Meteorological data analysis method, device, equipment and medium based on large model intelligent agent

Abstract

This application discloses a meteorological data analysis method, apparatus, device, and medium based on a large-scale model agent. These methods relate to the field of intelligent meteorological data management and are applied to a pre-set meteorological data analysis system. The method includes: when a master agent receives a weather message to be processed, it determines and calls a target agent from multiple extreme weather scenario agents; the master agent and multiple extreme weather scenario agents are pre-built using a target large language model; the target agent matches its own first prompt word template based on the received master agent instruction to determine and call a target meteorological data acquisition tool; the target meteorological data acquisition tool matches its own second prompt word template based on the received target agent instruction to acquire meteorological data based on the determined target interface parameters; the target agent determines a target answer based on the data acquisition result, allowing the master agent to reply to the message in the form of a dialogue. This method overcomes the limitations of existing meteorological data applications.

Description

Technical Field

[0001] The present invention relates to the field of intelligent management of meteorological data, and in particular to a meteorological data analysis method, device, equipment and medium based on a large model intelligent body. Background Art

[0002] Currently, ChatBI (Chat-based Business Intelligence) technology enables users to query and analyze data through natural language conversations, making data analysis more intuitive and user-friendly. However, due to difficulties and inefficiencies in obtaining meteorological data, as well as constraints on data confidentiality, its application in meteorological data is still in its early stages of exploration and has not been effectively implemented.

[0003] Therefore, how to overcome the limitations of existing meteorological data applications and achieve efficient use of meteorological data is an urgent problem to be solved. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a meteorological data analysis method, device, equipment, and medium based on a large-scale intelligent model, which can effectively overcome the limitations of existing meteorological data applications, improve the convenience and efficiency of meteorological data acquisition, and enhance the accuracy and reliability of answers. The specific scheme is as follows:

[0005] In a first aspect, the present application provides a meteorological data analysis method based on a large model agent, which is applied to a preset meteorological data analysis system, wherein the large model agent is a large model agent built based on a chatbot and business intelligence functions; wherein the method includes:

[0006] When a weather message to be processed is received from a client through a master agent, a target agent is determined and called from multiple extreme weather scenario agents based on the weather message to be processed; the master agent and the multiple extreme weather scenario agents are agents pre-built using a target large language model;

[0007] The target agent matches its own first prompt word template based on the received main agent instruction, and determines and calls the target meteorological data acquisition tool according to the obtained first template matching result;

[0008] Matching the target meteorological data acquisition tool with its own second prompt word template based on the received target agent instruction, and determining target interface parameters according to the obtained second template matching result, so as to acquire meteorological data based on the target interface parameters to obtain a data acquisition result;

[0009] The target agent performs analysis based on the received data acquisition result and the first template matching result to determine the corresponding target answer, and returns the target answer to the main agent so that the main agent can reply to the message using the target answer in the form of a dialogue.

[0010] Optionally, determining and calling a target agent from a plurality of extreme weather scenario agents based on the to-be-processed meteorological message includes:

[0011] Analyzing the to-be-processed meteorological message by the master agent to determine whether to call an extreme weather scenario agent;

[0012] When the first judgment result is yes, a target agent is determined from a plurality of extreme weather scenario agents based on the weather scenario type corresponding to the to-be-processed meteorological message, and the target agent is called.

[0013] Optionally, the target agent matches its own first prompt word template based on the received main agent instruction, and determines and calls the target meteorological data acquisition tool according to the obtained first template matching result, including:

[0014] receiving, through the target agent, a master agent instruction corresponding to the weather message to be processed;

[0015] Matching the first prompt word template of the master agent based on the master agent's instruction to obtain a corresponding first template matching result; the first prompt word template includes role prompt information, task prompt information, input parameter prompt information, and question-answer prompt information;

[0016] The corresponding target meteorological data acquisition tool name is determined by analyzing the first template matching result, and the target meteorological data acquisition tool name is used to call the target meteorological data acquisition tool.

