基于CLIPS的卫星任务规划专家系统设计  

Design of a CLIPS-Based Expert System for Satellite Scheduling

作　　者：杜莹[1] 陈浩[1] 李军[1] 唐宇[1] 陈宏盛[1] 

机构地区：[1]国防科技大学电子科学与工程学院

出　　处：《计算机工程与科学》2009年第12期138-141,共4页Computer Engineering & Science

基　　金：国家自然科学基金资助项目(60604035);国家863计划资助项目(2007AA12Z229;2007AA120202)

摘　　要：本文提出了基于CLIPS的卫星任务规划专家系统的设计方法,详细分析了系统的结构和功能,重点讨论了中文产生式系统的BNF范式、基于上下文的推理机制和集合运算符。中文产生式系统的BNF范式基于CLIPS标准BNF范式定义,并依据BNF范式进行规则表示和规则自定义获取;推理机采用上下文限制的规则控制策略,依据不同的上下文加载相关的事实和规则,提高推理机的运行效率;利用规则中的对象逻辑子式进行了集合运算符的设计,并对极值运算符、属性差值运算符和均值运算符等三类集合运算符进行了探讨。该系统解决了卫星任务规划中知识表示和知识获取问题,提高了卫星任务规划推理效率,为卫星任务规划人员提供有效的辅助决策功能。A method of designing the CLIPS-based expert system for satellite scheduling is proposed. The structure and functions of the systems are analyzed in detail. The BNF（Bachus_Naur Form） of Chinese production system , the reasoning mechanism based on context and the collection operators are mainly discussed. On the basis of CLIPS BNF, the BNF of Chinese production systems is defined for the rules representation and rules acquisition. An inference engine is realized via the context-control strategy. According to different contexts, the inference engine will load the relevant facts and rules to improve the efficiency of the reasoning mechanism. By means of the object logic patterns of rules, collection operators are designed and three kinds of collection operators are discussed. The method can solve the problems of knowledge representation and knowledge acquisition, which improve the reasoning efficiency of satellite scheduling and provides an effective decisionmaking function to the satellite scheduling people.

关 键 词：专家系统 卫星任务规划 BNF范式 集合运算符 

分 类 号：TP182[自动化与计算机技术—控制理论与控制工程]