A radar ranging method and device based on frequency domain interpolation

Abstract

The application provides a radar ranging method and device based on frequency domain interpolation, the ranging method comprising: acquiring time domain data of radar ranging, performing transformation processing on the time domain data to generate a first frequency spectrum; detecting a candidate peak value index set in the first frequency spectrum, and determining an initial peak value index based on a plurality of candidate peak value index sets; determining a narrow frequency band through the initial peak value index and a frequency resolution, performing frequency domain interpolation in the narrow frequency band based on linear frequency modulation convolution to obtain a second frequency spectrum; extracting a target frequency corresponding to a maximum frequency spectrum value from the second frequency spectrum, and inputting the target frequency into a frequency-distance relationship expression to obtain a detection distance. The radar ranging method provided by the application performs frequency domain interpolation in the narrow frequency band based on linear frequency modulation convolution, realizes local frequency spectrum refinement to lock the target frequency, and obtains a millimeter-level detection distance, thereby not only solving the reliability problem of the radar in measurement, but also improving the ranging accuracy of the radar from a centimeter level to a millimeter level.

Description

Technical Field

[0001] This invention relates to the field of level gauge measurement, and more particularly to a radar ranging method and apparatus based on frequency domain interpolation. Background Technology

[0002] In the field of level gauge measurement, achieving high-precision distance detection is a crucial prerequisite for ensuring production safety, accurate control, and material metering. High-frequency modulated continuous wave radar, with its short wavelength and narrower beam, provides the physical basis for achieving millimeter-level measurement accuracy. However, in actual spectrum analysis, the standard processing method based on Fast Fourier Transform (FFT) inherently suffers from the picket fence effect, causing the true peak frequency to often lie between two discrete sampling points. This inherent frequency deviation becomes the main bottleneck restricting the achievement of millimeter-level accuracy.

[0003] Currently, to overcome the frequency quantization error caused by the picket fence effect, various frequency domain interpolation methods have been developed, typically including parabolic fitting, the centroid method, and high-density fast Fourier transform (FFT). These methods aim to refine the estimation of initially detected spectral peaks. However, in complex real-world measurement scenarios, these traditional interpolation techniques have significant limitations: on the one hand, when the echo signal spectrum peaks are asymmetrical or there are adjacent interference peaks, the estimation errors of parabolic fitting and the centroid method amplify sharply, and stability decreases significantly; on the other hand, although the high-density FFT attempts to improve accuracy by increasing the spectral sampling density, it does not introduce new real signal information and cannot truly break through the theoretical limit of frequency resolution. Instead, the increased computational load makes it difficult to meet the processing requirements of embedded real-time systems, limiting its engineering practicality. Summary of the Invention

[0004] This invention provides a radar ranging method and apparatus based on frequency domain interpolation to solve the technical problems of low accuracy, inability to achieve millimeter-level measurement, poor interpolation stability, and poor practicality in existing radar ranging methods.

[0005] In a first aspect, the present invention provides a radar ranging method based on frequency domain interpolation, comprising: Acquire time-domain data of radar ranging and transform the time-domain data to generate a first spectrum; Detect the candidate peak index set in the first spectrum, and determine the corresponding initial peak index based on the candidate peak index set, wherein the candidate peak index set consists of the index values ​​corresponding to the maxima in the first spectrum; A narrow frequency band is determined based on the initial peak index and frequency resolution, and frequency domain interpolation is performed within the narrow frequency band by linear frequency modulation convolution to obtain a second spectrum, wherein the frequency resolution is determined based on the sampling frequency and the number of sampling points in the first spectrum. Extract the target frequency corresponding to the maximum spectral value from the second spectrum, and input the target frequency into the frequency-distance relationship expression to obtain the detection distance.

[0006] In one embodiment of the present invention, transforming the time-domain data to generate a first spectrum includes: performing a fast Fourier transform on the time-domain data to obtain the first spectrum.

[0007] In one embodiment of the present invention, detecting a set of multiple candidate peak indices in the first spectrum includes: dividing the first spectrum into multiple interval windows and calculating the window mean of each interval window; determining a peak height threshold based on the window mean; filtering out candidate peaks from each interval window according to the peak height threshold, and determining the candidate peak index corresponding to each candidate peak to generate the set of candidate peak indices.

