GIS dynamic grid management and control system based on intelligent question-answering model

Abstract

The present invention relates to the technical field of safe production of thermal power plants, and discloses a GIS dynamic grid management and control system based on an intelligent question-answering model. The system realizes the integration of spatial data such as equipment layout, pipeline information, and environmental monitoring of thermal power plants by constructing a multi-dimensional data module. The dynamic grid management unit adjusts the GIS grid according to actual needs to ensure the real-time and accuracy of the data. The GIS data processing unit provides functions such as map rendering and spatial query. The core of the system lies in the intelligent question-answering model, which receives, parses, and identifies management and control instructions, converts the user's natural language into standardized instructions, and recalls relevant information to generate candidate answers. The answer is optimized using a general large language model to generate a request reply text containing management and control suggestions and impact analysis. Finally, the management and control instruction confirmation unit ensures the accuracy of instruction execution. The present invention improves the convenience and flexibility of GIS system application in thermal power plants, and provides strong support for the safe operation of thermal power plants.

Description

Technical Field

[0001] The present invention relates to the technical field of safe production of thermal power plants, and in particular to a GIS dynamic grid management and control system based on an intelligent question-answering model. Background Art

[0002] With the continuous development and popularization of geographic information system (GIS) technology, it has shown strong application value in many fields such as urban planning, environmental protection, resource management and disaster warning. However, in the specific industry of thermal power plants, the application of GIS systems still faces many challenges, especially in terms of user convenience, intuitiveness and system flexibility.

[0003] As an important base for energy production, the operation and management of thermal power plants involve a large amount of spatial data, including equipment layout, pipeline direction, distribution of environmental monitoring points, etc. These data are crucial for the safe operation, efficient management and emergency response of thermal power plants. Although traditional GIS systems can provide basic spatial query and data analysis functions when processing these data, due to their design limitations, users often face many inconveniences during use. Specifically, most current GIS systems require users to follow strict grammatical rules and input instructions in a specific format, which undoubtedly increases the learning cost and usage threshold for non-professional users. In the field of thermal power plants, operation and maintenance personnel are often more concerned with the real-time status of equipment, the rapid acquisition of environmental monitoring data, and the rapid formulation of emergency responses, rather than spending a lot of time learning and mastering complex GIS operation commands.

[0004] In addition, the current GIS system lacks humanized design in terms of interaction. When users conduct spatial queries or data analysis, they must accurately remember and enter complex command codes instead of using more natural and intuitive language expressions. This non-humanized interaction method not only reduces the user experience, but also limits the application scope and flexibility of the GIS system in the daily work of thermal power plants. Especially in emergency situations, complex operating procedures and command input may delay the timing of emergency response, posing a potential threat to the safe operation of thermal power plants. Summary of the invention

[0005] The purpose of the present invention is to provide a GIS dynamic grid management and control system based on an intelligent question-answering model to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: a GIS dynamic grid management and control system based on an intelligent question-answering model, the system comprising:

[0007] The data module construction unit is used to pre-construct data modules of multiple dimensions including equipment layout, pipeline information, and environmental monitoring data according to the business data type of the thermal power plant GIS;

[0008] Dynamic grid management unit, which is used to dynamically adjust the size, shape and layout of the GIS grid according to the actual operation requirements of the thermal power plant, monitor the changes of geographic spatial data in the grid in real time, and update the grid information in real time;

[0009] A GIS data processing unit, connected to the control instruction receiving unit, for processing and analyzing GIS data, including map rendering, spatial query and path planning functions;

[0010] A control instruction receiving unit, used to receive a control instruction input by a user and obtain a first request text;

[0011] A control instruction parsing unit: connected to the control instruction receiving unit, and configured to parse the first request text to form a structured keyword list;

[0012] A control instruction recognition unit is used to input the structured keyword list into a pre-trained control instruction understanding model. The model is responsible for parsing the deep semantics of the keyword list, identifying the control instruction type and the completeness of the control instruction input conditions. If the model cannot identify the type of the control instruction, or the input conditions of the control instruction are incomplete, it will feedback the user to enter the correct control instruction; if the model recognizes the type of the control instruction and the input conditions of the control instruction are complete, it will output a standardized control instruction based on the built-in template to obtain the second request text;

[0013] A candidate answer recall unit, configured to recall relevant information from the data modules of multiple dimensions of the data module construction unit according to the second request text, and obtain multiple candidate answers;

[0014] A request reply generating unit is used to encapsulate the second request text and multiple candidate answers according to a preset prompt word template to obtain a third request text, further pass the third request text into the general large language model, and use the natural language understanding and generation capabilities of the general large language model to integrate, optimize and adjust the multiple candidate answers to form an answer including management and control suggestions and impact analysis, and obtain a request reply text;

[0015] The control instruction confirmation unit is used to feed back the request reply text to the user and ask the user whether to confirm the execution.

