System Identification Based on Experimental Technique Using Stability Boundary Locus Method for Linear Fractional Order Systems

Abstract

Fractional calculus is an important mathematical tool that is widely used in control systems. It is established in the literature that fractional order models are more accurate and more effective in system modelling. In this study, an alternative and novel technique is proposed to identify the fractional order time-delayed model of an unknown system. The method is based on obtaining the approximate stability boundary locus (SBL) curve of the unknown system by applying three different experimental tests. Three points on the SBL curve are determined by the experimental tests and then the parameters of the fractional order time-delayed model are computed by solving the nonlinear systems of equation. The system model with double fractional order element plus a time delay is obtained using the proposed method. The proposed method is explained through simulations on a twin rotor system. The proposed method is also used in model order reduction calculation of the higher order transfer functions.