Flow State Characterization of Horizontal Oil-gas-water Three-phase flow Using Independent Slow and Steady Feature Analysis
Linghan Li, Shumei Zhang, Feng Dong · 2022 IEEE International Instrumentation and Measurement Technology Conference (I2MTC) · 2022
To obtain gas-liquid and oil-water relationship from coupling information and characterize flow state of horizontal oil-gas-water three-phase flow, an improved blind source separation method based on independent slow and steady feature (IS2FA) analysis is developed. The pulse wave ultrasonic Doppler (PWUD) sensor is used to obtain information of gas-liquid interface, and the_resistance array sensor is used to obtain information of conductive phase. On the basis of maximizing non-Gaussianity, the temporal constraints are imposed on the optimization objective to extract temporal features since oil and gas will cause different changing speeds of signals, aiming at further clearly characterizing the oil phase and gas phase. The algorithm no longer relies solely on non-Gaussianity but also utilizes dynamics of different phase to extract features. The evolution and transition of the flow process are studied from the perspective of the general trend of interface fluctuation and phase separation information. The efficacy of the proposed algorithm is illustrated by analysis on typical oil-gas-water three-phase flow state and transition states from water-based dispersed plug flow to water-based dispersed slug flow.