Trajectory optimization for flat dynamic systems

A number of important chemical engineering processes are operated in a tran sient manner (e.g., batch processes) and cannot be considered to reach a st eady state. Optimizing the operations of such processes requires the soluti on of a dynamic optimization problem, producing time-based trajectories for process variables, A key characteristic of dynamic optimization problems i s that the process model contains differential equations. Numerical solutio n techniques, which are currently in widespread use, are usually based on d iscretization schemes and can be computationally expensive. This paper prop oses an alternative method for solving dynamic optimization problems in whi ch the nonlinear process model is flat. The approach exploits, as appropria te, either the differential flatness or the orbital flatness of the process model to explicitly eliminate the differential equations from the optimiza tion problem. The resulting optimization problem is solely algebraically co nstrained and can be solved using readily available optimization codes. The proposed approach is demonstrated on a range of benchmark problems taken f rom the literature.