Designing Databases with Fuzzy Data and Rules for Application to Discrete Control

Many real world systems and applications must deal with imprecise or vague data. For such systems, information management components are needed that provide support for managing this imprecise data. Fuzzy theory allows us to model imprecise or vague data. The use of fuzzy theory also allows us to model vague knowledge. There have been several proposals for extending relational database systems in order to represent as well as query fuzzy data. However, little work has been done in modeling uncertainty at the conceptual schema level or in developing higher level conceptual models for fuzzy rules. To fill the first gap, the authors have proposed a design methodology for fuzzy relational databases. This methodology prescribes a sequence of steps to implement a fuzzy relational database from a proposed extended fuzzy Entity-Relationship model. In this paper, we also propose a generic data model to represent fuzzy rules. Generic methods are also presented which allow decision making based on these fuzzy rules. A working prototype of the fuzzy database has been built for a discrete control system, namely a semiconductor manufacturing process supervisory control system, which needs to handle fuzzy data and rules.