Process optimization using neural networks

Optimization of chemical processes can result in decreased energy consumption, improved productivity, better product quality, and generally increased profits. Most chemical plants achieve optimal operation by gradually varying process conditions experimentally and exploring a localized feasible region around the current operating point. Major concerns are cost and time involved in attempting to achieve "optimal operation". This paper presents a systematic methodology using neural networks to improve the efficiency of a sequential search process for achieving optimal process operating conditions. An example, developed by Ultramax Corporation, Cincinnati, is presented to illustrate the approach