Feedforward control design for nonlinear stable minimum phase systems using a new state space model structure

This paper presents a new state space model structure for nonlinear systems together with an algorithm to identify its parameters, and an off-line feedforward control design method based on feedback linearization. The model structure combines a linear state space model and a feature space transformation and is therefore particularly suited for nonlinear systems with dominant linear characteristics. The identification algorithm is an iterative two-step procedure alternating the estimation of the linear and the nonlinear model part. The design of a feedforward signal to track a given output reference is based on feedback linearization, and requires that the system is single-input-single-output, stable and minimum phase. A compact implementation of the feedforward signal calculation is possible due to the special characteristics of the nonlinear model structure. Simulation results for a sixth order nonlinear single-input-single-output system are presented to illustrate the effectiveness of the model structure and feedforward control design.