A HYBRID EXPERT SYSTEMS ARCHITECTURE FOR YARN FAULT DIAGNOSIS
Nomusa Dlodlo, Lawrance Hunter, Cyprian Cele, Roger Metelerkamp, Anton F. Botha · Fibres and Textiles in Eastern Europe · 2007
This article describes a hybrid expert system architecture to support yarn fault diagnosis. The system uses a combination of rule-based and case-based techniques to achieve the diagnosis. Rule-based systems handle problems with well-defined knowledge bases, which limits the flexibility of such systems. To overcome this inherent weakness of rule-based systems (RBS), case-based reasoning (CBR) has been adopted to improve the performance of the expert system by incorporating previous cases in the generation of new cases. The idea of this research is to use rules to generate a diagnosis on a fault and to use cases to handle exceptions to the rules. The cases are represented using an object-oriented approach to support abstraction, re-use and inheritance features.