High-precision Level Measurement Method for Millimeter-wave Radar Using Measured Interpolation Calibration

Abstract

This invention provides a high-precision millimeter-wave radar level measurement method using measured interpolation calibration. This method coherently integrates two received echo signals of different ranges to obtain a spectrum signal with a signal-to-noise ratio (SNR) equal to the number of accumulated signals. Noise and interference signals are then removed using CAFR, and secondary interference removal is performed on the signal using a set threshold. A range-range matching method is then used to optimize the array of long-range peak points to obtain a high-precision array of peak points. Finally, an optimized cubic spline is used to interpolate to all possible target points to determine the target distance. This method not only improves target detection speed but also enhances ranging accuracy across different distance segments, enabling reliable measurement of target distance.

Description

Technical Field

[0001] The present invention belongs to the technical field of radar signal processing, and in particular relates to a high-precision level measurement method of a millimeter-wave radar using actual measurement interpolation calibration. Background Art

[0002] In actual use, the actual accuracy of the level meter will be reduced during the target distance measurement process due to factors such as application scenario requirements, insufficient hardware system performance, limited device power and external noise interference. In particular, when the device hardware only supports a certain number of FFT points, the measurement accuracy will decrease as the maximum measurement distance increases. Therefore, traditional methods find it difficult to maintain measurement accuracy for targets in the same distance range under different ranges.

[0003] Analysis of field measurement data reveals that, during actual level meter measurement, the target within the tank moves relatively slowly. Since the time it takes for a signal to be transmitted and the echo signal to be received is in the millisecond range, it can be assumed that the target has not moved from the start to the end of the signal transmission. While the limited number of FFT points doesn't guarantee consistent accuracy across the full range, as long as the target's distance segment is known, the target's ability to remain stationary for a short period of time can be exploited to achieve consistent accuracy across the entire range.

[0004] Traditional methods for measuring target distance using level meters often use either a range-based accuracy or a coarse-and-fine search approach. The former sets different accuracy levels for different ranges, resulting in the lowest accuracy for all distance segments within the range. This is because, for a given FFT number of points, as the maximum measurement range increases, the signal's maximum frequency also increases, widening the grid width, reducing the grid's precision, and consequently decreasing accuracy. The latter requires two signal transmissions per measurement: the first to search for the target's approximate distance and range, and the second to measure the target's specific location by transmitting a smaller signal. This method increases device power consumption and signal transmission and program execution time. Furthermore, in environments with high levels of noise and interference, it can easily misjudge the target's approximate distance, affecting the next fine search step and causing misjudgment. Furthermore, due to inherent hardware system influences, level meter measurement results generally do not change linearly with target movement. This means that the level meter's measurement result and the target's actual distance are not linearly related. This can result in actual measurement errors exceeding the theoretical error.

[0005] For this type of problem, traditional ranging methods cannot achieve high-precision measurement results. Therefore, it is very important to design and research a high-precision millimeter-wave radar level measurement method with excellent performance and robustness and using actual measurement interpolation calibration. Summary of the Invention

[0006] In order to solve the above problems existing in the prior art, the present invention provides a high-precision level measurement method of millimeter-wave radar using actual measurement interpolation calibration. The technical problem to be solved by the present invention is achieved through the following technical solutions:

[0007] The present invention provides a high-precision level measurement method of a millimeter-wave radar using actual measurement interpolation calibration, comprising:

[0008] Step 1: Use the radar transmitter to transmit two ranges of transmission signals according to preset transmission parameters, and receive the returned two ranges of echo signals;

[0009] Step 2: Send the echo signals of the two ranges to the hardware accelerator for FFT transformation respectively, and obtain the spectrum data of the two ranges after FFT transformation;

[0010] Step 3: The spectrum data of the two ranges are coherently accumulated to improve the signal-to-noise ratio of the spectrum, and the spectrum data of the two ranges after the signal-to-noise ratio is changed are obtained;

[0011] Step 4: Send the spectrum data of the two ranges after the signal-to-noise ratio changes to the hardware accelerator for CFRA detection, and obtain the CFAR curve arrays corresponding to the spectrum data of the two ranges;

