CALIN I. ANGHEL , MIRCEA V. CRISTEA PREDICTION OF THERMODYNAMIC PROPERTIES BY ARTIFICIAL INTELLIGENCE TECHNIQUES
Thermodynamic models are solid foundations for many theoretical investigations of multicomponent systems. Beside of cumbersome theoretical approaches necessary to develop new insights in this context of potential support is the integration of artificial intelligence innovative ideas. The paper implements a novel procedure of the artificial intelligence based on support vector machine in a minimax approach. The main goal of the paper is to compare the performance of the novel procedure with artificial neural networks or other theoretical approaches and to promote it as an effective technique for thermodynamic analyses. Comparative results demonstrate the capability of the proposed procedure to be used by engineers dealing with multicomponent systems analysis but also its potential application in other engineering fields.