Intelligent timing task execution method based on large language model

Abstract

The application particularly relates to an intelligent timing task execution method based on a large language model. In the intelligent timing task execution method based on the large language model, a user configures timing task information in a front-end interface and submits the timing task information to a back end, and the back end stores task data in a database storage module; when the system is started or a task is updated, a scheduling module reads task records and registers the task records in a timing scheduler; after a trigger time is reached, task description, prompt words and context are transmitted to a large language model for reasoning, and a task execution plan is output; after task execution is completed, task state and result data are analyzed, and a database storage module record is updated, and a task execution result is pushed to an external system. The intelligent timing task execution method based on the large language model realizes a change from traditional rule driving to intelligent decision driving, significantly reduces manual configuration cost, and improves the intelligentization, autonomy and execution efficiency of a task scheduling system.

Description

Technical Field

[0001] This invention relates to the field of database storage module operation and maintenance technology, and in particular to an intelligent scheduled task execution method based on a large language model. Background Technology

[0002] With the development of information systems and automated operation and maintenance, scheduled task technology has become a fundamental function in server management, data processing, and business process control. Existing scheduled task systems, such as those based on Cron expressions, the Quartz scheduling framework, or cloud platform task schedulers, primarily rely on fixed-time rules or simple task dependency configurations to execute periodic tasks. However, existing technologies have the following shortcomings: 1) Lack of intelligence and adaptability Traditional scheduled task execution mechanisms are typically based on static rules and cannot automatically adjust execution strategies according to system resource load, task execution history, or business priorities. For example, when the system is under high load, tasks may still be forced to execute, leading to performance degradation or task failure.

[0003] 2) Limited exception handling capabilities When scheduled tasks fail due to external dependencies, insufficient resources, or script errors, existing systems typically only log the failure or perform a fixed number of retries, failing to perform root cause analysis or generate targeted repair solutions.

[0004] 3) The task expression method is too simple. Traditional scheduling systems rely on complex rule configuration files or programming interfaces, requiring operators to have a strong technical background. They lack natural language-level interaction methods, making it difficult to achieve "semantic-level task definition and understanding".

[0005] 4) Insufficient task optimization and learning abilities. Existing systems cannot perform intelligent analysis and optimization based on task execution history, lack self-learning mechanisms, and cannot proactively discover patterns or automatically improve the efficiency and reliability of scheduling strategies.

[0006] With the rapid improvement of the capabilities of Large Language Models (LLMs) in natural language understanding, contextual reasoning, and adaptive optimization, introducing them into task scheduling systems is expected to enable functions such as task semantic understanding, dynamic scheduling optimization, and anomaly self-diagnosis and self-repair, thereby significantly improving the intelligence and automation level of scheduled task execution.

[0007] Therefore, there is an urgent need for an intelligent timed task execution method and system based on a large language model to solve the problems of insufficient intelligence, poor adaptability, and inadequate exception handling capabilities in existing technologies.

[0008] Based on the above, this invention proposes an intelligent timed task execution method based on a large language model. Summary of the Invention

[0009] To overcome the shortcomings of existing technologies, this invention provides a simple and efficient intelligent timed task execution method based on a large language model.

[0010] This invention is achieved through the following technical solution: A method for intelligent scheduled task execution based on a large language model includes the following steps: Step S1, Task Configuration and Storage Users configure scheduled task information in the front-end interface, the front-end submits the scheduled task information to the back-end, and the back-end stores the task data in the database storage module. In step S1, the scheduled task information includes Cron parameters, task description, prompt word, task execution status, task priority, and Webhook address.

[0011] The task description and prompts are in natural language, and the system generates structured task configuration data through semantic parsing.

