PERFORMANCE EVALUATION OF A PARAMETERIZED FUZZY PROCESSOR (PFP)

In the human-based world, the general-purpose and high-speed fuzzy pro cessor which supports more functions and capabilities plays a very imp ortant role. A parameterized fuzzy processor (PFP) has been proposed e arlier (B.-T. Chen et al., 1993) to demonstrate its structure and appl ications. In this paper, the general performance of the PFP will be sh own in detail by a series of analyzing results. First, fuzzy data pres entation adopted in the PFP is evaluated by the proposed criteria and compared with the other presentations. According to the versatile data presentations, the general-purpose operations are evaluated by the co mputation order and run time. Second, the emphasized characteristics o f PFP operations possess the feasibility of parallel and pipelined pro cessing. The designed circuit and execution cycles are confirmed in a simulation environment. Finally, the fuzzy applications extendable by the general-purpose PFP are shown by the solvable operations. Based on the PFP, a processing model for human-based problems is proposed to a nalyze the fuzzy applications. A practical application is proposed to demonstrate the capability of the PFP.