CN109238379B - Method and system for preventing pipeline vibration of vortex shedding flowmeter by combining frequency variance calculation and amplitude calculation - Google Patents

Method and system for preventing pipeline vibration of vortex shedding flowmeter by combining frequency variance calculation and amplitude calculation Download PDF

Info

Publication number
CN109238379B
CN109238379B CN201811275401.2A CN201811275401A CN109238379B CN 109238379 B CN109238379 B CN 109238379B CN 201811275401 A CN201811275401 A CN 201811275401A CN 109238379 B CN109238379 B CN 109238379B
Authority
CN
China
Prior art keywords
frequency
variance
vibration
flow
calculation
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201811275401.2A
Other languages
Chinese (zh)
Other versions
CN109238379A (en
Inventor
徐科军
黎翱
熊伟
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hefei University of Technology
Original Assignee
Hefei University of Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hefei University of Technology filed Critical Hefei University of Technology
Priority to CN201811275401.2A priority Critical patent/CN109238379B/en
Publication of CN109238379A publication Critical patent/CN109238379A/en
Application granted granted Critical
Publication of CN109238379B publication Critical patent/CN109238379B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01F—MEASURING VOLUME, VOLUME FLOW, MASS FLOW OR LIQUID LEVEL; METERING BY VOLUME
    • G01F1/00—Measuring the volume flow or mass flow of fluid or fluent solid material wherein the fluid passes through a meter in a continuous flow
    • G01F1/05—Measuring the volume flow or mass flow of fluid or fluent solid material wherein the fluid passes through a meter in a continuous flow by using mechanical effects
    • G01F1/20—Measuring the volume flow or mass flow of fluid or fluent solid material wherein the fluid passes through a meter in a continuous flow by using mechanical effects by detection of dynamic effects of the flow
    • G01F1/32—Measuring the volume flow or mass flow of fluid or fluent solid material wherein the fluid passes through a meter in a continuous flow by using mechanical effects by detection of dynamic effects of the flow using swirl flowmeters
    • G01F1/325—Means for detecting quantities used as proxy variables for swirl
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01F—MEASURING VOLUME, VOLUME FLOW, MASS FLOW OR LIQUID LEVEL; METERING BY VOLUME
    • G01F15/00—Details of, or accessories for, apparatus of groups G01F1/00 - G01F13/00 insofar as such details or appliances are not adapted to particular types of such apparatus
    • G01F15/02—Compensating or correcting for variations in pressure, density or temperature

Abstract

The invention relates to the field of flow detection, in particular to a method and a system for resisting pipeline vibration of a vortex shedding flowmeter, which aim at a single sensing structure and combine frequency variance calculation and amplitude calculation. In the output signals of the vortex street flow sensor, the frequency variance of the flow signals is large, and the frequency variance of the vibration of the pipe is small. However, when the signal-to-noise ratio is poor, vibration harmonics with frequency variance larger than the flow frequency variance may occur. When the flow signal and the pipe vibration frequency are the same, the amplitude of the signal intermittently and suddenly increases or decreases, resulting in measurement errors. Therefore, the output signal of the vortex street flow sensor is subjected to spectrum analysis, then the frequency variance of the frequency spectrum peak value is calculated, and meanwhile, the amplitude of the frequency spectrum peak value is calculated, so that the vibration harmonic interference is eliminated, the problem of measurement error of the same-frequency signal is solved, namely, the vortex street flow frequency is extracted by adopting a method of combining the frequency variance calculation with the frequency spectrum amplitude calculation, and the pipeline vibration resistance of the vortex street flowmeter is effectively improved.

