Estimation of the model parameters of ferroresonant circuits using jump voltage assignment

A novel approach is presented for the estimation of the model parameters of a ferroresonant circuit. Through the explicit use of readily available dynamic information about the onset and the collapse of the ferroresonant regime, the proposed method leads to models that are in excellent agreement with the measurements taken from corresponding experiments. Linear system identification commonly employs a linear least square criterion to find an estimate of the required parameters. However, simulation of the models thus obtained frequently exhibits a significant discrepancy from the observations made in experiments. The proposed method remedies this inconsistency by prescribing the locations of the two-fold bifurcation points of the circuit. The arising nonlinear constraint optimisation problem is solved using an adaptive simulated annealing algorithm. As an additional benefit, the new method also yields useful estimates for parameters that are otherwise difficult to determine, e.g. the magnetic losses of the ferromagnetic core or the collective resistive losses within the apparatus. The aim is to provide a robust and easy to use methodology for a dynamically consistent estimation of the model parameters of ferroresonant circuits.