Artificial Neural Network Prediction of Sulfur Content of Diesel fuel from its Physical Properties

2nd International Conference on Sustainable Engineering Techniques

工业技术（总论）

Younis, Younis Muhsin^1 ; Kayi, Hakan^2

Material Engineering Department, Technical Collage Baghdad, Middle Technical University, Baghdad, Iraq^1

Ankara University, Chemical Engineering Department, Ankara

06100, Turkey^2

关键词: ANN modeling;    Correlation coefficient;    Engineering calculation;    Experimental analysis;    Levenberg-Marquardt training algorithm;    Matlab- software;    Power station;    Sulfur contents;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/518/6/062008/pdf
DOI  :  10.1088/1757-899X/518/6/062008

学科分类：工业工程学

来源: IOP

【 摘 要 】

The sulfur content is important in engineering calculations, so this study has two major purposes. The first purpose of the study is to predict the sulfur content from its physical properties by using artificial neural network to decrease time and cost spent on experimental analysis of sulfur content, and the second purpose is to find the simplest formula to predict the sulfur content. Artificial Neural Network is applied as a black-box type modelling for sulfur content prediction of diesel fuel. The experimental data used in this study is obtained from Erbil power station. In this study, the Levenberg-Marquardt training algorithm is used to train the neural network and to predict the sulfur content. It was observed that the ANN model can predict the sulfur content of diesel quite well with correlation coefficient (R) 0.9813. The prediction Mean Square Error was between the targets values and the outputs values were obtained 0.000339 by the matlab software. The findings obtained in this study indicated that the designed neural network performs quite well in the prediction of sulfur content of diesel fuel from its physical properties.