Meteorological data analysis method and apparatus based on large-model agent, device, and medium

Abstract

A meteorological data analysis method and apparatus based on a large-model agent, a device, and a medium, relating to the field of intelligent management of meteorological data. The method is applied to a preset meteorological data analysis system, and comprises: when a main agent receives a meteorological message to be processed, determining and calling a target agent from among a plurality of extreme weather scenario agents, the main agent and the plurality of extreme weather scenario agents being agents constructed in advance using a target large language model; the target agent matching a first prompt template thereof on the basis of a received main-agent instruction, so as to determine and call a target meteorological data acquisition tool; the target meteorological data acquisition tool matching a second prompt template thereof on the basis of a received target-agent instruction, so as to acquire meteorological data on the basis of determined target interface parameters; and the target agent determining a target answer on the basis of a data acquisition result, such that the main agent makes a message reply in a conversational manner. The present invention can overcome limitations of existing meteorological data applications.

Description

Meteorological data analysis methods, devices, equipment, and media based on large-scale intelligent agents.

[0001] This application claims priority to Chinese Patent Application No. 202411830230.0, filed on December 12, 2024, entitled "Meteorological Data Analysis Method, Apparatus, Equipment and Medium Based on Large Model Intelligent Agent", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This invention relates to the field of intelligent management of meteorological data, and in particular to meteorological data analysis methods, devices, equipment and media based on large model intelligent agents. Background Technology

[0003] Currently, ChatBI (Chat-based Business Intelligence) technology enables users to query and analyze data through natural language dialogue, making data analysis more intuitive and user-friendly. However, due to difficulties in acquiring meteorological data, low acquisition efficiency, and constraints related to data confidentiality, its application in meteorological data is still in the initial exploratory stage and cannot be effectively utilized.

[0004] Therefore, how to overcome the limitations of existing meteorological data applications and achieve efficient utilization of meteorological data is an urgent problem to be solved. Summary of the Invention

[0005] In view of this, the purpose of this invention is to provide a method, apparatus, equipment, and medium for meteorological data analysis based on a large-scale intelligent agent model. This effectively overcomes the limitations of existing meteorological data applications, improves the convenience and efficiency of meteorological data acquisition, and enhances the accuracy and reliability of the results. The specific solution is as follows:

[0006] Firstly, this application provides a meteorological data analysis method based on a large-scale intelligent agent, applied to a pre-set meteorological data analysis system, wherein the large-scale intelligent agent is a large-scale intelligent agent constructed based on chatbot and business intelligence functions; wherein the method includes:

[0007] When the main agent receives a weather message to be processed from the client, it determines and calls the target agent from multiple extreme weather scenario agents based on the weather message to be processed; the main agent and the multiple extreme weather scenario agents are agents that are pre-constructed using the target large language model;

[0008] The target agent matches its own first prompt word template based on the received instructions from the main agent, and determines and calls the target meteorological data acquisition tool based on the obtained first template matching result;

[0009] 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 and obtain the data acquisition result.

[0010] The target 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 agent so that the main agent can reply to the message using the target answer in the form of a dialogue.

[0011] Optionally, determining and invoking the target agent from multiple extreme weather scenario agents based on the meteorological message to be processed includes:

[0012] The main intelligent agent analyzes the meteorological message to be processed to determine whether to invoke the intelligent agent for extreme weather scenarios.

[0013] When the first judgment result is yes, the target agent is determined from multiple extreme weather scenario agents based on the weather scenario type corresponding to the meteorological message to be processed, and the target agent is invoked.

[0014] Optionally, the step of matching the target agent's first prompt word template with the received master agent's instruction, and determining and invoking the target meteorological data acquisition tool based on the obtained first template matching result, includes:

[0015] The target agent receives instructions from the main agent corresponding to the meteorological message to be processed.

[0016] Based on the main intelligent agent's instructions, it matches its own first prompt word template 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 pair prompt information;

[0017] The 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.

[0018] Optionally, the step of using the target meteorological data acquisition tool to match its own second prompt word template based on the received target agent instruction, and determining the target interface parameters based on the obtained second template matching result, so as to acquire meteorological data based on the target interface parameters, includes:

[0019] The target meteorological data acquisition tool receives the corresponding target intelligent agent commands.

[0020] The target agent matches its own second prompt word template based on the target agent's instructions to obtain the second template matching result;

[0021] The target interface parameters and the fields related to the first target answer are determined by analyzing the matching results of the second template.

[0022] Based on the target interface parameters, access the corresponding data service interface, 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 the second target answer related fields and answer related data.

