The Application of Artificial Neural Network Model on the Establishment of Configuration Identification

Industrial Personal Computer (IPC) is a highly customized product, which customers can take part in and have their requirements from design to manufacture and shipment. It requires lots of special parts and components, and causes more difficulty and cost increase in parts management. Configuration Management is one of the commonly used methods to resolve this problem. The first part in Configuration Management is Configuration Identification that usually determines the system’s final result. Therefore, Configuration Identification is critical t for the Configuration Management. According to the past research, there are three major means of Configuration Identification: Work Breakdown Structure (WBS), Expert Questionnaire & Fuzzy Logic, and Expert Evaluation. In work breakdown structure, there would be some subjective choices for identification items because of some personnel issues; the bringing of identification items may be affected by the filling condition of questionnaire. In the expert evaluation, the correctness of identification items may be queried because they come from specific experts’ opinion. Thus, these three means all have weaknesses and they are unscientific to use. In view of this, this paper applied artificial neural Fuzzy-ART mode as Configuration Identification, bases on the grouping concept of Group Technology. In the end, it takes the screw used in industrial personal computer as target object to verify the theory in practice