基于HHT包络谱的油中水分含量检测研究

Study on the detection of the water content in oil based on the envelope spectrum of HHT

为了能准确检测油中水分含量,提出了一种基于希尔伯特-黄变换(HHT)包络谱的检测方法。首先针对含水油液的压力谐振信号是一个调制信号的特点,建立了压力谐振信号的表达式;然后进行经验模态分解,获取信号的各阶本征模态函数(IMF),将与原始信号相关性较大的IMF进行希尔伯特包络谱解调分析,获得包含水分含量的特征信息的解调信号,实验结果表明随着含水率的增加,油液的压力谐振频率的阶次在逐渐下降,不同含水率油液的谐振频率与含水率有一定的对应关系,此方法能够有效地实现油液水分含量的检测,为用油系统的安全运行提供理论支撑。

In order to accurately detect the water content in oil, a new detection approach based on Hilbert-Huang transform （HHT） envelope spectrum is presented. Firstly,according to the characteristic that the pressure resonance signal of the oil containing water is an amplitude-modulation signal, the expression of the pressure resonance signal is established; and the empirical mode decomposition （EMD） is performed to obtain the intrinsic mode functions （IMFs） of the signal. Then, the IMFs having strong correlation with the original signal are analyzed using the Hilbert transform envelope spectrum demodulation, and the accurate demodulation signal that contains the characteristic in- formation of water content in oil is obtained. The experiment results show that as the water content in oil increases, the order of the pressure resonance frequency decreases gradually, there is a corresponding relation between the reso- nance frequency of the oil with different water contents and the water content in oil. This method can effectively achieve the water content detection in oil, and provides a theoretical support for the safe operation of oil system.