Wavelet packet entropy feature extraction and characteristics analysis for gas/liquid two-phase flow regimes

Abstract

Identification of flow regime is the basis for measuring flow parameters in two-phase flow accurately. Because of the complexity of phase interaction in gas-liquid two-phase flow, it is difficult to discern its flow regime objectively. The 208 measured data from electrical resistance tomography (ERT) are arranged into series based on section to derive the wavelet packet entropy and investigate the characteristics of gas-liquid two-phase flow for flow regime. The results show that the wavelet packet entropy, which is sensitive to the flow regime transition, can characterize the nonlinear dynamics of gas-liquid two-phase flow. Wavelet packet entropy of flow regime may be as a parameter for better flow regime identification and analysis those wavelet packet entropy also provide an efficient supplementary to reveal the flow regime transition mechanism of gas/liquid two-phase flow.