Optimal Control of Distributed Processes Using Restricted Second Order Information

The open loop optimal control of combined lumped and distributed process systems is considered here. A method based on the Control Vector Parameterization approach, which uses exact second order information, is presented for the efficient computation of the control policy that optimizes a specific performance criterion. This approach makes use of first and second order sensitivities to obtain exact gradients and restricted second order information by the solution of an extended Initial Value Problem (IVP). We illustrate the use of this method considering the solution ot several case studies of increasing complexity. All these case studies were successfully solved with a modest computational effort, proving its capability to solve large scale, challenging problems in an efficient way.