|
The design of a fixture that is based largely on previous
experience is time-consuming work according to the traditional
approach. The development of a computer-aided design system in
this domain to assist designs, therefore, is very necessary.
The Expert system is a popular technology of artificial
intelligence and many fixture design expert systems have been
developed. However, the development of a rule-based expert
system is restricted because of some obstacles, specially in
design domain. A technology superior to the expert system,
called case-based reasoning (CBR), is employed in this study to
improve these obstacles. Three systems are involved in this
thesis. The first two systems employed the group technology
(GT) to index input problem and cases. They are similar in
system structure and reasoning except that one applied in
dedicated fixture and the other one in modular fixture. A
knowledge representation in CBR, called memory organization
package (MOP), constitutes the third system that is also
applied in modular fixture. An index system applied in modular
fixture is also proposed in this thesis to be referred when
characterizes the input problem and cases. It is believed that
the result of this study can provide a new way to improve the
capability of rule-based expert systems and establish a capable
computer-aided design system.
|