[0008] In one embodiment of the present invention, determining the peak height threshold based on the window mean includes: determining a reference value for each interval window according to the window mean and a preset window adjustment coefficient; and setting a peak height threshold within each interval window according to the reference value and a threshold coefficient.

[0009] In one embodiment of the present invention, determining the corresponding initial peak index based on the candidate peak index set includes: removing duplicate index values ​​from each of the candidate peak index sets to obtain an initial peak index set; and obtaining the index value corresponding to the largest amplitude value in the initial peak index set as the initial peak index.

[0010] In one embodiment of the present invention, after determining the corresponding initial peak index based on the candidate peak index set, the ranging method further includes: if the initial peak index is not unique, adjusting the segmentation parameters of the interval window to re-divide the interval window.

[0011] In one embodiment of the present invention, determining a narrow frequency band based on the initial peak index and frequency resolution includes: determining the half-width of the frequency band based on the frequency resolution and frequency band coefficient; determining the center frequency point based on the initial peak index and the frequency resolution; and dividing the frequency band range of the center frequency point based on the half-width of the frequency band to obtain the narrow frequency band.

[0012] In one embodiment of the present invention, frequency domain interpolation is performed within the narrow frequency band using linear frequency modulated convolution to obtain a second spectrum, comprising: within the narrow frequency band, taking the frequency point corresponding to the initial peak index as the center, determining the frequency band range according to the frequency band coefficient, and constructing a local spectrum vector; constructing a linear frequency modulated convolution kernel based on the index of the local spectrum vector; and multiplying the linear frequency modulated convolution kernel with the local spectrum vector to obtain the second spectrum.

[0013] In one embodiment of the present invention, extracting the target frequency corresponding to the maximum spectral value from the second spectrum, and inputting the target frequency into a frequency-distance relationship expression to obtain the detection distance includes: determining the target index value corresponding to the maximum spectral value in the second spectrum, and mapping the target index value to the corresponding frequency value on the frequency axis to obtain the target frequency; and inputting the target frequency into the frequency-distance relationship expression to obtain the detection distance.

[0014] Secondly, the present invention also provides a radar ranging device based on frequency domain interpolation, the device comprising: The data transformation module is used to acquire the time-domain data of radar ranging and transform the time-domain data to generate a first spectrum. An index determination module is used to detect a set of candidate peak indices in the first spectrum and determine a corresponding initial peak index based on the set of candidate peak indices, wherein the set of candidate peak indices consists of index values ​​corresponding to the maxima in the first spectrum; A frequency domain interpolation module is used to determine a narrow frequency band based on the initial peak index and frequency resolution, and to perform frequency domain interpolation within the narrow frequency band through linear frequency modulation convolution to obtain a second spectrum, wherein the frequency resolution is determined based on the sampling frequency and the number of sampling points in the first spectrum. The distance determination module is used to extract the target frequency corresponding to the maximum spectral value from the second spectrum, input the target frequency into the frequency-distance relationship expression, and obtain the detection distance.

[0015] The beneficial effects of this invention are as follows: This invention provides a radar ranging method and apparatus based on frequency domain interpolation. The ranging method includes: acquiring time-domain data for radar ranging; transforming the time-domain data to generate a first spectrum; detecting a set of candidate peak indices in the first spectrum and determining an initial peak index based on multiple candidate peak index sets; determining a narrow frequency band through the initial peak index and frequency resolution; performing frequency domain interpolation based on linear frequency-modulated convolution within the narrow frequency band to obtain a second spectrum; extracting the target frequency corresponding to the maximum spectral value from the second spectrum; inputting the target frequency into a frequency-distance relationship expression to obtain the detection range. The radar ranging method provided by this invention, within a narrow frequency band, performs frequency domain interpolation based on linear frequency-modulated convolution to refine the local spectrum, thereby locking the target frequency and obtaining a millimeter-level detection range. This ranging method can solve the reliability problem of radar ranging and improve the ranging accuracy from centimeter-level to millimeter-level, thus breaking through the accuracy limitations of traditional radar and upgrading radar from a coarse sensing instrument to a precision measuring instrument. Attached Figure Description

[0016] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application. It is obvious that the drawings described below are merely some embodiments of this application, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.