[0012] Step 5: Determine the value in each group of CFAR curve arrays that is higher than the CFRA spectrum as the effective value, determine the position sequence number of the effective value in the curve array as the subscript of the effective value, and obtain the one-dimensional peak point array corresponding to each range;

[0013] Step 6: In the one-dimensional peak point array corresponding to each measurement range, the peak points that are higher than the peak threshold are selected to form a peak point array, thereby obtaining a peak point array corresponding to each measurement range;

[0014] Step 7: Optimize the peak point array corresponding to the high range by using the distance range matching method and the peak point array corresponding to the low range in each range, and obtain the optimized peak point array corresponding to the high range;

[0015] Step 8: Use optimized cubic spline interpolation to interpolate the peak point array corresponding to the optimized high range to obtain all possible target points;

[0016] Step 9: Determine the actual distance between the actual target and the radar transmitter among all possible target points.

[0017] Beneficial effects of the present invention:

[0018] 1. The present invention provides a high-precision millimeter-wave radar level measurement method using measured interpolation calibration. This method coherently integrates the signals received by the IWR1443BOOST to obtain a spectrum signal with a signal-to-noise ratio equal to the accumulated number. Noise and interference signals are then removed using CAFR. A set threshold is used to perform secondary interference removal on the signal. A range-matching method is then used to establish a corresponding relationship and optimize the long-range peak point array to obtain a high-precision peak point array. Finally, an optimized cubic spline is used to interpolate to all possible target points to determine the target distance. This method not only improves the speed of target detection but also improves the ranging accuracy at different distances, achieving reliable measurement of target distance.

[0019] 2. This invention proposes an improved cubic spline interpolation method that can meet the needs of the low computing power of IWR1443BOOST and the industrial requirements of low energy consumption. It can quickly, accurately and conveniently interpolate and correct the measurement results, and accurately obtain the true distance result of the target.

[0020] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A flowchart of a high-precision level measurement method for millimeter-wave radar using measured interpolation calibration is provided in an embodiment of the present invention;

[0022] Figure 2 The present invention provides a method for high-precision level measurement using millimeter-wave radar interpolation calibration, which is a schematic diagram of the process. DETAILED DESCRIPTION

[0023] The present invention will be further described in detail below with reference to specific examples, but the embodiments of the present invention are not limited thereto.

[0024] like Figure 1 As shown, the present invention provides a high-precision millimeter wave radar level measurement method using measured interpolation calibration, which includes:

[0025] Step 1: Use the radar transmitter to transmit two ranges of transmission signals according to preset transmission parameters, and receive the returned two ranges of echo signals;

[0026] The two ranges of echo signals are respectively a first echo signal and a second echo signal, and the range of the first echo signal is higher than the range of the second echo signal.

[0027] It is worth noting that the parameters of IWR1443BOOST are configured according to the actual range requirements. The waveform parameters selected in the example of the present invention are: large range: 23m; starting carrier frequency: f = 77GHz, sampling rate Fs = 2MHz, bandwidth B = 3.865GHz, modulation frequency Kr = 13.03MHz / us, period Ts = 306.85us, number of sampling points: 593, number of waveform transmissions: 8; small range: 13m starting carrier frequency: f = 77GHz, sampling rate Fs = 2MHz, bandwidth B = 3.7653GHz, modulation frequency Kr = 23.03MHz / us, period Ts = 173.68us, number of sampling points: 327, number of waveform transmissions: 8. The relationship between the maximum measurement distance and sampling rate is as follows:

[0028]

[0029] Where K r K is the frequency modulation r =B / (T s -IdleTime), B is bandwidth, T s is the cycle, IdleTime is the waiting time, the waiting time, i.e. the idle time, is set by yourself. In the present invention, it can be set to 10e(-6)s), Fs is the sampling rate, C is the speed of light, which is 3e8m / s, and d_max is the maximum ranging range.

[0030] Step 2: Send the echo signals of the two ranges to the hardware accelerator for FFT transformation respectively, and obtain the spectrum data of the two ranges after FFT transformation;

[0031] The number of FFT points used in the present invention may be FFT_num=4096.

