Robust self tuning controller for electric furnace

In this paper, a continuous time self tuning controller is proposed for a electric furnace system which has large delay and unmodeled dynamics. The real time identification method in control algorithm is the continuous time exponentially weighted least squares method. The detuned model reference methods are adopted for the control algorithm. From simulation, one can conclude that the proposed algorithm is robust to unmodeled dynamics by choosing the control weighting factor and realization filters properly. The proposed controllers are implemented for an electric furnace system. The effectiveness of the proposed control algorithms are shown from experiments