CN-121996431-A - Resource allocation updating method and device, storage medium and electronic equipment

CN121996431ACN 121996431 ACN121996431 ACN 121996431ACN-121996431-A

Abstract

The application discloses a resource allocation updating method, a device, a storage medium and electronic equipment, and relates to the technical field of education, wherein the method comprises the steps of obtaining historical task execution data of an agent to be updated; the method comprises the steps of comprehensively evaluating historical task execution conditions of an agent to be updated according to historical task execution data to obtain execution capacity data representing task execution capacity of the agent to be updated, determining a grade update type of the agent to be updated, updating resource configuration grades of the agent to be updated according to the grade update type to obtain target resource configuration grades corresponding to the agent to be updated, carrying out resource configuration on the agent to be updated according to the target resource configuration grades, and determining target resource configuration information corresponding to the agent to be updated. The application updates the resource configuration of the intelligent agent by determining the class update type and updating the resource configuration class, realizes reasonable distribution of the computing power resource and ensures the service processing capability of the intelligent agent.

Inventors

Hua Chencheng
ZHANG JUNKAI
Xiang Linggang
YOU JUNJIE

Assignees

杭州海亮数字科技有限公司

Dates

Publication Date: 20260508
Application Date: 20260409

Claims (10)

1. A method for updating a resource allocation, comprising: acquiring historical task execution data of an agent to be updated, wherein the historical task execution data comprises agent operation data of the historical task executed by the agent to be updated and evaluation data of the historical task execution condition of the agent to be updated by a user initiating the historical task; Comprehensively evaluating the historical task execution condition of the to-be-updated intelligent agent according to the historical task execution data to obtain execution capacity data representing the task execution capacity of the to-be-updated intelligent agent; Determining a grade update type of the to-be-updated intelligent agent based on the execution capacity data, and updating the resource configuration grade of the to-be-updated intelligent agent according to the grade update type to obtain a target resource configuration grade corresponding to the to-be-updated intelligent agent; And carrying out resource allocation on the to-be-updated intelligent agent according to the target resource allocation grade, and determining target resource allocation information corresponding to the to-be-updated intelligent agent, wherein the target resource allocation information is used for the to-be-updated intelligent agent to execute a task to be executed.
2. The method according to claim 1, wherein the comprehensively evaluating the historical task execution condition of the agent to be updated according to the historical task execution data to obtain execution capability data representing the task execution capability of the agent to be updated includes: normalizing the resource consumption data of the to-be-updated intelligent agent according to the target resource consumption data of the to-be-updated intelligent agent to obtain normalized resource consumption data of the to-be-updated intelligent agent; respectively carrying out data processing on the performance data of the intelligent agent to be updated and the task complexity data of the historical task to obtain comprehensive performance data and comprehensive task complexity data; and carrying out fusion processing on the normalized resource consumption data, the comprehensive efficiency data, the comprehensive task complexity data and the evaluation data to obtain execution capacity data representing the task execution capacity of the agent to be updated.
3. The method according to claim 2, wherein the data processing the performance data of the agent to be updated and the task complexity data of the historical task to obtain comprehensive performance data and comprehensive task complexity data respectively includes: weighting response delay data, generation time of task reply information and success rate of the to-be-updated agent calling tool in the performance data generated by the to-be-updated agent execution history task to obtain comprehensive performance data; weighting the confidence coefficient and the complexity of each intention in the historical task to obtain basic task complexity data; And obtaining the comprehensive task complexity data based on the basic task complexity data and the context associated data of the historical task.
4. The method according to claim 2, wherein the fusing the normalized resource consumption data, the comprehensive performance data, the comprehensive task complexity data, and the evaluation data to obtain execution capacity data representing the task execution capacity of the agent to be updated includes: Correcting the comprehensive task complexity data based on a preset task difficulty correction coefficient to obtain task difficulty correction data, and processing the evaluation data and the task difficulty correction data to obtain evaluation difficulty fusion data; obtaining resource consumption constraint data according to the normalized resource consumption data, and obtaining resource constraint evaluation data according to the evaluation difficulty fusion data and the resource consumption constraint data; processing the resource constraint evaluation data and the comprehensive efficiency data to obtain last execution capacity data of the historical task executed by the to-be-updated agent; and based on a preset smoothing coefficient, carrying out fusion processing on the execution capacity data except the last execution capacity data in the last execution capacity data and the historical execution capacity data to obtain the execution capacity data representing the task execution capacity of the agent to be updated.
