Associate Professor, Mechanical Engineering Department, Faculty of Engineering, Assiut University, 71516, EGYPT
Abstract
This paper investigates how SISO nonlinear systems can be adaptively identified using fuzzy systems which are independent of human knowledge. The proposed methodology uses the on-line data to build up the fuzzy system which approximate the nonlinear dynamics. After filtering the input, the nonlinear system is approximated by a set of fuzzy rules that describes the local linear systems. The Lyapunov direct method is utilized to derive the adaptive law of the proposed identification procedure. Theoretical results are simulated on a one-link robot. Results show that the proposed on-line identifier can consistently track mechanical friction and pay-load variations.
View on SCiNiTO
Top