Description

Method and system for preventing pipeline vibration of vortex shedding flowmeter by combining frequency variance calculation and amplitude calculation
Technical Field
The invention relates to the field of flow detection, in particular to a method and a system for processing signals of a vortex shedding flowmeter, and particularly relates to a method and a system for resisting pipeline vibration of the vortex shedding flowmeter, which combine frequency variance calculation and amplitude calculation.
Background
The vortex shedding flowmeter is suitable for measuring the flow of various media such as liquid, gas, saturated steam and the like, and meanwhile, the vortex shedding flowmeter is widely applied due to the advantages of no mechanical movable part, simple and firm structure, long service life and the like. The vortex flowmeter is based on the fluid vibration principle, so the vortex flowmeter is particularly susceptible to the vibration interference of a pipeline. In the industrial field, devices such as motors, water pumps, valves and the like are generally arranged, and the vibration of pipelines can be caused by the operation of the devices. Especially, when the pipe vibration interference energy in the output signal of the vortex street sensor is larger than the flow signal energy, the general digital signal processing method cannot extract the flow information from the signal containing noise. Therefore, scholars at home and abroad research a pipeline vibration resisting method of the vortex shedding flowmeter. A general vortex shedding flowmeter is provided with only one sensor, the sensor senses flow information and vibration information, and the vortex shedding flowmeter is called a vortex shedding flowmeter based on a single sensor. In order to extract more information, some vortex shedding flowmeters adopt a double-sensor structure, wherein one sensor mainly senses flow information and the other sensor mainly senses vibration information, and the vortex shedding flowmeter is called as a vortex shedding flowmeter based on double sensors.
For the vortex shedding flowmeter based on the double-sensor structure, the Chinese invention patent discloses a digital vortex shedding flowmeter based on the double-sensor structure and resistant to strong interference (Xukejun, Roqinglin, Wanggang, Liushan, Kangyibo, Shihui, Xuyinjiang). In the two sensors of the vortex shedding flowmeter, one sensor senses a flow signal and a vibration signal, and the other sensor senses vibration noise and a weak flow signal. The circuit part takes a single chip microcomputer as a core, calculates instantaneous frequency by adopting a method of combining frequency domain subtraction and frequency variance calculation, switches according to different conditions on site, judges flow information and vibration noise and extracts the flow information. However, the digital vortex shedding flowmeter based on the double-sensor structure and resisting strong interference has the following defects: (1) the requirements on the electromechanical characteristics and the packaging process of the two vortex street flow sensors are extremely high, otherwise, when harmonic interference which cannot be removed by using cut-off frequency and cut-off amplitude occurs, misjudgment on the field condition can be caused, and further measurement errors are caused. (2) The sensitivity difference of the vortex street flow signals and the vibration noise sensed by the two vortex street sensors has higher requirements, and the design requirement of the sensitivity difference required by the method is difficult to meet in practical realization.
Aiming at the anti-vibration method of the vortex shedding flowmeter based on the single-sensor structure, the Chinese invention patent discloses a strong-interference-resistant vortex shedding flowmeter digital signal processing system based on a single sensor (Xukejun, Roqinglin, Wanggang, Liushan, kang-wave, Shiyilian, Xuyinjiang). The invention determines the frequency of the vortex street flow signal by performing frequency spectrum analysis, band-pass filtering and autocorrelation calculation on the output signal of the vortex street flow sensor according to the fact that the vortex street flow signal and the mechanical vibration noise have different frequency bandwidth characteristics and the autocorrelation function can reflect the bandwidth characteristics of the signal. The invention obtains better anti-vibration interference effect through verification experiments. However, the single-sensor-based anti-strong-interference vortex shedding flowmeter digital signal processing system also has the following defects: (1) the autocorrelation calculation amount is large, and in practical application, only the number of autocorrelation points can be reduced in order to meet the requirement of low power consumption of the vortex shedding flowmeter. (2) The autocorrelation calculation with a small number of points results in fluctuation of an autocorrelation function attenuation curve of a vortex street signal instead of complete one-way, so that the required ratio of the autocorrelation function amplitude is difficult to accurately obtain, and misjudgment of a result is easily caused.
The output signals of the sensors of the vortex shedding flowmeter with a single sensor structure are analyzed by shochunly et al in China, and the fact that the frequency of a vortex shedding flow signal fluctuates near an ideal frequency and fluctuates greatly due to the influence of flow noises such as turbulence, pulsation and unstable flow field of a flowing medium in a pipeline is found; the vibration noise of the industrial site is mainly generated by the work of equipment such as a motor, a water pump, a valve and the like, and the frequency of the vibration noise is biased to be fixed, so that the fluctuation is small; when the flow rate and the vibration have the same frequency (hereinafter, referred to as "same frequency signal"), the signal frequency fluctuates around the ideal frequency due to the combined action of mechanical vibration and flow noise, but the fluctuation range is between the fluctuation of the flow rate frequency and the fluctuation of the vibration frequency. Because the variance can reflect the fluctuation condition of data, the vortex street flow rate frequency can be extracted by adopting a method for calculating the frequency variance, and the interference of vibration noise is eliminated (1 Shaocheli. the strong vibration interference resisting method of the vortex street flow meter based on the frequency variance and the method for realizing D. the signal modeling and processing method of the vortex street flow meter under strong pipeline vibration 2014. The method has small calculated amount and lower requirements on the structure and the installation of the sensor, and comprises the following specific steps: firstly, carrying out spectrum analysis on signals output by a flow sensor, carrying out spectrum correction on the maximum 10 peak values in a spectrum, then comparing the corrected spectrum peak values with a cut-off frequency and a dynamic cut-off amplitude value, and keeping the spectrum peak values with the frequencies and the amplitudes simultaneously larger than the cut-off frequency and the dynamic cut-off amplitude value as actual peak values so as to eliminate partial interference; at the same time, the number of actual peaks is determined. When the actual number of the peak values is 0, the flow is not generated at the moment, and the frequency of the output vortex street flow is 0 Hz. When the number of the actual peaks is not 0, the frequency variance of each actual peak is calculated respectively, and the maximum frequency variance is compared with the set variance threshold. If the maximum frequency variance is larger than the variance threshold, the flow sensor output signal contains the vortex street flow signal, and the frequency corresponding to the maximum frequency variance is extracted as the frequency of the vortex street flow signal, otherwise, the flow sensor output signal does not contain the vortex street flow signal, and the output vortex street flow frequency is 0 Hz. Wherein the dynamic cut-off amplitude is obtained as follows: the method comprises the steps of setting flow rate of a vortex shedding flowmeter only, not adding vibration interference, selecting a plurality of different flow rate frequency points with certain intervals in the whole measuring range of the vortex shedding flowmeter to carry out experiments, respectively collecting the output of a vortex shedding flow sensor under each flow rate point, carrying out off-line spectrum analysis to obtain each flow rate frequency and a corresponding frequency spectrum amplitude, and fitting a relation between a dynamic cut-off amplitude and the flow rate frequency by half of the flow rate frequency and the corresponding frequency spectrum amplitude according to the approximate square relation between the frequency and the amplitude of a vortex shedding flow signal output by the vortex shedding flow sensor.
However, the method does not provide a specific method for calculating the frequency variance, and does not provide a description on how to determine whether the peak in the spectrum jumps, because when the peak jumps, the frequency of the actual peak used for calculating the frequency variance needs to be updated, otherwise, a calculation error of the frequency variance is caused.
In practical application, because vibration harmonic interference with frequency variance larger than flow frequency variance exists in an output signal of the vortex street flow sensor, the flow frequency is extracted to make mistakes according to the frequency variance. In addition, when the flow rate and the vibration frequency are the same, the amplitude of the output signal of the vortex street flow rate sensor intermittently and suddenly increases or suddenly decreases due to the reason that the phases of the flow rate and the vibration are not completely synchronous, in the data section of sudden attenuation of the signal amplitude, the number of actual peaks in the frequency spectrum is 0, and 0Hz is directly output as the flow rate frequency, so that measurement errors can be caused.
Therefore, to apply the method to the vortex shedding flowmeter product, the following problems need to be solved: (1) how to calculate the frequency variance quickly and accurately; (2) how to judge whether the frequency of each peak value in the frequency spectrum jumps or not; (3) how to eliminate the effect of vibration harmonic interference; (4) how to solve the measurement error caused by the reduction of the signal amplitude when the flow rate and the vibration have the same frequency.
Disclosure of Invention
The invention provides a method and a system for resisting pipeline vibration of a vortex shedding flowmeter, which can quickly output flow frequency, resist vibration harmonic interference and eliminate measurement errors when the flow and the vibration are in the same frequency, and combine frequency variance calculation and amplitude calculation.
The technical scheme of the invention is as follows: firstly, the output signal of the vortex street flow sensor is collected for spectrum analysis. Then, extracting flow frequency according to the frequency characteristic and amplitude calculation of the output signal of the vortex street flow sensor, namely, in the output signal of the vortex street flow sensor, because the flow signal is influenced by flow noises such as turbulence, pulsation and unstable flow field of a flowing medium in a pipeline, the frequency of the flow signal fluctuates near an ideal frequency, and the frequency variance of the flow signal is large; the vibration noise is mainly generated when mechanical devices such as a motor, a water pump, a valve and the like work, the frequency of the vibration noise is biased to be fixed, and the frequency variance is small; when the flow rate and the vibration have the same frequency (same frequency signal), the signal frequency fluctuates around the ideal frequency due to the combined action of the mechanical vibration and the flow noise, but the frequency variance of the flow rate and the vibration at the same frequency is between the flow rate frequency variance and the vibration frequency variance. Therefore, the vortex street flow signal can be extracted using the difference in frequency variance of the flow signal and the vibration noise interference. However, when the vortex street flow sensor is subjected to vibration interference, vibration harmonics may also exist in the output signal of the vortex street flow sensor, and particularly, when the frequency of a spectral peak in the frequency spectrum of the output signal of the vortex street flow sensor is small, the signal-to-noise ratio is poor, and vibration harmonics having a frequency variance larger than the flow frequency variance may occur. Meanwhile, for the working condition that the flow rate and the vibration frequency are the same, because the phases of the flow rate and the vibration signal are not completely the same, the amplitude of the output signal can be intermittently and suddenly increased or suddenly reduced, and a measurement error can occur in a data section in which the signal amplitude is suddenly attenuated. And finally, calculating the amplitude of the frequency spectrum peak value, extracting the flow frequency by combining frequency variance calculation, and eliminating measurement errors caused by pipeline vibration, vibration harmonic waves and attenuation of common-frequency signals.
