A simple algorithm for the time‐optimal control of chemical processes

Abstract

A simple algorithm for the time‐optimal control of chemical processes during setpoint changes, in processes which can be described by a second‐order lag plus dead time model, is described. Knowledge of the unsteady state model parameters is not required because the algorithm uses a dimensionless phase plane on which the switching curves are independent of system parameters for a given forcing function. The algorithm gives the parameters of a second‐order lag plus dead time model as a by product of the setpoint change. It is easily tuned and is relatively insensitive to changes in the process dynamics. The algorithm does not require a large computer or long computing times and has been implemented on both analog and digital computers in controlling computer simulated systems.

Time-optimal startup control algorithm for batch processes

“…Model-Independent Time-Optimal Control Beard (1971Beard ( , 1974 developed a time-optimal control algorithm that requires that the system be initially at steady state, that the steady-state process gain, K, be known, and that the forcing values, Ux and U2, be specified. However, a priori knowledge of the system dynamic parameters, tx and r2, is not required.…”