NSF Award
9761057
*** 9761057 Serio This Small Business Innovation Research Phase I project will demonstrate the feasibility of using Fourier Transform Infrared (FT-IR) spectroscopy measurements and artificial neural networks (ANNS) for adaptive control of chemical processes. Three recent events together provide an opportunity to develop a new generation of process monitoring and control systems: 1) the availability of low cost, high speed computation with personal computers; 2) the availability of user friendly software involving artificial intelligence techniques such as neural networks; 3) the availability of rugged, compact FT-IR spectrometers. The capabilities of FT-IR spectrometers to monitor multiple components of a gas phase process stream have been demonstrated by Advanced Fuel Research, .Inc. (AFR) and other researchers in numerous field demonstrations. However, the marriage of this technology with adaptive process control software such as ANNs to produce a superior process control algorithm remains to be proven. The objective of the Phase I program is to demonstrate the proof-of-concept of this approach for a model chemical process.. The Phase II program will concentrate on refinement and generalization of the methodology so that it will work in a full-scale plant as part of a distributed control system. The successful completion of this program would result in process control methodology for maximizing conversion, selectivity and/or environmental acceptability of many common chemical processes which involve gas phase intermediates and products. This technology would provide both economic and environmental benefits to the nation, and potential installations exist at thousands of sites in the U.S. and abroad. ***
