution result data; when the task type is one where the task logic exceeds the expressive power of table operators, program code is generated and a code interpreter tool is called to execute the program code to execute the structured subtask and obtain execution result data.

[0072] Among them, the task type "the task can be completed by simple tabular statistical operations" means that the planner judges and analyzes that the task can be completed by simple tabular statistical operations; the task type "the task logic exceeds the expressive power of tabular operators" means that the planner judges that the task logic exceeds the expressive power of tabular operators and that a data analysis script or plotting script needs to be written.

[0073] Among them, the execution result data is the execution result obtained after executing the subtask, also known as multi-source result data, which may include statistical indicators, calculation result tables and visualization image files.

[0074] It should be noted that the implementation of the steps in this embodiment depends on the toolbox in the intelligent agent, also known as the toolbox module.

[0075] Among them, the table processing tool is also known as the table processing MCP, and the code interpreter tool is also known as the code interpreter MCP.

[0076] For example, the agent-based power system simulation file analysis and processing method provided in this application further includes step S304: the toolbox performs sub-task execution, specifically including:

[0077] The toolbox module, as the system's underlying execution unit, is responsible for receiving subtask instructions from the planner and calling specific tools to execute those instructions. The toolbox contains two core components: the table processing module (MCP) and the code interpreter module (MCP).

[0078] The table processing tool MCP features table reading and writing, cell and row / column formula calculations, and basic statistical calculations such as extreme values ​​and means. Its functionality stems from a module registration mechanism during system initialization: a series of pre-written functions for manipulating table data are registered in MCP's internal operator library as operators. This operator library includes data reading operators, calculation operators, and statistical calculation operators, each encapsulated as an independent execution unit with standardized input / output interfaces. Upon receiving instructions from the planner, MCP dynamically matches the appropriate operator based on the task description and executes it through the intermediate instruction parsing and task scheduling module, automating the process from file parsing, cell calculation, and statistics. MCP can flexibly combine different operators without manual intervention to achieve table calculations and basic statistical analysis capabilities.

[0079] The code interpreter MCP provides a runtime environment for code generated from large models, executing the code and outputting results. Internally, it uses a containerized sandbox mechanism to ensure execution safety and resource control. The code interpreter environment comes pre-installed with mainstream data analysis and visualization libraries (Pandas, NumPy, Matplotlib, etc.), supporting complex statistical analysis and chart generation.

[0080] More specifically, when the planner determines that the analysis task can be completed by simple tabular statistical operations, it calls the table processing MCP; when the planner determines that the task logic exceeds the expressive power of the table operators and a data analysis script or plotting script needs to be written, the large model generates code and calls the code interpreter MCP to execute the code.

[0081] In another embodiment, in step S205, the template requirement data refers to the template data needed to generate the simulation analysis report, such as the report template structure definition. Execution in this embodiment depends on the reporter within the agent.

[0082] For example, the agent-based power system simulation file analysis and processing method provided in this application further includes step S305: the reporter integrates the results and generates a report, specifically including:

[0083] During runtime, the reporter receives task breakdown instructions and report template structure definitions from the planner, as well as multi-source result data (including statistical indicators, calculation result tables, and visualization image files) from the toolbox output. Based on this, the reporter uses the large language model's fundamental capabilities to automatically organize and present the results.

[0084] Figure 3 Here is an overall workflow diagram for one embodiment, which is described below in conjunction with... Figure 3 The technical research process and other technical details of this application are described with reference to a specific embodiment.

[0085] In traditional technologies, in recent years, power systems have heavily relied on numerical simulations of power flow, short circuits, and transient stability in planning, dispatching, and operation and maintenance. Mainstream commercial and open-source platforms typically export large numbers of CSV / XLSX result tables in batch processing. Engineers then use tools such as Excel and pandas for secondary statistical analysis, graphing, and writing analytical conclusions. This "simulation-export-manual compilation" technology chain has long suffered from the following pain points: cross-table data retrieval and derived index calculations lack unified semantics and reusable scripts, and the final report still heavily relies on manual organization and writing, limiting efficiency.