[0016] Preferably, the data module constructed by the data module construction unit includes:

[0017] GIS basic database module, used to store basic geographic spatial data;

[0018] Real-time data access module, used to receive and process real-time data from various sensors;

[0019] Grid management database module, used to store grid division, grid attributes and event information within the grid;

[0020] High-dimensional geographic feature vector database module, used to store high-dimensional vectors of feature-extracted geographic spatial data;

[0021] GIS business knowledge graph module builds geographic spatial knowledge graph and stores geographic entities, attributes and relationships.

[0022] Preferably, the control instruction understanding model is obtained by training based on a general language model.

[0023] Preferably, the method for obtaining the control instruction understanding model includes:

[0024] The control instruction types and control instruction conditions are defined using the extensible markup language XML syntax, wherein the control instruction types include map rendering instructions, spatial query instructions, path planning instructions, and data update instructions. Each instruction type corresponds to a set of specific control instruction conditions, and the control instruction conditions are described in the form of XML tags to form a control instruction definition template file with a hierarchical structure and attribute information;

[0025] Parsing the control instruction definition template file, extracting the control instruction type, control instruction conditions and the relationship between them through a parsing algorithm, and generating a control instruction definition data set for understanding by a general language model, wherein the data set represents the complete semantics of the control instruction in the form of natural language;

[0026] The control instruction definition data set is input into a general language model, and through training, the general language model obtains context information for understanding the control instruction structure and the relationship between control instructions, thereby having the ability to perform deep semantic analysis on the control instructions.

[0027] Preferably, the control instruction parsing unit parses the first request text in a manner including word segmentation, part-of-speech tagging, dependency syntactic analysis and keyword extraction.

[0028] Preferably, for different types of data modules, the management and control instruction identification unit generates corresponding second request texts respectively.

[0029] Preferably, the control suggestions output by the request response generation unit are specific operational suggestions given based on the control instructions input by the user and the current status of the geographic spatial data in the GIS grid, combined with the analysis capabilities of the general large language model.

[0030] Preferably, the step of parsing the control instruction definition template file to generate the control instruction definition data set includes:

[0031] Step 1: Select a programming language for parsing XML files, and import a toolkit for parsing XML files into the programming language;

[0032] Step 2: Use the toolkit to load the control instruction definition template file;

[0033] Step 3: Define an array for caching text fragments;

[0034] Step 4: Obtain each type of label through the label traversal method provided by the toolkit, and insert text fragments for creating control instructions, defining instruction parameters, setting instruction conditions, and specifying instruction actions into the cache array;

[0035] Step 5: After the traversal is completed, all text fragments in the cache array are converted into complete text strings and saved to generate a control instruction definition data set.

[0036] Preferably, the GIS data processing unit spatially segments and organizes the data according to the grid information provided by the control instruction receiving unit.

[0037] Preferably, the preset prompt word template includes a question type identifier, question keywords, a candidate answer source identifier, and an answer fusion requirement.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] By introducing the intelligent question-answering model, the system can understand and parse the control instructions entered by the thermal power plant operation or maintenance personnel in the form of natural language, without the need for users to master complex GIS operation commands and grammatical rules, greatly reducing the learning cost and usage threshold for non-professional users. In an emergency, the system can quickly parse the control instructions entered by the user, and based on the real-time updated GIS grid information, quickly generate answers containing specific control suggestions and impact analysis, providing timely and accurate information support for the emergency response of the thermal power plant, effectively shortening the emergency response time and reducing safety risks.

[0040] The dynamic grid management unit can flexibly adjust the size, shape and layout of the GIS grid according to the actual operation requirements of the thermal power plant, ensuring the real-time update of the grid information, making the system more adaptable to the complex and changeable operation environment of the thermal power plant. At the same time, the data module construction unit can pre-build data modules containing multiple dimensions to provide rich data support for the system and enhance the applicability and analysis capabilities of the system.

