Data-driven Response Prediction for Systems with Nonlinear Elements

This paper extends a data-driven response prediction method for linear systems into systems with nonlinear elements. By using two sets of measured input and output data, unknown parameters in models of nonlinear elements are tuned so that a residual related to a linear equation for linear systems is minimized. The effectiveness of the proposed method is illustrated through simulation results for some nonlinear elements and experimental results with a DC motor control system.