应用禁忌搜索算法的动态自适应网格调度

Dynamic self-adaptive tabu search algorithm for grid scheduling

提出基于禁忌搜索算法的动态网格调度算法,设计不同邻域结构,优化作业完成时间。兼顾网格动态特性,调度过程中采用分批调度,根据调度过程中上一次的部分调度信息动态调整下一次部分调度方案,自适应地修改算法参数。最后通过GridSim仿真环境和其他算法进行比较,获得较好结果。

A method based on tabu search （TS） for grid scheduling was proposed and different neighborhood architectures were designed to minimize the makespan. Considering the dynamic characteristic of grid, batch-mode scheduling was utilized and partial scheduling information was exploited to modify the next partial scheduling strategy and adaptively adjust the algorithm parameters. Finally, the algorithm was implemented in GridSim environment. The simulation results demonstrate that the proposed algorithm is superior to some heuristic classical algorithms.