摘要
针对实时系统中任务调度问题,提出了一种基于时间片的抢占控制模型.该模型以抢占次数上限为特征参数,在满足任务集可调度的前提下,由该特征参数计算出任务时间片并按片内不可抢占的限制条件优化任务抢占次数.采用遗传算法对该抢占控制模型进行了离线实现,同时使用惩罚函数来保证整个任务集的可调度性.通过仿真实验,验证了该模型的有效性.
A preemptive control model based on time slice was developed for the task scheduling in real-time systems. Based on the premise that the schedulability of task set should be satisfied, the upper limit of the occurrence of preemption was taken as characteristic parameter to compute the time slice, then the occurrence of task preemption was optimized according to the restriction that any preemption is forbidden in the time slice. The genetic algorithm was introduced to enable the model to implement off-line operation and, simultaneously, the schedulability of a task set was ensured entirely via a penalty function. Simulation results verified the effectiveness of the model proposed.
出处
《东北大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2009年第11期1570-1573,共4页
Journal of Northeastern University(Natural Science)
基金
教育部高等学校科技创新工程重大项目培育基金资助项目(708026)
关键词
调度算法
时间片
抢占调度
遗传算法
硬实时嵌入式系统
scheduling algorithm
time slice
preemptive scheduling
genetic algorithm
hard real-time embedded system