[0086] At the data analysis tool stack level, engineers can load multi-source tables through file reading interfaces and combine data cleaning, statistics, and visualization to form analysis records. However, existing workflows are still centered on "manual coding," and the understanding of power business semantics (such as system power / energy balance, external transmission / receiving, N-1 limit exceedance, reserve constraints, etc.) and the standardization of indicator definitions are not readily available, let alone automatically translating users' natural language questions into a stable "parsing-calculation-finalization" pipeline. Currently, the ecosystem of "tool access and multi-agent orchestration" for large models has matured rapidly in recent years. The Model Context Protocol (MCP) proposes an open standard for exposing external tools and data to the model in a unified protocol, enabling the model to call functions such as "table reading, calculation, and editing" in a controlled manner; cutting-edge multi-agent and workflow frameworks provide structured collaboration paradigms for different agents, which can decompose complex tasks into executable steps and form a traceable closed loop. These advancements provide a methodological foundation for standardizing industry tool functions for model calls and orchestrating complex analysis processes into observable workflows. However, general solutions still lack dedicated toolboxes and report generation mechanisms for the semantics of power simulation tables, making it difficult to directly meet the scenario requirements of "question-and-answer analysis and automatic drafting" for power system simulation documents.

[0087] For methods of analyzing power system simulation results files, some implementation paths mainly involve "simulation software + export + manual analysis / script compilation". This path has a low learning cost, but it heavily relies on handwritten scripts. Engineers need to manually search multiple worksheets and repeatedly check them, which limits efficiency and consistency.

[0088] Some technologies employ a chain of "simulation calculation → result export → manual analysis / script compilation → manual report writing." While this can accomplish routine statistics and graph generation, it has significant limitations when dealing with power system simulation tables. First, existing analysis methods heavily rely on manually calling Excel statistical tools or writing analysis code, resulting in a large workload and low efficiency. Second, due to inconsistencies in table structures and terminology generated by different analysis software or versions, manually written analysis code is often difficult to reuse; even slight changes in field names or column order require extensive refactoring. Third, although some cutting-edge general-purpose multi-agent or tool orchestration schemes exist that can call external tools, there is a lack of dedicated toolkits and knowledge injection mechanisms for power simulation scenarios, making it difficult to implement in the end-to-end closed loop from complex question answering to industry-specific reports. Therefore, existing technologies are generally characterized by complex and non-reusable processes, slow speed, and lack of scalability.

[0089] Based on this, this application provides an agent-based power system simulation file analysis and processing method, also known as agent-based power system simulation file analysis and processing technology, which constructs an end-to-end, high-efficiency, and scalable engineering solution. By standardizing and registering functions such as table reading, formula calculation, and region editing as MCP Server tools, large models can robustly parse different tables with a unified interface. When a simple toolchain is insufficient to cover complex requirements, it automatically switches to a code interpreter to generate and execute data analysis and visualization code, ensuring controllable performance of complex statistical and graphical outputs. The multi-agent deep research workflow orchestration of "coordinator-planner-toolbox-reporter" significantly improves the processing speed by "receiving requests-task decomposition-tool execution-report integration". At the same time, it introduces RAG (retrieval-enhanced generation) technology to inject knowledge and impose pre-reasoning constraints on power system terminology, indicator definitions, and calculation formulas, reducing errors caused by inconsistent understanding of terminology in the power system professional field.

[0090] Therefore, this application provides a fundamental technology for natural language question answering, automatic calculation and statistics, and professional drafting of power simulation results. While ensuring professional accuracy, it reduces maintenance costs, improves processing speed and overall efficiency, and meets the needs of high robustness and high efficiency in actual engineering.

[0091] This application proposes a technical solution that standardizes table processing functions for model calls, automatically switches code interpreters for complex requirements, and constructs an analysis workflow of "coordinator-planner-toolbox-reporter." By registering table processing and analysis tools as an MCP Server for unified use by large models, the system can read, calculate, and edit multi-source heterogeneous simulation results. When simple operations cannot meet the requirements, the system automatically switches to the code interpreter to generate and execute data analysis and visualization code, ensuring that complex statistical and plotting tasks can also be achieved. Simultaneously, an industry knowledge injection mechanism based on RAG (Retrieval Enhanced Generation) is introduced, constructing a knowledge base and terminology table of power system knowledge points and terms, and injecting relevant knowledge into the large model based on user questions before inference. Through the linkage mechanism of "tool standardization + code interpreter fallback + multi-agent orchestration + RAG industry semantic alignment," simulation result tables for power simulation can achieve question-and-answer style analysis and automatic drafting, significantly improving engineering efficiency and consistency of conclusions.

