Model-based control using interval type-2 fuzzy logic systems
- 93 Downloads
Type-2 fuzzy logic introduced new formalisms capable of overcoming the inherent uncertainties of approximating real-world processes by computational models. Yet, despite increasingly present in the nonlinear modeling literature, model-based control theory does not seem to be taking full advantage of the improvements that type-2 fuzzy logic provides. Therefore, the present work proposes the development of a new control methodology based on the generalized predictive control theory supported by interval type-2 Takagi–Sugeno fuzzy logic systems. The developed control system is based on locally linear approximations of the type-2 Takagi–Sugeno model and will be evaluated using a nonlinear process based on the yeast fermentation reaction. For comparison purposes, two additional generalized predictive control implementations based on a linear auto-regressive model with exogenous inputs and a type-1 Takagi–Sugeno fuzzy model will be used. The performance of the closed loop systems will be evaluated by subjecting the process to quick changes in the operation regime and to unmeasured external disturbances. The mean squared error, control effort and typical transient step response metrics (overshoot and settling time) will provide the benchmark criteria. The results achieved demonstrate that at the expense of a small increase in the computational effort, type-2 fuzzy logic systems improve the transient behavior of the closed loop system, presenting significant advantages when the controlled process is subject to unmodeled disturbances.
KeywordsInterval type-2 fuzzy sets Generalized predictive control Nonlinear system’s control
The authors acknowledge the support given by FCT—Foundation for Science and Technology, in the context of Ph.D. scholarship BD/71601/2010 and the IEETA Research Unit, funded by National Funds through FCT in the context of the projects UID/CEC/00127/2013 and Incentivo/EEI /UI0127/2014.
Compliance with ethical standards
Conflict of interest
The authors Rómulo Antão, Alexandre Mota and Rui Martins declare that have no conflict of interests.
- Åström K, Wittenmark B (2008) Adaptive control. Dover, LondonGoogle Scholar
- Chiu S (1994) Fuzzy model identification based on cluster estimation. J Intell Fuzzy Syst 2(3):267–278Google Scholar
- Hägglund T (1983) New estimation techniques for adaptive control, Ph.D. thesis, Department of Automatic Control, Lund Institute of Technology, SwedenGoogle Scholar
- Khanesar, M, Kayacan Teshnehlab M, Kaynak O (2011) Levenberg Marquardt algorithm for training of type-2 fuzzy neuro system with a novel type-2 fuzzy membership function. IEEE Symposium on advances in type-2 fuzzy logic systems, pp 88–93Google Scholar
- Kulhavý R (1985) Restricted exponential forgetting in real-time identification. IFAC Symposium on identification and system parameter estimation, pp 1143–1148Google Scholar
- Kumbasar T, Eksin I, Guzelkaya M, Yesil E (2011b) Type-2 fuzzy model inverse controller design based on bb-bc optimization method. In: 18th World congress of the international federation of automatic control (IFAC)Google Scholar
- Kumbasar T, Hagras H (2015) A gradient descent based online tuning mechanism for pi type single input interval type-2 fuzzy logic controllers. In: Proceedings of IEEE international conference on fuzzy systemsGoogle Scholar
- Mahfouf M, Linkens D, Abbod M (2000) Adaptive fuzzy TSK model-based predictive control using a CARIMA model structure. Chem Eng Res Des Process Control 78(4):590–596Google Scholar
- Melin P, Castillo O (2014) A review on type-2 fuzzy logic applications in clustering, classification and pattern recognition, applied soft computing, vol 21. Elsevier, AmsterdamGoogle Scholar
- Rossiter J (2004) Model-based predictive control: a practical approach, control series. CRC Press, Boca RatonGoogle Scholar
- Wu D, Tan W (2005) Type-2 FLC modeling capability analysis. In: Proceeding of the 2005 IEEE international conference on fuzzy systems, pp 242–247Google Scholar
- Zhao L (2010) Direct-inverse modeling control based on interval type-2 fuzzy neural network. In: Proceedings of the 29th Chinese control conference, pp 2630–2635Google Scholar