Time irreversibility analysis of two phase flow based on symbolic relative entropy

The relative entropy between the probability distributions of backward and forward series can be used to determine time asymmetry (or irreversibility). In the paper, we combine symbolic time series analysis method with relative entropy theory and propose the symbolic relative entropy method to quantify the time irreversibility of the time series of gas-liquid two-phase flow. By symbolic processing and relative entropy value calculation to the time series of three typical flow patterns, we find that the time irreversibility of three flow pattern signals are significantly different. The results suggest that the irreversibility analysis based on conductance fluctuating signals is an useful tool to characterize the dynamic characteristics of gas-liquid two-phase flow patterns and understand flow pattern transition mechanism.