面向路径搜索的多虚拟机协同计算关键技术

发布时间：2018-02-07 16:59

本文关键词： 路径搜索计算机辅助工程虚拟化协同计算任务分解　出处：《计算机集成制造系统》2011年10期 　论文类型：期刊论文

【摘要】：研究了一种面向路径搜索的多虚拟机协同计算系统,为计算机辅助工程领域路径搜索问题的求解提供了一种灵活高效的计算环境。针对任务规模和搜索路径的不确定性等,构建了虚拟化平台下的自适应性体系结构和协同工作流程,提出了所需关键技术。针对路径搜索引擎的特殊需求,设计了启发式探测性任务分解算法,以充分挖掘任务在数据、功能等方面的并行性。通过调度多个子任务在多台虚拟机上并行执行,最大限度地提高资源的利用率,实现以资源换取较低的运行时间,加快得到最优解或满意解。设计了启发式解空间优化算法,以优化计算过程和结果数据的管理。通过监控任务的状态,可以避免资源浪费和重复路径执行,进一步减少运行时间。开发了系统引擎模块原型,并以一个计算机辅助工程领域的板材切割路径搜索问题对所设计的算法进行了实验评估。进一步将多虚拟机协同计算系统与四种系统进行对比分析,实验结果表明,多虚拟机协同计算系统比其他系统更能提供平稳的运行状态,并具有较高的性能和效率,原型很好地实现了面向路径搜索任务的求解环境。
[Abstract]:In this paper, a multi-virtual machine collaborative computing system for path search is studied, which provides a flexible and efficient computing environment for the solution of path search problem in computer aided engineering field, aiming at the uncertainty of task size and search path, etc. The adaptive architecture and collaborative workflow under virtualization platform are constructed, and the key technologies are proposed. According to the special requirements of the path search engine, a heuristic exploratory task decomposition algorithm is designed to fully mine the data in the task. Function parallelism. By scheduling multiple subtasks to execute in parallel on multiple virtual machines, the utilization of resources can be maximized, and the resources can be exchanged for lower running time. The heuristic solution space optimization algorithm is designed to optimize the computation process and the management of the result data. By monitoring the state of the task, the waste of resources and repeated path execution can be avoided. Further reduce the running time. Developed the prototype of the system engine module, An experimental evaluation of the proposed algorithm is carried out with a computer-aided engineering field of plate cutting path search. The multi-virtual machine collaborative computing system is further compared with four systems. The experimental results show that, Compared with other systems, the multi-virtual machine cooperative computing system can provide stable running state, and has higher performance and efficiency. The prototype achieves the solution environment of path search task well.
【作者单位】：西安理工大学计算机科学与工程学院;西安交通大学机械制造系统工程国家重点实验室;
【基金】：国家863计划重点资助项目(2007AA010305) 西安理工大学优秀博士论文基金资助项目(102-211007)~~
【分类号】：TP302;TP391.7

【参考文献】