[0032] Step 3: The spectrum data of the two ranges are coherently accumulated to improve the signal-to-noise ratio of the spectrum, and the spectrum data of the two ranges after the signal-to-noise ratio is changed are obtained;

[0033] The large and small range spectrum data are retrieved from the memory separately. Because the echo energy of a single radar pulse is limited, a single received pulse is usually not used to detect and judge the target. Before judging, multiple pulse trains of a wave position are processed to improve the signal-to-noise ratio. The signal-to-noise ratio of the spectrum is improved through coherent accumulation. The so-called coherent accumulation takes into account the phase information of the received data of each cycle. The signal-to-noise ratio after coherent accumulation of multi-cycle signals is:

[0034]

[0035] M is the pulse accumulation number, A 2 is the single pulse echo signal power, σ 2is the noise power, that is, the signal-to-noise ratio is expanded M times after coherent integration. The spectrum data of the large and small range waveforms after accumulation are obtained;

[0036] Step 4: Send the spectrum data of the two ranges after the signal-to-noise ratio changes to the hardware accelerator for CFRA detection, and obtain the CFAR curve arrays corresponding to the spectrum data of the two ranges;

[0037] It is worth noting that the spectra of the large- and small-range waveforms were fed into the hardware accelerator to obtain their CFAR results. The CFAR detector used was a CA-CFAR detector with the following parameters: detection window length: 32, protection window length: 2, noise shift: 5, and nominal factor: 1.5.

[0038] Step 5: Determine the value in each group of CFAR curve arrays that is higher than the CFRA spectrum as the effective value, determine the position sequence number of the effective value in the curve array as the subscript of the effective value, and obtain the one-dimensional peak point array corresponding to each range;

[0039] It is worth noting that the result of CFAR is an array of CFAR curves with the same number of spectrum points. The obtained spectrum is screened by the CFAR curve, and the spectrum higher than the CFAR is recorded as a valid value. Its value and subscript are recorded to obtain a one-dimensional array of large and small ranges respectively.

[0040] Step 6: In the one-dimensional peak point array corresponding to each measurement range, the peak points that are higher than the peak threshold are selected to form a peak point array, thereby obtaining a peak point array corresponding to each measurement range;

[0041] As an optional embodiment of the present invention, step 6 includes:

[0042] Step 6-1: Determine whether the first point in the one-dimensional peak point array corresponding to each range is a peak point, whether the last point is a peak point, and for each current point except the first and last points, determine whether it is a peak point;

[0043] Step 6-11: In the one-dimensional peak point array corresponding to each range, determine whether the right adjacent point of the first point is smaller than the first point. If so, the first point is the peak point; otherwise, it is not a peak point.

[0044] Step 6-12: In the one-dimensional peak point array corresponding to each range, determine whether the peak value of the adjacent point on the left of the last point is smaller than the last point. If so, the last point is the peak point, otherwise it is not the peak point;

[0045] Step 6-13: For each current point except the first point and the last point, if there are no points larger than the current point on its left and right sides, and there are points smaller than the current point, then the current point is determined to be a peak point, otherwise it is not a peak point.

[0046] Step 6-2: Compare the determined peak point with the set peak threshold, remove the peak points that are smaller than the peak threshold, and obtain the peak point array corresponding to each range.

[0047] As an optional embodiment of the present invention, step 6-1 includes:

[0048] Step 7: Use the distance range matching method to establish the corresponding relationship and the peak point array corresponding to the low range in each range, optimize the peak point array corresponding to the high range, and obtain the optimized peak point array corresponding to the high range;

[0049] As an optional embodiment of the present invention, step 7 includes:

[0050] Step 7-1: Set the distance difference;

[0051] Step 7-2: Using the peak point array corresponding to the first echo signal as the standard group, select the first subscript of the first peak point in the standard group and the second subscript of the second peak point in the peak point array corresponding to the second echo signal respectively; if the distance between the first subscript and the second subscript is less than the distance interpolation value, use the distance between the first subscript and the second subscript as the distance difference to update the distance difference; and record the subscript of the selected second peak point in the second peak point array;

[0052] Step 7-3: Repeat step 7-2 until all first peak points in the second peak point array are traversed, and select the subscript of the second peak point with the smallest distance difference;

[0053] Step 7-4: Compare the minimum distance difference with the preset target threshold. If the minimum distance difference is greater than the inverse of the target threshold and less than the target threshold, replace the subscript of the first peak point corresponding to the minimum distance difference with the subscript of the second peak point until all first peak points in the first peak point array are replaced, thereby obtaining the optimized first peak point array.

