Abstract
A non-linear state feedback algorithm based on an OFS (orthonormal functional series) model combined with Volterra Functional Series is proposed to solve block-oriented non-linear systems. This algorithm has excellent approximation ability for the variance of system time-delay, and can guarantee the closed-loop robust stability and the zero steady-state error property. In addition, it has lower online optimization computational load when dealing with the hard input constraints in contrast to traditional NMPC (non-linear model predictive control) based on the OFS model. The analyses of the stability and robustness of this algorithm are put forward systemically, and the steady-state performance analysis is also given. Case studies involving simulations on a Wiener-type non-linear system and temperature experiments on a simulation chemical reactor are used to validate the efficiency and superiority of this proposed algorithm.
Get full access to this article
View all access options for this article.
