A relative time-of-flight estimation method based on ultrasonic flowmeter

Abstract

This invention relates to the field of ultrasonic flow measurement technology. This application discloses a relative time-of-flight estimation method based on an ultrasonic flow meter. The method includes: aligning the uplink and downlink echo signals in the time domain using prior absolute time of flight to prevent phase skipping; performing FFT transformation and frequency domain conjugate multiplication on the signals to extract amplitude and phase information; constructing a cost function using spectral amplitude weighting; and estimating the residual delay through least-squares fitting to finally obtain a high-precision relative time of flight. This method performs correlation operations in the frequency domain, significantly reducing computational complexity, improving estimation accuracy and noise immunity, and is suitable for embedded low-power systems.

Description

Technical Field

[0001] This invention relates to the field of ultrasonic flow measurement technology, and in particular to a relative time-of-flight estimation method based on an ultrasonic flow meter. Background Technology

[0002] When ultrasound propagates in a fluid medium, the propagation speed of the up and down echo signals is affected by the flow velocity of the fluid medium itself, resulting in a difference in the propagation time of the up and down ultrasound echo signals. The magnitude of the difference is related to the fluid flow velocity. Using the above principle, the fluid flow velocity can be calculated by the time difference (relative flight time) of the up and down propagation of the ultrasound signal in the fluid.

[0003] Existing methods for calculating fluid velocity are mainly divided into two types: TDC and ADC.

[0004] TDC method such as Figure 4 As shown, this method estimates the zero-crossing time of the first wave of the uplink and downlink echo signals by using high-frequency counting pulses, and records them as the absolute flight time of the uplink and downlink signals respectively. Then, the difference between the two is used to obtain the relative flight time. Since the TDC method is a method of directly estimating the arrival time of the first wave of ultrasound, it cannot effectively suppress the influence of noise or interference. At the same time, since the intensity of the first wave of ultrasound signal is usually weak, the existing TDC method has the disadvantages of insufficient estimation stability, possible "cycle skipping", and low estimation accuracy.

[0005] The ADC method is as follows: Figure 5 As shown, this method directly performs correlation processing on the uplink and downlink echo signals in the time domain. The specific signal processing flow is shown in the figure below: First, the uplink and downlink ultrasonic echo signals are correlated in the time domain, and then dToF is calculated based on the peak position of the time domain correlation. Although the ADC method directly performs correlation processing on the uplink and downlink echo signals in the time domain, which solves the problems of dToF estimation accuracy and stability, the huge amount of time domain correlation calculation results in increased hardware resources and power consumption.

[0006] Therefore, those skilled in the art need a relative time-of-flight estimation method that improves the accuracy of relative time-of-flight (dToF) estimation of uplink and downlink ultrasonic signals and reduces computational complexity during ultrasonic flow meter signal processing. Summary of the Invention

[0007] The purpose of this invention is to solve the above problems by designing a relative flight time estimation method based on an ultrasonic flow meter.

[0008] The technical solution of the present invention to achieve the above objectives is a relative time-of-flight estimation method based on an ultrasonic flow meter, which includes the following steps:

[0009] Step 1: Based on the estimated absolute flight time of the uplink and downlink echo signals, perform time-domain alignment preprocessing on the uplink and downlink echo signals;

[0010] Step 2: Perform Fast Fourier Transform (FFT) on the preprocessed uplink and downlink echo signals respectively to obtain the frequency domain signals;

[0011] Step 3: Perform conjugate multiplication on the frequency domain signal to obtain the frequency domain correlation result;

[0012] Step four: Extract the amplitude and phase information of the frequency domain correlation results and determine the effective frequency band;

[0013] Step 5: Construct a weighted cost function based on the frequency points and phase values ​​within the effective frequency band;

[0014] Step 6: Fit the cost function using the least squares method to estimate the slope;

[0015] Step 7: Calculate the residual delay estimate based on the slope, and then combine the coarsely estimated relative flight time with the residual delay estimate to obtain the final relative flight time estimate.

