摘要
针对工业应用,从控制的角度提出分层分布式的资源发现机制,研究网格环境下分布式控制系统的资源部署策略,建立不同应用需求下相应部署策略的数学模型。对于与属性无关资源的部署问题,由于数量充足,其部署策略视为静态匹配过程,可根据数学模型求出最优解;但受属性的限制,可供选择的资源数量不足,只能在数学模型的基础上,结合启发式算法,求得近优解。
In the sight of control, a hierarchical and distributed discovery model of resources is presented for application in industry. A deep study is made on the Grid-based Distributed Control System(GDCS) resource deploying policy with regard to resource requirement problems respectively in two different natures which is irrelevant and restrict to property, under the grid environment. As a result, deploying policy mathematical model is built up to meet various application requirements. As for resources deploying policy problem, which is irrelevant to property, an optimal solution can be worked out according to mathematical model since the deploying policy can be considered as a static matching process because of sufficient of the resources. However for the resource that is restrict to property, owing to the insufficient of the optional resources, the optimal solution can be obtained on the basis of mathematical model combined with heuristic algorithm.
出处
《计算机工程》
CAS
CSCD
北大核心
2008年第15期63-66,共4页
Computer Engineering
关键词
基于网格的分布式控制系统
资源部署
属性无关
Grid-based Distributed Control System(GDCS)
resource deploying
irrelevant to property