Modeling of wax deposition produced in the pipelines using PSO-ANFIS approach. (18th October 2017)

• Record Type: Journal Article
• Authors: Chu, Zheng-Qing; Sasanipour, Jafar; Saeedi, Mohammadhossein; Baghban, Alireza; Mansoori, Hamed
• Is Part Of: Petroleum science and technology. Volume 35:Number 20(2017)
• Page Start: 1974
• Page End: 1981
• Publication Date: 2017-10-18
• Subjects: wax deposition -- pipelines -- ANFIS -- PSO algorithm -- statistical analysis; Liquid fuels -- Periodicals; Petroleum -- Periodicals; 665.505
• DOI: 10.1080/10916466.2017.1374405
• Languages: English
• ISSNs: 1091-6466

Abstract

ABSTRACT: Wax deposition in petroleum industry is one of the major problems requiring accurate predictive procedures to reduce the deficiencies and effective designing of the process. An adaptive neuro fuzzy inference system (ANFIS) model is proposed to predict the wax deposition in oily systems. Parameters of the ANFIS model are optimized using the particle swarm optimization (PSO) method. Results are then compared to those previously reported by Kamari et al., demonstrating better performance of the proposed ANFIS model. Statistical and graphical approaches are employed to investigate the reliability of the proposed model, illustrating the model's capability of precise estimation of the wax deposition. Coefficient of determination (R 2 ) and mean squared error (MSE) values of 0.994 and 0.053 are obtained for the proposed ANFIS model, revealing the reliable prediction of wax deposition by the proposed ANFIS model.