基于分区思维分类下的单片机节能调度模型仿真

发布时间：2018-01-09 18:34

本文关键词：基于分区思维分类下的单片机节能调度模型仿真　出处：《科技通报》2015年01期 　论文类型：期刊论文

【摘要】：为了降低单片机系统调度过程的总能耗,提出一种分区思维分类下融合旅行商算法以及遗传算法的节能调度方法。基于分区思维分类方法,将单片机进程调度总能耗,划分成进程切换能耗、进程过渡调整能耗以及进程稳定调度能耗,将单片机进程节能调度问题,转化成单片机多进程调度的能耗优化问题,将单进程调度平稳模态作为节点、进程调度过渡模态作为支路,构建单片机进程调度总能耗的有向图模型,将单片机进程能耗优化过程看成旅行商问题,通过遗传算法的多目标逐层改进以及旅行商算法路径寻优原理,查询不同单片机进程调度下的最佳调度参数和多进程的最佳调度顺序,降低单片机进程调度的总能耗。实验结果说明,所提模型可提高单片机进程调度效率,降低调度能耗。
[Abstract]:In order to reduce the total energy consumption in the scheduling process of single-chip computer system, a kind of energy-saving scheduling method based on zonal thinking classification is proposed, which combines the traveling salesman algorithm and genetic algorithm under the zonal thinking classification. The total energy consumption of single-chip microcomputer process scheduling is divided into process switching energy consumption, process transition adjustment energy consumption and process stability scheduling energy consumption. The energy consumption optimization problem of single-chip microcomputer multi-process scheduling is transformed into the single-process scheduling steady mode as the node and the process scheduling transition mode as the branch to construct the single-chip microcomputer process scheduling total energy consumption directed graph model. The process energy consumption optimization process of single chip microcomputer is regarded as a traveling salesman problem, which is improved by the multi-objective and layer by layer genetic algorithm and the principle of path optimization of the traveling salesman algorithm. In order to reduce the total energy consumption of single-chip processor process scheduling, the optimal scheduling parameters and the optimal scheduling order of multi-process are queried. The experimental results show that the proposed model can improve the efficiency of single-chip microcomputer process scheduling. Reduce scheduling energy consumption.
【作者单位】：辽宁机电职业技术学院自控系;
【分类号】：TP368.1
【正文快照】： 当前随着计算机技术的发展,对嵌入式单片机系统的高性能和可靠性提出了更高的要求[1,2]。构建多个处理模块,实现进程的实时、高效和节能调度,成为当前相关人员分析的热点问题[3-5]。为了降低单片机系统调度过程的总能耗,本文提出了基于分区思维分类下的融合旅行商算法以及遗传

【参考文献】