Model Identification of Gas-Fired Industrial Furnace

Temperature control is a challenging task because it possesses high nonlinearity, complexity and dead time. Most of temperature process control requires a model which describes the behavior of the control system accuracy to construct a model-based controller for tracking a predefined temperature-time curve. This paper presents the identification of a gas-fired furnace to achieve a model using the experimental data. The identification model is obtained by two approaches: the bilinear model and first order plus dead time. Several experiment tests are conducted on the gas-fired industrial furnace with different steady-state inputs to reach the steady-state temperature output. The model parameters are estimated based on three experimental tests, while the other serves as the form of validation. The accuracy of identification models is validated through a comparison to the experimental tests on the gas-fired industrial furnace, with the objective of minimizing the mean square error.