基于双优先级队列的异构分布式控制系统容错调度算法

在对分布式控制系统进行分析的基础上,给出了任务模型和处理器模型.为了调度多种实时性的任务,提出了双优先级队列调度算法,用于调度每个处理器上的任务.该算法设置2个优先级队列,其中高优先级队列用于调度实时任务,低优先级队列用于调度非实时任务,高优先级队列中的任务可抢占低优先级队列中的任务.在此基础上,采用版本复制技术使系统具有容错能力,并分析了任务的容错可调度条件.基于此,采用首次适应的启发式任务分配策略,将任务分配到各个处理器上,在确保任务容错可调度的条件下使处理器负载均衡.仿真结果表明所提出的算法是有效的.

Fault-tolerant scheduling algorithm for heterogeneous distributed control systems based on dual priorities queues

Based on the analysis of distributed control system(DCS),the task models and processor model are given.In order to schedule all kinds of tasks,a novel tasks scheduling algorithm named dual-priority queues for uniprocessor is presented.Two priority queues,one is called high priority queue to be used to schedule real-time tasks and the other is called low priority queue to be used to schedule non-real-time tasks,are set in the algorithm.The task in high priority queue can preempt the running of the task in low priority queue.Based on this,task duplication technique is adopted to make DCS have fault-tolerance.The fault-tolerant and schedulable condition is investigated.Therefore,heuristic task assignment method named first-fit is applied to assign tasks to processor.The algorithm can equilibrate the load of every processor under the fault-tolerant scheduling condition of tasks.Finally,the simulation experiments show that the algorithm is effective.