سال انتشار: ۱۳۸۶

محل انتشار: پنجمین کنگره بین المللی مهندسی شیمی

تعداد صفحات: ۸

نویسنده(ها):

Zahedi – Simulation and AI Research Center, Department of Chemical Engineering, Razi University , Kermanshah, Iran
fatehifar – Department of Chemical Engineering, Sahand University of Technology, Tabriz, Iran

چکیده:

In this article artificial neural network modeling of hydrotreater plant has been the subject of study. In this case a typical refinery data has been collected. Data has been processed and mined and worse data had been disrgarded in modeling. 70% of data set have been used for traing of the network. The best network which provides the minimum error has been adopted for identification. Some parameters like number of hidden neurons, learning rate and momentum rate has been depicted for best network. Input of the network has been feed sulfur content,API , boiling poit. Outlet of network are sulfur content, light hydrocarbon yield and heavy hydrocarbon yields. The obtained network has been checked with 30% of the un_seen data. Excellent agreement between model evaluation and these data was observed. The obtained model is fast responding and easy to implement and can be used for optimization, control and specially for training staffs purposes. High accuracy is another features of the model