欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
针对信息物理融合系统典型的分布式异构并行计算结构存在的任务调度问题,以有向无环图为调度模型,最小化任务完成时间为目标,采用多变异位自适应遗传算法,通过依赖矩阵、改进的交叉和变异算子确保基因个体的有效性,同时采用多变异位和自适应的方法保证基因个体的多样性和算法的收敛性。仿真结果显示,该算法比基于任务表的启发式调度算法更有效。
英文摘要：
To solve the task scheduling problem of Cyber-Physical Systems （ CPSs）, which are typically distributed heterogeneous parallel computing architecture, the Directed Acyclic Graph （DAG） is adopted as the scheduling model, the minimum task implementation time is taken as the object, and multi-mutation adaptive genetic algorithm is used. The effectiveness of individual gene is ensured though dependency matrix, and improved crossover and mutation operators, while the diversity of individual gene and convergence of the algorithm are ensured by using multi-mutation and adaptive methods. Simulation results show that the algorithm is more efficient than list-scheduling algorithm.
同期刊论文项目
同项目期刊论文
期刊信息
位置：