[0054] It should be noted that: a target threshold K is set. Generally, the value of K is set as d_max / FFT_num, where d_max is the maximum ranging range and FFT_num is the number of points for FFT transformation; alternatively, large and small range waveforms can be used to measure the same target simultaneously to obtain the distance difference between the two, and this distance difference can be used as K, and the calculated difference value can be used as K. Compare the minimum distance difference dist_diff with K. If -K < dist_diff < K, update the peak point subscript of the large range to the subscript value recorded by the small range. If -K < dist_diff < K is not satisfied, no update is performed.

[0055] As an optional implementation manner of the present invention, the preset target threshold in step 7-4 is obtained through the following steps:

[0056] Step 7-41: Use the first peak point array and the second peak point array to measure the same target respectively, calculate the difference between the measurement results of the two peak point arrays, and determine this difference as the target threshold;

[0057] Step 7-42: Calculate the target threshold using the following formula:

[0058] where K = d_max / FFT_num, K represents the target threshold, d_max represents the maximum ranging range, and FFT_num represents the number of FFT points.

[0059] Step 8: Use optimized cubic spline interpolation for the peak point array corresponding to the optimized high range to obtain all possible target points;

[0060] As an optional implementation manner of the present invention, step 8 includes:

[0061] Step 8-1: Obtain a two-dimensional array for calibration;

[0062] where the first row y i , (i = 0, 1,..., n + 1) is the true distance of the target, and the second row x i , (i = 0, 1,..., n + 1) is the target distance measured by the device;

[0063] Step 8-2: Select the subscript g of any first peak point from the optimized first peak point array, and select four sets of data x0, x1, x2, x3 from the second row x i such that x0 < x1 < g < x2 < x3; and select four sets of data y0, y1, y2, y3 with the same column from the first row y i ;

[0064] Step 8-3: Calculate the step size h i = x i+1–x i , (i=0, 1, 2);

[0065] Step 8-4: Set the step length h i As the path point, the four sets of data y0, y1, y2, y3 are selected as breakpoints and substituted into the matrix equation to obtain the quadratic differential values m2 and m1;

[0066] Among them, the matrix equation is:

[0067]

[0068] Step 8-5: Using the quadratic differential values m2 and m1, calculate the cubic spline curve coefficients of the second segment;

[0069] Among them, the cubic spline coefficients are:

[0070] a1=y1

[0071]

[0072]

[0073]

[0074] The subscript g in step 8-6 is interpolated to:

[0075] f(g)=a1+b1(g-x1)+c1(g-x1) 2 +d1(g-x1) 3 .

[0076] Step 8-6: Use the cubic spline curve coefficients to calculate the interpolated value of subscript g.

[0077] The optimized cubic spline interpolation is used to interpolate the optimized large-scale peak point subscript array obtained in the above steps to obtain all possible target points. The error simulation results of the optimized cubic spline interpolation and the traditional cubic spline interpolation are shown in Table 1.

[0078] Table 1

[0079] Cubic spline interpolation results 0.398 0.4541 0.741 0.8082 0.9498 Optimized cubic spline interpolation results 0.398017223 0.454319 0.741245 0.808423 0.950155 error -1.7223E-05 -0.00022 -0.00025 -0.00022 -0.00035 Cubic spline interpolation results 1.0041 1.2417 3.5256 4.7301 5.1278 Optimized cubic spline interpolation results 1.004477 1.242028 3.525712 4.730353 5.12815 error -0.00038 -0.00033 -0.00011 -0.00025 -0.00035 Cubic spline interpolation results 6.327 7.9277 8.3315 9.9316 Optimized cubic spline interpolation results 6.327274 7.92812 8.331627 9.931726 error -0.00027 -0.00042 -0.00013 -0.00013

[0080] Step 9: Determine the actual distance between the actual target and the radar transmitter among all possible target points.