The specific technical scheme is as follows: a vortex shedding flowmeter pipeline vibration resisting system combining frequency variance calculation and amplitude calculation adopts direct current 24V for power supply, and the power consumption is less than 4mA when the system runs at full speed. The system comprises a power supply management circuit, a current output circuit, a pulse output circuit, a watchdog, a reset circuit, an undervoltage detection circuit, an FRAM (ferroelectric random access memory) circuit, a keyboard input circuit, a liquid crystal display, a communication circuit, a signal conditioning circuit, an MSP430F5418 single chip microcomputer and vortex shedding flowmeter pipeline vibration resisting algorithm software combining frequency variance calculation and amplitude calculation.
In the vortex shedding flowmeter pipeline vibration resisting algorithm combining frequency variance calculation and amplitude calculation, a relation between a dynamic cut-off amplitude and a flow frequency is fitted by half of the flow frequency and a corresponding frequency spectrum amplitude according to the approximate square relation between the frequency and the amplitude of a vortex shedding flow signal output by a vortex shedding flow sensor, and meanwhile, a variance threshold value Var is determinedthFrequency spectrum amplitude a of measurable flow lower limitminAnd a difference threshold D for judging the same frequency signalthAnd the like. In the actual signal processing process, the output signal of the vortex street flow sensor is collected, the frequency spectrum is calculated and corrected, then the frequency spectrum peak value in the corrected frequency spectrum is compared with the set cut-off frequency and the set dynamic cut-off amplitude value, and the frequency spectrum peak value with the frequency and the amplitude value which are simultaneously larger than the cut-off frequency and the dynamic cut-off amplitude value is extracted as the actual peak value. When the actual number of peak values is not 0, the actual number is compared with the actual number according to the spectrum amplitudeSorting the Peak values from large to small, reserving 10 actual Peak values with the largest spectrum amplitude, and sequentially writing the actual Peak values into a frequency array Peak _ fre [ i][j]In (1). Then, the frequency array Peak _ fre [ i ] is determined][j]Whether the actual peak value stored in the step (2) jumps or not is judged. If the frequency array Peak _ fre [ i ] is][j]If the actual Peak value stored in the middle and the actual Peak value stored in the front and the actual Peak value stored in the middle jump, the situation is changed, and the frequency array Peak _ fre [ i ] needs to be cleared][j]Otherwise, the calculation result of the frequency variance is wrong and cannot reflect the current actual working condition, so the frequency of the actual Peak value after the working condition is changed must be recalculated and stored in the frequency array Peak _ fre [ i [ ]][j](ii) a If no jump occurs, calculating the frequency variance of the actual peak value, and after calculating the frequency variance, determining that the frequency variance is greater than a variance threshold value VarthThe number of actual peaks in (2) is denoted as B _ num. If B _ num is 0, the output signal of the current sensor does not have a vortex flow signal, the output vortex flow frequency is 0Hz, and otherwise, the output signal of the current sensor has a vortex flow signal. For the working condition that the flow rate and the vibration frequency are the same, the amplitude of the output signal intermittently and suddenly increases or suddenly decreases because the phases of the flow rate and the vibration signal are not completely the same. The data segment with suddenly attenuated signal amplitude can cause the actual peak number in the frequency spectrum to be 0, and measurement errors are caused; meanwhile, because the frequency variance of the actual peaks of the same-frequency signals of the non-attenuation band is greater than the variance threshold, the number of the actual peaks of which the frequency variance is greater than the variance threshold is 1. Therefore, when B _ num is 1, it is necessary to determine whether or not the same frequency is present, and to set the maximum spectrum amplitude a of the actual peak valuemaxWith the smallest spectral amplitude AminDifference d between0＝Amax-AminAnd a set difference threshold DthA comparison is made. When d is0≥DthWhen the same-frequency flag is set, the same-frequency flag Cofre _ flag is set (Cofre _ flag is 1), otherwise, the same-frequency flag is cleared. If the maximum spectrum amplitude in the current actual peak value is a, when a is less than or equal to 3aminIn the process, the frequency of the spectral peak value is judged to be small, the signal-to-noise ratio is poor, vibration harmonic waves with frequency variance larger than flow frequency variance can occur, and at the moment, the frequency variance cannot be directly extractedThe frequency of the spectral peak which is greater than the variance threshold and the frequency variance is the largest is the flow frequency, otherwise, a measurement error occurs. However, since the dynamic cut-off amplitude curve is fit from the flow frequency and the corresponding spectral amplitude, it is independent of the frequency and spectral amplitude of the vibration harmonics. Meanwhile, as the flow rate increases, the frequency spectrum amplitude of the flow rate frequency also increases, but for the vibration harmonic generated by the same vibration, the frequency spectrum amplitude of the vibration harmonic tends to decrease as the frequency of the vibration harmonic increases. Therefore, for the vibration harmonic with smaller frequency and unable to be filtered, the ratio between the frequency spectrum amplitude and the dynamic cut-off amplitude should be smaller than the ratio between the frequency spectrum amplitude and the dynamic cut-off amplitude of the flow, so that the extracted frequency variance is larger than the variance threshold, and the frequency of the frequency spectrum peak value with the largest ratio between the frequency spectrum amplitude and the dynamic cut-off amplitude is the flow frequency; when a is>3aminIn the process, the frequency of the frequency spectrum peak value is large, the signal-to-noise ratio is good, and the vibration harmonic with the frequency variance larger than the flow frequency variance can not occur, so the frequency of the frequency spectrum peak value with the frequency variance larger than the variance threshold and the frequency variance maximum is directly extracted as the flow frequency. When the number of the actual peaks is 0, for a gas vortex flowmeter of DN50 (the diameter of the flow tube is 50mm), the sampling frequency does not exceed 3kHz, and when 2048-point FFT (fast Fourier transform) calculation is carried out, only 3 times of FFT calculation is needed to completely analyze one attenuation data segment of the same-frequency signal. Therefore, at this time, if the common-frequency flag Cofre _ flag is set, the times that the number of the actual peak values is continuously 0 are accumulated, when the times that the number of the actual peak values is continuously 0 does not exceed 3 times, it is determined that the current vortex street flow sensor outputs an attenuation data segment of the common-frequency signal, the flow frequency output last time is taken as the current flow frequency to output, but not the output flow frequency is 0Hz, and a measurement error caused by the signal attenuation data segment is eliminated. When the number of times that the number of the actual peak values is continuously 0 exceeds 3 times, the actual peak value number is 0 at this time, which is not caused by the attenuation data segment of the same-frequency signal, and the current flow frequency is directly output to be 0Hz, thereby eliminating the measurement error caused by the attenuation of the amplitude of the same-frequency signal.
The invention has the advantages that: the algorithm has small operand, and is beneficial to the realization of low power consumption of the vortex shedding flowmeter. Meanwhile, a vibration-proof experiment is carried out, and a better experiment result is obtained: when the flow rate and the vibration are different in frequency, the vortex shedding flowmeter can correctly display the flow rate frequency under the condition that the frequency spectrum amplitude of the vibration frequency is 3-8 times of the frequency spectrum amplitude of the flow rate frequency or the frequency spectrum amplitude of the vibration harmonic frequency is not more than 3 times of the frequency spectrum amplitude of the flow rate frequency; when the flow rate and the vibration are in the same frequency, the vortex shedding flowmeter can also correctly display the flow rate frequency under the condition that the ratio of the frequency spectrum amplitude of the vibration frequency to the frequency spectrum amplitude of the flow rate frequency is not more than 1: 1.
Drawings
Fig. 1 is a schematic view of the concept of the method proposed by the present invention.
Fig. 2 is a technical roadmap for implementing the invention.
FIG. 3 is a frequency domain average amplitude spectrum of the vortex street sensor output signal when the flow rate frequency is 53Hz and the vibration frequency is 10 Hz.
FIG. 4 is a frequency variance of the flow frequency 53Hz and vibration harmonics 50 Hz.
FIG. 5 is a ratio of spectral amplitude to dynamic cutoff amplitude for a flow frequency of 53Hz and a vibration harmonic of 50 Hz.
FIG. 6 is a frequency domain average amplitude spectrum of the vortex street sensor output signal at a flow rate frequency of 53Hz and a vibration frequency of 68 Hz.
FIG. 7 is a frequency variance of the flow frequency of 53Hz and the vibration frequency of 68 Hz.
Fig. 8 is a time domain waveform comparison of a traffic signal and a co-frequency signal.
Fig. 9 is an overall hardware block diagram of the inventive system.
FIG. 10 is a software block diagram of the inventive system.
FIG. 11 is a flow chart of a main monitoring program of the inventive system.
Detailed Description
The invention is further described below with reference to the accompanying drawings.
Fig. 1 is a schematic view of the concept of the method proposed by the present invention. And carrying out spectrum analysis on the output signal of the vortex street flow sensor, and then extracting a spectrum peak value with the frequency and the amplitude simultaneously larger than the cut-off frequency and the dynamic cut-off amplitude as an actual peak value. Because the flow signal of the vortex street flow sensor is influenced by flow noise such as turbulence, pulsation, unstable flow field and the like of a flowing medium in a pipeline, the frequency of the flow signal fluctuates around an ideal frequency and is large in fluctuation, and therefore the frequency variance of the flow signal is large; the vibration noise is mainly generated when electromechanical equipment such as a motor, a water pump, a valve and the like works, and the frequency of the vibration noise is biased to be fixed, so that the fluctuation is small, and the frequency variance is small; when the flow rate and the vibration frequency are the same, the signal frequency fluctuates around the ideal frequency due to the combined action of mechanical vibration and flow noise, but the fluctuation range is between the fluctuation of the flow rate frequency and the fluctuation of the vibration frequency, so that the frequency variance when the flow rate and the vibration have the same frequency is between the flow rate frequency variance and the vibration frequency variance. Therefore, after the actual peak in the frequency spectrum is determined, the frequency variance of the actual peak is calculated, and the vortex street flow rate frequency is extracted by using the difference between the flow rate signal and the frequency variance of the vibration noise interference. However, when the vortex street flow sensor is subjected to vibration interference, vibration harmonics may also exist in the output signal of the vortex street flow sensor, and particularly, when the frequency of a spectral peak in the frequency spectrum of the output signal of the vortex street flow sensor is small, the signal-to-noise ratio is poor, and vibration harmonics having a frequency variance larger than the flow frequency variance may occur. Meanwhile, for the working condition that the flow rate and the vibration frequency are the same, because the phases of the flow rate and the vibration signal are not completely the same, the amplitude of the output signal can be intermittently and suddenly increased or suddenly reduced, so that the number of actual peak values in the frequency spectrum can be 0 in the data section of the sudden attenuation of the signal amplitude, and measurement errors are caused.
