Neural network modelling of a depollution process

In this paper an artificial neural network is developed to model a new depo llution process that uses sequential cultures of anaerobic bacteria and yea sts to efficiently remove both carbon and nitrogen from wastewaters. A set of batch experimental runs are used to train and test various neural networ k topologies. It is shown that the neural network accurately tracks the dyn amics of the biological species of the yeast reactor in the process and acc ount for the influence of butyric acid, ammonia and pH on the overall effic iency of purification.