[0017] In the attached diagram: Figure 1 This is a flowchart of a radar ranging method based on frequency domain interpolation provided in one embodiment of the present invention; Figure 2 This is a schematic diagram of the specific process of a radar ranging method based on frequency domain interpolation provided in one embodiment of the present invention; Figure 3 This is a block diagram of a radar ranging device based on frequency domain interpolation provided in one embodiment of the present invention. Detailed Implementation

[0018] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0019] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. The drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0020] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0021] In the field of level gauge measurement, achieving high-precision distance detection is a crucial prerequisite for ensuring production safety, accurate control, and material metering. High-frequency modulated continuous wave radar, with its short wavelength and narrower beam, provides the physical basis for achieving millimeter-level measurement accuracy. However, in actual spectrum analysis, the standard processing method based on Fast Fourier Transform (FFT) inherently suffers from the picket fence effect, causing the true peak frequency to often lie between two discrete sampling points. This inherent frequency deviation becomes the main bottleneck restricting the achievement of millimeter-level accuracy.

[0022] Currently, to overcome the frequency quantization error caused by the picket fence effect, various frequency domain interpolation methods have been developed, such as parabolic fitting, the centroid method, and high-density fast Fourier transform. These methods aim to refine the estimation of initially detected spectral peaks. However, in complex actual measurement scenarios, the above-mentioned traditional interpolation techniques have obvious limitations: on the one hand, when the echo signal spectrum peaks are asymmetrical or there are adjacent interference peaks, the estimation errors of parabolic fitting and the centroid method will be amplified sharply, and the stability will decrease significantly; on the other hand, although the high-density fast Fourier transform method attempts to improve accuracy by increasing the spectral sampling density, it does not introduce new real signal information and cannot truly break through the theoretical limit of frequency resolution. Instead, the increased computational load makes it difficult to meet the processing requirements of embedded real-time systems, thus limiting its engineering practicality.

[0023] To solve the above problems, such as Figure 1 As shown, this application provides a radar ranging method based on frequency domain interpolation, which includes at least steps S110 to S140: S110. Acquire the time-domain data of radar ranging and transform the time-domain data to generate the first spectrum.

[0024] In detail, the time-domain data is transformed to generate the first spectrum, including performing a Fast Fourier Transform on the time-domain data to obtain the first spectrum. Specifically, as follows: Figure 2 As shown, in step S201, the radar acquires time-domain data in real time. In step S202, the time-domain data is transformed, that is, the time-domain data is transformed into frequency-domain data through fast Fourier transform to obtain discrete amplitude information. In step S203, it is determined whether the data transformation of the time-domain data is completed. If the frequency-domain transformation is not completed, the time-domain data transformation process continues. After the data transformation is completed, the first spectrum is obtained.

[0025] The expression for the first spectrum is as follows:

[0026] Where Y[k] is the spectral value corresponding to frequency point k in the first spectrum. It is time-domain data with sampling time n, where n is the sampling time in the time-domain data, N is the number of discrete points in the first spectrum, k is the index on the first spectrum, k is an integer, 0≤k≤N-1, and j represents the imaginary unit.

[0027] S120. Detect the candidate peak index set in the first spectrum, and determine the corresponding initial peak index based on multiple candidate peak index sets, wherein the candidate peak index set consists of the index value corresponding to the maximum value in the first spectrum.

[0028] For example, the first spectrum is divided into interval windows, peak detection is performed on each interval window to obtain multiple candidate peak index sets, and the initial peak index is locked based on the multiple candidate peak index sets.

[0029] In detail, detecting multiple candidate peak index sets in the first spectrum includes: dividing the first spectrum into multiple interval windows and calculating the window mean of each interval window; determining a peak height threshold based on the window mean; filtering candidate peaks from each interval window according to the peak height threshold, and determining the candidate peak index corresponding to each candidate peak to generate a candidate peak index set. Specifically, as... Figure 2 As shown, in step S204, the parameters of the interval window are set: the length of a single interval window is W, and the sliding step size of the interval window is S. The first spectrum is divided into multiple interval windows. The expression for determining the number of interval windows is as follows:

[0030] Where M is the number of interval windows, N is the number of discrete points in the first spectrum, W is the length of a single interval window, and S is the sliding step size between two adjacent interval windows.