[0017] Optionally, the target meteorological data acquisition tool matches its own second prompt word template based on the received target agent instruction, and determines target interface parameters according to the obtained second template matching result, so as to acquire meteorological data based on the target interface parameters, including:

[0018] Receiving corresponding target agent instructions through the target meteorological data acquisition tool;

[0019] Matching the second prompt word template of the target agent based on the target agent instruction to obtain a second template matching result;

[0020] Determine target interface parameters and first target answer related fields by analyzing the second template matching result;

[0021] Based on the target interface parameters, the corresponding data service interface is accessed, and the data service interface and the first target answer-related fields are used to connect to the meteorological big data platform and obtain meteorological data to obtain data acquisition results; the data acquisition results include the second target answer-related fields and answer-related data.

[0022] Optionally, the analyzing, by the target agent, the received data acquisition result and the first template matching result to determine a corresponding target answer includes:

[0023] receiving, through the target agent, the data acquisition result corresponding to the target agent instruction returned by the target meteorological data acquisition tool;

[0024] Answer information corresponding to the weather message to be processed is generated by analyzing the data acquisition result and the first template matching result to obtain a target answer.

[0025] Optionally, the master agent uses the target answer to reply to the message in the form of a dialogue, including:

[0026] The master agent determines that the target answer meets a preset condition to obtain a second judgment result;

[0027] If the second judgment result indicates that the preset condition is met, the to-be-processed weather message is replied to in a dialogue form based on the target answer.

[0028] Optionally, after obtaining the second judgment result, the method further includes:

[0029] If the second judgment result indicates that the preset condition is not met, the instruction information corresponding to the weather message to be processed is regenerated, and the new main intelligent agent instruction is sent to the target intelligent agent to jump again to the step of matching the first prompt word template of the target intelligent agent based on the received main intelligent agent instruction.

[0030] In a second aspect, the present application provides a meteorological data analysis device based on a large model agent, which is applied to a preset meteorological data analysis system. The large model agent is a large model agent built based on a chatbot and business intelligence functions; wherein the device includes:

[0031] An agent calling module is configured to, upon receiving a pending weather message inputted by a client via a master agent, determine and call a target agent from among multiple extreme weather scenario agents based on the pending weather message; the master agent and the multiple extreme weather scenario agents are agents pre-built using a target large language model;

[0032] a tool calling module, configured to match the target agent with its own first prompt word template based on the received master agent instruction, and determine and call a target meteorological data acquisition tool based on the obtained first template matching result;

[0033] a data acquisition module, configured to match its own second prompt word template based on the received target agent instruction through the target meteorological data acquisition tool, and determine target interface parameters based on the obtained second template matching result, so as to acquire meteorological data based on the target interface parameters to obtain a data acquisition result;

[0034] The message reply module is used to analyze the received data acquisition result and the first template matching result through the target agent to determine the corresponding target answer, and return the target answer to the main agent so that the main agent can use the target answer to reply to the message in the form of a dialogue.

[0035] In a third aspect, the present application provides an electronic device, comprising:

[0036] Memory, used to store computer programs;

[0037] A processor is used to execute the computer program to implement the steps of the aforementioned meteorological data analysis method based on a large model intelligent agent.

[0038] In a fourth aspect, the present application provides a computer-readable storage medium for storing a computer program, which, when executed by a processor, implements the steps of the aforementioned large-model intelligent agent-based meteorological data analysis method.

[0039] It can be seen that in the present application, in the preset meteorological data analysis system, when the meteorological message to be processed input by the client is received through the main intelligent agent, the target intelligent agent is determined and called from multiple extreme weather scene intelligent agents based on the meteorological message to be processed; the main intelligent agent and the multiple extreme weather scene intelligent agents are intelligent agents pre-constructed using the target large language model; the target intelligent agent matches its own first prompt word template based on the received main intelligent agent instruction, and determines and calls the target meteorological data acquisition tool based on the obtained first template matching result; the target meteorological data acquisition tool matches its own second prompt word template based on the received target intelligent agent instruction, and determines the target interface parameters based on the obtained second template matching result, so as to acquire meteorological data based on the target interface parameters to obtain data acquisition results; the target intelligent agent analyzes the received data acquisition results and the first template matching results to determine the corresponding target answer, and returns the target answer to the main intelligent agent, so that the main intelligent agent can reply to the message using the target answer in the form of a dialogue. That is, in this application, when the main agent in the preset meteorological data analysis system receives a meteorological message to be processed, it will trigger the corresponding extreme weather scene agent call operation, and then the called target agent will use the instruction sent by the main agent to match its own prompt word template to call the meteorological data acquisition tool, and then the called target meteorological data acquisition tool will use the instruction sent by the target agent to match its own prompt word template to collect meteorological data and return the result to the target agent. After the target agent generates an answer using the data acquisition result and returns it to the main agent, the main agent uses the target answer to reply to the message in the form of a dialogue. In this way, the limitations of existing meteorological data applications can be effectively overcome, the convenience and efficiency of meteorological data acquisition can be improved, and the accuracy and reliability of the answers can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0041] Figure 1 A flow chart of a meteorological data analysis method based on a large model agent provided in this application;