[0092] This application aims to enable users to generate tabular files of power system simulation results in natural language. The intelligent agent system automatically parses the files, performs calculations and statistics, and generates readable reports. The system relies on four key technologies: a table processing tool registered with the Model Context Protocol (MCP), a controlled sandbox code interpreter runtime environment, Retrieval Enhanced Generation (RAG) to inject power industry expertise and standardized metrics, and a multi-agent role system—coordinator, planner, toolbox, and reporter—that runs throughout the process, completing a closed loop from request acceptance to report generation.

[0093] like Figure 3 As shown, the agent-based power system simulation file analysis and processing method provided in this application is applied to an analysis system for generating simulation reports. The overall structure of the analysis system is as follows:

[0094] The system includes a user interface, coordinator, planner, toolbox, and reporter. It exposes two types of tools through a model context protocol server: table processing model context protocol tools and code interpreter runtime environment tools. These provide direct tools for table reading and processing, while also providing a code script execution environment for complex statistics and plotting.

[0095] The process involves the user interaction terminal establishing a session with the coordinator. The coordinator, based on user requests, invokes the enhanced knowledge base to extract relevant professional knowledge as guidance. It then rewrites the user request into a more easily understood, professionally standardized task description for the larger model, which, along with the uploaded table file, is parsed by the planner. The planner generates an executable task flow, breaking down the problem into several executable steps. Nodes involving table data analysis either directly call the table processing model context protocol tool to perform cross-table data retrieval and row / column formula calculations, or, when encountering complex statistics or professional plotting, write analysis code and execute it in the code interpreter runtime environment. The toolbox acts as a proxy for various tools, uniformly receiving instructions from the planner and returning standardized results. The reporter is responsible for binding the returned data, tables, and visualizations into a report.

[0096] In addition, the machine responsibilities and inputs / outputs of the key modules of this analysis system are shown in the table below.

[0097]

[0098] In addition, a typical way of operating the analysis system is illustrated in one specific embodiment.

[0099] When a user asks a question to "compare the power and energy balance, transmission and reception composition, and renewable energy absorption rate of two regions during peak hours on a certain day, and generate a report," the user interface receives the request and file and inputs them to the coordinator. The coordinator first uses retrieval enhancement to generate a knowledge base that retrieves entries such as "definitions of transmission and reception, balance conservation, reserve threshold, and absorption rate formula" to supplement industry knowledge. Then, the question is rewritten into a standardized task description. For example, a standardized task description could be:

[0100] The task is titled: Generation of a comparison report on the power and energy balance during peak hours in the two regions and an assessment report on the consumption of new energy sources.

[0101] The task objective is to conduct a comparative analysis of the power and electricity balance between Region A and Region B during peak hours on a specified date, provide the composition of power transmission and power reception, calculate the renewable energy consumption rate and related auxiliary indicators, and generate a readable report.

[0102] Regarding input constraints and sources of caliber:

[0103] The date and time period are: the user provides a specific date and a definition of "peak period" (such as the hourly interval corresponding to the top 10% of the load on the day; if not provided, the system will automatically identify it based on the top 10%).

[0104] The analysis areas are: Area A and Area B (the names must match the mapping in the table header).

[0105] Data source: Simulation result spreadsheet files (workbooks / sheets) uploaded by users, hourly resolution.

[0106] Regarding output delivery:

[0107] The main body of the report consists of: Problems and Scope, Explanation of Data and Definitions, Comparative Charts, Analysis Conclusions and Recommendations.