[0023] Optionally, the step of determining the corresponding target answer by analyzing the received data acquisition results and the first template matching results through the target agent includes:

[0024] The target agent receives the data acquisition result returned by the target meteorological data acquisition tool, which corresponds to the command of the target agent;

[0025] By analyzing the data acquisition results and the first template matching results, answer information corresponding to the weather message to be processed is generated to obtain the target answer.

[0026] Optionally, the main agent responds to the message using the target answer in the form of a dialogue, including:

[0027] The main intelligent agent determines that the target answer meets the preset conditions to obtain a second judgment result;

[0028] If the second judgment result indicates that the preset conditions are met, then the weather message to be processed is replied to based on the target answer in a dialogue format.

[0029] Optionally, after obtaining the second judgment result, the method further includes:

[0030] If the second judgment result does not meet the preset conditions, then the instruction information corresponding to the weather message to be processed is regenerated, and the new main agent instruction is sent to the target agent to jump back to the step of matching the first prompt word template of itself by the target agent based on the received main agent instruction.

[0031] Secondly, this application provides a meteorological data analysis device based on a large-scale intelligent agent, applied to a pre-set meteorological data analysis system. The large-scale intelligent agent is a large-scale intelligent agent constructed based on chatbot and business intelligence functions. The device includes:

[0032] The agent invocation module is used to determine and invoke a target agent from multiple extreme weather scenario agents based on the unprocessed meteorological message input by the client when the main agent receives the unprocessed meteorological message; the main agent and the multiple extreme weather scenario agents are agents pre-constructed using the target large language model;

[0033] The tool invocation module is used to match the first prompt word template of the target agent with the received instructions of the master agent, and to determine and invoke the target meteorological data acquisition tool based on the obtained first template matching result;

[0034] The data acquisition module is used to match its own second prompt word template based on the received target intelligent agent instruction through the target meteorological data acquisition tool, and determine the target interface parameters based on the obtained second template matching result, so as to acquire meteorological data based on the target interface parameters and obtain the data acquisition result;

[0035] The message reply module is used to analyze the data acquisition results and the first template matching results received by the target agent to determine the corresponding target answer, and return the target answer to the main agent so that the main agent can reply to the message in the form of a dialogue using the target answer.

[0036] Thirdly, this application provides an electronic device, comprising:

[0037] Memory, used to store computer programs;

[0038] 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.

[0039] Fourthly, this application provides a computer-readable storage medium for storing a computer program, which, when executed by a processor, implements the steps of the aforementioned meteorological data analysis method based on a large model intelligent agent.

[0040] As can be seen, in this application, in the pre-defined meteorological data analysis system, when the main intelligent agent receives a meteorological message to be processed input from the client, it determines and calls a target intelligent agent from multiple extreme weather scenario intelligent agents based on the meteorological message to be processed; the main intelligent agent and the multiple extreme weather scenario intelligent agents are intelligent agents pre-constructed using a target large language model; the target intelligent agent matches its own first prompt word template based on the received instructions from the main intelligent agent, and determines and calls a 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 instructions from the target intelligent agent, 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 in the form of a dialogue using the target answer. In other words, 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 scenario agent invocation operation. The invoked target agent will then use the instructions sent by the main agent to match its own prompt word template to invoke a meteorological data acquisition tool. The invoked target meteorological data acquisition tool will then use the instructions sent by the target agent to match its own prompt word template to collect meteorological data and return the results to the target agent. After the target agent generates an answer using the data acquisition results and returns it to the main agent, the main agent will reply with the target answer in a dialogue format. This effectively overcomes the limitations of existing meteorological data applications, improves the convenience and efficiency of meteorological data acquisition, and enhances the accuracy and reliability of the answers. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0042] Figure 1 is a flowchart of a meteorological data analysis method based on a large model intelligent agent provided in this application;

[0043] Figure 2 is a schematic diagram of a specific meteorological data analysis process based on a large model intelligent agent provided in this application;

[0044] Figure 3 is a schematic diagram of the structure of a meteorological data analysis device based on a large model intelligent agent provided in this application;

[0045] Figure 4 is a structural diagram of an electronic device provided in this application. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] Currently, ChatBI technology enables users to query and analyze data through natural language dialogue, making data analysis more intuitive and user-friendly. However, due to difficulties in acquiring meteorological data, low acquisition efficiency, and constraints related to data confidentiality, its application in meteorological data is still in the initial exploratory stage and cannot be effectively utilized. Therefore, this application provides a meteorological data analysis scheme based on a large-scale intelligent agent 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 the results.