[0042] Figure 2 A schematic diagram of a specific process of meteorological data analysis based on a large model agent provided in this application;

[0043] Figure 3 A schematic diagram of the structure of a meteorological data analysis device based on a large model intelligent agent provided in this application;

[0044] Figure 4 This is a structural diagram of an electronic device provided in this application. DETAILED DESCRIPTION

[0045] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0046] Currently, ChatBI technology enables users to query and analyze data through natural language conversations, making data analysis more intuitive and easy to use. However, due to the difficulties and low efficiency in obtaining meteorological data, as well as the constraints of data confidentiality, its application in meteorological data is still in the initial exploratory stage and cannot be effectively applied. To this end, this application provides a meteorological data analysis solution based on a large model intelligent agent, which can effectively overcome the limitations of existing meteorological data applications, improve the convenience and efficiency of meteorological data acquisition, and improve the accuracy and reliability of answers.

[0047] See also Figure 1 As shown, an embodiment of the present invention discloses a meteorological data analysis method based on a large model agent, which is applied to a preset meteorological data analysis system. The large model agent is a large model agent built based on a chatbot and business intelligence functions; wherein the method includes:

[0048] Step S11: When a weather message to be processed is received from a client through a main agent, a target agent is determined and called from a plurality of extreme weather scene agents based on the weather message to be processed; the main agent and the plurality of extreme weather scene agents are agents pre-constructed using a target large language model.

[0049] Combine Figure 2 As shown, in this embodiment, when the user inputs the weather information to be processed into the system, the main agent (i.e. Figure 2 That is, the weather message to be processed is received from multiple extreme weather scene agents (i.e., the master agent in the Figure 2The method includes determining and invoking a target agent from an independent extreme weather scenario agent in the master agent (e.g., an independent extreme weather scenario agent in the master agent), including: analyzing the pending weather message to be processed to determine whether to invoke the extreme weather scenario agent; and if a first determination result is yes, determining a target agent from multiple extreme weather scenario agents based on the weather scenario type corresponding to the pending weather message, and invoking the target agent. In other words, in this embodiment, the master agent first determines whether to invoke the extreme weather scenario agent, and if so, determines and invokes the target agent.

[0050] It should be understood that the extreme weather scenario agents include a blizzard agent, a severe convection agent, a rainstorm agent, and a cold wave agent. The master agent is responsible for routing and controlling the extreme weather scenario agents. Furthermore, each agent is supported by the same large language model (i.e., the target large language model), in specific permutations and combinations. The master agent is responsible for coordinating the autonomous extreme weather scenario agents, enabling them to efficiently collaborate and execute complex tasks.

[0051] Each extreme weather scenario agent can have its own prompts, LLM (Large Language Model), tools, and other custom code to collaborate with other agents. However, the same LLM can also play different roles based on the provided prompts. The prompt word template for each extreme weather scenario agent includes role prompts, task prompts, input parameter prompts, and question-and-answer prompts. Each extreme weather scenario agent has multiple data acquisition tools that connect to the meteorological data platform, which are divided into two types: data query tools and data statistical feature tools.

[0052] Taking the rainstorm agent as an example, its prompt word template can be as follows:

[0053] Role Tip: Rainstorm Scene

[0054] Task Tip: Accurate query and statistical analysis of rainstorm scene data

[0055] Input parameter prompts: time, location, ground data elements

[0056] Question and answer prompt: Basic situation / statistical characteristics of ground data elements related to heavy rain scenes at a certain time and place.

