摘要
压差波动信号由于其容易测量,而且包含流动的很多信息而被广泛用于气液两相流的研究中。在测量时由于难免受到各种干扰,这些干扰会影响非线性分析的精度和结果。因此,在利用压差波动信号进行非线性分析之前,通常要对其进行滤波处理。究竟多高频率以上的压差波动信号才是噪声,并如何对其进行定量检测的研究就非常重要。这里对水平管内气液两相流压差波动信号用小波包进行分解,得到按16Hz等间隔分段的16个不同频带信号,用自相关函数研究了各频段信号的线性相关性。通过理论证实了气液两相流压差波动信号中,64Hz以上信号不存在线性相关性,可以认为其是噪声信号。研究指出通过气液两相流压差波动信号来研究流型信息时,应考虑64Hz以下信号。
In the study of two-phase flow, differential pressure fluctuation signal is widely used, which is easy to be measured and includes much information of flow. As the differential pressure fluctuation signal is affected by ample noise during it is measured, it should be processed with filter before using it to analyze the nonlinear characteristic of flow. It is very important that how to define whether the signal is noise by its frequency and how to measure the noise quantitatively. The differential pressure fluctuation signal of gas-liquid two-phase in horizontal pipe is decomposed into sixteen different frequency band signals every other 16Hz by wavelet packet. The linear correlation characteristic of every frequency band signal is analyzed by using auto-correlation function. It is proved in theory that the signal above 64Hz has no linear correlation characteristic and the signal could be considered as noise. Result shows that the signal less than 64 Hz is useful in the study of two-phase flow regime.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2005年第6期636-639,共4页
Chinese Journal of Scientific Instrument
基金
吉林省科技发展计划(20040513)资助项目
关键词
两相流
小波包
自相关函数
相关性
压差波动
Two-phase flow Wavelet packet Auto-correlation function Correlation Differential pressure fluctuation
