A self-adapting time-optimal control algorithm for second order processes

A method for performing time-optimal control with only limited prior knowledge of specific process dynamics is described. The time constants of the process are identified during the transient period accompanying the desired set point change, and the valve switching times of the control algorithm are calculated based on these time constants. The adaptive features of this method make its use independent of process or operating variation. The effectiveness of the proposed control method was evaluated using digital simulation and experimental techniques. The experimental phase which involved random changes to the temperature set point of a batch reactor was carried out with the aid of an on-line digital computer.