[0108] Next, the planner generates a task flowchart and details based on the standardized task description, covering "document location—structure identification—cross-table aggregation—indicator calculation—visualization—final draft." It then calls the table processing model context protocol tool to complete multi-table reading analysis and indicator calculation. For relatively complex tasks such as confidence interval estimation, code is generated and executed by the code interpreter runtime environment. Results indicators and charts are entered into the report, and the reporter generates a report according to the template, achieving a traceable closed loop from problem to conclusion.

[0109] The key points of this application are: (1) Automated analysis and report generation of power system simulation result tables through a multi-agent architecture, replacing the process of "simulation results - export - manual compilation", significantly reducing manual workload and improving efficiency and consistency. (2) Multi-agent orchestration and task graph-driven operation of "coordinator - planner - toolbox - reporter": Natural language requests are rewritten into standardized task descriptions, and then translated into executable task graphs by the planner, so that each module can perform operations according to clear rules, forming a closed loop from parsing to finalization. (3) Automatic degradation and controlled execution of the code interpreter's runtime environment: When the table processing MCP is insufficient to cover complex statistical and visualization needs, analysis code is automatically generated and deterministically executed within the controlled sandbox of the code interpreter MCP. The flexible application of the two MCP tools can be controlled through the planner.

[0110] Compared to the mainstream approach of "simulation-export-manual compilation" in traditional technologies, this application achieves end-to-end automation of analysis and report generation through table processing tools based on the model context protocol, a code interpreter runtime environment, and a knowledge base generated with enhanced retrieval capabilities. The system remains stable under conditions of field changes, differences in worksheet structures, and the use of multiple software programs, significantly reducing manual intervention and script maintenance costs while improving analysis efficiency.

[0111] Figures 1 to 3 Any technical feature in the embodiments corresponding to any of the above items is also applicable to the embodiments of this application. Figures 4 to 6 The corresponding implementation examples will not be repeated hereafter.

[0112] The above describes a method for analyzing and processing power system simulation files based on intelligent agents in the embodiments of this application. The following describes the apparatus or device for performing the above method.

[0113] Figure 4 Here is a structural block diagram of an agent-based power system simulation file analysis and processing device in one embodiment, with reference to... Figure 4 The device includes:

[0114] The information acquisition module 401 is used to acquire natural language analysis and processing requirements information and simulation table files to be analyzed;

[0115] The retrieval module 402 is used to obtain standardized task description information and retrieval results corresponding to the knowledge base based on the natural language analysis and processing requirements information.

[0116] The task decomposition module 403 is used to decompose the overall analysis target corresponding to the natural language analysis and processing requirements information into at least two structured subtasks that can be executed in parallel, based on the simulation table file and the search results.

[0117] The task execution module 404 is used to execute structured subtasks based on the task type represented by the standardized task description information and obtain execution result data.

[0118] The report generation module 405 is used to integrate the results based on the execution result data and the template requirement data to generate a simulation analysis report associated with the simulation table file.

[0119] In this embodiment of the application, based on, as follows Figure 4 The connections between the modules shown in the diagram demonstrate how the cooperation between these modules can improve the reusability of simulation report output and increase report output efficiency.

[0120] In another embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 5 As shown, it includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores relevant data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements the various methods described in the above embodiments.

[0121] In yet another embodiment, a computer device is provided, such as a terminal, whose internal structure diagram may be as follows: Figure 6As shown, it includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements the various methods described in the above embodiments.

[0122] Those skilled in the art will understand that Figure 5 and Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. It may also include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements, in order to achieve functions such as a terminal or server.

[0123] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0124] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the systems, devices, equipment, modules or units described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0125] In the embodiments provided in this application, it should be understood that the disclosed systems, devices, equipment, or methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of devices or modules may be electrical, mechanical, or other forms.

[0126] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0127] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium.

[0128] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product.

[0129] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or data center that integrates one or more available media. The available medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., a solid-state disk (SSD)).

[0130] The technical solutions provided in the embodiments of this application have been described in detail above. Specific examples have been used in the embodiments of this application to illustrate the principles and implementation methods of the embodiments of this application. The description of the above embodiments is only for the purpose of helping to understand the methods and core ideas of the embodiments of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the embodiments of this application. Therefore, the content of this specification should not be construed as a limitation on the embodiments of this application.

