Journal/Conference: Proceedings of the 6th International Conference Mixed-Signal Design of Integrated Circuits and Systems (MIXDES'99), Kraków (Poland), June1999, pp. 389.
Abstract - Fuzzy systems theory proposes a systematic method for mapping human knowledge into a multidimensional input-output nonlinear relation. Several real-world control engineering tasks require this universal approximation characteristic provided by fuzzy inference engines. In the application presented herein, the need for nonlinear control surfaces arises in the area of high-frequency switching power converters, which are inherently nonlinear systems requiring highly nonlinear control laws. Thus, the need for high-frequency operation dictates the development of dedicated hardware implementation. The widely-accepted Takagi-Sugeno (TKS) [1] FKBC model is the best suited for implementation purposes, since it significantly reduces hardware complexity at the output defuzzyfication layer, using input-related output hyperplanes in place of output membership functions. The first-order TKS fuzzy controller is described by the following set of IF-THEN rules: