Design of QFT robust controller for fuzzy systems

Abstract

This paper introduces a new robust controller for fuzzy systems. Exact modeling of an industrial plant is not feasible for many processes due to nonlinearity and uncertainty in parameters associated with a plant. Fuzzy modeling is known as a suitable approach for modeling of systems including uncertainties. Fuzzy model of a system is a nonlinear combination of fuzzy rules. This paper exploits a robust controller design approach based on quantitative feedback theory (QFT) to design a controller for a fuzzy system. QFT provides an efficient robust controller design method since it introduces a robust stability analysis approach in the presence of structured and unstructured uncertainties. Employing a QFT approach for nonlinear systems based on a fixed-point theorem provides a method to design a QFT-fuzzy controller elaborated in this paper. To evaluate the QFT-fuzzy method, a fuzzy model is developed for a remotely operated vehicle (ROV). Then, a QFT-fuzzy controller is designed and applied to control the ROV. The results are compared with that of sliding mode controller applied to the KAVOSH. KAVOSH is an ROV, which was designed and implemented in the Submarine R&D Center at Isfahan University of Technology, Iran.