高含气率气液两相流差压信号时频特征分析  

Time-frequency Feature Analysis on Differential Pressure Signals of Gas-liquid Two-phase Flow with High Void Fraction

作　　者：石岗[1] 耿艳峰[1] 

机构地区：[1]中国石油大学信息与控制工程学院,山东东营257061

出　　处：《化工自动化及仪表》2010年第7期52-55,共4页Control and Instruments in Chemical Industry

摘　　要：为了更有效地揭示高含气率气液两相流流动特征,研究了一种新的时频特征分析方法。首先,对差压波动信号进行小波包变换并由变换系数计算信号能量的时频分布;然后,应用统计方法对时频分布进行特征提取得到一组时频特征量。应用类可分性测量准则分析该组特征量区分不同流型的效果,并与以往小波包特征分析方法相比较,结果表明:该组特征量具有更强的流型特征表征能力。最后,以该组特征量为输入向量,构建了集成多类支持向量机分类器实现了流型识别,其流型正确识别率可达97%。In order to characterize the gas-liquid two-phase flow,the wavelet packet transform was applied to analyze differential pressure signals of the slotted-orifice.Firstly,the time-frequency distribution of signal energy was calculated based on wavelet packet transform coefficients.Then,a new feature vector was extracted from the distribution.Research has showed that this feature vector could descript the character of flow regimes more effectively comparing with the features result in traditional methods.Finally,with the time-frequency features as input,designed a SVM multi-class classifier using the ensemble technology to realize flow regime recognition whose accuracy reach 97%.

关 键 词：气液两相流 时频分析 小波包变换 支持向量机 流型识别 

分 类 号：O359[理学—流体力学]