[0048] Referring to Figure 1, this embodiment of the invention discloses a meteorological data analysis method based on a large-scale intelligent agent, applied to a pre-set meteorological data analysis system. The large-scale intelligent agent is a large-scale intelligent agent constructed based on chatbot and business intelligence functions. The method includes:

[0049] Step S11: When the main agent receives the weather message to be processed from the client, the target agent is determined and invoked from multiple extreme weather scenario agents based on the weather message to be processed; the main agent and the multiple extreme weather scenario agents are agents pre-constructed using the target large language model.

[0050] Referring to Figure 2, in this embodiment, when a user inputs a weather message to be processed into the system, it is received and processed by the main intelligent agent (i.e., the main agent in Figure 2). That is, determining and invoking a target intelligent agent from multiple extreme weather scenario intelligent agents (i.e., independent extreme weather scenario agents in Figure 2) based on the weather message to be processed includes: analyzing the weather message to be processed by the main intelligent agent to determine whether to invoke an extreme weather scenario intelligent agent; when the first determination result is yes, determining the target intelligent agent from multiple extreme weather scenario intelligent agents based on the weather scenario type corresponding to the weather message to be processed, and invoking the target intelligent agent. In other words, in this embodiment, the main intelligent agent first determines whether it is necessary to invoke an extreme weather scenario intelligent agent; if it decides to invoke one, it determines and invokes the target intelligent agent.

[0051] It is important to understand that the extreme weather scenario agents include blizzard agents, severe convection agents, heavy rain agents, and cold wave agents. 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 a specific permutation and combination. The master agent is responsible for coordinating the autonomous extreme weather scenario agents, enabling them to collaborate efficiently and perform complex tasks.

[0052] Regarding the agents in the extreme weather scenarios, each 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 provided prompts. The prompt word templates for each agent include role prompts, task prompts, input parameter prompts, and question-answer pair prompts. Each agent possesses multiple data acquisition tools that interface with meteorological data platforms, which are divided into two types: data query tools and data statistical feature tools.

[0053] Taking the aforementioned rainstorm agent as an example, its prompt word template can be as follows:

[0054] Character Hint: Heavy Rain Scene

[0055] Task prompt: Accurate query and statistical analysis of rainstorm scenario data.

[0056] Input parameter prompts: time, location, ground data elements

[0057] Question and answer prompt: Basic information / statistical characteristics of ground data elements related to a rainstorm scene at a certain time and place.

[0058] Regarding the target large language model, both the main agent and the multiple independent extreme weather scenario agents use the target large language model, but each receives different prompts, thus forming a self-reflective AI (Artificial Intelligence) agent. This method of using the same large language model in multiple different roles in a cyclical manner is constructed using the Lang Graph framework. The Lang Graph framework can also be used to create multi-agent workflows. Just like in the 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.

[0059] Furthermore, it's important to understand that 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.

[0060] Step S12: The target agent matches its own first prompt word template based on the received instructions from the main agent, and determines and calls the target meteorological data acquisition tool based on the obtained first template matching result.

[0061] Referring to Figure 2, in this embodiment, after invoking the target agent, the target agent determines and invokes a meteorological data acquisition tool (i.e., the data access tool in Figure 2) based on the instructions of the master agent. That is, the process of the target agent matching its own first prompt word template based on the received master agent instructions, and determining and invoking the target meteorological data acquisition tool based on the obtained first template matching result, includes: receiving master agent instructions corresponding to the meteorological message to be processed; matching its own first prompt word template based on the master agent instructions 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 pair prompt information; determining the corresponding target meteorological data acquisition tool name by analyzing the first template matching result, and invoking the target meteorological data acquisition tool using the target meteorological data acquisition tool name. In other words, the invoked target agent matches the master agent instructions with the prompt word template to abstract the specific tool name of the data tool to be invoked. It is understood that agents in different extreme weather scenarios can possess different numbers of meteorological data acquisition tools.

