摘要
油水两相流温度信号受到外界噪声干扰,导致油水特性难以辨别,含油率测量范围小。为解决这一问题,采用一种新的信号处理方法——多小波相邻系数法,与多小波传统阈值、向量阈值去噪算法相比,较好地解决了变换系数之间的相关性,去噪效果较好。通过大量的油水实验数据处理发现:利用热电偶测温信号的电压幅值波动测含油率,测量范围为3%-70%,通过曲线拟合,得到热电偶电压幅值波动与含油率的数学表达式,测量误差为7.1%。
The temperature signal of oil/water two-phase flow is disturbed by noises from external interference which results in difficulty in distinguishing performance of oil/water and a small measuring range for oil holdup. In order to solve the problem, a new signal processing method based on the multiwavelet neighboring coefficient method was used. Compared with traditional threshold de-noising based on multiwavelet and vector threshold de-noising, the neighboring coefficient method can deal with the eorrelativity among the transformation coefficients and ensure better de-noising effect. Processing of a large amount of experiment data of oil/water have revealed that measurement of oil holdup based on voltage amplitude transform of thermoeouple gave a measuring range of 3%- 70% oil holdup. The mathematical expression between voltage ampitude fluctuation and oil holdup was obtained, and the measurement error of oil holdup was 7.1%.
出处
《化工学报》
EI
CAS
CSCD
北大核心
2009年第2期325-331,共7页
CIESC Journal
基金
国家高技术研究发展计划项目(2007AA04Z180)~~
关键词
油水两相流
去噪
热电偶
多小波
相邻系数
含油率
oil/water two-phase flow
de-noising
thermoeouple
multiwavelet
neighboring coefficient oil holdup