中文摘要：

研究了几种典型非线性时间序列的多尺度熵特征,在此基础上分析了由插入式阵列电导传感器采集的144种流动条件下的垂直上升气液两相流电导波动信号.研究结果表明：利用小尺度下样本熵的变化速率特征可以分辨三种典型流型（泡状流、段塞流、混状流）,而大尺度下样本熵的波动特征可以反映各种流型的动力学特性.泡状流随机可变特性表现为大尺度下样本熵的高值及振荡特征;段塞流气塞与液塞的间歇性运动表现为大尺度下样本熵的低值及平稳性;混状流极不稳定的振荡运动特性表现为介于泡状流及段塞流之间的熵值特点,并在更大尺度时熵值逐渐接近泡状流熵值.两相流多尺度熵分析有助于进一步理解流型转化动力学特性,多尺度熵值变化速率特征是流型辨识的新指示器.

英文摘要：

Muhiscale entropy is a new method to analyze nonlinear time series on multiple temporal and spatial scales. Firstly, multiscale entropy characteristics of several typical nonlinear series were studied, and then based on this, the fluctuant conductance signals of 144 two-phase flow conditions were analyzed, which were collected by using array conductance sensors in upward vertical gas-liquid two-phase flow. The results indicated that the changing rate of sample entropy at small scales could be used to classify the three typical flow patterns （bubble flow, slug flow and churn flow）, and the fluctuation of sample entropy of large scales reflected the dynamic characteristics of each flow pattern. The stochastic characteristic of bubble flow was shown as higher and oscillating sample entropy at large scales ; the intermittence of gas slug and liquid slug of slug flow was represented as lower and stable sample entropy of large scales ; the unstable and oscillating characteristics of churn flow behaved as the entropy between that of bubble flow and slug flow, and the entropy closed to that of bubble at larger scales. The multiscale entropy analysis of two-phase flow is helpful for understanding the dynamic characteristics of flow pattern transition, and the rate of muhiscale entropy is a new indicator of flow pattern identification.