[0041] The system adopts a humanized interactive mode. Users can interact with the system through natural language without having to accurately remember and enter complex command codes. The request-response generation unit uses the natural language processing and generation capabilities of the general large language model to integrate and optimize candidate answers, generate answers that are more in line with user needs and context, and significantly improve the user experience. The application of this system helps thermal power plants to achieve intelligent management of spatial data and improve operational efficiency and safety management. By real-time monitoring and analysis of spatial data such as equipment layout, pipeline direction, and environmental monitoring point distribution, the system can provide a strong basis for decision support in thermal power plants and promote the development of thermal power plants towards intelligent and refined management. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a work flow chart of the present invention;

[0043] Figure 2 A diagram of the steps to obtain the control instruction understanding model;

[0044] Figure 3 Diagram of the steps to define a data set for obtaining control instructions. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] See also Figure 1-3 The present invention provides a technical solution: a GIS dynamic grid management and control system based on an intelligent question-answering model, the system comprising:

[0047] The data module construction unit is responsible for pre-constructing data modules with multiple dimensions including equipment layout, pipeline information, and environmental monitoring data according to the data types of the thermal power plant GIS business; the data types include but are not limited to map data, spatial data, attribute data, etc. Through classification and modular processing, data management and access are made more efficient. The data modules constructed by the data module construction unit include:

[0048] GIS basic database module is used to store basic geographic spatial data.

[0049] Real-time data access module, used to receive and process real-time data from various sensors.

[0050] The grid management database module is used to store grid division, grid attributes and event information within the grid.

[0051] The high-dimensional geographic feature vector database module is used to store high-dimensional vectors of feature-extracted geographic spatial data.

[0052] GIS business knowledge graph module builds geographic spatial knowledge graph and stores geographic entities, attributes and relationships.

[0053] The dynamic grid management unit is used to dynamically adjust the size, shape and layout of the GIS grid according to the actual operation requirements of the thermal power plant, monitor the changes in the geographic spatial data within the grid, and update the grid information in real time. This unit can flexibly adjust the division of the grid according to actual needs to meet different spatial analysis requirements.

[0054] Specifically, the dynamic grid management unit adjusts the size, shape and layout of the grid by receiving instructions from the user or the system. At the same time, it continuously monitors the changes in the geospatial data within the grid, such as the increase or decrease of objects, changes in attributes, etc., and updates the grid information in real time to ensure the accuracy and timeliness of the data.

[0055] The GIS data processing unit is connected to the dynamic grid management unit and is responsible for processing and analyzing GIS data, including map rendering, spatial query, and path planning.

[0056] Specifically, the GIS data processing unit receives data from the dynamic grid management unit and performs operations such as map rendering, spatial query and path planning according to the user's request. For example, it can quickly locate the target object according to the user's query conditions and display its related attribute information; it can also plan the best path according to the user's travel needs.

[0057] The control instruction receiving unit is used to receive the control instruction input by the user and obtain the first request text. This unit is an interface for the user to interact with the system, and it can convert the user's natural language instruction into a format that the system can recognize.

[0058] The control instruction parsing unit is connected to the control instruction receiving unit and is used to parse the first request text to form a structured keyword list.

[0059] Specifically, the control instruction parsing unit first performs word segmentation processing on the first request text to split the text into individual words, and then uses natural language processing technology to identify and extract keywords to form a structured keyword list.

[0060] The control instruction recognition unit is used to input the structured keyword list into a pre-trained control instruction understanding model, which is responsible for parsing the deep semantics of the keyword list, identifying the control instruction type and the completeness of the control instruction input conditions.

[0061] Specifically, the control instruction recognition unit inputs the structured keyword list into the control instruction understanding model to identify the type of control instruction and the completeness of the input conditions. If the model cannot identify the type of control instruction or the input conditions are incomplete, it will feedback the user to enter the correct control instruction; if the model recognizes the type of control instruction and the input conditions are complete, it will output a standardized control instruction based on the built-in template to obtain the second request text.

[0062] The candidate answer recall unit is used to recall relevant information from the data modules of multiple dimensions of the data module construction unit according to the second request text to obtain multiple candidate answers.

[0063] Specifically, the candidate answer recall unit selects a corresponding data module from the data module construction unit for retrieval according to the content of the second request text, recalls information related to the request from the data module, and forms a plurality of candidate answers.

