分布式知识处理系统中基于知识的任务规划模型

A KNOWLEDGE BASED MODEL FOR TASK PLANNING IN DISTRIBUTED KNOWLEDGE PROCESSING SYSTEMS

分布式知识处理系统的求解主要由四步组成:任务规划、任务分布、协作求解,以及结果综合,其中任务的规划是整个求解过程的基础。目前设计的许多系统缺乏任务规划能力,任务的执行序列和执行任务的任务知识源主要由系统开发者指定,所以系统的求解结果带有很大的主观性,效率也难以提高,本文首先分析了分布式知识处理系统中任务、任务知识源和任务规划;其次采用逻辑-对象知识模型,来表示任务规划问题;最后提出了一种基于知识的任务规划模型。

The problem solving in Distributed Knowledge Processing System contains four steps: Task Planning,Task Distribution,Cooperative Problem Solving,and Results Synthesis.Task Planning is the key issue in DKPS.Based on the current theory and techniques of DAI,this paper analyses task and task planning in DKPS,adopts Logic-Object Knowledge Model(LOKM) to represent task planning problem,and proposes a knowledge-based model for Task Planning in DKPS.