中文摘要：

相位差测量是科氏流量计信号处理的核心内容。针对实际应用中信号缓变特性，基于SVD降噪和Hilbert变换，提出一种科氏流量计信号相位差高精度测量方法。首先对信号重叠分段，利用SVD对近似平稳的各信号段降噪并重构原信号，然后由重构信号与其Hilbert变换经三角运算得相位差的时间函数，从而测得相位差。阐述了方法思路与原理，给出了实现步骤和流程，仿真分析了方法收敛特性、测量精度、初始相位差影响及动态测量性能，现场实验验证了方法有效性。结果表明：与现有方法相比，该方法没有迭代运算，不存在收敛过程，无需预知信号频率，且测量精度高，能检测出相位差微小变化。

英文摘要：

Phase difference measurement is the core of Coriolis mass flowmeter （CMF） signal processing. Based on singular value decomposition （SVD） and Hilbert transform, aiming at the signal slow varying characteristic of CMF in practical application, a high-precision phase difference measurement method is proposed. In this method, the time-varying signal is firstly divided into overlapped segments, which can be treated as approximate stationary signals. Noise-reduction is performed on the segments and the original signal is reconstructed from these segments with the help of SVD algorithm. Then, the phase difference time function is obtained from the reconstructed signal and its Hilbert transform with trigonometric operation; therefore, the phase difference is figured out. The idea and principle of the proposed method are expounded, and the steps and flowchart of the method are given. The convergence, accuracy, initial phase difference effect and dynamic measuring performance are analyzed with simulations and the practicality of the method is validated with experiments. Results show that compared with existing algorithms, there is no convergence process in the proposed method, and the method does not require iteration operation and does not need to know the signal frequency; besides the method has high measurement accuracy and can detect tiny phase difference change.