摘要
建立了混合关联任务集描述模型,并引入人工智能领域的启发式搜索算法,以实现该模型任务集的最佳优先调度。在此基础上,进一步提出了基于回卷恢复机制的容错调度策略,通过对瞬时故障的容错恢复,提高数控系统运行的可靠性。与传统实时容错调度算法相比,基于回卷恢复机制和最佳优先调度算法的容错策略,能很好地适用于混合关联任务集。同时,通过降低错误恢复所需的时间开销,有效地提高了系统的资源利用率。实例分析和算法模拟研究证明了该算法应用于数控系统的有效性。
To guarantee the reliability of Numerical Control （NC） system, a model for hybrid dependent task set was established and the heuristic Best-First （BF） algorithm was introduced to realize optimized scheduling of this model. Then, a fault-tolerant scheduling strategy based on rollback recovery and BF algorithm was proposed to remove transient fault of system. Compared with traditional fault tolerate scheduling algorithm, the proposed strategy could be applied to scheduling in mixed and dependent tasks of NC system, and raise the level of resources utilization by reducing the cost of fault disposal. Validity of proposed approach applied to NC system was verified by case study and algorithm simulation.
出处
《计算机集成制造系统》
EI
CSCD
北大核心
2007年第4期768-776,共9页
Computer Integrated Manufacturing Systems
基金
国家863/CIMS主题资助项目(2002AA421150)~~
关键词
数控系统
混合任务系统
启发式搜索
容错调度
numerical control system
hybrid tasks system
heuristic search
fault-tolerant scheduling
