Resource scheduling method and device, storage medium and electronic device

By combining population optimization algorithms that integrate local and global search with reinforcement learning strategies, resource allocation schemes are generated and optimized, solving the problem of low resource utilization in traditional resource scheduling methods and achieving more efficient and flexible resource scheduling.

CN119902891BActive Publication Date: 2026-06-09CHINA TELECOM CORP LTD TECHNOLOGY INNOVATION CENTER +1
View PDF 1 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN Β· China
Patent Type
Patents(China)
Current Assignee / Owner
CHINA TELECOM CORP LTD TECHNOLOGY INNOVATION CENTER
Filing Date
2024-12-30
Publication Date
2026-06-09

AI Technical Summary

Technical Problem

Traditional static resource scheduling methods are difficult to adapt to the dynamic changes in computing resource demand and availability in cloud computing environments, resulting in low resource utilization.

Method used

A resource scheduling method is adopted, which combines population optimization algorithms (such as the spider monkey optimization algorithm) of local search and global search with reinforcement learning strategy to generate and optimize resource allocation schemes. It explores a broader solution space through global search capability and diversity preservation mechanism, and optimizes resource allocation by utilizing the environmental feedback and policy adjustment mechanism of reinforcement learning.

Benefits of technology

It effectively avoids falling into local optima traps, improves the efficiency and effectiveness of resource scheduling, enhances the flexibility and accuracy of resource scheduling, and adapts to dynamically changing computing needs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119902891B_ABST
    Figure CN119902891B_ABST
Patent Text Reader

Abstract

This disclosure provides a resource scheduling method, apparatus, storage medium, and electronic device, relating to the field of cloud computing technology. The method includes: generating an initial population comprising one or more candidate solutions, each candidate solution representing a resource allocation scheme; performing local and global searches based on the candidate solutions of the initial population to determine candidate solutions for an iteratively updated population; using the candidate solutions of the iteratively updated population as the initial state for reinforcement learning optimization, executing a reinforcement learning strategy to obtain reinforcement learning-optimized candidate solutions; and determining a final resource allocation scheme from the candidate solutions of the iteratively updated population and the reinforcement learning-optimized candidate solutions based on the fitness of the candidate solutions. By utilizing global search capabilities and diversity preservation mechanisms to explore a broader solution space, and simultaneously utilizing the environmental feedback and policy adjustment mechanisms of reinforcement learning algorithms to optimize the resource allocation strategy, the synergistic effect of these two approaches effectively avoids falling into local optima traps, improving the efficiency and effectiveness of resource scheduling.
Need to check novelty before this filing date? Find Prior Art