Informacoes Quantitativas E Qualitativas No Projeto De Controladores Nebulosos

This work presents a methodology for designing fuzzy logic controllers. The main objectives are concerned with rule base design and analysis. For this purpose we shall use quantitative information which is based on mathematical concepts as well as qualitative information provided by a human expert. We will try to show how the combination of this information can be used as a new way for designing and analyzing linear and nonlinear controllers which are analytically derived from linguistic rules of fuzzy logic controllers. The method proposed to design the control rules is based on the process step response and it contributes to rule base modification. Moreover, some metarules that relate the basic experience of a human operator are also used. An alternative approach to derive the analytical model of the linguistic control rules is proposed and it also contributes to local study of the rule base, membership function design, implication function and logical operator choice and scaling factors tuning. Subsequently, a new structure of a fuzzy controller is proposed. This structure is motivated by using an equivalent fuzzy PI controller rule base. The applicability of the concepts presented in this work is illustrated through simulation.82779