[0016] The time-domain alignment preprocessing involves shifting the downlink echo signal in the time domain based on the estimated absolute flight time of the uplink and downlink echo signals, so that it is roughly aligned with the uplink echo signal in the time domain.

[0017] The method for determining the effective frequency band is as follows: calculate the amplitude spectrum of the frequency domain correlation results, find the frequency point where the amplitude maximum value is located, and select a certain range of frequency points as the effective frequency band with this frequency point as the center.

[0018] The effective frequency band is selected within a range of 50 frequency points before and after the peak frequency point.

[0019] The weighted cost function is constructed as follows: for each frequency point, the square of the amplitude of the frequency domain correlation result is used as the weight to perform weighted least squares fitting on the linear relationship between phase and frequency.

[0020] The mathematical expression for the weighted cost function is:

[0021] ;

[0022] In the formula, The result of the conjugate multiplication. For phase value, For frequency index, The slope is denoted as .

[0023] The slope The calculation formula is:

[0024] .

[0025] The residual delay estimate The calculation formula is:

[0026] ;

[0027] In the formula, For FFT points, The sampling rate of the echo signal.

[0028] The final relative flight time estimate is the sum of the coarse relative flight time estimate and the residual delay estimate.

[0029] The formula for calculating the final relative flight time estimate is as follows:

[0030] ;

[0031] In the formula, This is a rough estimate of the relative flight time.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] 1. By performing correlation operations in the frequency domain, this application significantly suppresses the impact of noise on the dToF estimation accuracy compared to the TDC scheme, thereby improving the dToF estimation accuracy.

[0034] 2. In constructing the cost function, this application uses a spectral magnitude weighting method to maximize the use of each effective frequency component in the ultrasonic signal, thus achieving more reliable estimation accuracy;

[0035] 3. This application transforms the computationally complex time-domain correlation operations to be performed in the frequency domain, which significantly reduces the computational complexity compared to existing ADC schemes. Attached Figure Description

[0036] Figure 1 This is a flowchart of a relative time-of-flight estimation method based on an ultrasonic flow meter as described in this invention;

[0037] Figure 2 This is a schematic diagram of the cost function construction described in this invention;

[0038] Figure 3 This is the frequency-phase relationship diagram described in this invention;

[0039] Figure 4 This is a schematic diagram of the signal processing of the TDC method described in this invention;

[0040] Figure 5This is a flowchart of the signal processing of the ADC method described in this invention;

[0041] Figure 6 This is the amplitude information diagram described in Embodiment 1 of the present invention. Detailed Implementation

[0042] The present invention will now be described in detail with reference to the accompanying drawings, such as... Figure 1-3 As shown;

[0043] The main innovations of this application include: First, using frequency domain correlation operations instead of time domain correlation; Second, using prior information from absolute time-of-flight estimation to preprocess uplink and downlink ultrasonic echo signals to prevent "cycle skipping" of phase information during frequency domain correlation; Third, using the proportion of spectral amplitude information to construct a weighted cost function to fully utilize the effective frequency components of the echo signal; Fourth, using the least squares algorithm to calculate the relative time of flight.

[0044] When estimating the flow rate of a medium using ultrasound, it is necessary to estimate the relative time of flight (dToF) of the uplink and downlink echo signals. To improve the accuracy of dToF estimation, existing ADC methods directly perform correlation operations on the uplink and downlink echo signals in the time domain. However, this method requires significant computational resources, leading to increased power consumption and resource costs. This application innovatively proposes a frequency domain correlation method. The specific steps of this method are as follows:

[0045] Step 1: Based on the estimated absolute flight time of the uplink and downlink echo signals, adjust the time delays of the two echo signals to align them. This is done to prevent "cycle skipping" during phase calculation in the frequency domain.