[0057] The target large language model is used by both the main agent and the multiple independent extreme weather scenario agents, each with different prompts, thus forming a self-reflective AI (artificial intelligence) agent. This method of using the same large language model in a recurring manner with multiple different roles is constructed using the Lang Graph framework. The Lang Graph framework can also be used to create multi-agent workflows. Just like in a self-reflective AI agent, the target large language model can play multiple roles, each acting as a different AI agent. This is the concept of multi-agent.

[0058] Furthermore, it is important to understand that the connections between agents are represented by edges. Each edge can have a control condition that guides the flow of information from one agent to another. Each agent has a state that can be updated with information during each flow.

[0059] Step S12: The target agent matches its own first prompt word template based on the received main agent instruction, and determines and calls the target meteorological data acquisition tool according to the obtained first template matching result.

[0060] Combine Figure 2 As shown, in this embodiment, after calling the target agent, the target agent determines and calls a meteorological data acquisition tool (i.e., the Figure 2 ). That is, the target agent matches its own first prompt word template based on the received main agent instruction, and determines and calls the target meteorological data acquisition tool according to the obtained first template matching result, including: receiving the main agent instruction corresponding to the meteorological message to be processed through the target agent; matching its own first prompt word template based on the main agent instruction to obtain the corresponding first template matching result; the first prompt word template includes role prompt information, task prompt information, input parameter prompt information and question-answer prompt information; determining the corresponding target meteorological data acquisition tool name by analyzing the first template matching result, and using the target meteorological data acquisition tool name to call the target meteorological data acquisition tool. That is, the called target agent matches the main agent instruction with the prompt word template, and abstracts the specific tool name of the data tool that needs to be called. It can be understood that different extreme weather scenario agents can have different numbers of meteorological data acquisition tools.

[0061] It should be understood that, regarding the design of the meteorological data acquisition tool, the data acquisition tool for a specific extreme weather scenario connects the meteorological big data cloud platform of the corresponding region with the extreme weather scenario agent through the data service interface, and performs data acquisition and analysis according to the business scenario requirements. Each of the meteorological data acquisition tools has its own data tool name, data tool parameters, data tool description prompt words, and data source and interface for connecting to the meteorological big data cloud platform. The meteorological data acquisition tool implements field extraction of the returned data according to the data tool function definition, analyzes statistical calculations and processing, and then outputs the numerical value to the user in the form of a dialogue. Among them, the data tool name needs to be an agent data tool name that is easy for the large language model agent tool to call and recognize intent, the data tool parameters need to be data tool parameters that are easy for the large language model agent to realize parameter extraction based on conversational questions, and the data tool description prompt words need to be description prompt words that are easy for the large model agent tool to call and recognize intent, thereby reducing the misjudgment rate of the agent tool.

[0062] The design example of the meteorological data acquisition tool can be seen in Table 1 below:

[0063] Table 1

[0064]

[0065] Furthermore, in this embodiment, the prompt word templates can be designed for scenarios involving heavy rain, snowstorms, severe convection, and cold waves. The meteorological data platform can be used to organize meteorological big data interfaces, algorithms, data sources, data fields, and other disaster meteorological data elements. Prompt word templates and annotation specifications for disaster meteorological data elements are developed, and standardized prompt word annotation and management of disaster meteorological data elements are implemented to enable intelligent retrieval and recommendation services for disaster meteorological data elements in various forms, including data, calculations, components, and modules.

[0066] Step S13: The target meteorological data acquisition tool matches its own second prompt word template based on the received target intelligent agent instruction, and determines the target interface parameters according to the obtained second template matching result, so as to acquire meteorological data based on the target interface parameters to obtain data acquisition results.

[0067] In this embodiment, a designated target meteorological data acquisition tool is invoked, which matches the target agent instruction with the prompt word template, abstracts the specific parameter values ​​corresponding to the data service interface and the fields required for the answer, and connects to the meteorological big data platform by accessing the data service interface to obtain the fields and data required to answer the user's question. Specifically, the target meteorological data acquisition tool matches its own second prompt word template based on the received target agent instruction, and determines the target interface parameters based on the second template matching result, so as to acquire meteorological data based on the target interface parameters. The method includes: receiving the corresponding target agent instruction through the target meteorological data acquisition tool; matching its own second prompt word template based on the target agent instruction to obtain a second template matching result; determining the target interface parameters and the first target answer-related fields by analyzing the second template matching result; accessing the corresponding data service interface based on the target interface parameters, and using the data service interface and the first target answer-related fields to connect to the meteorological big data platform and acquire meteorological data to obtain a data acquisition result; the data acquisition result includes the second target answer-related fields and the answer-related data. It is understood that different meteorological data acquisition tools may have different numbers and functions of data service interfaces.