[0062] It is important to understand that, regarding the design of the meteorological data acquisition tools, the data acquisition tools for specific extreme weather scenarios connect the corresponding regional meteorological big data cloud platform with the extreme weather scenario intelligent agent through a data service interface, and perform data acquisition and analysis according to business scenario requirements. Each meteorological data acquisition tool has its own data tool name, data tool parameters, data tool description prompts, and the data source and interface for connecting to the meteorological big data cloud platform. The meteorological data acquisition tools extract fields from the returned data according to the data tool function definition, perform analysis, statistical calculation and processing, and then output the values ​​to the user in a dialogue format. Specifically, the data tool name must be an intelligent agent data tool name that is easy for the large language model intelligent agent to call and recognize the intent; the data tool parameters must be data tool parameters that are easy for the large language model intelligent agent to extract based on dialogue-based questions; and the data tool description prompts must be description prompts that are easy for the large model intelligent agent to call and recognize the intent, thereby reducing the intelligent agent tool's misjudgment rate.

[0063] A design example of the meteorological data acquisition tool can be found in Table 1 below:

[0064] Table 1

[0065] Furthermore, in this embodiment, regarding the prompt word template, the template design can be tailored to disaster weather scenarios such as heavy rain, heavy snow, severe convection, and cold waves. It also involves organizing meteorological big data interfaces, algorithms, data sources, and data fields for meteorological data platforms. Prompt word templates and annotation standards for disaster meteorological data elements are formulated, and standardized annotation and governance of prompt words for disaster meteorological data elements are carried out to achieve 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 based on the obtained second template matching result, so as to acquire meteorological data based on the target interface parameters and obtain the data acquisition result.

[0067] In this embodiment, a designated target meteorological data acquisition tool is invoked. This tool matches the target agent's instructions with a prompt word template, abstracts the specific parameter values ​​corresponding to the data service interface, and identifies the fields required for the answer. It then connects to the meteorological big data platform via the data service interface to obtain the fields and data needed to answer the user's question. Specifically, the process of the target meteorological data acquisition tool matching its own second prompt word template based on the received target agent's instructions and determining the target interface parameters based on the second template matching result, and then acquiring meteorological data based on these parameters, includes: receiving the corresponding target agent's instructions; matching its own second prompt word template based on the target agent's instructions 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; connecting to 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 the data acquisition result. The data acquisition result includes the second target answer-related fields and 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 results and the first template matching results 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 results returned by the router and a prompt word template. That is, the step of determining the corresponding target answer by analyzing the received data acquisition results and the first template matching results by the target agent includes: receiving the data acquisition results returned by the target meteorological data acquisition tool corresponding to the target agent's instruction; and generating answer information corresponding to the meteorological message to be processed by analyzing the data acquisition results and the first template matching results to obtain the target answer.

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

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

[0072] (1) It fills the technological gap in the application of ChatBI technology in the meteorological field;

[0073] (2) It can more accurately understand the user's natural language commands and convert them into a data query language that is easy to process. It also supports access to multiple data sources, provides diverse data analysis and visualization tools, meets the user's multidimensional needs for meteorological data, and makes the meteorological data analysis process more intuitive and efficient.

[0074] (3) Supports real-time data access, allowing users to query the latest meteorological data at any time and make timely decisions, thereby making meteorological data analysis more flexible and accurate;

[0075] (4) Allows users to query and analyze data through dialogue, improving the convenience and efficiency of meteorological data acquisition. It can not only answer questions quickly, 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 ensured through the private deployment of large models and data access control, so as to ensure data security and privacy protection.

[0077] (6) Through dialogue, intelligent data processing and analysis can be achieved, reducing manual intervention and improving the efficiency and accuracy of data analysis.

[0078] Therefore, in this embodiment of the application, in the preset meteorological data analysis system, when the main intelligent agent receives a meteorological message to be processed input from the client, it determines and calls a target intelligent agent from multiple extreme weather scenario intelligent agents based on the meteorological message to be processed; the main intelligent agent and the multiple extreme weather scenario intelligent agents are intelligent agents pre-constructed using a target large language model; the target intelligent agent matches its own first prompt word template based on the received instructions from the main intelligent agent, and determines and calls a 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 instructions from the target intelligent agent, 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 in the form of a dialogue using the target answer. In other words, 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 scenario agent invocation operation. The invoked target agent will then use the instructions sent by the main agent to match its own prompt word template to invoke a meteorological data acquisition tool. The invoked target meteorological data acquisition tool will then use the instructions sent by the target agent to match its own prompt word template to collect meteorological data and return the results to the target agent. After the target agent generates an answer using the data acquisition results and returns it to the main agent, the main agent will reply with the target answer in a dialogue format. This effectively overcomes the limitations of existing meteorological data applications, improves the convenience and efficiency of meteorological data acquisition, and enhances the accuracy and reliability of the answers.