[0031] In step S205, the spectrum within each interval window is scanned, that is, the mean of the frequency domain data corresponding to each interval window is calculated to obtain the window mean of each interval window. The expression for calculating the window mean is as follows:

[0032] in, Let S represent the mean of the t-th interval window, W be the length of the interval window, and S be the mean of the t-th interval window. t Y represents the starting index of the t-th interval window. i Let represent the i-th spectral data in the first spectrum, where t is a positive integer and 1 ≤ t ≤ M.

[0033] The peak height threshold is calculated based on the window mean. In step S206, candidate peaks within each interval window are identified, that is, the data within each interval window is filtered based on the peak height threshold to obtain candidate peaks within each interval window. A corresponding candidate peak index is determined for each candidate peak, and multiple candidate peak indices form a candidate peak index set.

[0034] More specifically, determining the peak height threshold based on the window mean includes: determining a baseline value for each interval window based on the average value of each window and a preset window adjustment coefficient; and setting the peak height threshold within each interval window based on the baseline value and a threshold coefficient. Specifically, based on the average value of each window... The baseline value of each interval window is calculated using the preset window adjustment coefficient α. And based on this benchmark value The threshold coefficient β sets the peak height threshold within each interval window. .

[0035] benchmark value and peak height threshold As shown below:

[0036]

[0037] in, Let be the baseline value for the t-th interval window. This is the preset window adjustment coefficient. Let be the window mean of the t-th interval window. Let β be the peak height threshold for the t-th interval window, and β be the threshold coefficient.

[0038] More specifically, determining the corresponding initial peak index based on the candidate peak index set includes: removing duplicate index values ​​from each candidate peak index set to obtain the initial peak index set; and obtaining the index value corresponding to the largest amplitude value in the initial peak index set as the initial peak index. Specifically, as... Figure 2 As shown, in step S207, the initial peak index is determined by the candidate peak index set, that is: removing duplicate index values ​​from the candidate peak index set to obtain the initial peak index set, and taking the index value with the largest amplitude in the initial peak index set as the initial peak index.

[0039] The formula for determining the candidate peak index set is as follows:

[0040] in, Represents the set of candidate peak indices. This represents the index value corresponding to the candidate peak within a single interval window. This indicates the number of candidate peaks within a single interval window. This represents the candidate peak value within the t-th interval window. This represents the peak height threshold of the t-th interval window, where q is a positive integer, 1 ≤ q ≤ m. t .

[0041] The formula for determining the initial peak index set is as follows:

[0042] in, Represents the initial set of peak indices. Let M represent the set of candidate peak indices, M represent the total number of interval windows, and t represent the t-th interval window.

[0043] In the initial peak index set Φ, if there are multiple candidate peak indices, the candidate peak index corresponding to the candidate peak with the largest amplitude is taken as the initial peak index; if there is only one candidate peak index, it is directly taken as the initial peak index.

[0044] The expression for determining the initial peak index is as follows:

[0045] in, Indicates the initial peak index. Let Φ represent the frequency amplitude of the candidate peak, Φ represent the initial peak index set, and p represent the index value in the initial peak index set. It is the initial peak index set The number of elements in the index, p1 indicates that there is only one candidate peak in the initial peak index.

[0046] More specifically, after determining the corresponding initial peak index based on the candidate peak index set, the ranging method further includes: if the initial peak index is not unique, adjusting the segmentation parameters of the interval window to re-divide the interval window. Specifically, as... Figure 2 As shown, in step S208, it is verified whether the initial peak index is unique. If there are two or more initial peak indices, in step S209, the window length W and sliding step size S of the interval window are adjusted to redivide the interval window so that the obtained initial peak index is unique.

