In this study, the growth of the indigenous Acidithiobacillus thiooxidans was predicted using artificial neural network (ANN). Four important variables of the growth medium: KH 2 PO 4 , (NH 4 ) 2 SO 4 , MgSO 4 , and elemental sulfur (S 0 ) were fed as input into the ANN model, while the dry cell weight (DCW) was the output. The ANN model adopted in this study, consisting of an input layer, a hidden layer, and an output layer, was found to give satisfactory results. Among different combinations of 10 mostly used transfer functions, Gaussian and Sigmoid transfer functions were selected for the hidden and the output layers, respectively, to minimize the error between the experimental results and the estimated outputs. Experimental data were randomly separated into a training set and a test set with 22 and 8 experimental runs, respectively. The resulting ANN shows satisfactory prediction of the DCW with R 2 = 0.991 and mean relative deviation (RD) = 0.026. The optimal medium composition of the indigenous A. thiooxidans was further predicted to be KH 2 PO 4 = 1.0 g/l, (NH 4 ) 2 SO 4 = 3.5 g/l, MgSO 4 = 0.65 g/l, and S 0 = 23 g/l with the optimal DCW being 0.722 g/l. The results of this study suggest that ANN provides a powerful tool in studying the nonlinear and time-variant biological processes.
ASJC Scopus subject areas
- Environmental Chemistry
- Chemical Engineering(all)
- Industrial and Manufacturing Engineering