On the infinite-dimensional model multiple-parameter estimation using feedback-relay identification test

Abstract

The objective of this contribution is twofold. First, it demonstrates a case study on applying the standard single-run relay-feedback parameter identification test to a representative of infinite-dimensional systems. Namely, a delayed mathematical model of a circuit heating laboratory appliance process is used. Second, an initial estimation of the model parameters is done via the parameter identification of another – simpler – model. The transition between these two models adopts the idea of dominant spectrum assignment that is solved by using the well-established Levenberg-Marquardt algorithm. Finally, the remaining model parameters are estimated by solving another nonlinear optimization problem in the frequency domains. As transfer function denominator parameters are set independently to the numerator ones, the proposed technique significantly reduces the number of additional relay experiments. Numerical results indicate that the method needs improvements regarding time-response as well as frequency-response accuracy. © 2021, The Author(s), under exclusive license to Springer Nature Switzerland AG.