الفهرس 
CHAPTER 1 IntroductionOnly 14 pages are availabe for public view
 |  | from | 170 |  |  |
| | | from | 170 | | |
Abstract
When traditional statistics and physics are unable to satisfactorily solve a complex problem, Artificial Intelligence (AI) techniques offer an alternative. But by comprehending how these AI methods work and deciphering the Blackbox, researchers may readily apply these machine learning approaches to discover empirical correlations without the requirement for coding languages like MATLAB. In order to employ machine learning (ML), which is regarded as an artificial intelligence (AI) technology, experimental data from published works was gathered and refined in order to estimate the Z-factor and pressure DROP resulting from multiphase flow in horizontal pipes. To create the best model that can accurately predict Z-factor and pressure drop, the gathered data was trained and tested. Results obtained by Adaptive neuro fuzzy inference and Artificial neural network system proved the best results among other methods to predict Z-factor with 0.12 and 0.159 average absolute percentage error (AAPE) respectively. Adaptive neuro fuzzy inference and Artificial neural network were able to estimate the pressure DROP in pipes with 3.4, and 3.7 average absolute percentage error respectively.