Modeling Chemical Engineering Processes Using Artificial Neural Networks
No Thumbnail Available
Date
2005-01
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
In this thesis the application of feed forward type artificial neural networks to model
chemical engineering processes are demonstrated with reference to five different
problems.
Neural network models are constructed and employed to predict vapor-liquid equilibrium
(VLE) data of twelve different binary systems having different chemical structures and
solution types (azeotrope-nonazeotrope) in various conditions (isothermal or isobaric). It
is observed that the data found by neural network model gives an excellent agreement
with the experimental data. In fact the neural network model can be treated as a powerful
means for VLE data prediction in a fast and reliable way.
This study has confirmed the feasibility of using a neural network to capture the
nonlinear and interacting relationships between the moisture content and different drying
conditions of potato. Simulating time series temperature profiles of adiabatic batch
reactor has also investigated. Neural network trained with a limited number of
experimental data were capable of predicting fresh data that were not used to train the
network. The results obtained in using the developed models are physically sound as
expected from experience.
Simulating a human operator controlling a chemical plant is also a good instance where
the advantage of using artificial neural networks is demonstrated in the thesis. This thesis
also describes the use of multilayer feed forward neural networks as a CO2 analyzer. It
was proved that MLP-type network of a relatively simple structure made it possible to
predict the CO2 effluent from a furnace.
Taking in to account the difficulties in experimental conditions, complicated
measurements and unavoidable errors of devices used, limit the precision of laboratory
measurement results. The accuracy of the results generated by the developed neural
network models may be considered satisfactory for engineering calculations.
Description
Keywords
Chemical Engineering