Reduced-order Extended Dissipative Filtering for Nonlinear Systems with Sensor Saturation via Interval Type-2 Fuzzy Model

Yi Zeng, Hak-Keung Lam, Bo Xiao, Ligang Wu, Ming Chen

Research output: Contribution to journalArticlepeer-review

2 Citations (Scopus)
144 Downloads (Pure)


The system nonlinearity, sensor saturation, and the uncertainty will hamper the analysis and affect the control performance. Filtering is a signal processing method which facilitates the system analysis and synthesis by signal estimation or noise suppression. To achieve generalized filtering problem for nonlinear systems with sensor saturations with lower computational burden, this article addresses the reduced-order extended dissipative filter design for nonlinear sensor-saturated system which is modeled by interval type-2 (IT2) T-S fuzzy system. For IT2 T-S fuzzy systems, the main challenge exists in the acquisition of the information in IT2 membership functions (MFs) for analysis and design. A membership-function-dependent (MFD) method is applied to capture the information of the MFs for reducing the conservativeness introduced by MFs not involved in the analysis. An extended dissipative filtering method, under imperfect premise matching (IPM) concept that the membership functions of the filter are different from those of the model, is proposed for sensor-saturated IT2 fuzzy systems. The proposed method has a high flexibility in parameters adjustment of both the filter and the design condition, including extended dissipative matrices, approximation MFs, and sensor saturation degree, one can freely choose the parameters according to the required performance of the fuzzy filter. A numerical example is given to demonstrate the effectiveness of the results.

Original languageEnglish
Pages (from-to)5058-5064
Number of pages7
JournalIEEE Transactions on Fuzzy Systems
Issue number11
Publication statusPublished - 1 Nov 2022


Dive into the research topics of 'Reduced-order Extended Dissipative Filtering for Nonlinear Systems with Sensor Saturation via Interval Type-2 Fuzzy Model'. Together they form a unique fingerprint.

Cite this