The acquisition and application of similarity knowledge based on consultation in engineering product design

摘要

In a knowledge-intensive engineering product design process, case retrieval based on partial problem descriptions (PPDs) becomes more important than the general case retrieval in normal conditions for product design and acts as a crucial role in decision-making. Within the consultation mechanism, similarity knowledge can be acquired expediently and effectively in the product case base described with partial and incomplete information and knowledge. At one time, PPDs can interact with domain knowledge in an appropriate manner to serve the acquisition of similarity knowledge based on consultation (SKC) and case retrieval based on PPDs effectively. So, the consultation approach to case knowledge retrieval is provided to support engineering product design in this paper and the consultation hierarchy is analyzed and discussed from the viewpoint of knowledge transfer. The similarity transformation matrix for similarity measures is presented to handle the relationship between the partial unknown features and the related features, and the determination of weights found on the semantic knowledge of the design domain. As the complementarities of the domain knowledge, explanation knowledge is utilized to explain design requirements for case knowledge reuse and to assist the acquisition of similarity knowledge. Finally, the PPD of an oil pump design is employed to demonstrate the above viewpoints.