节点文献

一个基于逻辑的多目标优先级排列专家系统MOPSES

A Multiobjective Priority Sorting Expert System (MOPSES)

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 李先国蒋立源杨保安张世杰

【Author】 Li Xianguo Jiang Liyuan Yang Baoan Zhang Shijie (Northwestern Polytechnical University)

【机构】 西北工业大学西北工业大学 讲师副教授教授硕士

【摘要】 鉴于多目标优化决策中目标优先级确定的困难,本文尝试用专家系统(ES)来辅助决策者确定优先级顺序.系统以企业经营决策,计划编制为对象,构造了一个目标规划模型的优先级排列ES,文中较详细阐述了系统的设计思想,结构,原理及具体实施.系统使用TURBO PROLOG语言开发,用户界面采用全屏幕多窗口下拉菜单.该系统在带有CH卡的IBM-pC兼容机上实现,操作系统为GWBIOS.由于能在一屏幕上显示25行汉字和十分友好的用户界面,使用该系统非常方便.

【Abstract】 Most research papers on MODM (Multiobjective Decision Making) failed to take into consideration the fact that the objectives are often in conflict with each other and that the problem of assigning priorities to objectives is an ill-structured problem. The problem generally requires a great amount of specialized knowlege of a particular field, which includes both the relevant theoretical knowlege and the valuable judgements of experts that come from years of practical experience. It is believed that Yang in 1990 was the first to explore the possibility of storing such a specialized knowlege in a knowlege base [4]. This paper presents MOPSES as the embodiment of Yang’s idea by applying knowlege engineering and artifical intelligence to MODM. MOPSES, employing TURBO PROLOG (2.0) as its language, has in its knowlege base rules formalized from relevant specialized knowlege, and has a program that can simulate the thought process of experts, i.e., the process of analysis, inference and judgement. Even though objectives are often in conflict, MOPSES can deal with such ill-structured problem and assign priorities to objectives. MOPSES has the following features: 1. It can assign priority ranks to objectives through combining "pair comparison"dialogue appropriately with "question and answer" dialogue. 2. MOPSES is flexible. The objectives (data), rules (knowlege), and question sentences (language) can be revised if desired. 3. All the modules of MOPSES are intergrated. The screen of MOPSES consists of many windows and pull-down menus. Thus it is easy to operate MOPSES. MOPSES has been applied to working out production plan of a factory. In this particular case, there are 9 objectives, 14 rules and 20 question sentences. It takes just a few minutes for MOPSES to get approximately the same result that requires human expert conversant with relevant specialized knowlege to take hours or even days to obtain.

【基金】 国家自然科学基金资助项目的一部分
  • 【文献出处】 西北工业大学学报 ,Journal of Northwestern Polytechnical University , 编辑部邮箱 ,1993年01期
  • 【下载频次】33
节点文献中: 

本文链接的文献网络图示:

本文的引文网络