Research on Feature Based IC Process Modeling and Clustering

Abstract

In order to make effective reuse of existing integrated circuit (IC) process cases, improve design efficiency, robustness of design resolutions and overcome knowledge famine and knowledge orientation problems in design process, a feature based hierarchical framework of knowledge modeling and similarity measurements based knowledge clustering method of IC process cases are proposed. In this paper, IC process cases knowledge is represented into four types: sequential type, textual type, crisp type and interval type. According to knowledge classification, sequential similarity measurement for process sequence knowledge and content similarity measurement for the rest types are listed. Finally, existing IC wafer handling process cases of an IC manufacturing enterprise are employed to verify the proposed methods.