[0068] Step S14: The target agent analyzes the received data acquisition result and the first template matching result to determine the corresponding target answer, and returns the target answer to the main agent so that the main agent can reply to the message using the target answer in the form of a dialogue.

[0069] In this embodiment, the target agent generates a corresponding answer based on the data acquisition result returned by the router and the prompt word template. Specifically, the target agent analyzes the received data acquisition result and the first template matching result to determine the corresponding target answer, including: receiving, by the target agent, the data acquisition result corresponding to the target agent's instruction, returned by the target meteorological data acquisition tool; and generating answer information corresponding to the to-be-processed meteorological message by analyzing the data acquisition result and the first template matching result to obtain the target answer.

[0070] It should be understood that, in this embodiment, if the target answer fed back by the target agent to the main agent is the final answer to the weather message to be processed, the main agent will feed back the target answer to the user. Otherwise, the main agent may send instructions to the extreme weather scene agent again. That is, the main agent uses the target answer to reply to the message in the form of a dialogue, including: judging by the main agent that the target answer meets the preset conditions to obtain a second judgment result; if the second judgment result shows that the preset conditions are met, replying to the weather message to be processed in the form of a dialogue based on the target answer. Moreover, after obtaining the second judgment result, it also includes: if the second judgment result shows that the preset conditions are not met, regenerating the instruction information corresponding to the weather message to be processed, and sending the obtained new main agent instruction to the target agent, so as to jump back to the step of matching its own first prompt word template based on the received main agent instruction by the target agent.

[0071] In summary, the technical solution described in this embodiment can have the following beneficial effects:

[0072] (1) Filling the technical gap in the application of ChatBI technology in the meteorological field;

[0073] (2) It can more accurately understand the user's natural language instructions and convert them into a data query language that is easy to process. It also supports access to multiple data sources and provides a variety of data analysis and visualization tools to meet the user's multi-dimensional needs for meteorological data, making the meteorological data analysis process more intuitive and efficient.

[0074] (3) Supporting real-time data access, users can query the latest weather data at any time and make timely decisions, making weather data analysis more flexible and accurate;

[0075] (4) Allow users to query and analyze data through dialogue, improving the convenience and efficiency of obtaining meteorological data. It can not only quickly answer questions, but also automatically generate various charts and reports to help users better understand meteorological data.

[0076] (5) The security of sensitive information that may be involved in meteorological data can be protected through the private deployment of large models and data permission control, thereby ensuring data security and privacy protection;

[0077] (6) Through dialogue, intelligent data processing and analysis can be achieved, human intervention can be reduced, and the efficiency and accuracy of data analysis can be improved.

[0078] It can be seen that in the embodiment of the present application, in the preset meteorological data analysis system, when the meteorological message to be processed input by the client is received through the main intelligent agent, the target intelligent agent is determined and called from multiple extreme weather scene intelligent agents based on the meteorological message to be processed; the main intelligent agent and the multiple extreme weather scene intelligent agents are intelligent agents pre-constructed using the target large language model; the target intelligent agent matches its own first prompt word template based on the received main intelligent agent instruction, and determines and calls the target meteorological data acquisition tool based on the obtained first template matching result; the target meteorological data acquisition tool matches its own second prompt word template based on the received target intelligent agent instruction, and determines the target interface parameters based on the obtained second template matching result, so as to acquire meteorological data based on the target interface parameters to obtain data acquisition results; the target intelligent agent analyzes the received data acquisition result and the first template matching result to determine the corresponding target answer, and returns the target answer to the main intelligent agent, so that the main intelligent agent can reply to the message using the target answer in the form of a dialogue. That is, in this application, when the main agent in the preset meteorological data analysis system receives a meteorological message to be processed, it will trigger the corresponding extreme weather scene agent call operation, and then the called target agent will use the instruction sent by the main agent to match its own prompt word template to call the meteorological data acquisition tool, and then the called target meteorological data acquisition tool will use the instruction sent by the target agent to match its own prompt word template to collect meteorological data and return the result to the target agent. After the target agent generates an answer using the data acquisition result and returns it to the main agent, the main agent uses the target answer to reply to the message in the form of a dialogue. In this way, the limitations of existing meteorological data applications can be effectively overcome, the convenience and efficiency of meteorological data acquisition can be improved, and the accuracy and reliability of the answers can be improved.