Claims

1. An agent-based power system simulation file analysis processing method, characterized by, The method comprises: acquiring natural language analysis processing requirement information and a simulation table file to be analyzed; obtaining standardized task description information and a retrieval result corresponding to a knowledge base according to the natural language analysis processing requirement information; according to the simulation table file and the retrieval result, the overall analysis target corresponding to the natural language analysis processing requirement information is disassembled into at least two structured sub-tasks that can be executed in parallel; according to the task type represented by the standardized task description information, the structured sub-tasks are executed to obtain execution result data; according to the execution result data and template requirement data, the results are integrated to generate a simulation analysis report associated with the simulation table file.

2. The method of claim 1, wherein, The acquisition of natural language analysis processing requirement information and a simulation table file to be analyzed comprises: the user interaction end is used as the external interface end of the analysis system to receive the natural language analysis processing requirement information and the simulation table file to be analyzed input by the target user.

3. The method of claim 1, wherein, The obtaining of standardized task description information and a retrieval result corresponding to a knowledge base according to the natural language analysis processing requirement information comprises: rewriting the natural language analysis processing requirement information based on a coordinator to obtain the standardized task description information; generating a knowledge base based on a coordinator and a retrieval enhancement, and determining a retrieval result according to the natural language analysis processing requirement information.

4. The method of claim 3, wherein, The generation of a knowledge base based on a coordinator and a retrieval enhancement, and the determination of a retrieval result according to the natural language analysis processing requirement information comprise: extracting structured elements from the natural language analysis processing requirement information; inputting the structured elements as retrieval vectors into the retrieval enhancement knowledge base for retrieval to obtain the retrieval result.

5. The method of claim 4, wherein, The inputting of the structured elements as retrieval vectors into the retrieval enhancement knowledge base for retrieval to obtain the retrieval result comprises: based on a hybrid architecture of vector indexing and keyword indexing, determining the industry specifications, term definitions, statistical indicators and calculation formulas related to the current task through vector similarity matching and semantic expansion retrieval; the industry specifications, the term definitions, the statistical indicators and the calculation formulas are used as the retrieval result.

6. The method of claim 1, wherein, The disassembly of the overall analysis target corresponding to the natural language analysis processing requirement information into at least two structured sub-tasks that can be executed in parallel according to the simulation table file and the retrieval result comprises: calling a table parsing tool to perform structured reading and data type identification on the simulation table file to obtain table reading and identification information; according to the table reading and identification information and the retrieval result, the overall analysis target is refined into atomic operator operations to determine the structured sub-tasks.

7. The method of claim 1, wherein, The execution of the structured sub-tasks according to the task type represented by the standardized task description information to obtain execution result data comprises: in the case where the task type is a task that can be completed by simple table statistical operation, a table processing tool is called to execute the structured sub-tasks to obtain the execution result data; In a case where the task type is a task logic beyond an expression capability of the table form operator, program code is generated and a code interpreter tool is called to execute the program code to execute the structured sub-tasks to obtain the execution result data.

8. An agent-based power system simulation file analysis processing apparatus characterized by comprising: The device comprises: An information acquisition module is configured to acquire natural language analysis processing requirement information and a simulation table form file to be analyzed. A retrieval module is configured to obtain standardized task description information and a retrieval result corresponding to a knowledge base according to the natural language analysis processing requirement information. A task decomposition module is configured to decompose an overall analysis target corresponding to the natural language analysis processing requirement information into at least two structured sub-tasks that can be executed in parallel according to the simulation table form file and the retrieval result. A task execution module is configured to execute the structured sub-tasks according to a task type represented by the standardized task description information to obtain execution result data. A report generation module is configured to integrate results according to the execution result data and template requirement data to generate a simulation analysis report associated with the simulation table form file.

9. A computer device, comprising: The computer device comprises: at least one processor, a memory; The memory is configured to store program code, and the processor is configured to call the program code stored in the memory to execute the method according to any one of claims 1 to 7.

10. A computer storage medium, characterized in that, The computer program product comprises instructions which, when executed on a computer, cause the computer to perform the method according to any one of claims 1 to 7.

Images