Kyoko Tsukano

Synthesis of operational transconductance amplifier-based analog fuzzy functional blocks and its application

A synthesis of analog fuzzy functional blocks (a complementary membership function circuit, a maximum circuit, a complementary maximum circuit, and a defuzzifier circuit) based on operational transconductance amplifiers (OTAs) is presented. The complementary membership function circuit and the maximum circuit are synthesized from the formulations using bounded-difference operations. The defuzzifier circuit is synthesized as the follower-aggregation circuit composed of multiplier-type OTAs, SPICE simulations showed that the proposed fuzzy functional blocks feature high-speed operations and low power consumption. The complementary membership function circuit, the maximum circuit, and the complementary maximum circuit were built with discrete components and commercially available OTAs, and the performances of these circuits were confirmed by experiments. As an application, a singleton fuzzy controller with 3/spl times/3 rules is synthesized using the proposed circuits. The simulation of this controller showed that the inference speed of the order of 15 MRPS (rules per second) is easily realizable. >

A design of current-mode analog circuits for fuzzy inference hardware systems

Current-mode analog circuits, i.e., a maximum (MAX) circuit, a minimum (MIN) circuit, and a defuzzified circuit, for fuzzy inference hardware systems, are proposed. The design of the proposed MAX and MIN circuits is based on simple bounded-difference equations. The defuzzifier circuit is synthesized using multipliers/dividers based on the translinear principle. The design considerations and the simulation results of a 2 /spl times/ 2 fuzzy controller using the proposed circuits are described.<<ETX>>

A current-mode chaos circuit with return map approximation using supervised learning

A new current-mode analog chaos circuit is proposed. The proposed circuit is composed of an operation block, a parameter block, and a delay block. The return maps of this circuit are generated in the neuro-fuzzy based operation block. These maps are determined by supervised learning. For the proposed circuit, the dynamics of the learning and the state of the chaos are analyzed by computer simulations. The theoretical analysis is carried out to predict the bifurcation diagrams under the existence of the nonidealities in the piecewise linear approximated return maps. The simulation results showed that the return maps can be realized with the precision of the order of several percent and that different kinds of bifurcation modes can be generated easily.