Workflow scheduling method for multi-target particle swarm optimization based on real-time state monitoring

Abstract

The invention discloses a workflow scheduling method for multi-target particle swarm optimization based on real-time state monitoring and relates to the field of cloud computing workflow scheduling. According to the method, first, workflow pre-scheduling is performed through a BHEFT algorithm, and the feasibility of the algorithm is improved; and a Pareto variance is introduced to perform real-time monitoring on an evolution state, wherein when the evolution state is in diversification, an external elite swarm self-optimization strategy is adopted to improve local search capability of the algorithm, and when the evolution state is in stagnation, an escape strategy is adopted, so that workflow scheduling solution spaces are diversified. In this way, development and exploitation of the algorithm in the evolution process are effectively balanced, and the convergence of workflow scheduling solutions and the diversity of scheduling solution space distribution are realized.

Description

technical field [0001] The invention relates to the field of cloud computing workflow scheduling, in particular to a workflow scheduling method based on real-time state monitoring and multi-objective particle swarm optimization. Background technique [0002] With the rapid development of cloud computing technology, more and more organizations migrate traditional business processes and applications to the cloud computing environment. Cloud computing is a computing model that uses the Internet to realize anytime, anywhere, on-demand, and convenient use of resources such as shared computing facilities, storage devices, and applications. The main features of cloud computing are high scalability and high reliability, which means that users can rent and release resources on demand, and providers use measures such as multi-copy data fault tolerance and homogeneous swapping of computing nodes to ensure high reliability of services. A workflow consists of a set of tasks with data de...

AI Technical Summary

Problems solved by technology

The existing multi-objective particle swarm optimization algorithm lacks a real-time state detection mechanism, and it is impossible to determine when and what strategy the algorithm adopts to a certain extent, which leads to problems of over-exploitation or incomplete development, and ultimately leads to the obtained workflow scheduling solution. The degree of convergence is not high or the diversity is not enough

Images