【作者】 赖朝安； 孙延明； 郑时雄；

【Author】 LAI Chao-an SUN Yan-ming Zheng Shi-xiong South China University of Technology, Guangdong Guangzhou, 510640, China

【机构】 华南理工大学机电系；

【摘要】 本文研究了敏捷虚拟企业(AVE)的决策间题及其GDSS的任务,提出了两种有多种用途的新方法并在该系统中应用:首先提出一种新的面向文字资料的知识获取方法,它包含一种统一的机械知识表示模型,并基于此模型设计XML标记未处理机械知识文本,使知识仍保留在原Web自然语言文本中而形成具有层次结构的文本知识库,使专家系统知识库的表现形式依据国际标推XML,可实现知识的异地共享;其次分析了基于实例推理技术当前研究的不足并提出用遗传算法进行组合优化来解决,为此提出一种实例关联知识的挖掘算法以寻找功能相似或作用原理相似的产品实例,并将相似实例作为遗传算法初始群体,克服传统遗传算法初始群体选择的盲目性,加快遗传算法运算的收敛。更多还原

【Abstract】 This paper analyzed the decision problems in agile virtual enterprise(AVE) and the tasks of its GDSS. The system is equipped with two new technologies for achieving various objectives: first, a new document-oriented knowledge acquisition methodology which includes a kind of united knowledge representation model was advanced, based on which model the mechanical knowledge XML tags was designed to label knowledge documents, so that knowledge still stay in primary Web document which become document knowledge base with hierarchy. Thus, the representation manner of ES knowledgebase bear the international standard XML and the remote share of knowledge can be achieved; second, it analyzed the disadvantage of the current study on CBR and proposed applying GA in combination and optimize to resolve them, and advanced a methodology of mining the relationship of cases to search those case being similar in the term of function or action principle for the purpose of original colony of GA, thus conquer the blindness of the produce of the original colony of GA so as to quicken the convergence of the operation of GA.更多还原