[0046] Among them, the uplink echo signal:

[0047] ;

[0048] The downlink echo signal can be approximated as:

[0049] ;

[0050] Step 2: Perform FFT transformation on the uplink and downlink signals:

[0051] ;

[0052] ;

[0053] Step 3: Perform conjugate multiplication on the two signals after the FFT transformation in Step 2:

[0054] ;

[0055] According to the properties of the Fourier transform, time-domain delay is equivalent to frequency-domain phase rotation. As shown in the formula above, the phase difference increases linearly with frequency (the effective frequency component), such as... Figure 3 As shown, this is the frequency-phase relationship diagram.

[0056] Step 4: Calculate the amplitude and phase information of the conjugate multiplication result in Step 3, and search for the position of the maximum amplitude, which is the position where the effective frequency component of the echo signal has the highest proportion.

[0057] Step 5: Construct the cost function.

[0058] According to step three, phase information increases linearly with frequency. Therefore, the relationship between frequency-domain correlated phase difference and frequency can be modeled as follows:

[0059] ;

[0060] Where x is the frequency index and y is the frequency domain correlation phase value. Slope and delay Positive correlation. The slope is the decimal sample delay estimate (or it may be a decimal + integer sample estimate, collectively referred to as the residual delay estimate).

[0061] Based on the frequency range near the relevant peak positions found in step four, the cost function is constructed as follows using a spectral magnitude weighting method:

[0062] ;

[0063] in This represents the frequency domain correlation amplitude.

[0064] Step 6: Use the least squares method to perform curve fitting estimation. By taking the partial derivative of the cost function, the slope can be obtained. The calculation formula is as follows:

[0065] ;

[0066] Step 7: Calculate the slope from Step 5. Converted to residual delay estimate:

[0067] ;

[0068] Where NFFT is the number of FFT points. The sampling rate of the echo signal.

[0069] The final delay estimate is:

[0070] ;

[0071] dTOF is a rough estimate of the uplink and downlink delay, which is the size of the uplink and downlink echo signal delay alignment time in step one.

[0072] The above steps can be found in the appendix. Figure 1 The signal processing flowchart in the diagram.

[0073] This application solves the problems of large estimation error and high computational resource complexity in the existing dToF technology, and significantly improves the estimation accuracy of dToF without consuming too many computational resources.

[0074] Example 1;

[0075] This embodiment uses a typical ultrasonic flow meter as an example, with a sampling frequency F. s =10MHz, FFT points NFFT=512, ultrasonic center frequency f c =0.5MHz, bandwidth is 0.1MHz.

[0076] 1. Signal Acquisition and Preprocessing:

[0077] The uplink echo signal usig and the downlink echo signal dsig are collected, each with a length of 512 points;

[0078] Based on prior absolute time-of-flight estimates (e.g., through thresholding or envelope detection), assume the uplink signal delay is t. up =200μs, downlink signal delay is t dn =205μs, then the rough estimate of dToF is:

[0079] ;

[0080] Shift dsig in the time domain to make it roughly aligned with usig in the time domain, thus avoiding phase jumps in the frequency domain.

[0081] 2. Frequency Domain Transformation and Conjugate Multiplication

[0082] Perform FFT on usig and dsig respectively to obtain the frequency domain representation U. x (k) and D x (k).

[0083] Calculate the conjugate multiplication:

[0084] ;

[0085] The complex result at each frequency point is obtained, including amplitude and phase information, such as Figure 3 and Figure 6 As shown.

[0086] 3. Find effective frequency bands

[0087] Calculate |R ud (k)∣, and find the frequency point k where its maximum value is located. peak .

[0088] Choose k peak Six frequency points before and after (a total of 13 points) are used as effective frequency bands for subsequent fitting.

[0089] 4. Construct the weighted cost function

[0090] Define frequency index x k =k, phase y k =∠R ud (k).

[0091] Construct the weighted least squares cost function:

[0092] ;

[0093] The weight is |R ud (k)∣ 2 It emphasizes frequencies with high signal-to-noise ratio.

