针对网格中的服务实体只具有局部视野,常导致服务的可用性受到限制的缺陷,提出了一种基于D—S证据理论的工作流宏观自组织演化机制。该机制以网络中实体的交互行为作为有序演化的基础,以D—S证据理论来解决实体间信任度的度量、传递和组合问题,并且扩展了D—S证据理论中对信任度的控制规则,提出了一种比例信任控制规则。实验结果表明,服务实体通过自适应学习,逐渐与自己的协作体形成了恰当的信任连接,所有这些连接形成加权连接图,有利于提高工作流的适用性。
The paper is concerned with the problem that the service entity only has the partial field of vision in the grid service, and causes the service to be restricted for application. A mechanism of the macro self-organization evolvement workflow based on the D-S evidence theory is proposed, which is based on the entity network interactive behavior, solves the entity confidence level measure, transmission and combination questions by the D-S evidence theory, and expands the control rule of the confidence level in the D-S evidence theory, proposes the proportion trust control rule. The experimental result shows that the workflow performance has been enhanced by the way.