Non-contact pipeline fluid flow velocity measurement method based on frequency wavenumber domain spatial spectrogram

Abstract

The invention relates to the technical field of pipeline detection, and provides a non-contact pipeline fluid flow velocity measurement method based on a frequency wavenumber domain spatial spectrogram, and the method comprises the steps: collecting a turbulence signal in a pipeline through a piezoelectric film sensor; carrying out snapshot number segmentation on the turbulence signal to obtain a plurality of segmented turbulence time domain signals; fourier transform and narrowband signal processing are carried out on the segmented turbulence time domain signals, and a spatial spectrum corresponding to each narrowband signal component is obtained; constructing a three-dimensional frequency-wave number domain spectrogram according to the spatial spectrum function; and according to the three-dimensional frequency-wavenumber domain spectrogram, calculating the flow velocity of the pipeline fluid. According to the method, a three-dimensional spectrogram characteristic space is constructed through conjoint analysis of the frequency and the wave number, the angle, distance and frequency characteristics of the turbulence signals are decoupled through spatial spectrum estimation methods such as the MUSIC algorithm, multipath interference and noise are effectively restrained, the flow speed of fluid in a pipeline can be accurately estimated, and limitation in a traditional method is overcome.

Description

Technical Field

[0001] The present invention relates to the technical field of pipeline detection, and in particular to flow velocity measurement of pipelines containing mud or other complex fluids. Background Art

[0002] During pipeline transportation, accurate flow velocity measurement is crucial for ensuring pipeline safety, improving transportation efficiency, and optimizing construction parameters. However, traditional flow velocity measurement methods (such as electromagnetic flowmeters or differential pressure sensors) are often affected by factors such as pipeline material, fluid composition, and damage to the measuring equipment, making them ineffective for measuring complex fluids. This is particularly true for fluids containing solid particles such as silt. Traditional methods suffer from large measurement errors and high maintenance costs. Furthermore, contact flow measurement devices, whose linings and electrodes are in contact with the measured medium, have a limited service life and require maintenance or replacement, increasing maintenance work and costs.

[0003] In recent years, turbulence signal location technology based on array sensors has become an emerging method for measuring flow velocity in pipes. This method estimates flow velocity by analyzing the signal characteristics of turbulent flow within the pipe. However, existing technologies are typically limited to time-domain or single-frequency domain analysis and lack the combined use of frequency and wavenumber, making it difficult to accurately identify broadband signals and estimate flow velocity.

[0004] Therefore, there is an urgent need for a method that can combine time domain and frequency domain information and accurately locate the source of turbulent signals through frequency-wavenumber domain spatial spectrum, thereby achieving high-precision flow velocity measurement. Summary of the Invention

[0005] The present invention provides a non-contact pipeline fluid flow velocity measurement method based on frequency-wavenumber domain spatial spectrum, which aims to accurately estimate the flow velocity of the fluid in the pipeline by jointly analyzing the frequency and wavenumber of the turbulence signal in the pipeline and combining the fitting algorithm of the three-dimensional frequency-wavenumber domain spectrum.

[0006] In order to achieve the above object, the present invention provides a non-contact pipeline fluid flow velocity measurement method based on frequency wavenumber domain spatial spectrum, comprising:

[0007] (1) Collect turbulence signals in the pipeline through piezoelectric film sensors;

[0008] (2) Perform snapshot splitting on the turbulence signal to obtain turbulence time domain signals of multiple sub-matrices;

[0009] (3) Perform Fourier transform and narrowband signal processing on the turbulence time domain signals of the submatrix to obtain the spatial spectrum corresponding to each narrowband signal component;

[0010] (4) Construct a three-dimensional frequency-wavenumber domain spectrum based on the spatial spectrum function;

[0011] (5) Calculate the flow velocity of the pipeline fluid based on the three-dimensional frequency-wavenumber domain spectrum.

[0012] Furthermore, in step (1), the piezoelectric film sensor is seamlessly attached to the outer wall of the pipeline. Assume that the piezoelectric film sensor contains M flexible film strips, of which The turbulence signal received by the flexible film strip is :

[0013] ;

[0014] in, For the The gain factor of a flexible film strip, is the signal emitted by the turbulence signal source, For the The noise of a flexible film strip.

[0015] Furthermore, in step (2), M flexible film strips are collected simultaneously, and each flexible film strip collects N data, totaling M×N matrix data. Then the N data are divided into L non-overlapping sub-matrices. The data length of a single sub-matrix is J, where J=N / L. Therefore, the turbulent time domain signal of each sub-matrix is for:

[0016] .