[0064] The request reply generation unit is used to encapsulate the second request text and multiple candidate answers according to a preset prompt word template to obtain a third request text, further pass the third request text into the general large language model, and use the natural language understanding and generation capabilities of the general large language model to integrate, optimize and adjust multiple candidate answers to form an answer including management and control suggestions and impact analysis, and obtain a request reply text.

[0065] The control instruction confirmation unit is used to feed back the request reply text to the user and ask the user whether to confirm the execution.

[0066] Specifically, the control instruction confirmation unit feeds back the request reply text to the user through a graphical interface. The user can judge whether the corresponding control instruction needs to be executed based on the content of the reply text. If the user confirms the execution, the system performs the corresponding operation according to the instruction; if the user does not confirm the execution, the system can further interact with the user to obtain more accurate instructions.

[0067] The present invention will be further described below in conjunction with Examples 1 to 3:

[0068] Embodiment 1:

[0069] The present invention provides a method for obtaining a control instruction understanding model, which is trained based on a general language model so that the model has the ability to perform deep semantic analysis on control instructions. The specific implementation methods include:

[0070] The control instruction types and control instruction conditions are defined using the Extensible Markup Language (XML) syntax. In this embodiment, the control instruction types include map rendering instructions, spatial query instructions, path planning instructions, and data update instructions. Each instruction type corresponds to a set of specific control instruction conditions, which are described in the form of XML tags to form a control instruction definition template file with a hierarchical structure and attribute information. An example of XML definition is as follows:

[0071]

[0072] In this example, two types of instructions are defined: Map Rendering and Spatial Query. For the Map Rendering instruction, two conditions are included: Layer and Zoom Level; for the Spatial Query instruction, two conditions are included: Query Type and Target.

[0073] The above control instruction definition template file is parsed, and the control instruction type, control instruction conditions and the relationship between them are extracted through the parsing algorithm.

[0074] The parsing algorithm first reads the root element of the XML file, then traverses each instruction type (InstructionType) element, extracts its name (name) and the contained condition (Condition) elements. For each condition element, extract its name (name), type (type) and other attribute information. Finally, a control instruction definition dataset for universal language model understanding is generated, which represents the complete semantics of the control instruction in the form of natural language.

[0075] The control instruction definition dataset generated above is input into the general language model for training. The goal of the training is to enable the general language model to understand the structure of the control instructions and the contextual information of the relationship between the control instructions.

[0076] During the training process, the general language model gradually masters the grammatical rules and semantic features of the instructions by learning a large number of control instruction examples, and learns how to generate responses that meet grammatical and semantic requirements based on the type and conditions of the instructions.

[0077] After the training is completed, the control instruction understanding model is applied to the actual GIS dynamic grid control system. After the system receives the control instruction input by the user, it first obtains the first request text through the control instruction receiving unit; then, the control instruction parsing unit parses the first request text to form a structured keyword list; then, the control instruction recognition unit inputs the structured keyword list into the trained control instruction understanding model, and the model performs deep semantic analysis on the keyword list to identify the type of control instruction and the integrity of the input conditions; finally, a standardized control instruction is generated based on the recognition result, and an answer containing control suggestions and impact analysis is generated through the candidate answer recall unit and the request reply generation unit, and fed back to the user.

[0078] In order to verify the effectiveness of the model, a series of test experiments can be conducted. For example, a set of representative control instructions can be selected as a test set and input into the model to observe whether the model can correctly identify the instruction type and conditions and generate reasonable control suggestions. By comparing the test results with the expected results, the performance and accuracy of the model can be evaluated.

[0079] Embodiment 2:

[0080] This embodiment provides a method for parsing a control instruction definition template file to generate a control instruction definition data set, and the specific steps are as follows:

[0081] Step 1: Choose a programming language and toolkit

[0082] Select a programming language for parsing XML files, use Python programming language, and import the tool package xml.etree.ElementTree for parsing XML files into the programming language.

[0083] Step 2: Load the control instruction definition template file

[0084] Use the toolkit to load the control instruction definition template file. Use the parse method of ElementTree to load the XML file.

[0085] Step 3: Define the cache array

[0086] Defines an array for caching text fragments to store instruction information extracted from XML files.

