Output tracking between operating points for nonlinear processes: Van de Vusse example

Inversion-based approach finds an input that exactly tracks a desired output trajectory - the method has been recently extended to nonlinear nonminimum phase systems using stable-inversion techniques. Exact-output tracking inputs can be found by the stable-inversion method provided the system can be maintained near (local to) an equilibrium point of the system. However, inversion for the nonlocal case is more challenging because inverse solutions may not exist, for example, when the equilibrium point (operating point) is changed. Such changes in operating points are necessary in process control applications, for example, during startups, shutdowns, and grade transitions. Nonlocal transfers of a system from one operating point to another operating point along pre-specified output trajectories are studied in this article for the Van de Vusse example, which is a benchmark problem in nonlinear, nonminimum-phase process-control.