Fig. 2 is a technical roadmap for implementing the invention.
The first step of fig. 2 is to determine the relevant parameters in the method of combining the frequency variance calculation with the magnitude operation.
When the vortex street frequency is extracted by a method combining frequency variance calculation and amplitude calculation, the vortex street frequency needs to be extractedDetermining the relevant parameters, wherein the method comprises the following steps: (1) and carrying out flow experiment on the vortex street flow sensor, and only setting flow without vibration interference. Selecting a plurality of different flow frequency points with certain intervals in the whole measuring range of the vortex shedding flowmeter for experiment, respectively collecting the output of a vortex shedding flow sensor under each flow point, carrying out off-line frequency spectrum analysis to obtain each flow frequency and a corresponding frequency spectrum amplitude, and fitting a relation between a dynamic cut-off amplitude and the flow frequency by half of the flow frequency and the corresponding frequency spectrum amplitude according to the approximate square relation of the frequency and the amplitude of a vortex shedding flow signal output by the vortex shedding flow sensor; at the same time, the frequency spectrum amplitude a of the lower limit of the measurable flow is determinedmin. (2) Determining the amplification factor of the signal conditioning circuit; meanwhile, flow and vibration signals with the same frequency of the vortex shedding flowmeter are given, data are collected and analyzed, and a difference threshold D for judging whether the flow and the vibration have the same frequency is calculatedth. (3) Under the condition of only giving flow, selecting a plurality of different flow frequency points with certain intervals in the whole measuring range of the vortex shedding flowmeter for experiment, calculating the variance value of each flow frequency, and extracting the minimum flow frequency variance as VarminI.e. the lower limit of the variance of the flow frequency. (4) The method comprises the steps of giving vibration to a vortex shedding flowmeter, selecting a plurality of scattered vibration frequency point signals within 10-150Hz of low-frequency vibration interference through a vibration table to perform experiments, respectively collecting the output of a vortex shedding flow sensor under each vibration frequency, performing off-line spectrum analysis, calculating the variance value of each vibration frequency, and extracting the largest vibration frequency variance as VarmaxI.e. the upper limit of the variance of the vibration frequency. Then the set variance threshold value Var of the vortex street flow sensorthThe calculation formula of (2) is as follows:
Figure BDA0001846859830000061
the second step of fig. 2 is to investigate the method of calculating the frequency variance.
After determining the relevant parameters in the method combining frequency variance calculation and amplitude calculation, the method for calculating the frequency variance needs to be researched to ensure that the frequency variance of the actual peak can be accurately calculated. In practical application, firstly, the signal output by the vortex street flow sensor is subjected to spectrum analysis, and a spectrum peak value with the frequency and the amplitude simultaneously larger than the cut-off frequency and the dynamic cut-off amplitude in the corrected spectrum is extracted as an actual peak value. When the actual number of peaks is not 0, the frequency variance of the actual peaks is calculated.
Saving the frequency of the actual Peak in a Peak frequency array Peak _ fre [ i ] [ j ], where i denotes the number of times the spectrum is calculated, and i ═ 0,1,2, … … 29; j represents the frequency of the j-th actual peak obtained by calculating the frequency spectrum, and j is 0,1,2, … … 9, that is, the frequency data of the actual peak is saved for at most 30 times, and the frequency of the 10 actual peaks with the maximum frequency spectrum amplitude is reserved for each calculation of the frequency spectrum. In other words, the frequency spectrum is calculated up to 30 times, each time yielding a maximum of 10 peak frequencies. When the frequency of the actual Peak value obtained by each calculation is saved, the actual Peak values are sorted according to the sequence of the spectrum amplitudes from large to small, and then the frequency is sequentially written into Peak _ fre [ i ] [ j ], namely Peak _ fre [0] [0] is the Peak frequency with the largest amplitude in the frequencies calculated for the first time, Peak _ fre [0] [1] is the Peak frequency with the second largest amplitude in the frequencies calculated for the first time, … …; peak _ fre [1] [0] is the Peak frequency with the largest amplitude in the frequencies calculated for the second time, and Peak _ fre [1] [1] is the Peak frequency with the largest amplitude in the frequencies calculated for the second time. When i is larger than or equal to 1, the frequency spectrum is calculated at least twice, and at the moment, the frequency variance is calculated under the condition that the frequency of the actual peak value obtained by the previous and subsequent calculations does not jump, so that the rapid output flow frequency is ensured; and when i is 29, circularly updating the frequency data in the array Peak _ fre [ i ] [ j ] according to a first-in first-out mode, and at the moment, calculating the frequency variance by using 30 frequency data so as to ensure the accuracy of the calculated frequency variance.
The general idea of judging whether the frequency of the actual peak value obtained by two previous and next calculations jumps or not is as follows: and under the condition that the frequency of the actual peak value obtained by two previous and subsequent calculations is not hopped, the frequency variance is calculated to ensure that the calculated frequency variance is accurate. If the frequency of the actual Peak value obtained by two times of calculation is jumped, the data in the Peak _ fre [ i ] [ j ] needs to be cleared, and the data in the Peak _ fre [ i ] [ j ] is accumulated again. Firstly, judging whether the number of actual Peak values obtained by two times of calculation is equal, if not, directly judging that the frequency of the actual Peak values obtained by two times of calculation jumps, and clearing frequency data in Peak _ fre [ i ] [ j ]. If the frequency of one actual peak is equal to the frequency of all actual peaks obtained by the previous calculation, subtracting the frequency of the current actual peak from the frequency of all actual peaks obtained by the previous calculation to obtain a difference value, taking the absolute value of the difference value, and then comparing the minimum value of the absolute values of all the difference values of the frequency of the actual peak with 3 Δ f (the minimum value is selected to find a frequency closest to the current frequency from the frequency of the last calculated actual peak), wherein Δ f is the frequency resolution. And if the minimum difference is not more than 3 delta f, namely, a frequency within +/-3 delta f of the frequency of the actual peak exists in the frequency of the actual peak obtained by the previous calculation, judging that the frequency of the actual peak does not jump, and otherwise, judging that the frequency of the actual peak does not jump. And judging whether the frequencies of other actual peak values obtained by current calculation jump or not by analogy, and judging that the frequencies of the actual peak values obtained by two times of calculation do not jump only under the condition that the frequencies of all the actual peak values obtained by current calculation do not jump. And if the frequency of one actual Peak value jumps, judging that the frequency of the current actual Peak value jumps, and clearing the frequency data in the Peak _ fre [ i ] [ j ].
The specific method for judging whether the frequency of the actual peak value obtained by two previous and next calculations jumps or not is as follows: the actual number of peaks obtained by the i +1 th calculation is recorded as Num, and the actual number of peaks obtained by the i-th calculation is recorded as P _ Num. When Num is not equal to P _ Num, namely the number of the actual Peak values in the two times of before and after changes, the frequency of the actual Peak values in the two times of before and after is judged to jump, and Peak _ fre [ i ≠ P _ Num][j]And clearing the medium frequency data. When Num is P _ Num, that is, the number of the actual Peak values in the two previous and subsequent times does not change, because the frequency of the actual Peak value calculated each time is sorted according to the sequence of the spectrum amplitude from large to small, and then is stored in the array Peak _ fre [ i ]][j]Therefore, the absolute value of the difference between the frequency of the first actual peak obtained by the i +1 th calculation and the frequency of the first actual peak obtained by the i-th calculation is calculated first and is denoted as D11. Let f[i+1][n]Is an array Peak _ fre [ i ]][j]Frequency of the nth actual peak, f, obtained by the (i + 1) th calculation[i][n]Is an array Peak _ fre [ i ]][j]The frequency of the nth actual peak obtained in the ith calculation, where n is 1,2,3, … … 10, then D11＝f[i+1][1]-f[i][1]L. Then, the absolute value of the difference between the frequency of the actual peak obtained by the i +1 th calculation and the frequency of the first actual peak obtained by the i-th calculation is calculated and recorded as D1n，D1n＝|f[i+1][n]-f[i][1]And | n ═ 2,3, … … 10. Get D1nThe minimum value of (1) is denoted as D1mDenotes D1nThe minimum value in (1) is the absolute value of the difference between the frequency of the mth actual peak obtained by the calculation of the (i + 1) th time and the frequency of the first actual peak obtained by the calculation of the ith time, that is: d1m＝min(D1n) M is more than or equal to 2 and less than or equal to 10. For spectral analysis with a sampling frequency of Fs and a data length of N, the frequency resolution Δ f is Fs/N. When D is present1m≥D11And D is11<3 deltaf, the frequency of the first actual peak value obtained by the i +1 th calculation is judged not to jump. When D is present1m<D11But D is1m<3 Δ f, it is also determined that the frequency of the first actual peak obtained by the i +1 th calculation has not hopped, but at this time, the frequency interval between the mth frequency in the frequency sequence of the actual peak obtained by the i +1 th calculation and the first actual peak obtained by the i th calculation is smaller, and the position of the mth frequency in the frequency sequence of the actual peak obtained by the i +1 th calculation needs to be exchanged with the first frequency. In addition to the above, we consider that the frequency of the first actual peak calculated at the (i + 1) th time jumps. And (5) judging whether the frequency of other actual peak values jumps or not, and so on. Only under the condition that all current actual peak values do not jump, judging that the actual peak values obtained by two times of calculation do not jump; before the decision, only one actual peak value jumpsJumping the actual Peak value obtained by the last two times of calculation, and adding Peak _ fre [ i [ ] to the Peak value][j]And clearing the medium frequency data.
The third step of fig. 2 is to study the method of calculating the frequency variance in combination with calculating the ratio of the spectral amplitude and the dynamic cut-off amplitude to eliminate harmonic interference.