[0081] It is worth noting that the present invention can select an appropriate peak selection decision. Since the measured target may have a relatively small dielectric constant, that is, the target peak is not necessarily the largest, and it may also be the first peak in the peak array. Therefore, it is necessary to select an appropriate peak selection decision to find the distance represented by the peak of the real target. The final measured results are shown in Table 2.

[0082] Table 2

[0083]

[0084]

[0085] As can be seen from Tables 1 and 2, the present invention provides a high-precision millimeter-wave radar level measurement method using measured interpolation calibration, which realizes segmented high-precision measurement of the target. The method simultaneously sends two waveforms of different ranges within a certain period of time, and obtains two sets of peak arrays after receiving the echo. Using the distance range matching method, with the measurement results of the large range as a reference, after the signal processing results, the accuracy of the large-range results is optimized by the waveform results of the small range, and finally the improved cubic spline interpolation is used to obtain the true distance of the target. The effectiveness of this method was verified by simulation experiments, which not only improved the speed of target measurement, but also improved the measurement accuracy of a large range and short distance. Finally, it was successfully put into practical use and achieved the desired effect.

[0086] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature identified as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0087] Although the present application is described herein with reference to various embodiments, those skilled in the art will be able to understand and implement other variations of the disclosed embodiments in practicing the claimed application by reviewing the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude a plurality.

[0088] The above is a further detailed description of the present invention in conjunction with specific preferred embodiments, and the specific implementation of the present invention should not be considered to be limited to these descriptions. For those skilled in the art of the present invention, without departing from the concept of the present invention, several simple deductions or substitutions can be made, which should be considered to fall within the scope of protection of the present invention.

Claims

1. A high-precision level measurement method of millimeter-wave radar using measured interpolation calibration, characterized in that: include: Step 1: Use the radar transmitter to transmit two ranges of transmission signals according to preset transmission parameters, and receive the returned two ranges of echo signals; Step 2: Send the echo signals of the two ranges to the hardware accelerator for FFT transformation respectively, and obtain the spectrum data of the two ranges after FFT transformation; Step 3: The spectrum data of the two ranges are coherently accumulated to improve the signal-to-noise ratio of the spectrum, and the spectrum data of the two ranges after the signal-to-noise ratio is changed are obtained; Step 4: Send the spectrum data of the two ranges after the signal-to-noise ratio changes to the hardware accelerator for CFRA detection, and obtain the CFAR curve arrays corresponding to the spectrum data of the two ranges; Step 5: Determine the value in each group of CFAR curve arrays that is higher than the CFRA spectrum as the effective value, determine the position sequence number of the effective value in the curve array as the subscript of the effective value, and obtain the one-dimensional peak point array corresponding to each range; Step 6: In the one-dimensional peak point array corresponding to each measurement range, the peak points that are higher than the peak threshold are selected to form a peak point array, thereby obtaining a peak point array corresponding to each measurement range; Step 7: Optimize the peak point array corresponding to the high range by using the distance range matching method and the peak point array corresponding to the low range in each range, and obtain the optimized peak point array corresponding to the high range; Step 8: Use optimized cubic spline interpolation to interpolate the peak point array corresponding to the optimized high range to obtain all possible target points; Step 9: Determine the actual distance between the actual target and the radar transmitter among all possible target points.

2. The high-precision level measurement method of millimeter-wave radar using measured interpolation calibration according to claim 1 is characterized in that: The two ranges of echo signals are respectively a first echo signal and a second echo signal, and the range of the first echo signal is higher than the range of the second echo signal.

3. The high-precision level measurement method of millimeter-wave radar using measured interpolation calibration according to claim 2 is characterized in that: The step 6 comprises: Step 6-1: Determine whether the first point in the one-dimensional peak point array corresponding to each range is a peak point, whether the last point is a peak point, and for each current point except the first and last points, determine whether it is a peak point; Step 6-2: Compare the determined peak point with the set peak threshold, remove the peak points that are smaller than the peak threshold, and obtain the peak point array corresponding to each range.

