The attributes represented in the FORM differ from each other for similar parts. For this reason, face attributes play an important role for part recognition. For example, FORMs of the

 

 

An expert system or intelligent knowledge based system, which is a branch of artificial intelligence (AI), is able to automate the reasoning activities to capture logic, heuristic knowledge and experience in a computer environment. It is capable of utilizing domain-specific knowledge and making intelligent decisions. In this study, an expert system module is developed and embedded within the part recognition system. It plays a key role and performs part recognition process by reasoning FORM and the heuristic knowledge that is composed of the rules in the knowledge base. The expert system consists of a knowledge base, an optional and automatic rule writing module, an inference engine and a knowledge acquisition. Since part recognition procedure is fully automatic, the user interface is not required.

 

5.1. Knowledge base

 

The knowledge base is a database that holds domain-specific information and is composed of rules and facts. Rules define logical relations between concepts of the problem domain. Since an expert system approach is employed in the developed approach, a knowledge base should be constructed. Windows based a text editor is used for knowledge base in the system. A rule for each part has been written in the knowledge base. Rule belonging to each part in the knowledge base is represented with face adjacency relations and face attributes by writing in following format:

 

‘‘RULE’’ number: ‘‘IF’’

 

‘‘the’’ + surface type + ‘‘has neighbour’’ or ‘‘neighbours’’ + neighbouring face1, neighbouring

 

face2, neighbouring face3 . . . neighbouring faceN ‘‘AND’’ ‘‘the’’ + attribute ‘‘=’’ value of the attribute + unit of the attribute

 

‘‘THEN’’

 

‘‘the part is a’’ or ‘‘an’’ + part name

 

According to this format, a component can be described with face adjacency relations and face attributes in IF section of the rule. While adjacency relations of a face are defined, the face and neighbouring face(s) belonging to that face are written as shown in above rule format. The user must define face

 

740 A. C¸ ic¸ek, M. Gu¨lesin / Computers in Industry 58 (2007) 733–746

 

 

adjacency relations for rule of any part to be represented in the knowledge base. If face adjacency relations are enough to define a part, then face attributes may not be described. If face adjacency relations are not enough to define a part, then face attributes must be described.

 

The rule represented in the knowledge base of the retaining ring for bores in Fig. 10 is given below. In the rule of the retaining ring for bores, face adjacency relations are only defined. This has been adequate for recognition of the retaining ring.

 

RULE 26:

 

IF

 

the plane_face has neighbours plane_face, plane_face, plane_face, plane_face,

 

cylindrical_face, cylindrical_face, cylindrical_face, cylindrical_face, cylindrical_face

 

cylindrical_face AND

 

the plane_face has neighbours plane_face, plane_face, cylindrical_face, cylindrical_face AND

 

the cylindrical_face has neighbours plane_face, plane_face, plane_face, plane_face AND

上一篇:拉丁美洲海岸侵蚀的现状英文文献和中文翻译
下一篇:地下车库的污染物扩散英文文献和中文翻译

加速度传感器ZIGBEE协议的...

声音识别技术的焊接机器...

知识工程的汽车覆盖件冲...

级进模零部件的知识库系...

液压泵系统识别方法比较英文文献和中文翻译

CAD知识的模具设计英文文献和中文翻译

知识管理及其对隐性知识...

网络语言“XX体”研究

互联网教育”变革路径研究进展【7972字】

ASP.net+sqlserver企业设备管理系统设计与开发

麦秸秆还田和沼液灌溉对...

老年2型糖尿病患者运动疗...

张洁小说《无字》中的女性意识

新課改下小學语文洧效阅...

安康汉江网讯

我国风险投资的发展现状问题及对策分析

LiMn1-xFexPO4正极材料合成及充放电性能研究