SVD-based reduction to MISO TS models
Péter Bárányi, Yeung Yam, Annamária R. Várkonyi-Kóczy, Ron J. Patton · IEEE Transactions on Industrial Electronics · 2003
The main objective of this paper is to expound the singular-value-decomposition (SVD)-based reduction technique proposed to single-input-single-output Takagi-Sugeno (TS) fuzzy models to multivariable cases. The use of higher order singular value decomposition is proposed in this paper for the complexity reduction of multiple-input-single-output TS fuzzy model approximation. A detailed illustrative example of a nonlinear dynamic model is also discussed.