基于遗传算法的网格服务工作流调度的研究

Scheduling of grid workflow for grid services based on genetic algorithm

网格服务的提出为网格工作流的研究提供了新的契机与挑战,由网格服务组成的工作流(GSF)的调度问题是一个典型的NP问题。利用遗传算法所具有的并行性和全局解空间搜索的特点,针对网格服务调度问题,提出基于遗传算法的网格服务工作流调度算法GSFGA,并改进了遗传算法的收敛特性,通过分析实验结果证明该算法优于传统的调度算法。

The service oriented grid work_flow, GSF, has been a research focus in grid technology. As an NP problem, grid service scheduling is difficult to solve by means of classic algorithms. Featured in searching concurrently and globally, genetic algorithm can be a better option for solving GSF scheduling problem. Therefore, a GA-based grid service scheduling algorithm, GSFGA, was provided for obtaining the best GSF instance with highest fitness, Experiment results prove it available and better than some traditional algorithms. As a conclusion, the further work was also pointed out.