Identification method for nonlinear LFR block-oriented models with multiple inputs and outputs

Abstract

Recently, the nonlinear LFR model has been proposed as a candidate model with high potential, due to its surprising flexibility and parsimony. It is a quite general block-oriented model consisting of a static nonlinearity (SNL) and multiple-input-multiple-output (MIMO) dynamics. It can cope with both nonlinear feedforward and nonlinear feedback effects and does not postulate the SNL's location prior to the identification. This contribution extends the model from single-input-single-output (SISO) to MIMO. Starting from two classical frequency response measurements of the system, the method delivers the best possible MIMO dynamics and estimates the SNL in an automated, user-friendly, non-iterative way, with an improved computational efficiency. The method is successfully applied on a numerical simulation example to illustrate the theory.