ICA and Its Application in the Phase Information Extraction of Oil-Gas-Water Three-Phase Flow
Lingfu Kong, Fan Gao, Yingwei Li, Liu Xingbin, Jinhai Hu · 2010
The independent component analysis(ICA)is applied to studying the oil-gas-water three-phase flow noisy signals responded by conductance sensor, and extracting the flow signals which characterized the gas and liquid phase respectively so as to obtain gas/liquid phase information. The algorithm of FastICA which based on kurtosis is applied to separating the field simulation well logging oil-gas-water three-phase signals responded by multi-channel sensors. And then identify the gas/liquid phase signals between the separated signals. Compare and analysis indicate that ICA is effective in obtaining the gas/liquid phase flow characteristics of oil-gas-water three-phase flow, and the extracted independent components have explicit physical meaning corresponded to the gas phase and the liquid phase separated.