After calculating the frequency variance of the actual peak, it is determined that the variance is greater than the variance threshold value VarthThe number of actual peaks in (2) is denoted as B _ num. If B _ num is 0, it indicates that there is no vortex street flow signal in the output signal of the current vortex street flow sensor, and the output vortex street flow frequency is 0Hz, otherwise, it indicates that there is a vortex street flow signal in the output signal of the current sensor. Under the condition that B _ num is not equal to 0, if the maximum spectrum amplitude in the current actual peak value is a, when a is not more than 3aminIn the process, the frequency of the actual peak value is judged to be small, the signal-to-noise ratio is poor, and vibration harmonics with frequency variance larger than the flow frequency variance can occur.
FIG. 3 is a frequency domain average amplitude spectrum of the output signal of the vortex shedding flow sensor when the flow rate frequency is 53Hz and the vibration frequency is 10 Hz. Wherein the solid line is the frequency spectrum, the dashed line is the dynamic cut-off amplitude, and the dotted line is the set cut-off frequency. As can be seen from the average amplitude spectrum, in addition to the flow rate frequency of 53Hz and the vibration frequency of 10Hz, vibration harmonic frequencies of 20Hz, 30Hz, 50Hz, and 60Hz are also present in the output signal of the vortex street flow sensor. The 10Hz vibration interference and the 20Hz and 30Hz vibration harmonic interference are filtered by the set cut-off frequency, the 60Hz vibration harmonic interference is filtered by the dynamic cut-off amplitude, but the frequency spectrum amplitude of the 50Hz vibration harmonic is close to the frequency spectrum amplitude of the flow 53Hz, therefore, the dynamic cut-off amplitude can not filter the 50Hz vibration harmonic interference, at this time, the flow 53Hz and the vibration harmonic 50Hz are selected as the actual peak value in the frequency spectrum, and the frequency spectrum amplitude of the flow frequency in the current frequency spectrum is the largest, but is less than 3amin。
At this time, the frequency variance of the flow rate frequency 53Hz and the vibration harmonic 50Hz was calculated.
FIG. 4 is a frequency variance of the flow frequency 53Hz and vibration harmonics 50 Hz. Wherein the solid line is the set variance threshold. It can be seen from the figure that the variance of both frequencies is greater than the variance threshold, and that the variance of the harmonic frequencies of the vibration is generally greater than the variance of the flow frequency. If the extracted frequency variance is larger than the variance threshold value and the frequency of the frequency spectrum peak value with the largest frequency variance is the flow frequency, the vibration harmonic 50Hz is selected as the flow frequency, and a measurement error occurs. The dynamic cut-off amplitude curve is obtained by fitting the flow frequency and the corresponding frequency spectrum amplitude and is irrelevant to the frequency and the frequency spectrum peak value of the vibration harmonic. Meanwhile, as the flow rate increases, the peak value of the frequency spectrum of the flow rate also increases, but for the vibration harmonic generated by the same vibration, the amplitude of the frequency spectrum of the vibration harmonic tends to decrease as the frequency of the vibration harmonic increases. Therefore, for the vibration harmonic with small frequency and unable to be filtered out, the ratio between the frequency spectrum amplitude and the dynamic cut-off amplitude is smaller than the ratio between the frequency spectrum amplitude and the dynamic cut-off amplitude of the flow. Therefore, the ratio of the spectral amplitude of the flow frequency 53Hz and vibration harmonic 50Hz to the corresponding dynamic cut-off amplitude is calculated.
FIG. 5 is a ratio of spectral amplitude to dynamic cutoff amplitude for a flow frequency of 53Hz and a vibration harmonic of 50 Hz. As can be seen from the figure: at this time, the ratio of the spectral amplitude of the flow rate frequency to the dynamic cut-off amplitude is larger than the ratio of the spectral amplitude of the vibration harmonic spectrum to the dynamic cut-off amplitude. Therefore, by selecting the frequency of the spectrum peak value with the frequency variance larger than the variance threshold value and the maximum ratio of the spectrum amplitude to the dynamic cut-off amplitude as the flow frequency, the correct flow frequency of 53Hz can be extracted, and the influence of the vibration harmonic of 50Hz on the measurement is reduced.
FIG. 6 is a frequency domain average amplitude spectrum of the vortex street sensor output signal at a flow rate frequency of 53Hz and a vibration frequency of 68 Hz. Wherein the solid line is the frequency spectrum, the dashed line is the dynamic cut-off amplitude, and the dotted line is the set cut-off frequency. As can be seen, in addition to a flow rate frequency of 53Hz and a vibration frequency of 68Hz, there is a vibration harmonic frequency of 136 Hz. However, the spectral amplitude of the harmonic frequency of the 136Hz vibration is filtered out to be less than the dynamic cut-off amplitude, so that the actual peaks in the spectrum are 53Hz flow and 68Hz vibration. The frequency spectrum amplitude of vibration 68Hz in the current frequency spectrum is maximum and is larger than 3aminAt this time, the frequency variance of the flow rate frequency 53Hz and the vibration 68Hz was also calculated.
FIG. 7 isThe flow frequency is 53Hz and the frequency variance of the vibration is 68 Hz. Wherein the solid line is the set variance threshold. As can be seen from the figure, at this time, only if the variance of the flow rate frequency is greater than the variance threshold, the correct flow rate frequency of 53Hz can be extracted by selecting the frequency of the spectral peak with the frequency variance greater than the variance threshold and the frequency variance maximum as the flow rate frequency. Thus, in summary: the frequency spectrum amplitude of the lower limit frequency of the measurable flow is aminIf the maximum spectrum amplitude in the current actual peak value is a, the vortex shedding flowmeter is used for measuring the frequency spectrum amplitude of the current actual peak value. When a is less than or equal to 3aminAt the moment, the frequency of the actual peak value is small, the signal-to-noise ratio is poor, and vibration harmonics with frequency variance larger than the flow frequency variance may occur, so that the frequency of the frequency spectrum peak value with the frequency variance larger than the variance threshold value and the largest ratio of the frequency spectrum amplitude and the dynamic cut-off amplitude is extracted as the flow frequency to eliminate harmonic interference; when a is>3aminAt the moment, the frequency of the actual peak value is larger, the signal-to-noise ratio is good, vibration harmonic waves with frequency variance larger than the flow frequency variance cannot occur, and the frequency of the frequency spectrum peak value with frequency variance larger than the variance threshold and the maximum frequency variance can be directly extracted as the flow frequency.
The fourth step of fig. 2 is to study the method of calculating the frequency variance and calculating the difference of the spectral amplitude to eliminate the measurement error caused by the co-frequency of the flow and the vibration.
Because the frequency of the flow and the vibration are the same (same frequency signal), the phases of the flow and the vibration signal are not completely the same, the amplitude of the output signal will be intermittently and suddenly increased or suddenly decreased, and in the data segment where the amplitude of the same frequency signal is suddenly attenuated, the number of the actual peak values in the frequency spectrum will be 0. If 0Hz is directly output as the flow rate frequency, the measurement is wrong.
Fig. 8 is a time domain waveform comparison of a traffic signal and a co-frequency signal. Wherein, the flow frequency and the vibration are both 73 Hz. As can be seen from the figure: when the flow rate and the vibration frequency are the same, the amplitude of the output signal of the vortex street flow sensor intermittently and suddenly increases or suddenly decreases. Since the frequency variance of the actual peaks of the non-attenuated segments of the same-frequency signal is greater than the variance threshold, the number of actual peaks whose frequency variance is greater than the variance threshold at this time is 1. Therefore, when the variance is greater than the variance threshold value VarthIf the number B _ num of actual peaks in (a) is 1, there is a possibility that the frequency is the same, and it is necessary to determine whether the frequency is the same. The judging method comprises the following steps: maximum spectral amplitude A of actual peakmaxWith the smallest spectral amplitude AminDifference d between0＝Amax-AminAnd a set difference threshold DthMaking a comparison when d0≥DthWhen the same frequency flag is set, the same frequency flag Cofre _ flag is set (Cofre _ flag is 1), otherwise, the same frequency flag is cleared. If the common-frequency signal firstly appears as non-attenuation data, before the common-frequency signal appears as attenuation data, determining the maximum difference value between the number of the actual peak values with the frequency variance larger than the variance threshold value and all the spectrum amplitude values of the actual peak values, and setting a common-frequency flag Cofre _ flag. For the gas vortex shedding flowmeter of DN50, the sampling frequency does not exceed 3kHz, and when 2048-point FFT calculation is carried out, one attenuation data segment of the same-frequency signal can be completely analyzed by only 3 times of FFT calculation. Therefore, under the condition that the common-frequency flag Cofre _ flag is set, when the number of times that the number of actual peak values is continuously 0 does not exceed 3 times, the current vortex street flow sensor is judged to output the attenuation data segment of the common-frequency signal, the last output flow frequency is taken as the current flow frequency to be output, but not 0Hz, and the influence of the attenuation data segment of the common-frequency signal on the measurement is eliminated. When the number of times that the actual peak value number is continuously 0 exceeds 3 times, the fact that the actual peak value number is 0 at the moment is not caused by the attenuation data section of the same-frequency signal is judged, and the current flow frequency is directly output to be 0 Hz. If the same frequency signal firstly appears in attenuation data, the actual peak value number of the frequency spectrum is 0 at this moment, the output wrong flow frequency is 0Hz, but after the non-attenuation data of the same frequency signal appears, the method for analyzing and processing the non-attenuation data of the same frequency signal firstly appears is directly adopted, so that the measurement error can be caused by only one section of attenuation data at most, and the capability of the vortex shedding flowmeter for resisting the same frequency vibration interference is greatly improved. In practical application, the time for debugging the vortex shedding flowmeter exists; when the calibration experiment is carried out, the calibration is started after the working condition is stable, so that the wrong flow frequency is output only within a period of attenuation dataThe rate has no influence on the practical application and calibration of the vortex shedding flowmeter.
The fifth step of fig. 2 is to develop a hardware system of the vortex shedding flowmeter with low power consumption.
Fig. 9 is an overall hardware block diagram of the inventive system. The power supply is powered by direct current 24V, the power consumption is less than 4mA when the power supply runs at full speed, and the power supply comprises a power supply management circuit, a current output circuit, a pulse output circuit, a watchdog, a reset circuit, under-voltage detection, an FRAM (ferroelectric random access memory) circuit, a keyboard input circuit, a liquid crystal display, a communication circuit, a signal conditioning circuit and an MSP430F5418 singlechip.
The working process of the invention is as follows: the output signal of the vortex street flow sensor is sent to an ADC input port on the singlechip for sampling after passing through the signal conditioning circuit. Then, a vortex shedding flowmeter pipeline vibration resisting algorithm combining frequency variance calculation and amplitude calculation is operated on the single chip microcomputer, and the flow is calculated. Finally, a software program which is run and programmed by the singlechip converts the flow information into corresponding digital quantity and transmits the digital quantity to the pulse output circuit, the current output circuit and the communication circuit, converts the flow into pulses and standard current signals and outputs the signals, and meanwhile, completes the communication with external equipment.
