Method for mutual correlation phase difference measurement based on Hilbert transform

Abstract

The invention relates to a method for mutual correlation phase difference measurement based on the Hilbert transform and the field of signal processing, in particular to a method for phase difference measurement. The method is suitable for a plurality of signal segments with the same center frequency. The method comprises the following steps that firstly, three signal segments are randomly selected from the signal segments with the same center frequency, wherein the three signal segments comprise two signal segments containing the phase difference to be calculated and one reference signal segment; secondly, the Hilbert transform is conducted on the reference signal segment or 90-degree phase shift is conducted on the reference signal segment so that a second reference signal segment can be obtained; thirdly, mutual correlation is conducted on the two reference signal segments and the two signal segments with the phase difference to be calculated so that four kinds of mutual correlation relationships can be obtained; finally, a phase difference value of tested signals can obtained according to the four kinds of mutual correlation relationships. According to the method for phase difference measurement, the phase difference of the signals can be accurately measured under the condition that the center frequency of the signals is unknown, and the accuracy and the robustness are good.

Description

technical field [0001] The invention relates to the field of signal processing, in particular to a method for measuring cross-correlation phase difference based on Hilbert transform. Background technique [0002] Phase difference is a very important parameter in the analysis and research of sinusoidal steady-state circuits, and it is an important index for comparing the phase relationship between two sinusoidal signals of the same frequency. With the development of science and technology, the measurement of phase difference has extensive and important applications in the fields of electronic measurement, communication, instrumentation, equipment fault diagnosis, radar, automatic control, navigation and positioning, such as the detection of fluid velocity and flow, underwater Positioning of targets, measurement of power factor of power system, detection of mechanical and physical parameters of shallow seabed, flow measurement of Coriolis mass flowmeter, detection of phase-sen...

AI Technical Summary

Problems solved by technology

[0005] (2) Hilbert transform method (reference [2]: Manuel Varlet, MichaelJ.Richardson. Computation of continuous relative phase and modulation of frequency of human movement [J]. Journal of Biomechanics, 2011, 44 (2): 1200-1204. Reference [3]: Zhang Tao, Lin Chengyan, Zhang Xianguo, etc. High-resolution processing method of well logging curve based on Hilbert transform [J]. Journal of China University of Petroleum (Natural Science Edition), 2012, 36( 1): 68-72.): The Hilbert transform method can reflect the phase change of the whole process, so it has a good application prospect in the phase dynamic measurement, but the instantaneous phase measured by this method has a sudden change, which leads to the existence of phase measurement In addition, this method has high requirements on the storage capacity and real-time performance of the computer, which limits the wide application of the Hilbert transform dynamic phase measurement method.

[0006] (3) Correlation method (reference [4]: ​​Weng Lijing, Zhang Xiyun. Phase difference calculation method based on DSP and signal correlation analysis [J]. Information Technology Research, 2010, 36(6): 22-25. Reference [ 5]: Huang Youpeng, Cao Wanhua, Li Lin, Zhao Xiaolan. A Robust Correlation Algorithm for Infrared Sensors and Passive Radar Tracks [J]. Journal of Instrumentation, 2012, 33(11): 2629-2634. Reference [6]: Jing Li, Zhang Huijuan. Detection and location of copy tampering in image area based on phase correlation [J]. Computer Applications, 2013, 33(3): 660-662, 687.): Correlation method for interference such as DC offset and noise in the signal It has a strong suppression ability, and its error is mainly due to the fact that the measured signal and its own noise are not completely irrelevant, coupled with the strict requirement of full-cycle sampling, the accuracy of its phase difference measurement needs to be improved

[0007] (4) Discrete Burier transform method (Reference [7]: Guo Xiangyu, Fei Shumin. A method for detecting cloth defects based on phase transformation [J]. Industrial Control Computer, 2012, 25(6): 64-65 , 69. Reference [8]: Wang Zhengjun, Pan Lei, Ji Guorui. Frequency measurement method based on discrete phase difference detection [J]. Journal of Instrumentation, 2013, 34(5): 994-1000.): Discrete Burier The transformation method is to transform the signal into the frequency domain for phase processing. Since the measured signal is a finite-length sequence and the sampling frequency cannot be guaranteed to be an integer multiple of the signal frequency, there are spectrum leakage and fence effects, resulting in a large measurement error of the phase difference.

[0008] (5) Least squares method (reference [9]: Yin Aihan, Zhang Qingmiao, Jiang Huiming. Research on an improved weighted least squares phase unwrapping method [J]. Computer Engineering and Application, 2013, 49(3): 159-162 .): The basic idea of ​​this method is to use the mathematical model of curve fitting to solve the phase. Its essence is to minimize the overall deviation between the gradient of the phase and the gradient of the unfolded phase, but the iterative algorithm based on the least squares method has a slow convergence speed. Even convergence cannot be guaranteed in some cases, and its ability to deal with high-frequency noise is poor, thus limiting its scope of use

The Hilbert transform method can measure the instantaneous phase of the signal, but the instantaneous phase measured by this method has a sudden change, which leads to a large error in the phase difference measurement, and the "Gibbs" effect makes the phase difference between the two ends of the signal The measurement is not accurate enough, which affects the stability of the method; the correlation method has a good noise suppression ability, but the accuracy needs to be improved in the case of non-full-period sampling

The discrete Fourier transform method is to transform the signal into the frequency domain for phase processing. Since the measured signal is a finite-length sequence and the sampling frequency cannot be guaranteed to be an integer multiple of the signal frequency, there are spectrum leakage and fence effects, resulting in phase difference measurement errors. bigger

Images