控制系统中实时任务的动态优化调度算法

Optimal dynamic scheduling algorithm for real-time tasks in digital control systems

提出一种新的调度算法——带有非周期服务器的EDF调度算法.分析了所有任务的可调度性,给出了可调度条件,并给出一种新的周期性任务模型以及主优先级和辅助优先级的概念.它们在保证任务可调度的前提下,对周期性任务的采样频率和控制延时进行优化.仿真结果表明,该算法可以提高周期性任务的采样频率,并降低控制延时,即能优化系统的性能.

A novel scheduling algorithm, named Earliest Deadline First (EDF) with sporadic server, is presented. The schedulability of all tasks including the periodic tasks and the sporadic tasks is investigated, and the schedulability test is given. A new periodic task model and concepts of main priority and assistant priority are presented. By using them, the sampling frequencies of periodic tasks are optimized under maintained schedulability and the control delay of periodic task is reduced. The result of simulation shows that the sampling frequencies of periodic tasks can be increased and the control delays of periodic tasks can be reduced.