[0079] Referring to Figure 3, this application also discloses a meteorological data analysis device based on a large-scale intelligent agent, applied to a preset meteorological data analysis system. The large-scale intelligent agent is a large-scale intelligent agent built based on chatbot and business intelligence functions. The device includes:

[0080] The agent invocation module 11 is used to determine and invoke a target agent from multiple extreme weather scenario agents based on the weather message to be processed when the main agent receives the weather message to be processed input from the client; the main agent and the multiple extreme weather scenario agents are agents pre-constructed using the target large language model.

[0081] The tool invocation module 12 is used to match the first prompt word template of the target intelligent agent based on the received instructions of the main intelligent agent, and to determine and invoke the target meteorological data acquisition tool according to the obtained first template matching result;

[0082] The data acquisition module 13 is used to match its own second prompt word template based on the received target intelligent agent instruction through the target meteorological data acquisition tool, and determine the target interface parameters according to the obtained second template matching result, so as to acquire meteorological data based on the target interface parameters and obtain the data acquisition result.

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

[0084] For more detailed information on the working process of each of the above modules, please refer to the relevant content disclosed in the foregoing embodiments, which will not be repeated here.

[0085] Therefore, 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 scenario agent invocation operation. The invoked target agent will then use the instructions sent by the main agent to match its own prompt word template to invoke the meteorological data acquisition tool. The invoked target meteorological data acquisition tool will then use the instructions sent by the target agent to match its own prompt word template to collect meteorological data and return the results to the target agent. After the target agent generates an answer using the data acquisition results and returns it to the main agent, the main agent will reply with the target answer in a dialogue format. This effectively overcomes the limitations of existing meteorological data applications, improves the convenience and efficiency of meteorological data acquisition, and enhances the accuracy and reliability of the answers.

[0086] In some specific embodiments, the agent invocation module 11 can be used to: analyze the weather message to be processed by the main agent to determine whether to invoke the extreme weather scenario agent; when the first determination result is yes, determine the target agent from multiple extreme weather scenario agents based on the weather scenario type corresponding to the weather message to be processed, and invoke the target agent.

[0087] In some specific embodiments, the tool invocation module 12 can be used to: receive a master agent instruction corresponding to the meteorological message to be processed through the target agent; match its own first prompt word template based on the master 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-answer pair prompt information; determine the corresponding target meteorological data acquisition tool name by analyzing the first template matching result, and invoke the target meteorological data acquisition tool using the target meteorological data acquisition tool name.

[0088] In some specific embodiments, the data acquisition module 13 can be used to: receive corresponding target agent instructions through the target meteorological data acquisition tool; match its own second prompt word template based on the target agent instructions to obtain a second template matching result; determine target interface parameters and 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 used to: receive the data acquisition result returned by the target meteorological data acquisition tool corresponding to the instruction of the target intelligent agent through the target intelligent agent; and 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 used to: determine whether the target answer meets the preset conditions through the main intelligent agent to obtain a second judgment result; if the second judgment result indicates that the preset conditions are met, then reply to the weather message to be processed based on the target answer and in the form of a dialogue.

[0091] In some specific embodiments, the meteorological data analysis device based on a large model intelligent agent can also be used to: if the second judgment result indicates that the preset conditions are not met, regenerate 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 back 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, this application also discloses an electronic device. FIG4 is a structural diagram of an electronic device 2 / 30 according to an exemplary embodiment. The content in the figure should not be considered as any limitation on the scope of use of this application.

[0093] Figure 4 is a schematic diagram of the structure of an electronic device 2 / 30 provided in an embodiment of this application. Specifically, the electronic device 2 / 30 may 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 stores 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 intelligent agent disclosed in any of the foregoing embodiments. Furthermore, the electronic device 2 / 30 in this embodiment may specifically be a 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 external devices, and the communication protocol it follows can be any communication protocol applicable to the technical solution of this application, and is not specifically limited here; the input / output interface 2 / 35 is used to acquire external input data or output data to the outside world, and 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 optical disk, etc. The resources stored thereon can include operating system 2 / 321, 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 various hardware devices on the electronic device 2 / 30 and the computer program 2 / 322, which may be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program capable of performing the meteorological data analysis method based on a large model intelligent agent, which is executed by the electronic device 2 / 30 as disclosed in any of the foregoing embodiments, the computer program 2 / 322 may further include computer programs capable of performing other specific tasks.