[0079] See also Figure 3 As shown, the embodiment of the present application also discloses a meteorological data analysis device based on a large model agent, which is applied to a preset meteorological data analysis system. The large model agent is a large model agent built based on a chatbot and business intelligence functions; wherein the device includes:

[0080] The agent calling module 11 is configured to, upon receiving a weather message to be processed from a client via a master agent, determine and call a target agent from among multiple extreme weather scenario agents based on the weather message to be processed; the master agent and the multiple extreme weather scenario agents are agents pre-built using a target large language model;

[0081] The tool calling module 12 is configured to match the first prompt word template of the target agent based on the received master agent instruction, and determine and call the target meteorological data acquisition tool according to the obtained first template matching result;

[0082] The data acquisition module 13 is configured to match the second prompt word template of the target intelligent agent based on the received target intelligent agent instruction through the target meteorological data acquisition tool, determine target interface parameters based on the obtained second template matching result, and acquire meteorological data based on the target interface parameters to obtain a data acquisition result;

[0083] The message reply module 14 is used to analyze the received data acquisition result and the first template matching result through the target agent to determine the corresponding target answer, and return the target answer to the main agent so that the main agent can use the target answer to reply to the message in the form of a dialogue.

[0084] Among them, for more specific working processes of the above modules, please refer to the corresponding contents disclosed in the aforementioned embodiments, which will not be repeated here.

[0085] It can be seen that in this application, when the main agent in the preset meteorological data analysis system receives a meteorological message to be processed, it will trigger the corresponding extreme weather scene agent call operation, and then the called target agent will use the instruction sent by the main agent to match its own prompt word template to call the meteorological data acquisition tool, and then the called target meteorological data acquisition tool will use the instruction sent by the target agent to match its own prompt word template to collect meteorological data and return the result to the target agent. After the target agent generates an answer using the data acquisition result and returns it to the main agent, the main agent uses the target answer to reply to the message in the form of a dialogue. In this way, the limitations of existing meteorological data applications can be effectively overcome, and the convenience and efficiency of meteorological data acquisition are improved, and the accuracy and reliability of the answers are improved.

[0086] In some specific embodiments, the agent calling module 11 can be specifically used to: analyze the meteorological message to be processed through the main agent to determine whether to call the extreme weather scene agent; when the first judgment result obtained is yes, determine the target agent from multiple extreme weather scene agents based on the weather scene type corresponding to the meteorological message to be processed, and call the target agent.

[0087] In some specific embodiments, the tool calling module 12 can be specifically used to: receive the main agent instruction corresponding to the meteorological message to be processed through the target agent; match its own first prompt word template based on the main agent instruction to obtain a corresponding first template matching result; the first prompt word template includes role prompt information, task prompt information, input parameter prompt information and question-and-answer prompt information; determine the corresponding target meteorological data acquisition tool name by analyzing the first template matching result, and use the target meteorological data acquisition tool name to call the target meteorological data acquisition tool.

[0088] In some specific embodiments, the data acquisition module 13 can be specifically used to: receive corresponding target intelligent agent instructions through the target meteorological data acquisition tool; match its own second prompt word template based on the target intelligent agent instructions to obtain a second template matching result; determine the target interface parameters and the first target answer related fields by analyzing the second template matching result; access the corresponding data service interface based on the target interface parameters, and use the data service interface and the first target answer related fields to connect to the meteorological big data platform and acquire meteorological data to obtain data acquisition results; the data acquisition results include second target answer related fields and answer related data.

[0089] In some specific embodiments, the message reply module 14 can be specifically used to: receive the data acquisition result corresponding to the target intelligent agent instruction returned by the target meteorological data acquisition tool through the target intelligent agent; generate answer information corresponding to the meteorological message to be processed by analyzing the data acquisition result and the first template matching result to obtain the target answer.

