基于小波包能量特征的气液二相流流型识别方法

Identification method of gas-liquid two-phase flow pattern based on wavelet packet energy feature

论述了小波包分解及其能量谱处理压差信号的原理与方法,根据小波包变换能将信号按任意时频分辨率分解到不同频段的特性,提出小波包能量特征的概念及算法,并对水平管内空气-水二相流的压差信号进行特征提取,得到各流型的小波包能量特征,然后与BP神经网络相结合,提出一种新的流型识别方法,并用实验数据验证了该方法的正确性和有效性。

The theory and method of wavelet packet decomposition and its energy spectrum used for dealing with differential pressure fluctuation signals were presented. The concept and algorithm of wavelet packet energy feature were put forward, for the wavelet packet transform has the characteristic of decomposing signals to any frequency bands. At the same time, the features of differential pressure fluctuation signals were extracted and the wavelet packet energy features of various flow patterns were obtained. By combining wavelet packet energy features with BP network, a new flow pattern identification method of two-phase flow was proposed. The method was demonstrated by experimental data.