论文部分内容阅读
分布式计算系统中的一个根本问题是任务模块在处理器上的合理分配，以使总费用最小。针对随机试探法对初始条件敏感的不足，本文利用改进的遗传算法，通过设计合理的遗传算子寻求该任务分配问题的最优解。实验结果表明，本文的方法对初始条件不敏感，对具有不同拓扑结构的一致性及非一致性任务分配问题，其平均总费用降低约２％ ，此外，在大多数情况下也能使完成费用降低。
A fundamental problem in distributed computing systems is the reasonable allocation of task modules on the processor to minimize the total cost. Aiming at the shortcomings of stochastic heuristic sensitive to the initial conditions, this paper uses an improved genetic algorithm to find the optimal solution of the task assignment problem by designing a reasonable genetic operator. The experimental results show that the proposed method is insensitive to the initial conditions and the average total cost is reduced by about 2% for the consistency and inconsistency task assignment with different topologies. Moreover, in most cases, the completion cost reduce.