5. The method of claim 1, wherein the performing resource allocation on the to-be-updated agent according to the target resource allocation level, and determining target resource allocation information corresponding to the to-be-updated agent, comprises: When the target resource allocation level is higher or lower than the current resource allocation level, adjusting the resource allocation information according to the target allocation resource level to obtain the target resource allocation information; And under the condition that the target resource allocation level is equal to the current resource allocation level, determining the resource allocation information corresponding to the current resource allocation level as the target resource allocation information.
6. The method according to claim 1, wherein determining the class update type of the to-be-updated agent based on the execution capability data, and updating the resource configuration class of the to-be-updated agent according to the class update type, to obtain the target resource configuration class corresponding to the to-be-updated agent, includes: Determining that the class update type of the to-be-updated intelligent agent is a lifting class type under the condition that the execution capacity data is larger than or equal to a lifting class threshold value, and upgrading the resource configuration class of the to-be-updated intelligent agent based on the execution capacity data to obtain a first resource configuration class after the to-be-updated intelligent agent is upgraded; determining that the level update type of the agent to be updated is a maintenance level type when the execution capacity data is greater than or equal to a lower level threshold and less than an upper level threshold, and determining the resource configuration level of the agent to be updated as a second target resource configuration level based on the execution capacity data; And under the condition that the execution capacity data is smaller than a reduction level threshold, determining that the level update type of the agent to be updated is a reduction level type, and degrading the resource configuration level of the agent to be updated based on the execution capacity data to obtain a third resource configuration level after the agent to be updated is degraded.
7. The method of claim 6, wherein determining that the level update type of the agent to be updated is a reduced level type if the performance capability data is less than a reduced level threshold, and downgrading the resource configuration level of the agent to be updated based on the performance capability data to obtain a third resource configuration level of the agent to be updated after downgrading, comprises: Reducing the number of tasks and task authorities of the to-be-updated agent for executing the to-be-executed tasks according to the execution capacity data, and monitoring the execution condition of the to-be-executed tasks; In response to monitoring that the level update type of the to-be-updated agent executing the task to be executed is a lifting level type or a maintaining level type, determining the current resource configuration level of the to-be-updated agent as a target resource configuration level; And in response to the fact that the grade update type of the to-be-updated agent executing the task to be executed is the degradation type, degrading the resource configuration grade of the to-be-updated agent to obtain a third resource configuration grade after the to-be-updated agent is degraded.
8. The method of claim 1, wherein prior to the acquiring historical task execution data of the agent to be updated, the method further comprises: obtaining scoring data determined by the user based on historical task execution results, response speed and interaction experience of the agent to be updated; Acquiring interactive text information of the user and the agent to be updated in the execution process of the historical task, and carrying out semantic analysis and emotion analysis on the interactive text information to obtain emotion change data; And carrying out weighted fusion processing on the grading data and the emotion change data to obtain the grading data.
9. A resource allocation updating apparatus, comprising: The system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is configured to acquire historical task execution data of an agent to be updated, the historical task execution data comprises agent operation data of the historical task executed by the agent to be updated and evaluation data of the historical task execution condition of the agent to be updated by a user initiating the historical task; The evaluation module is configured to comprehensively evaluate the historical task execution condition of the to-be-updated intelligent agent according to the historical task execution data to obtain execution capacity data representing the task execution capacity of the to-be-updated intelligent agent; the updating module is configured to determine a grade updating type of the to-be-updated intelligent agent based on the execution capacity data, and update the resource configuration grade of the to-be-updated intelligent agent according to the grade updating type to obtain a target resource configuration grade corresponding to the to-be-updated intelligent agent; The determining module is configured to perform resource configuration on the to-be-updated agent according to the target resource configuration grade, and determine target resource configuration information corresponding to the to-be-updated agent, wherein the target resource configuration information is used for the to-be-updated agent to execute a task to be executed.
10. An electronic device comprising a storage medium, a processor and a computer program stored on the storage medium and executable on the processor, characterized in that the processor implements the method of any one of claims 1 to 8 when executing the computer program.