[0047] S130. Determine the narrow frequency band based on the initial peak index and frequency resolution, and perform frequency domain interpolation within the narrow frequency band using linear frequency modulation convolution to obtain the second spectrum, wherein the frequency resolution is determined based on the sampling frequency and number of sampling points in the first spectrum.

[0048] In detail, determining the narrow frequency band based on the initial peak index and frequency resolution includes: obtaining the sampling frequency and the number of sampling points, and determining the frequency resolution based on the sampling frequency and the number of sampling points; determining the half-width of the bandwidth based on the frequency resolution and the bandwidth coefficient; determining the center frequency point based on the initial peak index and the frequency resolution; and dividing the bandwidth range of the center frequency point based on the half-width of the bandwidth to obtain the narrow frequency band. Specifically, when only one initial peak index exists, such as... Figure 2 As shown, in step S210, a narrow frequency band is generated and the sampling frequency is obtained. And the number of sampling points G, the sampling frequency The ratio of the frequency resolution to the number of sampling points G is used as the frequency resolution. Where the number of sampling points G is the number of discrete points in the first spectrum, G=N; the frequency resolution and frequency band coefficient Multiplication calculation of half-width of the frequency band ;Initial peak index and frequency resolution Multiplication to calculate center frequency point ; with the center frequency point Centered at the frequency spectrum, half-width of the bandwidth The upper and lower limits of the spectrum are adjusted to obtain a narrow frequency band.

[0049] Frequency resolution The deterministic expression is as follows:

[0050] in, For frequency resolution, G represents the sampling frequency, and G represents the number of sampling points.

[0051] Bandwidth half-width The deterministic expression is as follows:

[0052] in, For half the bandwidth, This refers to the frequency resolution.

[0053] Center frequency point The deterministic expression is as follows:

[0054] in, Center frequency point For the initial peak index, This refers to the frequency resolution.

[0055] The formula for determining a narrow frequency band is as follows:

[0056] in, For a narrow frequency band, Center frequency point It is half the bandwidth.

[0057] More specifically, the second spectrum is obtained by frequency domain interpolation through linear frequency modulation convolution within a narrow frequency band, including: within the narrow frequency band, taking the frequency point corresponding to the initial peak index as the center, determining the frequency band range according to the frequency band coefficient, and constructing a local spectrum vector; constructing a linear frequency modulation convolution kernel based on the index of the local spectrum vector; and multiplying the linear frequency modulation convolution kernel with the local spectrum vector to form the second spectrum.

[0058] Specifically, γ spectral data points located within a narrow frequency band are extracted from the first spectrum, and the frequency band coefficients are calculated using the initial peak index as the center of the frequency point. The number of data points extending to both sides of the frequency point center is used to construct a local spectrum vector.

[0059] The formula for determining the γ spectral data within the narrow frequency band is as follows: γ=

[0060] Where γ is the number of spectral data points extracted within the narrow frequency band. .

[0061] Local spectral vector The deterministic expression is as follows:

[0062] in, , For the initial peak index, Let T be the local spectrum vector, and T be the transpose.

[0063] Construct a complex matrix R modulated by both a quadratic phase function and a linear phase function as the linear frequency modulation convolution kernel. Linear frequency modulated convolution kernel The definite expressions are shown in the following three expressions:

[0064]

[0065]

[0066] in, It is the frequency modulation factor, used to control the frequency domain interpolation accuracy and sidelobe level. These are parameters that control the starting frequency and frequency step. These are the parameters that control linear frequency modulation, and D is the total number of spectral data points in the local spectral vector. It is a data index in the local spectrum vector. It is the index of the second spectrum. This represents the total number of spectral data points in the second spectrum, i.e., the number of frequency points included after refining the local spectral vector. Here, j represents the imaginary unit. It is a linear frequency modulated convolution kernel.

[0067] like Figure 2 As shown, in step S211, spectral interpolation is performed based on linear frequency modulation convolution to transform the local spectral vector. With linear frequency modulated convolution kernel Matrix multiplication is performed, and the result is equivalent to inserting H high-resolution frequency points within a narrow frequency band. In step S212, a second spectrum Z is generated. The expression for determining the second spectrum Z is shown in the following three expressions:

[0068]

[0069]

[0070] in, In the local spectrum vector, the first... One spectrum data, Indicates the index in the second spectrum Z as The corresponding spectrum data, For the initial peak index, j represents the imaginary unit, and D is the total number of spectral data points in the local spectral vector. These are parameters that control the starting frequency and frequency step. These are the parameters that control linear frequency modulation. It is a linear frequency modulated convolution kernel.

