electromagnetic flowmeter supplier for sales

electromagnetic flowmeter supplier for sales

Flow is a very important detection parameter in the field of industrial measurement and control, and the measurement accuracy of phase difference is one of the key factors affecting the accuracy of flow measurement. Firstly, the basic principles of the phase difference measurement methods commonly used in three intelligent electromagnetic flowmeters: spectrum analysis method, correlation analysis method and Hilbert transform method are introduced. For the problem of full-period sampling, a method to suppress non-full-period sampling is proposed, and the specific implementation steps of the improvement measures are described in detail. The simulation analyzes the advantages and disadvantages of various methods, and compares and analyzes the methods before and after the improvement to verify the effectiveness of the improvement measures.
0 Introduction
Flow detection is an indispensable technical foundation in the industrial field. The safety, reliability, and accuracy of flow measurement are closely related to national defense construction, economic development, and scientific research. There are many types of intelligent electromagnetic flowmeters, among which Coriolis flowmeters are the most commonly used type of flowmeters. The Coriolis flowmeter measures the fluid mass flow by calculating the phase difference (time difference) of the output signals of the sensors on both sides of the vibrating tube. Because of its high measurement accuracy, not affected by medium parameters, wide application range, easy installation, easy maintenance and maintenance, etc., it has been widely used in all aspects of production and life. With the continuous development of science and technology, the requirements for flow measurement accuracy are also constantly improving. For this reason, how to measure the phase difference quickly, accurately and accurately, and then achieve the purpose of improving the accuracy of flow measurement, is a hot topic of current research [1-2].
There are many phase difference measurement methods, among which the correlation analysis method calculates the phase difference through the cross-correlation function value of the two channels of the same frequency signal at zero delay [3], which requires the entire cycle of the signal to be sampled. In order to improve the accuracy of phase difference measurement under non-periodic sampling, a variety of new algorithms have been proposed, such as interpolation method [4], multiple cross-correlation method [5], etc., but they all have a large amount of calculation and cannot be completely overcome. Issues affected by non-full-cycle sampling. The spectrum analysis method obtains the discrete spectrum through Fourier transform, respectively calculates the phases of the maximum spectral line positions of the two signals, and subtracts them to obtain the phase difference [6], and also requires the entire cycle of the signal to be sampled. In order to improve the accuracy of phase difference measurement under non-periodic sampling, measures such as interpolation or windowing are usually used to suppress the influence of spectrum leakage, such as ratio method [7], phase difference correction method [8] and so on. Although these methods have achieved certain results, they have not really solved the fundamental problem of spectrum leakage. The root cause of spectrum leakage is that the number of sampling points in the DFT calculation window is not the entire period length. The Hilbert transform method calculates two sinusoidal signals with the same frequency and the signal after Hilbert transform. Only the time function of the signal phase angle is obtained, and the phase difference can be obtained after subtraction [9]. The purpose of improving the accuracy of the phase difference measurement is usually achieved through multiple operations or transformations, but these methods ignore the end effect, and the fundamental reason lies in the impact of non-full-period sampling.
This article mainly introduces three phase difference measurement methods commonly used in intelligent electromagnetic flowmeters: spectrum analysis method, correlation analysis method and Hilbert transform method. Based on the measurement principle, the phase difference measurement error is further analyzed, and then Propose corresponding improvement measures. The specific implementation steps of suppressing non-periodic sampling are described in detail, and the phase difference measurement methods before and after the improvement are simulated and analyzed to verify the effectiveness of the improvement measures.
1 Three commonly used phase difference measurement methods
1.1 Correlation analysis method

When N is sampling in a whole cycle, the underlined parts in formulas (6)～(8) are all 0, and substituting into formula (5) can accurately obtain the phase difference; when N is not sampling in a whole cycle, formulas (6)～ (8) The underlined parts in (8) are not 0. Substituting into equation (5) will produce large errors, indicating that non-full-period sampling is the main reason that affects the phase difference measurement accuracy of the correlation method.
1.2 Spectrum analysis method

When the signal frequency is equal to the analysis frequency & omega; k0, the gain through the k0 filter bank is 1, and the gain through the remaining filter banks is 0. At this moment, the DFT output can accurately reflect the actual frequency components of the signal. If the signal frequency is not equal to any analysis frequency & omega; k0, the gain through each equivalent filter bank is not 0, which will cause spectrum leakage [10]. When obtaining the initial phase at the maximum spectral line, there will be a large error, and then there will be a large error in obtaining the phase difference. The fundamental reason is still that the number of sampling points in the DFT calculation window is not the length of the entire period, indicating that non-full period sampling is also the main reason that affects the accuracy of the phase difference measurement of the FT method.