The sixth step of fig. 2 is to develop a software system of the low power consumption vortex shedding flowmeter.
FIG. 10 is a system software block diagram of the inventive system. The intelligent monitoring system mainly comprises a main monitoring program, an interruption module, a protection module, an initialization module, an FRAM module, a calculation module, a communication module, a current output module (including a pulse output program) and a man-machine interface module. And calling each module through the main monitoring program to complete system tasks. The initialization module mainly comprises two parts of liquid crystal initialization and singlechip initialization, and the initialization of the whole system is carried out by calling an initialization function. The calculation module calculates the vortex street frequency by using a vortex street flowmeter pipeline vibration resisting algorithm combining frequency variance calculation and amplitude calculation, and further calculates the flow. And the current output module converts the calculated flow into corresponding digital quantity and sends the digital quantity into a current output hardware circuit to finish current output. The man-machine interface module comprises a liquid crystal display module and a keyboard monitoring subprogram, the singlechip sends data information such as frequency, flow and the like to the liquid crystal display module, and flow information display, password setting and instrument parameter setting are completed through keys. The FRAM module sends the modified parameters to the external expansion iron for storage, and when the system encounters emergency such as power failure, information such as frequency and flow during power failure is stored in the FRAM for query. The protection module is mainly composed of an external watchdog program and a power supply monitoring interrupt program. The communication module completes communication with an external communication device.
FIG. 11 is a flow chart of a main monitoring program of the inventive system. The main monitoring program calls each module program to coordinate and complete the system task, and is an endless loop program. After the system is powered on, an initialization module program is called first to complete the initialization of the single chip microcomputer and the liquid crystal module, and meanwhile, the relation between the dynamic cut-off amplitude and the flow frequency and the variance threshold value Var are setthSpectral amplitude a of the lower flow limitminAnd a difference threshold D for judging the same frequency signalth. After initialization is completed, data collected by an ADC (analog to digital converter) on the single chip are read, spectrum analysis and spectrum correction are carried out, cut-off frequency set by a spectrum peak value in a corrected spectrum is extracted and compared with a dynamic cut-off amplitude value, and the spectrum peak value with the frequency and the amplitude value simultaneously larger than the dynamic cut-off frequency and the cut-off amplitude value is reserved as an actual peak value. If the number of the actual peak values is not 0, calculating the frequency variance and amplitude value operation, and calculating the frequency variance of the actual peak values, the ratio of the amplitude value of the actual peak values to the dynamic cut-off amplitude value and the maximum spectrum amplitude value A of the actual peak valuesmaxWith the smallest spectral amplitude AminDifference d between0Wherein d is0＝Amax-Amin. And then judging the same frequency condition. The method for judging the same frequency condition comprises the following steps: according to the calculation result of the frequency variance of the actual peak value, the variance is determined to be larger than a variance threshold value VarthThe number of actual peaks in (2) is denoted as B _ num. When the variance is greater than the variance threshold value VarthWhen the number B _ num of the actual peak values is 1, the maximum spectrum amplitude a of the actual peak values is determinedmaxWith the smallest spectral amplitude AminDifference d between0And a set difference threshold DthMaking a comparison when d0≥DthTime, judgeThe signal for vortex street flow sensor output is the same frequency signal, set same frequency flag Cofre _ flag (Cofre _ flag is 1), otherwise, the same frequency flag is clear. Next, the flow rate frequency is extracted. If B _ num is 0, it indicates that there is no vortex street flow signal in the output signal of the current vortex street flow sensor, and the extracted vortex street flow frequency is 0Hz, otherwise, it indicates that there is a vortex street flow signal in the output signal of the current sensor, and it is necessary to eliminate the vibration harmonic interference by combining the frequency variance of the actual peak value and the ratio of the amplitude of the actual peak value to the dynamic cut-off amplitude, and accurately extract the flow frequency. And finally, calculating instantaneous flow and accumulated flow, calling a current output module, changing the instantaneous flow into a standard current signal and outputting pulses, further refreshing the liquid crystal, and re-entering a new cycle after calling a communication module program to complete related data communication. And if the actual number of the peak values is 0, directly judging whether the current same-frequency mark is set. If the same frequency flag is set, processing the same frequency condition, wherein the processing method comprises the following steps: accumulating the times that the number of the actual peak values is continuously 0, and when the times that the number of the actual peak values is continuously 0 does not exceed 3 times, considering that the current vortex street flow sensor outputs an attenuation data section of a same-frequency signal, outputting the flow frequency output last time as the current flow frequency instead of outputting 0Hz, and eliminating measurement errors caused by the signal attenuation data section; when the number of times that the actual peak number is continuously 0 exceeds 3 times, the actual peak number is considered to be 0 at the moment and is no longer caused by the attenuation data segment of the same-frequency signal, and the output flow frequency is 0 Hz. If the same-frequency flag is not set, the direct output flow frequency is 0 Hz. Then, the step of calculating the instantaneous flow and the accumulated flow is carried out, and then after the current and pulse output, the liquid crystal refreshing and the data communication are completed, the circulation is returned and enters a new round again.
The seventh step of fig. 2 is to perform a verification experiment of the vortex shedding flowmeter against vibration.
The vortex shedding flowmeter anti-strong vibration interference experiment platform consists of an air blower, an upstream straight pipe section, a vortex shedding flowmeter, a downstream straight pipe section, a vibration table controller, a notebook computer, a single chip microcomputer simulator, a linear power supply and an oscilloscope.
The experimental process of the vortex shedding flowmeter anti-vibration verification experiment is as follows: a DN25 gas vortex flowmeter (the diameter of the flow tube is 25mm) is fixed on a vortex flowmeter anti-strong vibration interference experiment platform, a blower of the vortex flowmeter anti-strong vibration interference experiment platform is powered on, gas flow exists in a pipeline, the vortex flowmeter starts to measure the gas flow, the frequency of the vortex flow is calculated, and the frequency is displayed on liquid crystal. Then, the liquid crystal display vortex street flow rate frequency of the vortex street flowmeter is read and recorded. Meanwhile, data before entering the single chip microcomputer are collected through the oscilloscope, Matlab is used for carrying out spectrum analysis, and the size of the air outlet of the pipeline is changed for many times so as to change the flow. And comparing the flow frequency obtained by Matlab analysis with the flow frequency displayed by the liquid crystal of the vortex shedding flowmeter, wherein the two results are the same, and the result shows that the vortex shedding flowmeter can correctly calculate the vortex shedding flow frequency, so that the subsequent experiment can be carried out. Because the common vibration frequency range of the industrial field is 10-150Hz, the subsequent experiments investigate the anti-vibration performance of three most representative flow points in the measuring process of the vortex shedding flowmeter under the interference of vibration with the frequency of 10-150Hz, wherein the three flow points are respectively the flow point which is most frequently measured by the vortex shedding flowmeter in no-flow, measurable flow lower limit and actual working condition. (1) Turning off the air blower, turning on the vibration table, turning on the power supply of the vibration table under the condition of no flow, giving periodic vibration interference within 10-150Hz through the vibration table, reading and recording the reading of the vortex shedding flowmeter under each vibration frequency, and simultaneously collecting data. (2) And (3) turning on the blower, and changing the size of the air outlet of the pipeline to stabilize the frequency of the gas flow in the pipeline to be about 110Hz, wherein the flow with the frequency of 110Hz is the lower measurement limit of the DN25 gas vortex flowmeter. Then, a power supply of the vibration table is started, periodic vibration interference within 10-150Hz is given through the vibration table, reading and recording the reading of the vortex shedding flowmeter under each vibration frequency, and meanwhile, data are collected. (3) And (3) turning on the blower, and changing the size of the air outlet of the pipeline to stabilize the frequency of the gas flow in the pipeline to be about 160Hz, wherein the flow with the frequency of 160Hz is the most frequently measured flow point of the DN25 gas vortex shedding flowmeter in the actual working condition. Then, a power supply of the vibration table is started, periodic vibration interference within 10-150Hz is given through the vibration table, the reading (output frequency and corresponding amplitude) of the vortex shedding flowmeter at each vibration frequency is read and recorded, and data are collected at the same time.
The test result of the vortex shedding flowmeter anti-vibration verification experiment is as follows: the reading conditions of the vortex shedding flowmeter under the vibration interference of different frequencies when no flow exists are shown in table 1.
TABLE 1 results of vibration-free experiments
Figure BDA0001846859830000131
When the flow is not available, under the influence of vibration interference of different frequencies, the liquid crystal display flow frequency of the vortex shedding flowmeter is 0Hz, which shows that under the working condition, the system can completely and correctly measure the flow, and the vibration frequency cannot be extracted as the vortex shedding flow frequency.
The reading conditions of the vortex shedding flowmeter under the vibration interference of different frequencies when the flow signal frequency is about 110Hz are shown in Table 2.
TABLE 2 flow rate of 110Hz anti-vibration experiment results
Figure BDA0001846859830000132
When the frequency of the flow signal is about 110Hz, under the influence of vibration interference of different frequencies, the frequency spectrum amplitude of the flow frequency fluctuates between 47 and 80mV, and at the moment, under the periodic vibration interference with the maximum frequency spectrum amplitude of 1230mV and the frequency of 120Hz or the vibration harmonic interference with the maximum frequency spectrum amplitude of 352mV and the frequency of 40Hz, the liquid crystal of the vortex shedding flowmeter displays the correct flow frequency.
The reading of the vortex shedding flowmeter under vibration interference conditions of different frequencies when the flow signal frequency is about 160Hz is shown in Table 3.
TABLE 3 flow rate of 160Hz anti-vibration experiment results
Figure BDA0001846859830000141
When the frequency of the flow signal is about 160Hz, under the influence of vibration interference of different frequencies, the frequency spectrum amplitude of the flow frequency fluctuates between 132 and 179mV, and at this time, under the periodic vibration interference with the maximum frequency spectrum amplitude of 1500mV and the frequency of 100Hz or the vibration harmonic interference with the maximum frequency spectrum amplitude of 364.7mV and the frequency of 40Hz, the liquid crystal of the vortex shedding flowmeter displays correct flow frequency.

