Integrating multi-granularity model and similarity measurement for transforming process data into different granularity knowledge

Abstract

The core of intelligent manufacturing is to incorporate the expert knowledge in manufacturing process, and knowledge transformation is the key to knowledge accumulation and application. In this paper, the research carried on transformation for different granularity knowledge from the cases of sheet metal parts in process planning. First of all, this paper analyzes the difference of organization structure between process data and knowledge in the base. The multi-granularity model of process knowledge is established in the form of tuple, which helps to clarify the hierarchy structure and internal relations. Thereafter, the concrete process is presented to transform single granularity process data into multi-granularity process knowledge, i.e., process data extraction, state determination and knowledge construction. With respect to state determination, similarity measure methods for different granularity knowledge are established to reduce the redundancy in the transformation process. As a novel approach, sequence alignment based on edit distance is proposed to calculate similarity exactly between two process flows. Finally, the knowledge transformation tool for different granularity knowledge is developed to enhance knowledge acquisition and improve the strength of knowledge reuse in fabrication order design for sheet metal parts through application of the above method. Also an example is given to illustrate the usefulness of the proposed method.