On feedback linearization in LMI-based nonlinear MPC

Feedback linearization has proven to be very valuable in nonlinear predictive control. It is, however, only applicable to a small class of systems and it has the drawback of degrating the original performance index. In this paper an approximate feedback linearization technique makes it possible to extend the method to a wider class of nonlinear systems. A recovery of the original performance index is suggested, all in an LMI-based setting. Finally, tuning the feedback linearization parameters can be important to obtain optimal robustness. A simulation example shows the effect of the performance index recovery and different parameter settings