This study investigates the prediction of biodegradation of polycyclic aromatic hydrocarbons using a mixture of naphthalene; anthracene and pyrene in a continuously stirred tank reactor by an artificial neural network. Artificial neural networks are relatively crude electronic networks of "neurons" whose operations are based on the neural structure of the brain. They process records one at a time, and "learn" by comparing their prediction of the record (which, at the onset, is largely arbitrary) with the known actual record. Experimental data were employed in the design of the feed forward neural networks for modeling the prediction of biodegradation process. Comparatively, results showed that predictions from the feedforward neural network closely fitted the measured values. The degradation pattern was characterized by an exponential decline in the concentrations of the polycyclic aromatic hydrocarbons, and this was followed by a ‘plateau’ concentration signifying the attainment of endpoint of the degradation process.
Keywords: Model, Neuron, Feed forward, Training, Input, Hidden and Output layers
Journal of the Nigerian Association of Mathematical Physics, Volume 19 (November, 2011), pp 395 – 398