Phasor approach to continuous-time system identification

In this work, continuous-time systems are considered, and an innovative system identification approach is applied in the frequency domain, where phasor measurements, as provided by frequency analyzers, are used. This algorithm is obtained through a careful analysis of the physics of the measurement situation, following which the method of minimum variance (MV) estimation is iteratively applied in a linear regression framework. This process renders an efficient pseudominimum variance algorithm that is of linear prediction type and is superior to alternative methods. Thus, it is shown that the parameter estimate is a fixed point of a nonlinear mapping associated with the MV calculation, rather than the solution of an optimization problem. The simulation experiments are carefully performed, and the conditions are fully documented.