成批处理工作流动态分组调度优化方法

摘要

针对现有成批处理工作流调度方法的不足，建立考虑活动实例对执行者执行能力需求等约束的动态分组调度优化模型，提出一种解决该问题的实现算法。算法主要思想是利用微粒群算法的智能优化原理，同时优化最小化活动实例的停留时间总和与执行开销总和这两个目标函数，最终产生一组满足约束条件的Pareto优化调度方案。仿真实验说明了算法的有效性。%Aiming at the shortcomings in existing scheduling methods for batch processing workflow, a dynamic grouping scheduling optimization model for considering the constraints of activity instances on executive capacity de- mand was established, and an implementation algorithm to solve this problem was presented. In this algorithm, in- telligent optimization theory of particle swarm optimization was utilized, and two objective functions such as total residence time of minimum activity instance and total execution costs were optimized, thus a set of Pareto optimal scheduling scheme was produced. The result of simulation experiment showed the effectiveness of this algorithm.