[0012] Step S2, Task Loading and Registration When the system starts up or a task is updated, the scheduling module reads the task record from the database storage module, registers the task to the timer scheduler, and generates a task triggering plan. Step S3: Task Triggering and Large Language Model Invocation When the task reaches its trigger time, the scheduling module automatically calls the LangGraph interface to pass the task description, prompts, and context to the large language model for inference. In step S3, natural language task definition is supported, and users configure tasks through dialogue.

[0013] Step S4: Task plan generation and execution The large language model understands the semantics of the task and outputs a task execution plan; The LangGraph interface calls the MCP (Model Context Protocol) tool module to perform the task steps of generating large language models, including calling external interfaces, executing scripts, or triggering system actions.

[0014] In step S4, an automatic retry mechanism for task exceptions is supported. When the MCP tool module fails to execute, the large language model automatically generates repair instructions and re-executes the task. The MCP tool module supports the registration and invocation of multiple types of tools, enabling cross-system and cross-service operations.

[0015] Step S5: Result Analysis and Status Update After the task is completed, receive the AI ​​information returned by the large language model, parse the task status and result data, and update the database storage module records. Step S6: Result Push and Closed-Loop Feedback Based on the Webhook address configured for the task, the task execution results are pushed to an external system, forming an intelligent closed-loop task execution process of "configuration-execution-feedback-optimization".

[0016] An intelligent timed task execution system based on a large language model, used to implement the above method, includes: The task configuration module is responsible for receiving the scheduled task information entered by the user on the front-end interface and submitting it to the back-end, which then stores the task information in the database storage module. The database storage module is responsible for storing basic information and execution status data of tasks, supporting tasks to add, delete, modify, and query operations and updating task status, and ensuring persistent management of task configuration and execution information. The scheduling module is responsible for loading task configurations from the database storage module and registering scheduled tasks; generating task plans based on the task's Cron parameters and triggering task execution at regular intervals; parsing the AI ​​information returned by the large language model and extracting task execution results and status. At the same time, based on the Webhook address configured for the task, the execution result is pushed to an external system or service as structured data to achieve closed-loop feedback of task results; The LangGraph agent module is responsible for passing the task description, prompts, and context information to the large language model through the LangGraph interface when a task is scheduled and triggered; understanding, reasoning, and planning the task semantics through the large language model to generate a structured task execution plan; and calling the corresponding MCP tool module according to the plan generated by the large language model. If an execution exception or error is detected, a retry or repair plan is automatically generated based on the error information returned by the large language model.

[0017] A smart timed task execution device based on a large language model includes a memory and a processor; the memory is used to store a computer program, and the processor is used to execute the computer program to implement the above-described method steps.

[0018] A readable storage medium storing a computer program that, when executed by a processor, implements the above-described method steps.

[0019] The beneficial effects of this invention are: the intelligent scheduled task execution method based on a large language model, by introducing a large language model into the scheduled task execution process, enables the system to have the ability of task semantic understanding, dynamic planning and tool invocation, and realizes the transformation of scheduled tasks from traditional rule-driven to intelligent decision-driven. It can automatically generate execution plans based on task descriptions and complete the actual operations through the MCP tool module, which significantly reduces the cost of manual configuration and improves the intelligence, autonomy and execution efficiency of the task scheduling system. Attached Figure Description

[0020] 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 some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Appendix Figure 1 This is a schematic diagram of the intelligent timed task execution method based on a large language model according to the present invention.

[0022] Appendix Figure 2 This is a schematic diagram of the intelligent timed task execution system based on a large language model according to the present invention. Detailed Implementation

[0023] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions in the embodiments of this invention will be clearly and completely described below in conjunction with the embodiments of this invention. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.

[0024] This intelligent scheduled task execution method based on a large language model includes the following steps: Step S1, Task Configuration and Storage Users configure scheduled task information in the front-end interface, the front-end submits the scheduled task information to the back-end, and the back-end stores the task data in the database storage module. In step S1, the scheduled task information includes Cron parameters, task description, prompt word, task execution status, task priority, and Webhook address.