[0087] Step 4: Loop through tags and insert text snippets

[0088] Use the tag traversal method provided by the toolkit to obtain each type of tag, and insert text snippets that create control instructions, define instruction parameters, set instruction conditions, and specify instruction actions into the cache array. Specifically, you can traverse the root element of the XML file, then traverse each instruction type (InstructionType) element and its child elements (Condition), extract relevant information, and generate corresponding text snippets.

[0089] Step 5: Convert and save the text string

[0090] After the traversal is completed, all text fragments in the cache array are converted into complete text strings and saved to generate a control instruction definition dataset. This dataset can be used for subsequent general language model training.

[0091] The following is a brief Python code example for parsing an XML file and generating a control instruction definition dataset:

[0092]

[0093] In the above code, we first load the XML file named control_instructions.xml and get its root element. Then, we traverse each InstructionType element, extract the instruction type name and condition information, and generate the corresponding text fragment. Finally, we convert all the text fragments into complete text strings and save them in a file named instruction_dataset.txt.

[0094] Embodiment 3:

[0095] The control instruction parsing unit is responsible for receiving the first request text input by the user and performing in-depth parsing based on the specific context of the thermal power plant. The parsing steps include:

[0096] Word segmentation: Using a word segmentation algorithm optimized for the industrial field, the first request text, such as "Please check the real-time temperature of generator set No. 2", is split into independent vocabulary units, namely "please", "check", "No. 2", "generator set", "real-time", and "temperature".

[0097] Part-of-speech tagging: tag each lexical unit with its part of speech, for example, "check" is a verb, "No. 2" is an ordinal number, "generator set" is a noun, "real-time" is an adjective, and "temperature" is a noun.

[0098] Dependency syntactic analysis: Through the dependency syntactic analysis algorithm in the industrial context, the grammatical relationship between lexical units is identified, such as the subject-object relationship between "inspection" and "generator set", and the modification relationship between "real-time" and "temperature".

[0099] Keyword extraction: Based on the word part, frequency of occurrence, and importance in the industrial context, the keywords "inspection", "generator unit 2", and "real-time temperature" are extracted. These keywords are directly related to the user's query intention.

[0100] The control instruction recognition unit recognizes the user's intention to check the real-time status of a specific device based on the parsed first request text, and generates a corresponding second request text for the equipment layout module of the thermal power plant, such as "obtain the current temperature data of generator set No. 2".

[0101] The request reply generation unit receives the second request text, combines the geospatial data of the location of Generator Unit 2 in the GIS grid and the real-time monitoring data, and uses the analysis capabilities of the general large language model to generate a reply text containing specific operation suggestions and current status descriptions, such as "The current temperature of Generator Unit 2 is 85 degrees Celsius. It is recommended to check the cooling system immediately to ensure safe operation."

[0102] The GIS data processing unit spatially divides and organizes the equipment in the thermal power plant according to the grid information provided by the control and instruction receiving unit. For example, the area where the No. 2 generator set is located is divided into fine grids and a unique identifier is assigned to facilitate efficient management and query of real-time data in the area.

[0103] The preset prompt word templates are designed for common queries in thermal power plants, such as: "Please check the [real-time / historical] [temperature / pressure / current] data of [equipment number / name]", where [equipment number / name], [real-time / historical] and [temperature / pressure / current] are keywords filled in by users according to specific needs, thereby guiding users to enter control instructions more accurately.

[0104] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0105] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. GIS dynamic grid management and control system based on intelligent question-answering model, characterized by: The system comprises: The data module construction unit is used to pre-construct data modules of multiple dimensions including equipment layout, pipeline information, and environmental monitoring data according to the business data types of the thermal power plant GIS; Dynamic grid management unit, which is used to dynamically adjust the size, shape and layout of the GIS grid according to the actual operation requirements of the thermal power plant, monitor the changes of geographic spatial data in the grid in real time, and update the grid information in real time; A GIS data processing unit, connected to the dynamic grid management unit, for processing and analyzing GIS data, including map rendering, spatial query and path planning functions; A control instruction receiving unit, configured to receive a control instruction input by a user and obtain a first request text; a control instruction parsing unit, connected to the control instruction receiving unit, and configured to parse the first request text to form a structured keyword list; A control instruction recognition unit is used to input the structured keyword list into a pre-trained control instruction understanding model. The model is responsible for parsing the deep semantics of the keyword list, identifying the control instruction type and the completeness of the control instruction input conditions. If the model cannot identify the type of the control instruction, or the input conditions of the control instruction are incomplete, it will feedback the user to enter the correct control instruction; if the model recognizes the type of the control instruction and the input conditions of the control instruction are complete, it will output a standardized control instruction based on the built-in template to obtain the second request text; A candidate answer recall unit, configured to recall relevant information from the data modules of multiple dimensions of the data module construction unit according to the second request text, and obtain multiple candidate answers; A request reply generating unit is used to encapsulate the second request text and multiple candidate answers according to a preset prompt word template to obtain a third request text, further pass the third request text into the general large language model, and use the natural language understanding and generation capabilities of the general large language model to integrate, optimize and adjust the multiple candidate answers to form an answer including management and control suggestions and impact analysis, and obtain a request reply text; The control instruction confirmation unit is used to feed back the request reply text to the user and ask the user whether to confirm the execution.

