制造网格工作流任务调度技术

The technology of task of the manufacturing grid workflow

根据制造网格资源的独立性、离散性等特点,设计了一种改进的微粒群算法。优化算法是基于原微粒群算法具有记忆,根据速度来决定搜索特性的特点,并考虑了资源的各种不同性质,对速度搜索方式加以的进化修改。优化后算法具有随机匹配资源,动态查找资源的特性,并将原算法单粒子多元素同方向同速度飞行,修改成单粒子多元素多方向多速度飞行,搜索范围明显大于其他算法。在与其他算法比较、实验分析后,得到比较满意的结果,更具可用性。

Designed an improved particle swarm algorithm,according to the Manufacturing Grid resources＇s independence,discrete and so on.The algorithm is based on the original PSO algorithm has a memory and the basis on determine by searching for properties depend on the speed,and taking into account the different nature of the resources and,changing the searching methods of speed approach.This optimized algorithm has some features,like random matching of resources,and news search resources,modified to single-particle multi-element multi-speed flight over the direction of,based on original algorithm and multi-element single-particle speed flight with the same direction,the bounds of the searching was stronger than the other algorithms.After the experimental analysis in comparison with other algorithms,obtained more satisfactory results,clearly,it has more availability than the others.