【作者】 郭盈； 李永光； 李崇详； 陈亮； 马立新；

【Author】 Guo Ying Li Yong-guang Li Chong-xiang Chen Liang Ma Li-xin (Xi’an Jiaotong University ,Xi’an 710049,China Shanghai University of Electrical Power, Shanghai 200090,China)

【机构】 西安交通大学； 上海电力学院；

【摘要】 通过试验,获取了大量气液两相绕流中漩涡脱落物体的信号数据,对这些信号数据进行三层小波分解,并提取第三层小波分解的8个重构信号的能量作为人工神经网络的输入特征向量,训练了人工BP神经网络并应用该网络实现了气液两相绕流中含气率大小的准确定位。试验表明:小波分解的特征提取技术和人工神经网络的模式识别技术的联合应用可以作为测量两相流组分的一种新方法。更多还原

【Abstract】 Through experiments of gas-liquid two-phase flow wavelet decomposition artificial neural networkliquid two-phase flow we obtained a lot of signal data of the detachment of vortex from object, and decomposed the signal data to the 3rd layer by wavelet function, and extracted the 8 recomposed signal. Afterward, we designed a BP neural network and trained it by the power feature vector of the 8 recomposed signal and used the network to identify the void fraction in the gas-liquid two-phase flow precisely. The experiment shows: the technique of feature extraction based on the wavelet decomposition and the technique of pattern recognition based on the artificial neural network can be combined as a new method which can be used to identify the components in gas-liquid two-phase flow.更多还原