摘要
为了充分利用网格的大规模计算能力,提高其计算效率,提出了一种改进的遗传算法来解决网格任务调度问题。由于任务之间具有依赖关系,将任务按高度值进行划分,高度值小的任务优先进行处理,从而可以提高种群的初始质量,减少遗传算法的执行时间。实验结果表明,此算法提高了种群的初始质量,获得了较优的调度效果。
In order to take full advantage of the massive computing power and improve computational efficiency of grid, an improved genetic algorithm is proposed to solve the grid task scheduling problem. Due to dependencies between tasks, tasks are divided by height and task with height small is priority processed, which can improve the quality of initial population and reduce the execution time of genetic algorithm. Experimental results show that the improved algorithm improves the quality of the initial population and reduce the execution time of the task that is in a shorter period of time to get a better effect.
出处
《信息通信》
2016年第3期56-58,共3页
Information & Communications
关键词
网格
任务调度
遗传算法
初始种群
Grid
task scheduling
Genetic Algorithm（GA）
initial population
