一种可扩展型CLP系统的统一建模方法  

Modeling for Extended Constraint Logic Programming

作　　者：易军凯[1] 翟性泉 王雪晶[1] 

机构地区：[1]北京化工大学信息科学与技术学院,北京100029 [2]海军后勤装备技术研究所,北京101200

出　　处：《计算机工程与应用》2004年第30期116-119,共4页Computer Engineering and Applications

摘　　要：约束推理是人工智能中主要组成部分之一,可以解决实际优化调度和规划过程中的约束求解问题。这里在解释了约束逻辑程序设计的原理和过程基础上,打破封闭式约束逻辑程序设计系统,从软件工程上采用统一建模语言,提出一种新的开放的可扩展型约束逻辑程序设计结构系统。为实现可扩展的约束推理搜索系统,引进UML建模语言中用例图、类图和协作图。在建模基础上详细说明了可扩展约束逻辑程序设计中数学模型,搜索引擎和搜索驱动三者间的关系以及它们内部的工作内容。最后在描述系统结构后,提出了可扩展的内容。根据扩展因素,外界为满足更多的需要可扩展本系统的约束过滤器。Constraint logic programming is one important part of artificial intelligence and can solve constraint problems such as optimization and scheduling.Here,the theory and process of extended constraint logic programming are introduced,which broke normal closed system structure and built an open system with UML in software engineering.In order to realize constraint combinatorial search,use case,class and collaboration diagram in UML are involved.On the basis of modeling,the relation among constraint model,search engine and search solver is detailedly described.Also,the content of three parts is analyzed.At last,after description of system structure,extensive algorithm is shown.According to factors of extension,outside can extend constraint filtering to be satisfied with more constraint requirements.

关 键 词：约束逻辑程序设计 数学建模 搜索引擎 搜索求解器 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论]