基于并行遗传算法的对称多处理器任务调度策略研究  

Research on a Parallel Genetic Algorithm for Tasks Scheduling in SMP System

作　　者：苗硕[1] 马光思[1] 

机构地区：[1]西安建筑科技大学信控学院,陕西西安710055

出　　处：《微电子学与计算机》2006年第6期181-184,共4页Microelectronics & Computer

摘　　要：任务分配问题是公认的NP难问题。文章在以往有关多处理机任务分配算法的基础上,提出了一种适用于SMP系统结构的并行遗传调度算法。仿真结果表明,该算法具有较好的效果和收敛性。In general the tasks scheduling problem is a NP-hard problem. Based on the algorithms related with multiprocessor scheduling problem, this paper presents a parallelism genetic scheduling algorithm which well-adapted to the SMP system architecture. Simulation results comparing with other two genetic algorithms, both from the Literatures, show that our algorithm produces encouraging results in terms of quality of solutions and execution speed and even has more advantages in convergence.

关 键 词：任务分配与调度 并行遗传算法 SMP 

分 类 号：TP322[自动化与计算机技术—计算机系统结构]