Enhanced Local Stabilization of Constrained N-TS Fuzzy Systems With Lighter Computational Burden
Xiangpeng Xie, Fuyi Yang, Lei Wan, Jianwei Xia, Kaibo Shi
Abstract
The constrained Takagi–Sugeno fuzzy systems with local nonlinear models (N-TS fuzzy systems) are much more practical than the conventional T–S fuzzy models, but their applications are seriously affected by either the great conservatism or the heavy computational burden. In this short article, an enhanced local stabilization of constrained N-TS fuzzy systems is investigated with the purpose of not only reducing conservatism but also relieving computational burden. To do this, a new class of threshold functions are established in order to achieve real-time online assessment of the variation of the normalized fuzzy weighting functions (NFWFs) at different sampling instants, i.e., the one-step-past one and the current one. Thus, a key alternative is proposed for eliminating all the involved terms containing the one-step-past NFWFs while their main information is retained. Since our obtained design criteria are merely dependent on the current NFWFs, the conservatism can be reduced at the expense of lighter computational burden. The benefits of our proposed results are validated by implementing several objective comparisons with those recent results.