Application of Resilient Back-Propagation Neural Networks for Generating a Universal Pressure Drop Model in Pipelines
This study aims at generating and validating a universal pressure drop model at pipelines under three-phase flow conditions. There is a pressing need for estimating the pressure drop in pipeline systems using a simple procedure that would eliminate the tedious and yet the non accurate and cumbersome methods. In this study resilient back-propagation Artificial Neural Network technique will be utilized as a powerful modeling tool to establish the complex relationship between input parameters and the pressure drop in pipeline systems under wide range of angles of inclination. A total number of data points consists of 335 sets has been used for generating, validating, and testing the ANN model. A model performance has been evaluated against the best empirical correlations and mechanistic models (Xiao et al., Gomez et al., and Beggs and Brill). A series of statistical and graphical analysis were conducted to show the significance of the generated model. The new developed model outperforms all investigated models with correlation coefficient reaches 98.82%.
Keywords: Artificial Neural Networks; resilient back-propagation; Multiphase flow; pressure drop; universal model.
- There are currently no refbacks.
Copyright (c) 2015 Khartoum University Engineering Journal