【摘要】 从仿真系统工作流的关键路径分析,确定关键路径上的联邦成员节点,使网格资源调度方面优先获得保证。鉴于仿真网格系统运行的复杂性和不确定性,引入了随机规划理论,提出了仿真网格的动态关键路径概念以及基于动态关键路径的资源调度算法,共同解决当资源有限以及任务相关情况下仿真网格的资源调度问题。具体实现,主要使用遗传算法以及 MCP 算法。仿真实验结果表明,基于动态关键路径的资源调度算法能够优化仿真网格资源的调度。更多还原

【Abstract】 From the critical path analysis of simulation system work flow,federation members nodes of the critical path are confirmed.They are assigned to grid resource priority.In view of the complexity and uncertainty of simulation grid,stochastic programming is introduced.The dynamic critical path concepts for simulation grid and resource schedu- ling algorithm based on dynamic critical path are brought up to solve the problem of resource scheduling for simulation grid when simulation grid resources are limited and its tasks are dependent.The concrete implementation on the part of resource scheduling is made by mainly using the genetic algorithm and MCP algorithm.Simulation experiment results indicate that resource scheduling algorithm based on dynamic critical path can optimize the scheduling of simulation grid resources.更多还原