[0017] Furthermore, the specific process of step (3) is as follows:

[0018] 31) The first The flexible film strip receives the The turbulence signal is transformed into the turbulence frequency domain signal by Fourier transform : ;

[0019] 32) Turbulence frequency domain signal The frequency is segmented through a bandpass filter to obtain multiple narrowband components, which are expressed as follows:

[0020] X i,l f =[ x i,l f 1 x i,l f 2 ··· x i,l f j ] ;

[0021] in, ;

[0022] 33) Use spatial spectrum estimation (such as MUSIC algorithm) to The signal received by a flexible film strip is divided into narrowband components at j frequency points Analyze and get the corresponding spatial spectrum: for each frequency on the narrowband component , and obtain its corresponding spatial spectrum function through spatial spectrum estimation :When the turbulence of the signal source in the pipe is distributed in different frequency bands and wave numbers, the spatial spectrum function for:

[0023] ;

[0024] in, is the array manifold vector, which represents the signal incident angle generated by the turbulence signal source and the distance between the signal source and the flexible film strip and wave number function;

[0025] It is The noise subspace matrix of the narrowband component at frequency points;

[0026] It is The conjugate transpose of the noise subspace matrix of the narrowband component at the frequency point;

[0027] Represents the conjugate transpose of the array manifold.

[0028] Furthermore, the specific process of step (4) is as follows:

[0029] (41) Obtain the spatial spectrum of each narrowband component: For each frequency on the narrowband component The corresponding spatial spectrum function is obtained by spatial spectrum estimation ;

[0030] (42) Wave number conversion: According to the relationship between wave number and frequency, the spatial spectrum of each frequency band is converted to Spatial spectrum converted to wavenumber domain :

[0031] For near-field signals, the array manifold vector depends on the wave number , then the array manifold vector Expressed as:

[0032] ;

[0033] in, For signal source to The distance between the flexible film strips;

[0034] For the The wave number of the frequency point, =2π / c, where c is the propagation speed of the signal;

[0035] The frequency of each narrowband component Each corresponds to a wave number , then the spatial spectrum function is transformed from wave number :

[0036] ;

[0037] (43) Construct a three-dimensional spectrum: By traversing the frequency and wave number, all the calculated Aggregated together, a three-dimensional frequency-wavenumber domain spectrum is constructed.

[0038] Furthermore, the specific process of step (5) is as follows:

[0039] Based on the three-dimensional frequency-wavenumber domain spectrum, a fitting algorithm (such as the least squares method) is used to calculate the slope of the wavenumber change with frequency. , wave number With frequency The change relationship is approximately linear, and the relationship is:

[0040] ;

[0041] The flow rate The calculation formula is:

[0042] .

[0043] Furthermore, the Kalman filter is used to perform real-time corrections to the velocity estimate. With each snapshot data update, the velocity is calculated and the prediction results are fed back in real time to ensure high accuracy. The optimal estimation of the Kalman filter effectively reduces the impact of non-stationary noise in the turbulent signal (such as pipe vibration and sensor drift) on the velocity measurement.

[0044] The present invention has the following beneficial effects:

[0045] (1) The present invention constructs a three-dimensional spectrum feature space through the joint analysis of frequency and wave number, and uses spatial spectrum estimation methods such as the MUSIC algorithm to decouple the angle, distance and frequency characteristics of the turbulence signal, effectively suppressing multipath interference and noise, and can accurately estimate the flow velocity of the fluid in the pipeline, overcoming the limitations of traditional methods;

[0046] (2) The present invention collects signals through array sensors without direct contact with the fluid, thus avoiding the wear and maintenance problems of traditional flow meters;

[0047] (3) The present invention can update and dynamically correct the flow velocity in the pipeline in real time through dynamic feedback and Kalman filtering;

[0048] (4) The present invention can effectively process broadband signals and is applicable to various pipeline fluids, especially complex fluids such as mud and sewage. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a flow chart of the fluid flow rate measurement in the present invention.

[0050] Figure 2 It is a flow chart of calculating the spatial spectrum in the present invention.