Description

Resource allocation updating method and device, storage medium and electronic equipment Technical Field The present application relates to the field of education technologies, and in particular, to a method and apparatus for updating resource allocation, a storage medium, and an electronic device. Background With the development of artificial intelligence technology, AI agents are widely used in educational business scenarios such as teaching assistance, educational administration, and student management. At present, an agent hosting platform configures an initial computing power resource for each AI agent in the creation stage of the AI agent, and then adjusts the resource according to parameters such as Token consumption and the like generated by each AI agent executing task. The AI agent with large Token consumption will allocate sufficient computational resources, while the AI agent with small Token consumption allocates less computational resources. However, the use frequency of the AI agent in different educational service scenes is different, and the use requirement of the AI agent in the same educational service scene can dynamically change along with service time, so that the prior art cannot adapt and adjust resources according to the difference of the use frequency and the change of the use requirement of the AI agent, and the waste of computational resources or the reduction of service level of the AI agent can be caused. Disclosure of Invention In view of this, the present application provides a method, an apparatus, a storage medium and an electronic device for updating resource allocation, which are mainly aimed at improving the technical problems that the use frequency of AI agents in different educational service scenarios is different, the use requirement of AI agents in the same educational service scenario can dynamically change along with service time, and the prior art cannot adapt and adjust resources according to the difference of the use frequency and the change of the use requirement of AI agents, which can lead to the waste of computational resources of AI agents or the reduction of service level. In a first aspect, the present application provides a method for updating resource configuration, including: acquiring historical task execution data of an agent to be updated, wherein the historical task execution data comprises agent operation data of the historical task executed by the agent to be updated and evaluation data of the historical task execution condition of the agent to be updated by a user initiating the historical task; Comprehensively evaluating the historical task execution condition of the to-be-updated intelligent agent according to the historical task execution data to obtain execution capacity data representing the task execution capacity of the to-be-updated intelligent agent; Determining a grade update type of the to-be-updated intelligent agent based on the execution capacity data, and updating the resource configuration grade of the to-be-updated intelligent agent according to the grade update type to obtain a target resource configuration grade corresponding to the to-be-updated intelligent agent; And carrying out resource allocation on the to-be-updated intelligent agent according to the target resource allocation grade, and determining target resource allocation information corresponding to the to-be-updated intelligent agent, wherein the target resource allocation information is used for the to-be-updated intelligent agent to execute a task to be executed. Optionally, the comprehensively evaluating the historical task execution situation of the to-be-updated agent according to the historical task execution data to obtain execution capacity data representing the task execution capacity of the to-be-updated agent, including: Normalizing the resource consumption data according to the target resource consumption data of the to-be-updated intelligent agent to obtain normalized resource consumption data of the to-be-updated intelligent agent; respectively carrying out data processing on the performance data of the intelligent agent to be updated and the task complexity data of the historical task to obtain comprehensive performance data and comprehensive task complexity data; and carrying out fusion processing on the normalized resource consumption data, the comprehensive efficiency data, the comprehensive task complexity data and the evaluation data to obtain execution capacity data representing the task execution capacity of the agent to be updated. Optionally, the performing comprehensive processing on the performance data of the to-be-updated agent and the task complexity data of the historical task to obtain comprehensive performance data and comprehensive task complexity data respectively includes: weighting response delay data, generation time of task reply information and success rate of the to-be-updated agent calling tool in the performance data genera