Frequency-Based Estimation of the B-H Curve of Steel Strips in a Continuous Induction Furnace*

Abstract

A parameter estimation strategy for continuous inductive heating of ferromagnetic steel strips is developed and investigated. During strip processing in an induction furnace, the parameters of the magnetization curve (B-H curve) of the strip material are identified by a moving horizon estimator (MHE). This estimator uses measurements of the resonance frequency of the electric circuit. A tailored 2D harmonic model of the electromagnetic subsystem is used to calculate the complex coil impedance. Other model parameters are identified in a scenario, where the magnetization curve of the strip is known. The approach does not require a thermal model of the induction furnace in favor of the computing time and the achieved accuracy. For model validation, the simulated resonance frequency is compared with measurements. To avoid a high computational effort, the effective magnetization approach is applied to approximate the solution of the nonlinear Maxwell equations. The effective B-H curve is mathematically expressed in terms of a linear and an exponential function and parametrized by two parameters. The MHE is validated in a test scenario using a strip with a known magnetization curve.