[0051] Figure 3 It is a flow chart of constructing a three-dimensional frequency-wavenumber domain spectrum in the present invention. DETAILED DESCRIPTION

[0052] The technical solution of the present invention is further described in detail below in conjunction with specific embodiments, but this embodiment is not intended to limit the present invention. All similar structures and similar variations of the present invention should be included in the scope of protection of the present invention. The semicolons in the present invention represent the relationship of and, and the English letters in the present invention are case-sensitive.

[0053] like Figure 1 As shown, the present invention provides a non-contact pipeline fluid flow velocity measurement method based on frequency wavenumber domain spatial spectrum, comprising:

[0054] S1, collects turbulence signals in the pipeline through piezoelectric film sensors;

[0055] The piezoelectric film sensor is seamlessly attached to the outer wall of the pipeline. Assume that the piezoelectric film sensor contains M flexible film strips, of which The turbulence signal received by the flexible film strip is :

[0056] ;

[0057] in, For the The gain factor of a flexible film strip, is the signal emitted by the turbulence signal source, For the The noise of a flexible film strip.

[0058] S2, performing snapshot number segmentation on the turbulence signal to obtain turbulence time domain signals of multiple sub-matrices;

[0059] For each flexible film strip, the received turbulence signal includes the turbulence signal source and background noise. After sampling, the turbulence signal is divided into multiple time domain snapshot data. M flexible film strips are collected simultaneously, and each flexible film strip collects N data, totaling M×N matrix data. Then the N data is divided into L non-overlapping sub-matrices, then the data length of a single sub-matrix is J, where J=N / L. Therefore, the turbulence time domain signal of each sub-matrix is for:

[0060] .

[0061] S3, performing Fourier transform and narrowband signal processing on the segmented turbulence time domain signals to obtain the spatial spectrum corresponding to each narrowband signal component;

[0062] like Figure 2 As shown, the specific process is:

[0063] S31, the turbulent time domain signal The flexible film strip receives the The turbulence signal is transformed into the turbulence frequency domain signal by Fourier transform :

[0064] ;

[0065] S32, since pipeline turbulence signals are usually broadband signals containing multiple frequency components, turbulence frequency domain signals Through a bandpass filter, frequency segmentation is performed to obtain multiple narrowband components, which are expressed as follows:

[0066] X i,l f =[ x i,l f 1 x i,l f 2 ··· x i,l f j ] ;

[0067] in, Each narrowband component It can be considered as a narrowband signal, and its bandwidth is much smaller than the total bandwidth of the signal. Through this decomposition method, the individual frequency components of the signal can be analyzed more accurately, thereby improving the accuracy of flow velocity estimation.

[0068] S33, use spatial spectrum estimation (such as MUSIC algorithm) to The signal received by a flexible film strip is divided into narrowband components at j frequency points Analyze and get the corresponding spatial spectrum: for each frequency on the narrowband component , and obtain its corresponding spatial spectrum function through spatial spectrum estimation :When the turbulence of the signal source in the pipe is distributed in different frequency bands and wave numbers, the spatial spectrum function for:

[0069] ;

[0070] in, is the array manifold vector, which represents the signal incident angle generated by the turbulence signal source and the distance between the signal source and the flexible film strip and wave number function;

[0071] It is The noise subspace matrix of the narrowband component at frequency points;

[0072] It is The conjugate transpose of the noise subspace matrix of the narrowband component at the frequency point;

[0073] Represents the conjugate transpose of the array manifold.

[0074] After decomposing the turbulence frequency domain signal (broadband signal) into multiple narrowbands, each narrowband can be processed separately to obtain more precise frequency and spatial spectrum characteristics, thereby obtaining more accurate flow velocity estimation in subsequent steps.

[0075] like Figure 3 As shown in S4, a three-dimensional frequency-wavenumber domain spectrum is constructed based on the spatial spectrum function; the specific process is:

[0076] S41, obtain the spatial spectrum of each narrowband component: for each frequency on the narrowband component The corresponding spatial spectrum function is obtained by spatial spectrum estimation ;

[0077] S42, wave number conversion: According to the relationship between wave number and frequency, the spatial spectrum of each frequency band is converted to Spatial spectrum converted to wavenumber domain Through this wavenumber conversion, the signal source can be located and estimated in the frequency-wavenumber domain, thereby calculating the flow velocity in the pipeline.