Claims (4)

1. The utility model provides a vortex shedding flowmeter anti-pipe vibration method that frequency variance calculation and amplitude operation combined together, contains power management circuit, current output circuit, pulse output circuit, watchdog, reset circuit, undervoltage detection, FRAM circuit, keyboard input, LCD display, communication circuit, signal conditioning circuit, MSP430F5418 singlechip and the vortex shedding flowmeter anti-pipe vibration algorithm software that frequency variance calculation and amplitude operation combined together which characterized in that:
carrying out spectrum analysis on an output signal of the vortex street flow sensor, and then extracting a spectrum peak value with the frequency and the amplitude simultaneously larger than a cut-off frequency and a dynamic cut-off amplitude as an actual peak value; after the actual peak value in the frequency spectrum is determined, calculating the frequency variance of the actual peak value, and extracting the vortex street flow frequency by using the difference between the flow signal and the frequency variance of vibration noise interference, however, when the vortex street flow sensor is subjected to vibration interference, vibration harmonic waves can also exist in the output signal of the vortex street flow sensor, and when the frequency of the frequency spectrum peak value in the frequency spectrum of the output signal of the vortex street flow sensor is smaller, the signal-to-noise ratio is poor, and the vibration harmonic waves with the frequency variance larger than the flow frequency variance can appear; meanwhile, for the working condition that the flow rate and the vibration frequency are the same, because the phases of the flow rate and the vibration signal are not completely the same, the amplitude of the output signal can be intermittently and suddenly increased or suddenly reduced, so that the actual peak number in the frequency spectrum is 0 in the data section of the suddenly attenuated signal amplitude, and the measurement error is caused; therefore, the frequency spectrum amplitude of the frequency spectrum peak value needs to be calculated, namely, the frequency variance calculation is combined with the amplitude calculation, the vortex street flow frequency is extracted, the measurement error caused by vibration harmonic interference and attenuation of common-frequency signals is eliminated, and the vortex street flow frequency is accurately extracted;
for lower limiting frequency of measurable flowHas a spectral amplitude of aminIf the largest frequency spectrum amplitude in the current actual peak value is a, when a is less than or equal to 3aminAt the moment, the frequency of the actual peak value is small, the signal-to-noise ratio is poor, and vibration harmonics with frequency variance larger than the flow frequency variance may occur, so that the frequency of the frequency spectrum peak value with the frequency variance larger than the variance threshold value and the largest ratio of the frequency spectrum amplitude and the dynamic cut-off amplitude is extracted as the flow frequency to eliminate harmonic interference; when a is>3aminAt the moment, the frequency of the actual peak value is larger, the signal-to-noise ratio is good, vibration harmonic waves with frequency variance larger than the flow frequency variance cannot occur, and the frequency of the frequency spectrum peak value with frequency variance larger than the variance threshold and the maximum frequency variance can be directly extracted as the flow frequency;
when the variance is greater than the variance threshold value VarthIf the number B _ num of actual peaks in (2) is 1, there is a possibility of a co-frequency situation, and it is necessary to determine whether the co-frequency situation is present, where the determination method is: maximum spectral amplitude A of actual peakmaxWith the smallest spectral amplitude AminDifference d between0＝Amax-AminAnd a set difference threshold DthComparing; when d is0≥DthSetting a same-frequency flag Cofre _ flag, and otherwise, resetting the same-frequency flag; under the condition that a common-frequency flag Cofre _ flag is set, when the number of times that the number of actual peak values is continuously 0 does not exceed 3 times, judging that the current vortex street flow sensor outputs an attenuation data segment of the common-frequency signal, outputting the last output flow frequency as the current flow frequency instead of outputting 0Hz, and eliminating the influence of the attenuation data segment of the common-frequency signal on measurement; when the number of times that the actual peak value number is continuously 0 exceeds 3 times, the fact that the actual peak value number is 0 at the moment is not caused by the attenuation data section of the same-frequency signal is judged, and the current flow frequency is directly output to be 0 Hz.
2. The method for resisting pipeline vibration of the vortex shedding flowmeter by combining the frequency variance calculation and the amplitude calculation as claimed in claim 1, wherein:
under the condition of only giving flow rate, a plurality of different materials are selected in the whole measuring range of the vortex shedding flowmeter,Flow frequency points with certain intervals are tested, the variance value of each flow frequency is calculated, and the minimum flow frequency variance is extracted as VarminI.e. the lower limit of the flow frequency variance; the method comprises the steps of giving vibration to a vortex shedding flowmeter, selecting a plurality of scattered vibration frequency point signals within 10-150Hz of low-frequency vibration interference through a vibration table to perform experiments, respectively collecting the output of a vortex shedding flow sensor under each vibration frequency, performing off-line spectrum analysis, calculating the variance value of each vibration frequency, and extracting the largest vibration frequency variance as VarmaxI.e. the upper limit of the variance of the vibration frequency, the set variance threshold value Var of the vortex street flow sensorthThe calculation formula of (2) is as follows:
Figure FDA0002284724510000011
3. the method for resisting pipeline vibration of the vortex shedding flowmeter by combining the frequency variance calculation and the amplitude calculation as claimed in claim 1, wherein:
saving the frequency of the actual Peak in a Peak frequency array Peak _ fre [ i ] [ j ], where i denotes the number of times the spectrum is calculated, and i ═ 0,1,2, … … 29; j represents the frequency of the j-th actual peak obtained by calculating the frequency spectrum each time, and j is 0,1,2, … … 9, that is, the frequency data of the actual peak is saved at most 30 times, the frequency of the 10 actual peaks with the maximum spectrum amplitude is reserved for calculating the frequency spectrum each time, the frequency spectrum is calculated at most 30 times, and at most 10 peak frequencies are obtained each time; when the frequency of the actual Peak obtained by each calculation is saved, sequencing is carried out according to the sequence of the spectrum amplitudes from large to small, and then Peak _ fre [ i ] [ j ] is sequentially written in, namely Peak _ fre [0] [0] is the Peak frequency with the largest amplitude in the frequencies calculated for the first time, Peak _ fre [0] [1] is the Peak frequency with the largest amplitude in the frequencies calculated for the first time, Peak _ fre [1] [0] is the Peak frequency with the largest amplitude in the frequencies calculated for the second time, Peak _ fre [1] [1] is the Peak frequency with the largest amplitude in the frequencies calculated for the second time, when i is larger than or equal to 1, the frequency spectrum is calculated for at least two times, at the moment, under the condition that the frequency of the actual Peak obtained by the previous and subsequent calculations does not jump, the frequency variance is started to be calculated, so as to ensure the rapid output of the flow rate frequency; and when i is 29, circularly updating the frequency data in the array Peak _ fre [ i ] [ j ] according to a first-in first-out mode, and at the moment, calculating the frequency variance by using 30 frequency data so as to ensure the accuracy of the calculated frequency variance.
4. The method for resisting pipeline vibration of the vortex shedding flowmeter by combining the frequency variance calculation and the amplitude calculation as claimed in claim 1, wherein:
the number of actual Peak values obtained by the i +1 th calculation is recorded as Num, the number of actual Peak values obtained by the i th calculation is recorded as P _ Num, when Num is not equal to P _ Num, namely the number of actual Peak values in the two times before and after changes, the frequency of the actual Peak values in the two times before and after is judged to jump, and Peak _ fre [ i _ fre ] is processed][j]Clearing the medium frequency data; when Num is P _ Num, that is, the number of the actual Peak values in the two previous and subsequent times does not change, because the frequency of the actual Peak value calculated each time is sorted according to the sequence of the spectrum amplitude from large to small, and then is stored in the array Peak _ fre [ i ]][j]Therefore, the absolute value of the difference between the frequency of the first actual peak obtained by the i +1 th calculation and the frequency of the first actual peak obtained by the i-th calculation is calculated first and is denoted as D11(ii) a Let f[i+1][n]Is an array Peak _ fre [ i ]][j]Frequency of the nth actual peak, f, obtained by the (i + 1) th calculation[i][n]Is an array Peak _ fre [ i ]][j]The frequency of the nth actual peak obtained in the ith calculation, where n is 1,2,3, … … 10, then D11＝|f[i+1][1]-f[i][1]L, |; then, the absolute value of the difference between the frequency of the actual peak obtained by the i +1 th calculation and the frequency of the first actual peak obtained by the i-th calculation is calculated and recorded as D1n，D1n＝|f[i+1][n]-f[i][1]L, n ═ 2,3, … … 10; get D1nThe minimum value of (1) is denoted as D1mDenotes D1nThe minimum value in (1) is the absolute value of the difference between the frequency of the mth actual peak obtained by the calculation of the (i + 1) th time and the frequency of the first actual peak obtained by the calculation of the ith time, that is: d1m＝min(D1n) M is more than or equal to 2 and less than or equal to 10; for the spectrum analysis with the sampling frequency of Fs and the data length of N, the frequency resolution delta f is Fs/N; when D is present1m≥D11And D is11<3 deltaf, judging that the frequency of the first actual peak value obtained by the (i + 1) th calculation does not jump; when D is present1m<D11But D is1m<3 deltaf, and judging that the frequency of the first actual peak value obtained by the (i + 1) th calculation does not jump; however, in this case, the frequency interval between the mth frequency in the frequency sequence of the actual peak obtained by the i +1 th calculation and the first actual peak obtained by the i +1 th calculation is smaller, and the position of the mth frequency in the frequency sequence of the actual peak obtained by the i +1 th calculation needs to be exchanged with the first frequency; except the above situation, the frequency of the first actual peak value obtained by the (i + 1) th calculation is considered to jump, and the method for judging whether the frequency of other actual peak values jumps is analogized in turn; only under the condition that all current actual peak values do not jump, judging that the actual peak values obtained by two times of calculation do not jump; if only one actual Peak value jumps, the actual Peak value obtained by two times of calculation before and after the actual Peak value jumps, and Peak _ fre [ i ] is determined to jump][j]And clearing the medium frequency data.
CN201811275401.2A 2018-10-30 2018-10-30 Method and system for preventing pipeline vibration of vortex shedding flowmeter by combining frequency variance calculation and amplitude calculation Active CN109238379B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811275401.2A CN109238379B (en) 2018-10-30 2018-10-30 Method and system for preventing pipeline vibration of vortex shedding flowmeter by combining frequency variance calculation and amplitude calculation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811275401.2A CN109238379B (en) 2018-10-30 2018-10-30 Method and system for preventing pipeline vibration of vortex shedding flowmeter by combining frequency variance calculation and amplitude calculation