[0071] S140. Extract the target frequency corresponding to the maximum spectral value from the second spectrum, input the target frequency into the frequency-distance relationship expression, and obtain the detection range.

[0072] In detail, the target frequency corresponding to the maximum spectral value is extracted from the second spectrum, and the target frequency is input into the frequency-distance relationship expression to obtain the detection range. This includes: determining the target index value corresponding to the maximum spectral value in the second spectrum, mapping the target index value to the corresponding frequency value on the frequency axis to obtain the target frequency; and inputting the target frequency into the frequency-distance relationship expression to obtain the detection range.

[0073] Specifically, peak detection is performed on the frequency data in the second spectrum to obtain the maximum value of the spectral data in the second spectrum, and the target frequency index is locked based on the maximum value of the spectral data. The deterministic expression is as follows:

[0074] in, Indicates the target frequency index. Indicates the index in the second spectrum is The corresponding amplitude value, This is the index value in the second spectrum. For integers, 0 ≤ ≤H-1, where H is the total number of indices in the second spectrum.

[0075] Input the target frequency index into the index-frequency conversion formula to map the target frequency index onto the frequency axis, and obtain the target frequency. .

[0076] The index-frequency conversion formula is shown in the following expression:

[0077] in, Represents the spectrum transformation expression function. Indicates the center frequency point. For half the bandwidth, U represents the index value in the second spectrum, where U represents the second spectrum Z[ The total number of frequency points in the [].

[0078] Target frequency The definite expression is shown in the following expression:

[0079] in, Represents the spectrum transformation expression function. Indicates the target frequency index.

[0080] like Figure 2 As shown, in step S213, the mapping calculation from target frequency to distance is performed, and the target frequency is... Input the frequency-distance relationship expression, and in step S214, obtain the detection distance. Detection distance The deterministic expression is as follows:

[0081] in, It's the speed of light. It is the frequency modulation period. It is frequency modulation bandwidth. .

[0082] like Figure 3 As shown, the present invention also provides a radar ranging device based on frequency domain interpolation, the device comprising: The data transformation module 310 is used to acquire the time-domain data of radar ranging and transform the time-domain data to generate the first spectrum; The index determination module 320 is used to detect the candidate peak index set in the first spectrum and determine the corresponding initial peak index based on multiple candidate peak index sets, wherein the candidate peak index set is composed of the index value corresponding to the maximum value in the first spectrum; The frequency domain interpolation module 330 is used to determine a narrow frequency band based on an initial peak index and frequency resolution, and to perform frequency domain interpolation within the narrow frequency band by linear frequency modulation convolution to obtain a second spectrum, wherein the frequency resolution is determined based on the sampling frequency and the number of sampling points in the first spectrum; The distance determination module 340 is used to extract the target frequency corresponding to the maximum spectral value from the second spectrum, input the target frequency into the frequency-distance relationship expression, and obtain the detection distance.

[0083] It should be noted that the radar ranging device based on frequency domain interpolation provided in the above embodiments and the radar ranging method based on frequency domain interpolation provided in the above embodiments belong to the same concept. The specific way of performing each step has been described in detail in the system embodiments, and will not be repeated here.

[0084] For example, in specific testing, a complete test and verification system conforming to GB / T38620-2020 "Performance Evaluation Method for Level Gauges" is constructed, and the test index requirements of frequency-modulated radar level gauges are used as the judgment criteria. The device includes a vertical column with precisely controllable height, a standard reflective target plate, a temperature and humidity controllable environment, and a dual-frequency laser interferometer. The device (with micrometer-level accuracy) is used as a length measurement standard in the experiment, and the test environment temperature is... (degrees Celsius), relative humidity is The radar prototype under test and the laser interferometer target are mounted on the same optical platform.

