wangwanglianhe/Hybrid_Fuzzy_Kalman_Filter

Mixed Kalman-Fuzzy Sliding Mode State Observer in Disturbance Rejection Control of a Vibrating Smart Structure Atta Oveisi*1, Tamara Nestorović1 ​1Ruhr-Universität Bochum, Mechanik adaptiver Systeme, Institut Computational Engineering, D-44801, Bochum, Germany. E-Mail: atta.oveisi@rub.de ABSTRACT In the controllers that are synthesized on a nominal model of the nonlinear plant, the parametric matched uncertainties and nonlinear/unmodeled dynamics of high order nature can significantly affect the performance of the closed-loop system. In this note, owing to the robust character of the sliding mode observer against modeling perturbations, measurement noise, and unknown disturbances and due to the non-fragile behavior of the Kalman filter against process noise, a mixed Kalman sliding mode state-observer is proposed and later enhanced by the addition of an intelligent fuzzy agent. In light of the proposed technique, the chattering phenomena and the conservative boundary neighboring layer of the high gain sliding mode observer are addressed. Then, a robust active disturbance rejection controller is developed by using static feedback of the estimated states using direct Lyapunov quadratic stability Theorem. The reduced order plant for control design purposes is subjected to some simulated square-integrable disturbances and is assumed to have mismatch uncertainties in system matrices. Finally, the robust performance of the closed-loop scheme with respect to the mentioned perturbation signals and modeling imperfections is tested by implementing the control system on a mechanical vibrating smart cantilever beam. Keywords: Fuzzy system; Nonlinear control; Active disturbance rejection; Kalman Filter; Vibration suppression.

ScilabBSD-2-Clause