[0078] For near-field signals, the array manifold vector depends on the wave number , then the array manifold vector Expressed as:

[0079] ;

[0080] in, For signal source to The distance between the flexible film strips;

[0081] For the The wave number of the frequency point, =2π / c, where c is the propagation speed of the signal;

[0082] The frequency of each narrowband component Each corresponds to a wave number , then the spatial spectrum function is transformed from wave number :

[0083] ;

[0084] Since the position and angle of the signal source vary in each frequency band, the array manifold is different at different wave numbers. The variation in the array manifold vector is crucial for calculating the spatial spectrum. Unlike traditional angle-distance spatial spectrograms, frequency-wavenumber domain spectrograms simultaneously consider the effects of both frequency and wavenumber, providing richer information about the signal source. Using three-dimensional spectrograms, we can effectively identify turbulence sources in different frequency bands and estimate their corresponding wavenumbers and flow velocities.

[0085] S43, construct a three-dimensional spectrum: by traversing the frequency and wave number, all the calculated Aggregated together, a three-dimensional frequency-wavenumber domain spectrum is constructed, which contains the spatial distribution information of the signal source at different frequencies and wavenumbers.

[0086] S5, calculates the flow velocity of the pipeline fluid based on the three-dimensional frequency-wavenumber domain spectrum. The specific process is as follows:

[0087] Based on the three-dimensional frequency-wavenumber domain spectrum, a fitting algorithm (such as the least squares method) is used to calculate the slope of the wavenumber change with frequency. , wave number With frequency The change relationship is approximately linear, and the relationship is:

[0088] ;

[0089] The flow rate The calculation formula is:

[0090] .

[0091] By fitting different frequency and wave number points, the estimated flow velocity of the fluid in the pipeline is obtained. In order to improve the measurement accuracy, multiple snapshot data can be averaged to obtain more stable and accurate flow velocity results.

[0092] Preferably, a Kalman filter is used to perform real-time corrections to the velocity estimate. With each snapshot data update, the velocity is calculated and the prediction results are fed back in real time to ensure high accuracy of the velocity estimate. The optimal estimation of the Kalman filter effectively reduces the impact of non-stationary noise in the turbulent signal (such as pipe vibration and sensor drift) on the velocity measurement.

[0093] First, define the state variables:

[0094] State variables: = ,in, is the flow rate of the pipeline fluid at the kth snapshot moment;

[0095] State transition equation: = + ;

[0096] Where F = 1, F is the state transfer matrix, assuming that the flow rate changes slowly in the short term; is the process noise (obeying the normal distribution N(0,Q), Q is the process noise covariance);

[0097] Measurements: = = , obtained by fitting the frequency-wavenumber domain spectrum of the current snapshot;

[0098] Measurement Model: =H + ;

[0099] Among them, H is the measurement matrix, H+1, is the measurement noise (obeying the normal distribution N(0,R), R is the measurement noise covariance);

[0100] The specific steps are:

[0101] 1) Initial state setting: Determine the initial flow rate based on previous measurements or experience ;

[0102] 2) Extracting measurement values from the frequency-wavenumber domain spectrum: After each snapshot data is processed, the velocity measurement value is obtained by fitting the three-dimensional frequency-wavenumber domain spectrum. , ,in, is the wave number - frequency slope of the current snapshot;

[0103] 3) Predicting flow rate: Using the state transition equation to predict the current state , ;

[0104] 4) Update the flow velocity estimate: Use the current flow velocity measurement obtained from the spatial spectrogram and predicted values , calculate the Kalman gain , and update the velocity estimate and the corresponding covariance;

[0105] Kalman gain for:

[0106] = ; The updated flow rate is:

[0107] = + ( );

[0108] Updated covariance for:

[0109] ;

[0110] in, The prediction covariance matrix; R is the observation noise covariance matrix; H is the observation matrix;

[0111] 5) Real-time and iterative feedback: The updated velocity estimate is used as the initial value for the next prediction, and the flow velocity in the pipeline is continuously monitored and corrected in real time.

[0112] The present invention constructs a three-dimensional spectrum feature space through a joint analysis of frequency and wavenumber, and uses spatial spectrum estimation methods such as the MUSIC algorithm to decouple the angle, distance, and frequency characteristics of the turbulence signal, effectively suppressing multipath interference and noise. It can accurately estimate the flow velocity of the fluid in the pipeline, overcoming the limitations of traditional methods.

[0113] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present application.

