Controle Fuzzy adaptativo de sistemas não lineares incertos representados por modelos Takagi-Sugeno aplicado a um sistema de rotor duplo

Abstract

In this dissertation, a new adaptive approximation approach based on the backstepping technique is proposed for the control of the pitch angle (SISO operating mode) of an electromechanical system known as Twin-Rotor MIMO System – TRMS. Universal approximants in the control law are used to estimate the unknown part of the dynamics using Takagi-Sugeno (TS) fuzzy models. In order to reduce the complexity in the controller design by backstepping, command filters are used to avoid the explicit calculation of the time derivatives of the virtual control actions. A dead zone modification, turning off the adaptation law in the region where the convergence of the Lyapunov function cannot be guaranteed, is used in order to avoid the problem of parameters drift. The design and analysis of backstepping control using the adaptive approximation approach applied to the double rotor system is presented, considering that not all states are measured. Therefore fuzzy systems are used to approximate the unknown nonlinear functions and a fuzzy state observer is constructed to estimate unmeasured states. The effectiveness of the controller design is investigated through numerical simulations, emphasizing the reduction in the number of parameters to be estimated.