A Multi-task Scheduling Method with Multiple Competing Agents

Abstract

The invention discloses a multi-task scheduling method with multiple competing agents, which comprises the following steps: set m agents, and agent i has a group of jobs; both machines and jobs are available at zero time, and the machines are at most Process a job; among m agents, agent 1 is a long-term critical agent with high priority, while other agents are short-term one-time agents, each agent has its own objective function to be optimized that only depends on the completion time of its own job; A problem is constructed for the above objective function, and the problem is solved to obtain a multi-task scheduling strategy with multiple competing agents. By constructing a multi-agent and multi-task scheduling model, the present invention can obtain a multi-agent multi-task scheduling schedule to fill the research gap of multi-agent multi-task scheduling and adapt to the scheduling requirements of the cloud manufacturing platform environment.

Description

technical field [0001] The invention relates to the technical field of cloud manufacturing platforms, in particular to a multi-task scheduling method with multiple competing agents. Background technique [0002] In the existing scheduling technology field, multi-agent scheduling and multi-task scheduling are usually adopted. Among them, multi-agent scheduling refers to the process of assigning services over time to perform a set of tasks (jobs) requested by two or more competing agents, while satisfying certain constraints; where each agent wishes to optimize only the The objective function of the completion time of its own task. Since the goals of agents are often conflicting, they need to negotiate on how to share limited resources to optimize their own goal functions. Extensive research has been done on a variety of multi-agent scheduling models with various constraints and performance criteria from different production environments; that is, the prior art studies the s...

AI Technical Summary

Problems solved by technology

Extensive research has been done on a variety of multi-agent scheduling models with various constraints and performance criteria from different production environments; that is, the prior art studies the scheduling problem of multi-agents, but does not address the problem of solving multi-tasking; Moreover, the existing multi-task scheduling problems only consider the case of a single agent and a single objective function

[0003] However, the above scheduling model cannot be applied to the multi-task scheduling with multiple competing agents; especially for the multi-task scheduling technology under the cloud manufacturing platform, which contains multiple competing agents, the distributed resources are encapsulated in the cloud The centralized cloud manufacturing platform in manufacturing is managed; the above-mentioned existing models that only consider multi-agent single task or only consider multi-task single agent cannot effectively solve the problem.

Images