基于改进的混洗蛙跳算法的异构多核处理器任务调度

发布时间：2018-05-07 01:25

本文选题：异构多核处理器 + 任务调度　；参考：《计算机应用研究》2015年12期

【摘要】：针对异构多核环境下的任务调度问题,为充分发挥异构多核平台优势,提出一种基于改进的混洗蛙跳算法的异构多核处理器任务调度算法——ISFLA算法。ISFLA算法通过建立合适的个体编码方案和位置更新方法实现个体搜索空间到离散空间的映射,通过调整子群更新方法规避在离散空间下标准算法中子群更新易于停滞在最差值的情况,使得标准SFLA算法可以应用于异构多核环境下的任务调度问题。同时算法引入了遗传算法的交叉操作,通过将全局较优解群体的信息扩散到种群,增加了算法跳出局部极优值的能力,优化了算法的全局搜索能力。最后通过实验证明,对比常见的遗传算法,ISFLA算法具有更好的求解能力,可以提供更少执行时间的任务调度方案,具有较好的应用价值。
[Abstract]:Aiming at the task scheduling problem in heterogeneous multi-core environment, in order to give full play to the advantages of heterogeneous multi-core platform, A task scheduling algorithm for heterogeneous multi-core processors based on improved shuffling leapfrog algorithm (ISFLA) is proposed. ISFLA algorithm can map individual search space to discrete space by establishing appropriate individual coding scheme and location updating method. By adjusting the subgroup updating method to avoid the problem that the subgroup update is easy to stagnate at the worst value in the discrete space, the standard SFLA algorithm can be applied to the task scheduling problem in heterogeneous multi-core environment. At the same time, the crossover operation of genetic algorithm is introduced. By diffusing the information of the global optimal solution to the population, the ability of the algorithm to jump out of the local optimal value is increased, and the global search ability of the algorithm is optimized. Finally, it is proved by experiments that compared with the common genetic algorithm (GA), the ISFLA algorithm has better solving ability and can provide a task scheduling scheme with less execution time, and has better application value.
【作者单位】：桂林理工大学信息科学与工程学院;
【基金】：国家自然科学基金资助项目(61063001,61262075) 广西高校重大科研项目(201201ZD012) 广西信息科学实验中心经费资助项目(20130206)
【分类号】：TP332;TP18

【相似文献】