The process of generating 900 conjugate signals by Hilbert transform is obtained through Fourier transform; double-sided spectrum folded into single-sided spectrum — inverse Fourier transform; this transformation process is obtained. If the signal is subjected to DFT transformation under non-full-period sampling, the problem of spectrum leakage will occur. When the unilateral spectrum is subjected to the inverse DFT transformation, the error effect caused by the spectrum leakage cannot be offset, and it will be reflected in the time domain waveform, which will lead to the calculation. The distortion phenomenon of the conjugate signal is mainly reflected in the two ends of the signal, which is the so-called end effect; the problem, and further shows that the non-full-period sampling is also the main reason that affects the phase difference measurement accuracy of the Hillbert transform method.
2 Improvement measures
Noise is everywhere, and noise will affect the accuracy of parameter estimation. In the above error analysis, the effect of noise is not discussed. This article only discusses the problem of non-periodic sampling.
As mentioned in Section 1, the three phase difference measurement methods are all affected by non-full-period sampling. For this reason, this paper proposes a data continuation method to suppress non-full-period sampling and improve the accuracy of phase difference measurement. The specific implementation steps are as follows:

The above data continuation method is the method proposed in this article to suppress the non-periodic sampling. The three measurement methods introduced in this article can all be implemented according to the above process, and then calculated using traditional calculation formulas respectively, which can significantly improve the phase difference measurement accuracy of the original method.
3 Simulation analysis
In order to test the actual effect of the proposed method, in the Matlab environment, the three methods were tested and analyzed before and after the improvement. The two sinusoidal signals with the same frequency are:

In the formula, the signal frequency is 100 Hz, and the sampling frequency is 2000 Hz.
3.1 Analysis of the results before and after the continuation of the correlation method
In order to analyze the accuracy of the phase difference measurement before and after the continuation of the correlation method under the non-full-period sampling points, the simulation experiment was carried out under the condition that the signal-to-noise ratio SNR=20dB and the number of sampling points N varies from 30 to 60. The simulation results are shown in Figure 1. Show.

It can be seen from Figure 1 that the phase difference measurement accuracy before the correlation method continuation is affected by whether the number of sampling points is the entire cycle length. The phase difference measurement accuracy at the entire cycle length is the highest, and the curve presents a periodic oscillation attenuation trend, with the number of sampling points The increase of will gradually improve the phase difference measurement accuracy, but when the number of sampling points increases to a certain number, increasing the number of sampling points will not significantly improve the phase difference measurement accuracy. The phase difference measurement accuracy after the continuation of the correlation method is approximately a straight line, which is always below the measurement curve of the pre-extended method, indicating that the phase difference measurement accuracy after the continuation is not affected by whether the number of sampling points is the entire cycle length, and has a higher phase. Poor measurement accuracy.
3.2 Analysis of the results before and after FT method extension
In order to analyze the accuracy of the phase difference measurement before and after the extension of the FT method under the non-full-period sampling points, a simulation experiment was carried out under the condition that the signal-to-noise ratio SNR=20dB and the number of sampling points N=64. The simulation results are shown in Figure 2.
It can be seen from Figure 2 that there are large errors in the frequency estimation before the FT method extension, and the spectrum leakage is more serious; after the data extension, the influence of the spectrum leakage is better suppressed, and the frequency estimation accuracy is higher. The phase difference obtained at the maximum spectral line before continuation is 31.440, and the phase difference obtained at the maximum spectral line after continuation is 30.080. Furthermore, it shows that after continuation, the phase difference measurement accuracy is higher, and the phase difference measurement accuracy is not affected by whether the number of sampling points is the whole cycle length.
3.3 Analysis of the results before and after the continuation of the Hilbert transform method
In order to analyze the measurement accuracy of the phase difference before and after the extension of the Hilbert transform method, the simulation experiment was carried out under the condition that the signal-to-noise ratio SNR=20dB and the number of sampling points N=64. The simulation results are shown in Figure 3.

It can be seen from Figure 3 that the phase difference measurement value before the continuation of the Hilbert transform method has a large error at both ends, and there is an end effect. The phase difference measurement value after the continuation tends to be stable, and there is no end effect. The reason is: The method of data continuation adjusts the number of sampling points to the full period length, which overcomes the influence of the phase difference measurement accuracy on the non-full period length of the sampling points.
4 Conclusion
In order to improve the phase difference measurement accuracy of the traditional phase difference measurement method in the intelligent electromagnetic flowmeter, starting from the basic principles of the spectrum analysis method, correlation analysis method, and Hilbert transform method, the main reason for the error in the phase difference measurement is sampling. The number of points is affected by the length of the non-integral period, and then a method to suppress the non-integral sampling by using the existing data for data continuation is proposed. The specific steps of the method implementation are given in detail, and the simulation analysis of the phase difference measurement before and after the continuation is performed. .
The results show that the method proposed in this article does not require prior information, is not affected by whether or not the entire period is sampled, and always maintains a high phase difference measurement accuracy. The three phase difference measurement methods introduced in the article are all applicable and have a certain universality. In addition, the spectrum analysis method has more prominent advantages when the signal-to-noise ratio is low; the dynamic characteristics of the Hillbert transform method are more obvious; the correlation method takes into account the advantages of both and is suitable for occasions with a high signal-to-noise ratio.