This paper investigates the implementation of start-up policies for an exothermic chemical batch reactor. The study includes the development of a rule-based expert system which will "learn" the "best" trajectory through experience as it controls the reactor. In this way, the start-up time can be minimized while maintaining stable operating conditions despite uncertainties in the kinetics and design parameters. Integrated into the expert system is either an adaptive or a robustly tuned PID algorithm employed for fine-tuning the temperature control under normal operating conditions, which removes the need for re-tuning the controller from batch to batch. The control scheme is tested on a simulation of a batch reactor carrying out two consecutive irreversible reactions with the product of the first reaction being desired.
|Number of pages||6|
|State||Published - 1 Jan 1989|
|Event||1989 IEEE International Conference on Control and Applications, ICCON 1989 - Jerusalem, Israel|
Duration: 3 Apr 1989 → 6 Apr 1989
|Conference||1989 IEEE International Conference on Control and Applications, ICCON 1989|
|Period||3/04/89 → 6/04/89|