Nonlinearity Compensation Based on Identified NARX Polynomials Models.

This paper deals with the compensation of nonlinearities in dynamical systems using Nonlinear polynomial AutoRegressive models with eXogenous inputs (NARX) identified from data. The compensation approach is formulated for static and dynamical contexts for the general case and is also adapted for systems with hysteresis. Both simulated and experimental results are presented to illustrate the method. In the experimental case, the proposed method is compared to other approaches and was found to be competitive. This method yielded a maximum tracking error of $$3.9\%$$ while the corresponding value for the uncompensated system was $$21.0\%$$ . Furthermore, the presented technique typically results in compensation signals with lower energy requirements. The results also show that the proposed methodology can provide compensation signals that practically linearize the systems using simple nonlinear models with very few terms.