A Methodology for Analysis and Prediction of Volume Fraction of Two-Phase Flow Using Particle Swarm Optimization and Group Method of Data Handling Neural Network

Determining the volume percentages of flows passing through oil transmission lines is one most essential problems in oil, gas, and petrochemical industries. This article proposes a detecting system made Pyrex-glass pipe between an X-ray tube NaI detector to record photons. geometry was modeled using MCNP version X algorithm. Three liquid-gas two-phase flow regimes named annular, homogeneous, stratified were simulated ranging from 5 95%. Five time characteristics, three frequency five wavelet characteristics extracted signals obtained simulation. radiation-based flowmeters’ accuracy has been improved by PSO choose best case among thirteen characteristics. The proposed feature selection method introduced seven features as combination. void fraction inside could be predicted GMDH neural network, with given inputs network. novel aspect current study application PSO-based calculate percentages, which yields outcomes such following: (1) presenting suitable time, frequency, for calculating percentages; (2) presented accurately components RMSE MSE less than 0.30 0.09, respectively; (3) dramatically reducing amount calculations applied detection system. research shows that simultaneous use well system, can significantly help improve