سال انتشار: ۱۳۹۰
محل انتشار: کنفرانس بین المللی فرآورش پلیمرها
تعداد صفحات: ۴
Maryam Sadi – Research Institute of Petroleum Industry, Process Engineering Development Division
Abbas Shahrabadi – Research Institute of Petroleum Industry, EOR Research Department, Tehran, Iran
In this study, an adaptive neuro fuzzy inference system (ANFIS) has been developed to determine the solubility of nitrogen in polypropylene, high density polyethylene, polystyrene, polyvinyl acetate, low density polyethylene and toluene-polyvinyl acetate mixture. The model investigated the effects of the wide ranges of temperature and pressure on the nitrogen solubility in polymers. At the first, to identify the structure and parameters of neuro-fuzzy modeling technique, the optimum number and shape of membership functions were determined.ANFIS model designed in this paper was based on the partitioning algorithm with temperature and pressure as input parameters and solubility of nitrogen in several polymers as output. After adjusting membership functions, developing the structure and training the neuro-fuzzy system, model predictions were compared to experimental data that extracted from literatures. The ability and accuracy of developed neuro fuzzy model to predict the nitrogen solubility in polymers, was evaluated by calculation of statistical parameters such as R-square and mean relative error. The obtained results show that ANFIS can predict the nitrogen solubility in different polymers by high accuracy and there are good agreement between modeling results and experimental data.