[0090] In some specific embodiments, the message reply module 14 can be specifically used to: determine through the main intelligent agent whether the target answer meets the preset conditions to obtain a second judgment result; if the second judgment result indicates that the preset conditions are met, reply to the pending weather message based on the target answer and in the form of a dialogue.

[0091] In some specific embodiments, the meteorological data analysis device based on the large model intelligent agent can also be used to: if the second judgment result indicates that the preset condition is not met, regenerate the instruction information corresponding to the meteorological message to be processed, and send the obtained new main intelligent agent instruction to the target intelligent agent, so as to jump again to the step of matching the first prompt word template of the target intelligent agent based on the received main intelligent agent instruction.

[0092] Furthermore, the embodiment of the present application also discloses an electronic device, Figure 4This is a structural diagram of an electronic device 2 / 30 according to an exemplary embodiment. The content in the diagram cannot be considered as any limitation on the scope of use of this application.

[0093] Figure 4 This is a schematic diagram of the structure of an electronic device 2 / 30 provided in an embodiment of the present application. The electronic device 2 / 30 may specifically include: at least one processor 2 / 31, at least one memory 2 / 32, a power supply 2 / 33, a communication interface 2 / 34, an input / output interface 2 / 35, and a communication bus 2 / 36. The memory 2 / 32 is used to store a computer program, which is loaded and executed by the processor 2 / 31 to implement the relevant steps in the meteorological data analysis method based on a large model agent disclosed in any of the aforementioned embodiments. In addition, the electronic device 2 / 30 in this embodiment may specifically be an electronic computer.

[0094] In this embodiment, the power supply 2 / 33 is used to provide operating voltage for each hardware device on the electronic device 2 / 30; the communication interface 2 / 34 can create a data transmission channel between the electronic device 2 / 30 and the external device. The communication protocol it follows is any communication protocol that can be applied to the technical solution of this application and is not specifically limited here; the input and output interface 2 / 35 is used to obtain external input data or output data to the outside world. Its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0095] In addition, the memory 2 / 32, as a carrier for resource storage, can be a read-only memory, random access memory, disk or CD, etc. The resources stored thereon can include an operating system 2 / 321, a computer program 2 / 322, etc., and the storage method can be temporary storage or permanent storage.

[0096] The operating system 2 / 321 is used to manage and control the hardware devices and computer program 2 / 322 on the electronic device 2 / 30, and can be Windows Server, NetWare, Unix, Linux, etc. In addition to including a computer program capable of implementing the large-model agent-based meteorological data analysis method disclosed in any of the aforementioned embodiments and executed by the electronic device 2 / 30, the computer program 2 / 322 may further include computer programs capable of performing other specific tasks.

[0097] Furthermore, this application discloses a computer-readable storage medium for storing a computer program; wherein, when executed by a processor, the computer program implements the aforementioned large-model agent-based meteorological data analysis method. The specific steps of this method can be found in the corresponding contents disclosed in the aforementioned embodiments and will not be repeated here.

[0098] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from the other embodiments. Reference can be made to the descriptions of the identical or similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and the relevant parts can be referred to the descriptions of the methods.

[0099] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0100] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0101] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0102] The above is a detailed introduction to the technical solution provided by the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for those skilled in the art, according to the ideas of the present application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A meteorological data analysis method based on a large model agent, characterized in that: Applied to a preset meteorological data analysis system, the large model agent is a large model agent built based on a chatbot and business intelligence functions; wherein the method includes: When a weather message to be processed is received from a client through a master agent, a target agent is determined and called from multiple extreme weather scenario agents based on the weather message to be processed; the master agent and the multiple extreme weather scenario agents are agents pre-built using a target large language model; The target agent matches its own first prompt word template based on the received main agent instruction, and determines and calls the target meteorological data acquisition tool according to the obtained first template matching result; Matching the target meteorological data acquisition tool with its own second prompt word template based on the received target agent instruction, and determining target interface parameters according to the obtained second template matching result, so as to acquire meteorological data based on the target interface parameters to obtain a data acquisition result; The target agent performs analysis based on the received data acquisition result and the first template matching result to determine the corresponding target answer, and returns the target answer to the main agent so that the main agent can reply to the message using the target answer in the form of a dialogue.

