Quality problem oriented extension knowledge representation and diagnostic reasoning

Abstract:

To reasonably and timely solve the quality problem lacking of correlation detection information, a case-based reasoning method was proposed based on extension knowledge representation by taking the limitations of formal representation and reasoning method into consideration. Firstly, the knowledge representations of matter-element characteristic items and instance model in quality problem domain were put forward based on the matter-element model. Secondly, the problems of feature adjustment in the process of case retrieving, reusing, revising and retaining were solved by extension transformation. And then, combining with hierarchical organization form for cases based on domain knowledge and nearest neighbor strategy, the case retrieval algorithm was adapted for quality problems in the production process. With the application in vibration problem of 608-2RS ball bearing, the result indicates that the method has good practicability and feasibility.