4. The high-precision level measurement method of millimeter-wave radar using measured interpolation calibration according to claim 3 is characterized in that: The step 6-1 includes: Step 6-11: In the one-dimensional peak point array corresponding to each range, determine whether the right adjacent point of the first point is smaller than the first point. If so, the first point is the peak point; otherwise, it is not a peak point. Step 6-12: In the one-dimensional peak point array corresponding to each range, determine whether the peak value of the adjacent point on the left of the last point is smaller than the last point. If so, the last point is the peak point, otherwise it is not the peak point; Step 6-13: For each current point except the first point and the last point, if there are no points larger than the current point on its left and right sides, and there are points smaller than the current point, then the current point is determined to be a peak point, otherwise it is not a peak point.

5. The high-precision level measurement method of millimeter-wave radar using measured interpolation calibration according to claim 2, characterized in that: The step 7 comprises: Step 7-1: Set the distance difference; Step 7-2: Using the peak point array corresponding to the first echo signal as the standard group, select the first subscript of the first peak point in the standard group and the second subscript of the second peak point in the peak point array corresponding to the second echo signal respectively; if the distance between the first subscript and the second subscript is less than the distance interpolation value, use the distance between the first subscript and the second subscript as the distance difference to update the distance difference; and record the subscript of the selected second peak point in the second peak point array; Step 7-3: Repeat step 7-2 until all first peak points in the second peak point array are traversed, and select the subscript of the second peak point with the smallest distance difference; Step 7-4: Compare the minimum distance difference with the preset target threshold. If the minimum distance difference is greater than the inverse of the target threshold and less than the target threshold, replace the subscript of the first peak point corresponding to the minimum distance difference with the subscript of the second peak point until all first peak points in the first peak point array are replaced, thereby obtaining the optimized first peak point array.

6. The high-precision level measurement method of millimeter-wave radar using measured interpolation calibration according to claim 3 is characterized in that: The target threshold value preset in step 7-4 is obtained through the following steps: Step 7-41: using the first peak point array and the second peak point array to measure the same target, respectively, calculating the difference between the measurement results of the two sets of peak point arrays, and determining the difference as the target threshold; Step 7-42: Calculate the target threshold using the following formula: Wherein, K=d_max / FFT_num, K represents the target threshold, d_max represents the maximum ranging range, and the number of FFT points.

7. The high-precision level measurement method of millimeter-wave radar using measured interpolation calibration according to claim 3 is characterized in that: The step 8 comprises: Step 8-1: Obtain a set of two-dimensional arrays for calibration; Among them, the first row y of the two-dimensional array i , (i=0, 1, ..., n+1) is the actual distance of the target, the second row x i , (i=0, 1, ..., n+1) is the target distance measured by the device; Step 8-2: Select the subscript g of any first peak point from the optimized first peak point array, and select four groups of data x0, x1, x2, x3 from the second row x i such that x0 < x1 < g < x2 < x3; and select four groups of data y0, y1, y2, y3 with the same columns from the first row y i ; Step 8-3: Calculate the step length h between the four sets of data x0, x1, x2, and x3 i =x i+1 –x i , (i=0, 1, 2); Step 8-4: Set the step length h i As the path point, the four sets of data y0, y1, y2, y3 are selected as breakpoints and substituted into the matrix equation to obtain the quadratic differential values m2 and m1; Step 8-5: Use the quadratic differential values m2 and m1 to calculate the cubic spline curve coefficients of the second segment: Step 8-6: Use the cubic spline curve coefficients to calculate the interpolated value of subscript g.

8. The high-precision level measurement method of millimeter-wave radar using measured interpolation calibration according to claim 7, characterized in that: The matrix equation of step 8-4 is: The cubic spline curve coefficients in step 8-5 are: a1=y1 The subscript g in step 8-6 is interpolated to: f(g)=a1+b1(g-x1)+c1(g-x1) 2 +d1(g-x1) 3 。

Images