[0025] The task description and prompts are in natural language, and the system generates structured task configuration data through semantic parsing.

[0026] Step S2, Task Loading and Registration When the system starts up or a task is updated, the scheduling module reads the task record from the database storage module, registers the task to the timer scheduler, and generates a task triggering plan. Step S3: Task Triggering and Large Language Model Invocation When the task reaches its trigger time, the scheduling module automatically calls the LangGraph interface to pass the task description, prompts, and context to the large language model for inference. In step S3, natural language task definition is supported, and users configure tasks through dialogue.

[0027] Step S4: Task plan generation and execution The large language model understands the semantics of the task and outputs a task execution plan; The LangGraph interface calls the MCP (Model Context Protocol) tool module to perform the task steps of generating large language models, including calling external interfaces, executing scripts, or triggering system actions.

[0028] In step S4, an automatic retry mechanism for task exceptions is supported. When the MCP tool module fails to execute, the large language model automatically generates repair instructions and re-executes the task. The MCP tool module supports the registration and invocation of multiple types of tools, enabling cross-system and cross-service operations.

[0029] Step S5: Result Analysis and Status Update After the task is completed, receive the AI ​​information returned by the large language model, parse the task status and result data, and update the database storage module records. Step S6: Result Push and Closed-Loop Feedback Based on the Webhook address configured for the task, the task execution results are pushed to an external system, forming an intelligent closed-loop task execution process of "configuration-execution-feedback-optimization".

[0030] This intelligent timed task execution system based on a large language model is used to implement the above method, including: The task configuration module is responsible for receiving the scheduled task information entered by the user on the front-end interface and submitting it to the back-end, which then stores the task information in the database storage module. The database storage module is responsible for storing basic information and execution status data of tasks, supporting tasks to add, delete, modify, and query operations and updating task status, and ensuring persistent management of task configuration and execution information. The scheduling module is responsible for loading task configurations from the database storage module and registering scheduled tasks; generating task plans based on the task's Cron parameters and triggering task execution at regular intervals; parsing the AI ​​information returned by the large language model and extracting task execution results and status. At the same time, based on the Webhook address configured for the task, the execution result is pushed to an external system or service as structured data to achieve closed-loop feedback of task results; The LangGraph agent module is responsible for passing the task description, prompts, and context information to the large language model through the LangGraph interface when a task is scheduled and triggered; understanding, reasoning, and planning the task semantics through the large language model to generate a structured task execution plan; and calling the corresponding MCP (ModelContext Protocol) tool module according to the plan generated by the large language model. If an execution exception or error is detected, a retry or repair plan is automatically generated based on the error information returned by the large language model.

[0031] The intelligent timed task execution device based on a large language model includes a memory and a processor; the memory is used to store computer programs, and the processor is used to execute the computer programs to implement the above-described method steps.

[0032] The readable storage medium stores a computer program that, when executed by a processor, implements the above-described method steps.

[0033] Compared with existing technologies, this intelligent timed task execution method based on a large language model has the following characteristics: 1) Intelligent task configuration Users can configure scheduled tasks via natural language. The system uses a large language model to automatically parse and generate structured task parameters, reducing the difficulty of task creation and maintenance.

[0034] 2) Intelligent scheduling and execution By embedding a large language model into the task scheduling process, the system can automatically generate an execution plan and dynamically adjust the execution strategy after a task is triggered, thereby improving the flexibility and intelligence of task execution.

[0035] 3) Collaboration between intelligent agents and tools The system is based on the collaborative work of LangGraph agents and MCP tool modules, realizing a fully automated closed loop from task understanding and plan generation to actual execution.

[0036] 4) Real-time feedback of results The system automatically updates the database storage module status after the task is completed and pushes the execution results through the Webhook address, realizing real-time monitoring of task execution and linkage with external systems.

