Prediction of Surface Tension of Organic Liquids Using Artificial Neural Networks

摘要

The theoretical prediction of the surface tension of organic compounds is required in many chemical engineering calculations.In view of this and due to lack of predictability of surface tension by available theoretical models,the capability of artificial neural network is tested for this purpose.A forward-feed back propagation neural network,based on the Levenberg-Marquardt optimization and gradient descent with momentum weight and bias method was used.The input parameters,e.g.,density,refractive index and parachor,to the neural network were chosen from the previous studies on theoretical prediction of surface tension.The trained neural network predicted the surface tension of various polar,non-polar,saturated and unsaturated organic compounds at 20 deg C with a reasonable accuracy for the training data set (overall absolute percent deviation ([AAD%] 0.31) and for the test compounds (%AAD 3.24).These predictions are an improvement over various predictions from presently used models for the same set of compounds.