Publications (2)

Publication Number Publication Date
CN109238379A CN109238379A (en) 2019-01-18
CN109238379B true CN109238379B (en) 2020-03-24

Family

ID=65079327

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811275401.2A Active CN109238379B (en) 2018-10-30 2018-10-30 Method and system for preventing pipeline vibration of vortex shedding flowmeter by combining frequency variance calculation and amplitude calculation

Country Status (1)

Country Link
CN (1) CN109238379B (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110455465B (en) * 2019-08-14 2020-07-03 合肥工业大学 Frequency fluctuation-based sodium bubble detection signal processing method
CN112781723B (en) * 2021-01-27 2023-09-12 南京微动智测信息技术有限公司 Harmonic component detection method based on frequency spectrum variance
CN113008336B (en) * 2021-02-18 2022-07-12 重庆川仪自动化股份有限公司 Method for realizing permanent magnet sodium flowmeter cross-correlation method in-situ calibration based on error correction
CN113959508A (en) * 2021-10-14 2022-01-21 安徽大学 Method for identifying output transient impact interference of vortex street sensor based on optimal estimation of objective function
CN116608917B (en) * 2023-07-19 2023-09-22 成都秦川物联网科技股份有限公司 Gas ultrasonic metering instrument metering anti-interference method and intelligent gas Internet of things system

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3946608A (en) * 1974-02-26 1976-03-30 Fischer & Porter Company Vortex flowmeter with external sensor
CN1912550A (en) * 2006-08-18 2007-02-14 江苏伟屹电子有限公司 Vibration inteference control method of vortex street flowmeter
CN101644590A (en) * 2009-09-08 2010-02-10 合肥工业大学 Anti-strong interference vortex street flowmeter digital signal processing system based on single sensor
CN103148897A (en) * 2013-02-01 2013-06-12 浙江大学 Digital intelligent vortex shedding flowmeter based on DSP and metering method of same
CN104729591A (en) * 2015-01-16 2015-06-24 合肥工业大学 Signal processing method for vortex shedding flowmeter for resisting low-frequency strong transient impact vibration based on data replacement

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107167195B (en) * 2017-05-24 2019-04-19 西北工业大学 A kind of Coriolis mass flowmeter vibration-starting method based on digital driving method

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3946608A (en) * 1974-02-26 1976-03-30 Fischer & Porter Company Vortex flowmeter with external sensor
CN1912550A (en) * 2006-08-18 2007-02-14 江苏伟屹电子有限公司 Vibration inteference control method of vortex street flowmeter
CN101644590A (en) * 2009-09-08 2010-02-10 合肥工业大学 Anti-strong interference vortex street flowmeter digital signal processing system based on single sensor
CN103148897A (en) * 2013-02-01 2013-06-12 浙江大学 Digital intelligent vortex shedding flowmeter based on DSP and metering method of same
CN104729591A (en) * 2015-01-16 2015-06-24 合肥工业大学 Signal processing method for vortex shedding flowmeter for resisting low-frequency strong transient impact vibration based on data replacement

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
基于恒定磁场的数字式电磁涡街流量计;许伟等;《电子测量与仪器学报》;20180331;第32卷(第3期);第163-168页 *

Also Published As

Publication number Publication date
CN109238379A (en) 2019-01-18

Similar Documents

Publication Publication Date Title
CN109238379B (en) Method and system for preventing pipeline vibration of vortex shedding flowmeter by combining frequency variance calculation and amplitude calculation
US10082249B2 (en) Determining fluid leakage volume in pipelines
CN106679741B (en) Processing method and system based on vortex-shedding meter anti-jamming signal
WO2017143649A1 (en) Kalman filter-based vortex flowmeter anti-transient shock interference signal processing method and system
CN101644590B (en) Anti-strong interference vortex street flowmeter digital signal processing system based on single sensor
CN106090626B (en) A kind of water supply network exception detecting method
CN110162864B (en) Method for predicting system stability of low-lift bidirectional operation pump station
CN104729591A (en) Signal processing method for vortex shedding flowmeter for resisting low-frequency strong transient impact vibration based on data replacement
CN103148899B (en) A kind of liquid micro-flow detection method
CN106123973B (en) Vortex-shedding meter based on cross-correlation analysis measurement low reynolds number flow
CN110631646B (en) Vortex flowmeter supporting flow instability detection
Meng et al. A flow stability evaluation method based on flow-pressure correlation
KR20010056616A (en) A real-time efficiency monitoring system of hydro-turbine generator and pump motor
CN112943639B (en) Method for detecting cavitation failure of pump
Shao et al. Segmented Kalman filter based antistrong transient impact method for vortex flowmeter
CN213808033U (en) Metering pump performance testing device with flow calibration function
JP2907921B2 (en) Valve leak monitoring device
CN114707438A (en) Method, device, apparatus and storage medium for evaluating check valve surface pressure
CN115355449A (en) Monitoring method of pipeline flow, electromagnetic flowmeter and system
EP2009407B1 (en) Converter pulse width shaping circuit and excessive flow rate meter
KR100666491B1 (en) A Frequency Measuring Method with Digital Filtering
CN105973350A (en) Dynamic response characteristic self-checking device of fluid flowmeter
CN205027419U (en) Flow sensor testing platform
JP2765456B2 (en) Pipeline leak detection method
Xu et al. Leakage Monitoring Technology of Oil Pipeline and Its Application

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant