基于多自主元的柔性工作流研究  

Flexible workflow based on multi autonomic objects

作　　者：王东勃[1] 王润孝[1] 阎秀天 盛义军[1] 

机构地区：[1]西北工业大学制造自动化软件与信息研究所,陕西西安710072 [2]斯特莱斯克莱德大学设计制造工程管理系

出　　处：《计算机集成制造系统》2007年第5期955-960,共6页Computer Integrated Manufacturing Systems

基　　金：国家自然科学基金资助项目(70272068)~~

摘　　要：为提高柔性工作流的智能性和不确定性变化响应能力,基于自主计算技术,提出了自主元概念,柔性工作流的智能性是通过嵌入工作流活动的自主元及自主元间的协作实现,给出了基于多自主元的柔性工作流形式化定义及体系结构,进而研究了多自主元的协作推理算法。使用加权不确定知识表达方法描述了自主元推理性知识,应用决策实验室分析方法给出了一种获取综合知识重要性和独立性的权重方法。最后,通过一个转包企业生产管理系统实例,证明了基于多自主元的柔性工作流系统能够显著提高柔性工作流的智能性。To improve the intelligence of flexible workflow and its responsiveness to uncertainties, the concept of Autonomic Object （AO） based on the autonomic computation technology was proposed. Intelligence of flexible workflow was realized by embedding AO in workflow activities and collaborations among AOs. The definition and architecture of flexible workflow activities based on multi-AO were presented. Then collaborative reasoning algorithm based on multi automatic objects was studied. The reasoning knowledge of AO was represented by weighted uncertainty knowledge. And a weight determination algorithm considering both the importance and independence of knowledge was presented, which was based on the decision making trial and evaluation laboratory approach. Final- ly, the proposed architecture was validated by a ease study, result showed that the multi-AO could remarkably improve the intelligence of flexible workflow.

关 键 词：自主计算 柔性工作流 多自主元 不确定性 

分 类 号：TH166[机械工程—机械制造及自动化] TP18[自动化与计算机技术—控制理论与控制工程]