摘要
资源调度算法是网格计算研究的一个重要研究方向。文章讨论了树型网格的逻辑调度模式,提出了基于树型网格的混合遗传算法HGATG。算法利用启发式操作算子加快最优值的收敛速度,并采用随机法和唯一法来避免算法的早熟。实验结果表明HGATG算法在求解速度、成功率和求解问题的规模等方面有较好的效果。
The research on resource scheduling algorithms becomes a major direction in grid computing. A logical scheduling mode based on the tree grid is firstly discussed. Then a hybrid genetic algorithm for the tree grid is designed. Some heuristic operators are used to speed the convergence of the algorithm. The precocity is effectively prevented by the random and sole methods. The experimental results show that the algorithm performs well in the aspects of speed, the success rate and the solvable problem size.
出处
《微电子学与计算机》
CSCD
北大核心
2006年第9期123-125,128,共4页
Microelectronics & Computer
基金
粤港关键领域重点突破项目(2005A10307007)
关键词
网格计算
遗传算法
收敛
资源调度
Grid computing, Genetic algorithm, Convergence, Resource scheduling