2. The GIS dynamic grid management and control system based on the intelligent question-answering model according to claim 1 is characterized in that: The data module constructed by the data module construction unit includes: GIS basic database module, used to store basic geographic spatial data; Real-time data access module, used to receive and process real-time data from various sensors; Grid management database module, used to store grid division, grid attributes and event information within the grid; High-dimensional geographic feature vector database module, used to store high-dimensional vectors of feature-extracted geographic spatial data; GIS business knowledge graph module builds geographic spatial knowledge graph and stores geographic entities, attributes and relationships.

3. The GIS dynamic grid management and control system based on the intelligent question-answering model according to claim 2 is characterized in that: The control instruction understanding model is obtained by training based on a general language model.

4. The GIS dynamic grid management and control system based on the intelligent question-answering model according to claim 3 is characterized in that: Methods for obtaining the control instruction understanding model include: The control instruction types and control instruction conditions are defined using the extensible markup language XML syntax, wherein the control instruction types include map rendering instructions, spatial query instructions, path planning instructions, and data update instructions. Each instruction type corresponds to a set of specific control instruction conditions, and the control instruction conditions are described in the form of XML tags to form a control instruction definition template file with a hierarchical structure and attribute information; Parsing the control instruction definition template file, extracting the control instruction type, control instruction conditions and the relationship between them through a parsing algorithm, and generating a control instruction definition data set for understanding by a general language model, wherein the data set expresses the complete semantics of the control instruction in the form of natural language; The control instruction definition data set is input into a general language model, and through training, the general language model obtains context information for understanding the control instruction structure and the relationship between control instructions, thereby having the ability to perform deep semantic analysis on the control instructions.

5. The GIS dynamic grid management and control system based on the intelligent question-answering model according to claim 4 is characterized in that: The control instruction parsing unit parses the first request text in the following ways: word segmentation, part-of-speech tagging, dependency syntactic analysis, and keyword extraction.

6. The GIS dynamic grid management and control system based on the intelligent question-answering model according to claim 5 is characterized in that: For different types of data modules, the management and control instruction identification unit generates corresponding second request texts respectively.

7. The GIS dynamic grid management and control system based on the intelligent question-answering model according to claim 6 is characterized in that: The control suggestions output by the request response generation unit are specific operational suggestions given based on the control instructions input by the user and the current status of the geographic spatial data in the GIS grid, combined with the analysis capabilities of the general large language model.

8. The GIS dynamic grid management and control system based on the intelligent question-answering model according to claim 4 is characterized in that: The steps of parsing the control instruction definition template file to generate the control instruction definition data set include: Step 1: Select a programming language for parsing XML files, and import a toolkit for parsing XML files into the programming language; Step 2: Use the toolkit to load the control instruction definition template file; Step 3: Define an array for caching text fragments; Step 4: Obtain each type of label through the label traversal method provided by the toolkit, and insert text fragments for creating control instructions, defining instruction parameters, setting instruction conditions, and specifying instruction actions into the cache array; Step 5: After the traversal is completed, all text fragments in the cache array are converted into complete text strings and saved to generate a control instruction definition data set.

9. The GIS dynamic grid management and control system based on the intelligent question-answering model according to claim 1 is characterized in that: The GIS data processing unit performs spatial segmentation and organization on the data according to the grid information provided by the dynamic grid management unit.

10. The GIS dynamic grid management and control system based on the intelligent question-answering model according to claim 1, characterized in that: The preset prompt word template includes a question type identifier, question keywords, a candidate answer source identifier, and an answer fusion requirement.

Images