Optimization of bioethanol production from sorghum grains using artificial neural networks integrated with ant colony

In this study, an artificial neural networks (ANN) model is developed to investigate the relationship between bioethanol production and the operating parameters of enzymatic hydrolysis and fermentation processes. The operating parameters of the hydrolysis process which influence the reducing sugar concentration are the substrate loading, α-amylase concentration, amyloglucosidase concentration and strokes speed. The operating parameters of the fermentation process which influence the ethanol concentration are the yeast concentration, reaction temperature and agitation speed. The desirability function of the model is integrated with ant colony optimization (ACO) in order to determine the optimum operating parameters which will maximize reducing sugar and ethanol concentrations. The optimum substrate loading, α-amylase concentration, amyloglucosidase concentration and strokes speed is determined to be 20% (w/v), 109.5 U/g, 36 U/mL and 50 spm, respectively. The reducing sugar obtained at these optimum conditions is 175.94 g/L, which is close to the average value from experiments (174.29 g/L). The optimum yeast concentration, reaction temperature and agitation speed is found to be 1.3 g/L, 35.6 °C and 181 rpm, respectively. The ethanol concentration obtained from the fermentation of sorghum starch by Saccharomyces cerevisiae yeast at these optimum conditions is 82.11 g/L, which is in good agreement with the average value from experiments (81.52 g/L). Based on the results, it can be concluded that the model developed in this study model is an effective method to optimize bioethanol production, and it reduces the cost, time and effort associated with experimental techniques.