Stable Learning Mechanism for Novel Takagi-Sugeno-Kang Type Interval-valued Fuzzy Systems
Yi Han Lee, Ching Hung Lee, Member Iaeng · 2014
Abstract—In this paper, we propose a novel Takagi-Sugeno-Kang type interval-valued neural fuzzy system with asymmetric fuzzy membership functions (called TIVNFS-A). In addition, the corresponding type reduction procedure is integrated in the adaptive network layers to reduce the amount of computation in the system. Based on the Lyapunov stability theorem, the TIVNFS-A system is trained by the back-propagation (BP) algorithm having an optimal learning rate (adaptive learning rate) to guarantee the stability and faster convergence. Finally, the TIVNFS-A with the optimal BP algorithm is applied in nonlinear system identification to demonstrate the effectiveness and performance. Index Terms—interval-valued fuzzy system, Lyapunov stability theorem, asymmetric membership function, TSK type, nonlinear system I I.