Skip to Main content Skip to Navigation

Réduction de la complexité des contrôleurs flous : applications à la commande multivariable

Abstract : Fuzzy control enables to avoid the use of mathematical models sometimes difficult to obtain. Its capability to translate the human operator knowhow into expertise rules expressed in a simple language makes it a very promising technique. Nevertheless, when the number of variables to be considered becomes too high, the rule base rapidly explodes, and problems linked to its implementation follow. This thesis resides within the framework of the actual research works on fuzzy control and treats the problem of the combinatorial explosion in the number of rules. In the first part, the basic principles of fuzzy logic and fuzzy control are recalled. In the second part, solutions, in order to make easier the synthesis of a fuzzy logic controller, are presented. Two cases are considered : the rule base already exists, the rule base is not available and we have to develop a fuzzy logic controller of reduced complexity. In practice, we generally do not have this rule base. Therefore, we lay stress on the second point. Once the structure of the fuzzy logic controller has been decided, the number of parameters to be tuned may reach a high level. It becomes then interesting to apply some learning techniques in order to automatically tune the numerous parameters (gains and membership functions for both inputs and outputs) of the fuzzy logic controller. These parameters are optimized using the very simple gradient descent method. In the third part, the methodology proposed in this dissertation is applied with success to the control of two multivariable processes : a mixing tank and a biological waste-water treatment process.
Document type :
Complete list of metadata
Contributor : Emilie Marchand <>
Submitted on : Friday, September 2, 2005 - 3:32:21 PM
Last modification on : Sunday, October 25, 2020 - 7:07:43 AM
Long-term archiving on: : Tuesday, September 7, 2010 - 5:16:08 PM


  • HAL Id : tel-00010030, version 1


Véronique Lacrose. Réduction de la complexité des contrôleurs flous : applications à la commande multivariable. Automatique / Robotique. INSA de Toulouse, 1997. Français. ⟨tel-00010030⟩



Record views


Files downloads