Prediction of density and kinematic viscosity of biodiesel by artificial neural networks

Environmental pollution is one of the biggest issues all over the world. For this reason, researchers try to find alternative fuels for diesel engines, and biodiesel is the most profitable alternate fuel for diesel engines. In this study, biodiesel produced from cotton oil was used. The produced cotton oil biodiesel was mixed with diesel fuel at volumetric fraction of 20, 30, 40, 50, and 75%. Viscosity and density values at different temperatures for each fuel and blends were determined experimentally. Then, artificial neural network technique was used to predict viscosity and density. In this way, temperature and blend ratio were used as input for prediction of fuel properties. To train network, 85% of total data were used, and the remaining 15% of data were used to test prediction performance of structure. Results were compared with linear regression modelling. As a result, artificial neural network gave more accurate results than linear regression and can be suggested as good a prediction method.