Nonlinear Adaptive Inverse Control Synthesis Based on RLS Volterra Model

In this paper, a new methodology of nonlinear Direct Adaptive Inverse Control (DAIC) based on Volterra model is proposed. Using the model based on Volterra series, it is possible to track the plant inverse dynamic through a structure composed of linear and nonlinear terms. The proposed formulations for the control methodology were developed for a Volterra series truncated up to the second order kernel. The update of the estimate of the weights vector of Volterra model was performed by the Recursive Least Square (RLS) algorithm. Since the Volterra model is able to represent nonlinear dynamics of polynomial type, the evaluation of the proposed control methodology was performed on a system described by a Nonlinear AutoRegressive with eXogenous inputs (NARX) model.