[0037] 5) Possesses self-recovery and self-optimization capabilities. When a task fails, the large language model can generate a repair plan and automatically retry; the system can also continuously optimize prompts and strategies based on historical execution results to improve overall performance and stability.

[0038] The embodiments described above are merely one specific implementation of the present invention. Ordinary changes and substitutions made by those skilled in the art within the scope of the technical solution of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent timed task execution based on a large language model, characterized in that: Includes the following steps: Step S1, Task Configuration and Storage Users configure scheduled task information in the front-end interface, the front-end submits the scheduled task information to the back-end, and the back-end stores the task data in the database storage module. Step S2, Task Loading and Registration When the system starts up or a task is updated, the scheduling module reads the task record from the database storage module, registers the task to the timer scheduler, and generates a task triggering plan. Step S3: Task Triggering and Large Language Model Invocation When the task reaches its trigger time, the scheduling module automatically calls the LangGraph interface to pass the task description, prompts, and context to the large language model for inference. Step S4: Task plan generation and execution The large language model understands the semantics of the task and outputs a task execution plan; The LangGraph interface calls the MCP tool module to perform the task steps of generating large language models, including calling external interfaces, executing scripts, or triggering system actions. Step S5: Result Analysis and Status Update After the task is completed, receive the AI ​​information returned by the large language model, parse the task status and result data, and update the database storage module records. Step S6: Result Push and Closed-Loop Feedback Based on the Webhook address configured for the task, the task execution results are pushed to an external system, forming an intelligent closed-loop task execution process of "configuration-execution-feedback-optimization".

2. The intelligent timed task execution method based on a large language model according to claim 1, characterized in that: In step S1, the scheduled task information includes Cron parameters, task description, prompt words, task execution status, task priority, and Webhook address. The task description and prompts are in natural language, and the system generates structured task configuration data through semantic parsing.

3. The intelligent timed task execution method based on a large language model according to claim 1, characterized in that: In step S3, natural language task definition is supported, and users configure tasks through dialogue.

4. The intelligent timed task execution method based on a large language model according to claim 1, characterized in that: In step S4, an automatic retry mechanism for task exceptions is supported. When the MCP tool module fails to execute, the large language model automatically generates repair instructions and re-executes the task.

5. The intelligent timed task execution method based on a large language model according to claim 4, characterized in that: The MCP tool module supports the registration and invocation of multiple types of tools, enabling cross-system and cross-service operations.

6. An intelligent timed task execution system based on a large language model, characterized in that: To implement the method according to any one of claims 1 to 5, comprising: The task configuration module is responsible for receiving the scheduled task information entered by the user on the front-end interface and submitting it to the back-end, which then stores the task information in the database storage module. The database storage module is responsible for storing basic information and execution status data of tasks, supporting tasks to add, delete, modify, and query operations and updating task status, and ensuring persistent management of task configuration and execution information. The scheduling module is responsible for loading task configurations from the database storage module and registering scheduled tasks; generating task plans based on the task's Cron parameters and triggering task execution at regular intervals; parsing the AI ​​information returned by the large language model and extracting task execution results and status. At the same time, based on the Webhook address configured for the task, the execution result is pushed to an external system or service as structured data to achieve closed-loop feedback of task results; The LangGraph agent module is responsible for passing the task description, prompts, and context information to the large language model through the LangGraph interface when a task is scheduled and triggered; understanding, reasoning, and planning the task semantics through the large language model to generate a structured task execution plan; and calling the corresponding MCP tool module according to the plan generated by the large language model. If an execution exception or error is detected, a retry or repair plan is automatically generated based on the error information returned by the large language model.

7. An intelligent timed task execution device based on a large language model, characterized in that: It includes a memory and a processor; the memory is used to store a computer program, and the processor is used to execute the computer program to implement the method as described in any one of claims 1 to 5.

8. A readable storage medium, characterized in that: The readable storage medium stores a computer program that, when executed by a processor, implements the method as described in any one of claims 1 to 5.

Images