A new adaptive Mamdani-type fuzzy modeling strategy for industrial gas turbines

Author

Yu Zhang ; Jun Chen ; Bingham, Chris ; Mahfouf, Mahdi

Author_Institution

Sch. of Eng., Univ. of Lincoln, Lincoln, UK

fYear

2014

fDate

6-11 July 2014

Firstpage

1599

Lastpage

1603

Abstract

The paper presents a new system identification methodology for industrial systems. Using the original Mamdani fuzzy rule based system (FRBS), an adaptive Mamdani fuzzy modeling (AMFM) is introduced in this paper. It differs from the original Mamdani FRBS in that it applies different membership functions and a denazification mechanism that is `differentiable´ with respect to the membership function parameters. The proposed system also includes a back error propagation (BEP) algorithm that is used to refine the fuzzy model. The efficacy of the proposed AMFM approach is demonstrated through the experimental trails from a compressor in an industrial gas turbine system.

Keywords

fuzzy set theory; gas turbines; knowledge based systems; power engineering computing; AMFM; BEP algorithm; FRBS; Mamdani fuzzy rule based system; adaptive Mamdani-type fuzzy modeling strategy; back error propagation algorithm; denazification mechanism; industrial gas turbines; membership functions; Adaptation models; Computational modeling; Engines; Prediction algorithms; Predictive models; Training data; Turbines; Mamdani fmzy rule based system; adaptive Mamdani fmzy modeling; back error propagation; industrial gas turbine;

fLanguage

English

Publisher

ieee

Conference_Titel

Fuzzy Systems (FUZZ-IEEE), 2014 IEEE International Conference on

Conference_Location

Beijing

Print_ISBN

978-1-4799-2073-0

Type

conf

DOI

10.1109/FUZZ-IEEE.2014.6891815

Filename

6891815