Rapid Prototyping and Optimization based on Conceptual Specification for Fuzzy Applications

In this paper, a new framework for rapid system design and implementation for fuzzy systems is proposed. The given system specification is separated into two components: a conceptual specification and a parameter specification. The conceptual specification defines the system core, which is rarely changed during the system adjustment and may be implemented in hardware at an early stage of the design process. The parameter specification defines parameters that are often changed and are implemented in software for easy adjustment. Such a partitioning approach integrates both hardware and software capability. Two models, a Behavioral Network, which describes the overall system behavior, and a Conceptual State Graph, which represents the conceptual specification, are discussed. The presented methodology gives a rapid prototyping and efficient system tuning process which, therefore, can reduce the development cycle time and increase flexibility when designing fuzzy systems. This work was supported by NSF grant MIP 95-01006 and Royal Thai Government Scholarship. 1