[0094] 5. Least squares fitting

[0095] Calculate the slope a:

[0096] ;

[0097] Assume that the calculated value of a is 0.023.

[0098] 6. Residual Delay Estimation

[0099] Convert the slope 'a' to residual delay:

[0100] ;

[0101] The final dToF estimate is: τ = dTOF + τ1 = -5 + 0.46855 = -4.53145 μs

[0102] 7. Effect Comparison

[0103] Compared with traditional time-domain correlation methods, this method improves the dToF estimation accuracy by about 300% under the same signal-to-noise ratio and at low flow velocities compared with the aToF subtraction method; it can also prevent cycle skipping by 100%; and the computational complexity is reduced by about 40% compared with the direct time-domain correlation method for calculating dToF.

[0104] This embodiment is applicable to embedded system implementation and can run in a low-power microcontroller.

[0105] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, the phrase "comprising an element defined as..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0106] The above technical solutions only embody the preferred technical solutions of the present invention. Any modifications that may be made by those skilled in the art to certain parts thereof embody the principles of the present invention and fall within the protection scope of the present invention.

Claims

1. A method of relative time-of-flight estimation based on an ultrasonic flow meter, characterized in that, The method comprises the following steps: Step one, according to the absolute time of flight estimation value of the uplink and downlink echo signals, the uplink and downlink echo signals are preprocessed in time domain alignment; Step two, the preprocessed uplink and downlink echo signals are respectively subjected to fast Fourier transform to obtain frequency domain signals; Step three, the frequency domain signals are subjected to conjugate multiplication operation to obtain frequency domain correlation results; Step four, the amplitude and phase information of the frequency domain correlation results are extracted, and an effective frequency band is determined; Step five, based on the frequency points and phase values in the effective frequency band, a weighted cost function is constructed, and the construction method of the weighted cost function is that, for each frequency point, the amplitude square of the frequency domain correlation result is taken as the weight, and the linear relationship between the phase and the frequency is subjected to weighted least square fitting; The mathematical expression of the weighted cost function is: ； wherein is a conjugate multiplication result, is a phase value, is a frequency index, is a slope; Step six, the least square method is used to fit the cost function to estimate the slope; Step seven, according to the slope, a residual delay estimation value is calculated, then the coarse relative time of flight is combined with the residual delay estimation value to calculate the final relative time of flight estimation result.

2. A relative time of flight estimation method based on an ultrasonic flow meter according to claim 1, characterized in that, The time domain alignment preprocessing is that, according to the absolute time of flight estimation value of the uplink and downlink echo signals, the downlink echo signal is subjected to time domain translation to make it roughly aligned with the uplink echo signal in time domain.

3. A relative time of flight estimation method based on an ultrasonic flow meter according to claim 1, wherein, The determination method of the effective frequency band is that, the amplitude spectrum of the frequency domain correlation result is calculated, and the frequency point where the amplitude maximum value is located is found, and the frequency points within a certain range are selected as the effective frequency band with the frequency point as the center.

4. A relative time of flight estimation method based on an ultrasonic flow meter according to claim 3, characterized in that, The selected range of the effective frequency band is that, 50 frequency points before and after the peak frequency point.

5. A relative time of flight estimation method based on an ultrasonic flow meter as defined in claim 1, wherein, The slope The calculation formula is: 。 6. A relative time of flight estimation method based on an ultrasonic flow meter as defined in claim 1, wherein, said residual delay estimate The formula for calculating is: ； In the formula, is the number of FFT points, is the sampling rate of the echo signal.

7. A relative time of flight estimation method based on an ultrasonic flow meter as defined in claim 1, wherein, The final relative time of flight estimation value is the sum of the coarse relative time of flight and the residual delay estimation value.

8. A relative time of flight estimation method based on an ultrasonic flow meter according to claim 7, characterized in that, The calculation formula of the final relative time of flight estimation value is: ； In the formula, is a rough estimate of the relative flight time.

Images