一类分布式控制系统中带有优先约束的周期性任务容错调度方法  

Fault-tolerant Scheduling Algorithm for Periodic Tasks with Precedence Constraints in a Class of Distributed Control Systems

作　　者：刘怀[1] 黄建新 

机构地区：[1]南京师范大学电气与自动化工程学院,南京210042

出　　处：《小型微型计算机系统》2016年第4期830-834,共5页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61273114)资助;教育部留学回国人员基金项目(2014[1685])资助

摘　　要：随着分布式控制系统复杂性的增加,系统出现故障的可能性增大,然而系统中的任务在任何情况下都必须在其时限前完成,因此需要为分布式控制系统提供容错能力.本文针对回路任务之间具有优先约束的分布式控制系统,先给出了系统模型、任务模型和容错模型.然后基于版本复制技术设计了可同时容忍软件故障和硬件故障容错的容错调度算法.算法通过设计一个调度序列来满足回路任务之间存在的优先约束关系和主副版本之间串行执行需求.考虑到系统的采样周期与系统的控制性能有关,采用改进的粒子群算法对调度算法进行了优化.仿真实验表明本算法是有效的.With the increase of the complexity of the distributed control systems（ DCS）,they are subject to hardware failures and software failures.However,loop tasks in DCS must be finished before their deadlines in any case.Therefore,it is necessary to offer faulttolerance to DCSS.For DCSs,the system model,task models and fault-tolerant model are presented in these paper firstly.Then a faulttolerant scheduling algorithm,which can tolerant hardware failures and software failures simultaneously,is designed based on task duplication technique.The precedence constraints among loop tasks and among the primary copy and backup copy of each loop task are met by designing a scheduling sequence.Because of the relationship between the sampling period and the control performance,an improved particle swarm optimization algorithm is employed to optimize scheduling sequences.Results of computer simulation showthat the algorithm presented in this paper can tolerate hardware failures and software failures and obtain optimal scheduling sequence in order to acquire the optimal sampling period and the minimal processors that is necessary for the system.

关 键 词：分布式控制系统 容错调度 调度序列 粒子群优化 

分 类 号：TP316[自动化与计算机技术—计算机软件与理论]