Claims

1. A non-contact pipeline fluid flow velocity measurement method based on frequency wavenumber domain spatial spectrum, characterized in that: The measuring method comprises: (1) Collect turbulence signals in the pipeline through piezoelectric film sensors; (2) Perform snapshot splitting on the turbulence signal to obtain turbulence time domain signals of multiple sub-matrices; (3) Perform Fourier transform and narrowband signal processing on the turbulence time domain signals of the submatrix to obtain the spatial spectrum corresponding to each narrowband signal component; (4) Construct a three-dimensional frequency-wavenumber domain spectrum based on the spatial spectrum function; (5) Calculate the flow velocity of the pipeline fluid based on the three-dimensional frequency-wavenumber domain spectrum.

2. The non-contact pipeline fluid flow velocity measurement method based on frequency wavenumber domain spatial spectrum according to claim 1 is characterized in that: In step (1), a piezoelectric film sensor is seamlessly attached to the outer wall of the pipeline. Assume that the piezoelectric film sensor contains M flexible film strips, of which The turbulence signal received by the flexible film strip is : ； in, For the The gain factor of a flexible film strip, is the signal emitted by the turbulence signal source, For the The noise of a flexible film strip.

3. The non-contact pipeline fluid flow velocity measurement method based on frequency wavenumber domain spatial spectrum according to claim 2 is characterized in that: In step (2), M flexible film strips are collected simultaneously, and each flexible film strip collects N data, totaling M×N matrix data. Then the N data are divided into L non-overlapping sub-matrices. The data length of a single sub-matrix is J, where J=N / L. Therefore, the turbulent time domain signal of each sub-matrix is for: 。 4. The non-contact pipeline fluid flow velocity measurement method based on frequency wavenumber domain spatial spectrum according to claim 2 is characterized in that: The specific process of step (3) is: 31) The first The flexible film strip receives the The turbulence signal is transformed into the turbulence frequency domain signal by Fourier transform : ； 32) Turbulence frequency domain signal Through a bandpass filter, frequency segmentation is performed to obtain multiple narrowband components, which are expressed as follows: ； in, Indicates the The signal received by each flexible film strip is divided into narrowband components at j frequency points; 33) Use spatial spectrum estimation to The signal received by a flexible film strip is divided into narrowband components at j frequency points Analyze and get the corresponding spatial spectrum: for each frequency on the narrowband component , and obtain its corresponding spatial spectrum function through spatial spectrum estimation :When the turbulence of the signal source in the pipe is distributed in different frequency bands and wave numbers, the spatial spectrum function for: ； in, is the array manifold vector, which represents the signal incident angle generated by the turbulence signal source and the distance between the signal source and the flexible film strip and wave number function; It is The noise subspace matrix of the narrowband component at frequency points; It is The conjugate transpose of the noise subspace matrix of the narrowband component at the frequency point; Represents the conjugate transpose of the array manifold.

5. The non-contact pipeline fluid flow velocity measurement method based on frequency wavenumber domain spatial spectrum according to claim 1 is characterized in that: The specific process of step (4) is: (41) Obtain the spatial spectrum of each narrowband component: For each frequency on the narrowband component The corresponding spatial spectrum function is obtained by spatial spectrum estimation ; (42) Wave number conversion: According to the relationship between wave number and frequency, the spatial spectrum of each frequency band is converted to Spatial spectrum converted to wavenumber domain : For near-field signals, the array manifold vector depends on the wave number , then the array manifold vector Expressed as: ； in, For signal source to The distance between the flexible film strips; For the The wave number of the frequency point, =2π / c, where c is the propagation speed of the signal; The frequency of each narrowband component Each corresponds to a wave number , then the spatial spectrum function is transformed from wave number : ； (43) Construct a three-dimensional spectrum: By traversing the frequency and wave number, all the calculated Aggregated together, a three-dimensional frequency-wavenumber domain spectrum is constructed.

6. The non-contact pipeline fluid flow velocity measurement method based on frequency wavenumber domain spatial spectrum according to claim 1 is characterized in that: The specific process of step (5) is: According to the three-dimensional frequency-wavenumber domain spectrum, the slope of the wavenumber change with frequency is calculated using the fitting algorithm , wave number With frequency The change relationship is approximately linear, and the relationship is: ； The flow rate The calculation formula is: 。 7. The non-contact pipeline fluid flow velocity measurement method based on frequency wavenumber domain spatial spectrum according to claim 1 is characterized in that: It also includes the use of Kalman filtering to make real-time corrections to flow velocity estimates. Every time the snapshot data is updated, the flow velocity is calculated in real time and the prediction results are fed back to ensure high accuracy of the flow velocity estimation.

Images