Artificial neural network model for predicting production of Spirulina platensis in outdoor culture

Process variables contributing to describe the growth of Spirulina platensis in outdoor cultures were evaluated. Mathematical models of the process using inputs which were simple and easy to collect in any operating plant were developed. Multiple linear regression (MLR) and artificial neural network (ANN) modelling procedures were evaluated. The dataset contributing to the growth prediction model were biomass concentration, nitrate concentration, pH and dissolved oxygen concentration of culture fluid, light intensity and days in culture, measured once a day. Datasets of 12 days were sufficient to develop a model to predict the succeeding day’s biomass concentration with a coefficient of determination of greater than 0.98 under outdoor growth conditions. Insufficient number of datasets resulted in overestimation of the predicted output value.