[0097] Furthermore, this application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, it implements the aforementioned meteorological data analysis method based on a large model intelligent agent. Specific steps of this method can be found in the corresponding content disclosed in the foregoing 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 other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

[0099] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can 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 can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, 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" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0102] The technical solutions provided in this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A meteorological data analysis method based on a large-scale intelligent agent model, characterized in that, The method is applied to a pre-defined meteorological data analysis system, wherein the large-scale intelligent agent is a large-scale intelligent agent built based on chatbot and business intelligence functions; wherein the method includes: When the main agent receives a weather message to be processed from the client, it determines and calls the target agent from multiple extreme weather scenario agents based on the weather message to be processed; the main agent and the multiple extreme weather scenario agents are agents that are pre-constructed using the target large language model; The target agent matches its own first prompt word template based on the received instructions from the main agent, 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 and obtain the data acquisition result. The target 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 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-scale intelligent agent according to claim 1, characterized in that, The step of determining and invoking the target agent from multiple extreme weather scenario agents based on the meteorological message to be processed includes: The main intelligent agent analyzes the meteorological message to be processed to determine whether to invoke the intelligent agent for extreme weather scenarios. When the first judgment result is yes, the target agent is determined from multiple extreme weather scenario agents based on the weather scenario type corresponding to the meteorological message to be processed, and the target agent is invoked.

3. The meteorological data analysis method based on a large-scale intelligent agent according to claim 1, characterized in that, The step of matching the target agent's first prompt word template with the received instructions from the master agent, and determining and invoking the target meteorological data acquisition tool based on the obtained first template matching result, includes: The target agent receives instructions from the main agent corresponding to the meteorological message to be processed. Based on the main intelligent agent's instructions, it matches its own first prompt word template 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 pair prompt information; The 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-scale intelligent agent according to claim 1, characterized in that, The step of using the target meteorological data acquisition tool to match its own second prompt word template based on the received target intelligent agent command, and determining the target interface parameters based on the obtained second template matching result, to acquire meteorological data based on the target interface parameters, includes: The target meteorological data acquisition tool receives the corresponding target intelligent agent commands. The target agent matches its own second prompt word template based on the target agent's instructions to obtain the second template matching result; The target interface parameters and the fields related to the first target answer are determined by analyzing the matching results of the second template. Based on the target interface parameters, access the corresponding data service interface, 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 the second target answer related fields and answer related data.

5. The meteorological data analysis method based on a large-scale intelligent agent according to claim 1, characterized in that, The step of determining the corresponding target answer by analyzing the received data acquisition results and the first template matching results through the target agent includes: The target agent receives the data acquisition result returned by the target meteorological data acquisition tool, which corresponds to the command of the target agent; By analyzing the data acquisition results and the first template matching results, answer information corresponding to the weather message to be processed is generated to obtain the target answer.

6. The meteorological data analysis method based on a large-scale intelligent agent according to claim 1, characterized in that, The main intelligent agent responds to messages using the target answer in a dialogue format, including: The main intelligent agent determines that 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, then the weather message to be processed is replied to based on the target answer in a dialogue format.

7. The meteorological data analysis method based on a large-scale intelligent agent according to claim 6, characterized in that, After obtaining the second judgment result, the process also includes: If the second judgment result does not meet the preset conditions, then the instruction information corresponding to the weather message to be processed is regenerated, and the new main agent instruction is sent to the target agent to jump back to the step of matching the first prompt word template of itself by the target agent based on the received main agent instruction.

8. A meteorological data analysis device based on a large-scale intelligent agent, characterized in that, The device is applied to a pre-defined meteorological data analysis system, wherein the large-scale intelligent agent is a large-scale intelligent agent built based on chatbot and business intelligence functions; wherein, the device includes: The agent invocation module is used to determine and invoke a target agent from multiple extreme weather scenario agents based on the unprocessed meteorological message input by the client when the main agent receives the unprocessed meteorological message; the main agent and the multiple extreme weather scenario agents are agents pre-constructed using the target large language model; The tool invocation module is used to match the first prompt word template of the target agent with the received instructions of the master agent, and to determine and invoke the target meteorological data acquisition tool based on the obtained first template matching result; The data acquisition module is used to match its own second prompt word template based on the received target intelligent agent instruction through the target meteorological data acquisition tool, and determine the target interface parameters based on the obtained second template matching result, so as to acquire meteorological data based on the target interface parameters and obtain the data acquisition result; The message reply module is used to analyze the data acquisition results and the first template matching results received by the target agent to determine the corresponding target answer, and return the target answer to the main agent so that the main agent can reply to the message in the form of a dialogue using the target answer.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program to implement the meteorological data analysis method based on a large model intelligent agent as described in any one of claims 1 to 7.

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

Images