2. The meteorological data analysis method based on a large model agent according to claim 1 is characterized in that: The determining and calling a target agent from a plurality of extreme weather scenario agents based on the to-be-processed meteorological message includes: Analyzing the to-be-processed meteorological message by the master agent to determine whether to call an extreme weather scenario agent; When the first judgment result is yes, a target agent is determined from a plurality of extreme weather scenario agents based on the weather scenario type corresponding to the to-be-processed meteorological message, and the target agent is called.

3. The meteorological data analysis method based on a large model agent according to claim 1 is characterized in that: The target agent matches its own first prompt word template based on the received main agent instruction, and determines and calls the target meteorological data acquisition tool according to the obtained first template matching result, including: receiving, through the target agent, a master agent instruction corresponding to the weather message to be processed; Matching the first prompt word template of the master agent based on the master agent's instruction to obtain a corresponding first template matching result; the first prompt word template includes role prompt information, task prompt information, input parameter prompt information, and question-answer prompt information; The corresponding target meteorological data acquisition tool name is determined by analyzing the first template matching result, and the target meteorological data acquisition tool name is used to call the target meteorological data acquisition tool.

4. The meteorological data analysis method based on a large model agent according to claim 1 is characterized in that: The target meteorological data acquisition tool matches its own second prompt word template based on the received target agent instruction, and determines the target interface parameters according to the obtained second template matching result, so as to acquire meteorological data based on the target interface parameters, including: Receiving corresponding target agent instructions through the target meteorological data acquisition tool; Matching the second prompt word template of the target agent based on the target agent instruction to obtain a second template matching result; Determine target interface parameters and first target answer related fields by analyzing the second template matching result; Based on the target interface parameters, the corresponding data service interface is accessed, and the data service interface and the first target answer-related fields are used to connect to the meteorological big data platform and obtain meteorological data to obtain data acquisition results; the data acquisition results include the second target answer-related fields and answer-related data.

5. The meteorological data analysis method based on a large model agent according to claim 1 is characterized in that: The target agent performs analysis based on the received data acquisition result and the first template matching result to determine a corresponding target answer, including: receiving, through the target agent, the data acquisition result corresponding to the target agent instruction returned by the target meteorological data acquisition tool; Answer information corresponding to the weather message to be processed is generated by analyzing the data acquisition result and the first template matching result to obtain a target answer.

6. The meteorological data analysis method based on a large model agent according to claim 1 is characterized in that: The master agent uses the target answer to reply to the message in the form of a dialogue, including: The master agent determines that the target answer meets a preset condition to obtain a second judgment result; If the second judgment result indicates that the preset condition is met, the to-be-processed weather message is replied to in a dialogue form based on the target answer.

7. The meteorological data analysis method based on a large model agent according to claim 6 is characterized in that: After obtaining the second judgment result, the method further includes: If the second judgment result indicates that the preset condition is not met, the instruction information corresponding to the weather message to be processed is regenerated, and the new main intelligent agent instruction is sent to the target intelligent agent to jump again to the step of matching the first prompt word template of the target intelligent agent based on the received main intelligent agent instruction.

8. A meteorological data analysis device based on a large model agent, characterized in that: Applied to a preset meteorological data analysis system, the large model intelligent agent is a large model intelligent agent built based on a chatbot and business intelligence functions; wherein the device includes: An agent calling module is configured to, upon receiving a pending weather message inputted by a client via a master agent, determine and call a target agent from among multiple extreme weather scenario agents based on the pending weather message; the master agent and the multiple extreme weather scenario agents are agents pre-built using a target large language model; a tool calling module, configured to match the target agent with its own first prompt word template based on the received master agent instruction, and determine and call a target meteorological data acquisition tool based on the obtained first template matching result; a data acquisition module, configured to match its own second prompt word template based on the received target agent instruction through the target meteorological data acquisition tool, and determine target interface parameters based on the obtained second template matching result, so as to acquire meteorological data based on the target interface parameters to obtain a data acquisition result; The message reply module is used to analyze the received data acquisition result and the first template matching result through the target agent to determine the corresponding target answer, and return the target answer to the main agent so that the main agent can use the target answer to reply to the message in the form of a dialogue.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the large model agent-based meteorological data analysis method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that Used to store a computer program, which, when executed by a processor, implements the meteorological data analysis method based on a large model intelligent agent as described in any one of claims 1 to 7.

Images