科氏流量计相位差估计的ap-Hilbert法

Ap-Hilbert method for phase difference estimation of Coriolis mass flowmeter

相位差估计是科氏流量计精确计量的基础,现有方法计算量较大,估计精度受频谱泄漏影响,且需要预知信号频率,难以满足复杂流量环境的测量需求。为此,提出一种科氏流量计相位差估计的ap-Hilbert法,首先对2路传感器信号进行全相位预处理,即对采样信号循环分段后移位对齐相加得到全相位序列,以消除频谱泄漏导致的端点效应;然后,对全相位序列进行Hilbert变换获得其解析信号,分别提取2路解析信号的实信号和虚信号部分;再通过三角变换,利用三角函数和差公式即可求出相位差。实验结果表明,所提方法克服了Hilbert变换的端点效应,与现有方法相比较,较大提高了相位差估计精度。

Phase difference estimation is the basis of accurate measurement of Coriolis flowmeter. The existing phase difference estimation methods have a large amount of calculation, the estimation accuracy is affected by spectrum leakage, and the signal frequency needs to be predicted. So these methods are difficult to meet the measurement requirements of complex flow environment. To address these issues, an ap-Hilbert method is proposed for phase difference estimation of Coriolis flowmeter. Firstly, all-phase pre-processing is performed on the two-channel sensor signals to eliminate the endpoint effect caused by spectrum leakage, i.e. all-phase sequence is obtained by adding the shifts and alignments of the sampled signals after cyclic segmentation. Then, Hilbert transform is used to analyze all-phase sequence. The real signal and the virtual signal of the two analytic signals are extracted, respectively. Finally, the phase difference can be calculated by using the tangent difference formula of the trigonometric function through the trigonometric transformation. Experimental results show that the proposed method overcomes the end effect of Hilbert transform and greatly improves the phase difference estimation accuracy compared with the existing methods.