Identification of equivalent dynamic model parameters in eddy current problems

Abstract

PurposeThe goal of the paper is to introduce a new method of obtaining equivalent dynamic model of electromagnetic field quantities. Proposed algorithm allows approximation of the frequency and step response by a simple inertial element model, with adjustable rank and delay. The values of the model parameters may be also used to describe the dynamics of considered system.Design/methodology/approachThe dynamics of interesting field quantity in certain space location may be represented by an equivalent model of inertial element. Parameters of the model are identified using the solution of the problem in quasi‐stationary conditions for very limited number of excitation frequencies. These solutions are further used to build a matching approximation of real frequency response function (FRF).FindingsThe proposed method allows fast approximation of transient states of linear vector field. It may be useful with fast and relatively precise estimation of dynamic parameters of the electromagnetic field, e.g. in screening and eddy current problems.Research limitations/implicationsThe main limitation of the method is the assumption of linearity of the problem. However, many practical tasks similar to the examples presented in the paper can be considered highly linear.Practical implicationsThe main advantage of the method is that it allows fast estimation of the field dynamics without either solving the problem for whole range of frequencies or computing the transient state in time domain. It does not need the access to the original mass and stiffness matrices. Therefore, it may be used with commercial FEM software, which usually restricts access to its internal data.Originality/valueThe method is based on well known concept of moments, but the use of existing stationary FEM solutions for approximating transient states is a novel approach. Proposed procedure may be easily automated for the simulation environments with scripting capabilities.