[0085] Align the radar prototype and the laser interferometer parallel to the same target reflector. Move the target reflector using a precision guide rail, with each test point approximately 1 meter apart, ranging from 1 to 30 meters. At each test point, after the display readings stabilize, record three measurements from the radar prototype and the standard values ​​from the laser interferometer, as shown in Tables 1 and 2. These tables represent the ranging accuracy results of the two prototype radars based on the frequency domain interpolation provided by this invention. The units of the values ​​in the tables are... (millimeters), the first data of each prototype radar is taken from data located after zero.

[0086] Table 1 Ranging accuracy results of prototype radar 1

[0087] Table 2 Ranging accuracy results of prototype radar 2

[0088] Based on Tables 1 and 2, it can be seen that, using the frequency domain interpolation-based radar ranging technology provided by this invention, the maximum error is within the full 30m range. .

[0089] This invention provides a radar ranging method and apparatus based on frequency domain interpolation. The ranging method includes: acquiring time-domain data for radar ranging; transforming the time-domain data to generate a first spectrum; detecting a set of candidate peak indices in the first spectrum and determining an initial peak index based on multiple candidate peak index sets; determining a narrow frequency band using the initial peak index and frequency resolution; performing frequency domain interpolation based on linear frequency-modulated convolution within the narrow frequency band to obtain a second spectrum; extracting the target frequency corresponding to the maximum spectral value from the second spectrum; inputting the target frequency into a frequency-range relationship expression to obtain the detection range. The radar ranging method provided by this invention, within a narrow frequency band, performs frequency domain interpolation based on linear frequency-modulated convolution to refine the local spectrum, thereby locking the target frequency and obtaining a millimeter-level detection range. This ranging method can solve the reliability problem of radar ranging and improve the ranging accuracy from centimeter-level to millimeter-level, thus breaking through the accuracy limitations of traditional radar and upgrading radar from a coarse sensing instrument to a precision measuring instrument.

[0090] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A radar ranging method based on frequency domain interpolation, characterized in that, include: Acquire time-domain data of radar ranging and transform the time-domain data to generate a first spectrum; Detect the candidate peak index set in the first spectrum, and determine the corresponding initial peak index based on the candidate peak index set, wherein the candidate peak index set consists of the index values ​​corresponding to the maxima in the first spectrum; A narrow frequency band is determined based on the initial peak index and frequency resolution, and frequency domain interpolation is performed within the narrow frequency band by linear frequency modulation convolution to obtain a second spectrum, wherein the frequency resolution is determined based on the sampling frequency and the number of sampling points in the first spectrum. Extract the target frequency corresponding to the maximum spectral value from the second spectrum, and input the target frequency into the frequency-distance relationship expression to obtain the detection distance.

2. The radar ranging method based on frequency domain interpolation according to claim 1, characterized in that, The time-domain data is transformed to generate a first spectrum, including: The first spectrum is obtained by performing a fast Fourier transform on the time-domain data.

3. The radar ranging method based on frequency domain interpolation according to claim 1, characterized in that, Detecting a set of multiple candidate peak indices in the first spectrum, including: The first spectrum is divided into multiple interval windows, and the window mean of each interval window is calculated; The peak height threshold is determined based on the window mean. Based on the peak height threshold, candidate peaks are selected from each interval window, and the candidate peak index corresponding to each candidate peak is determined to generate the candidate peak index set.

4. The radar ranging method based on frequency domain interpolation according to claim 3, characterized in that, Determining the peak height threshold based on the window mean includes: The baseline value of each interval window is determined based on the mean value of each window and the preset window adjustment coefficient. The peak height threshold within each interval window is set according to each of the aforementioned benchmark values ​​and threshold coefficients.

5. The radar ranging method based on frequency domain interpolation according to claim 3, characterized in that, Determining the corresponding initial peak index based on the candidate peak index set includes: By removing duplicate index values ​​from each of the candidate peak index sets, an initial peak index set is obtained; Obtain the index value corresponding to the largest amplitude value in the initial peak index set, and use it as the initial peak index.

6. The radar ranging method based on frequency domain interpolation according to claim 5, characterized in that, After determining the corresponding initial peak index based on the candidate peak index set, the ranging method further includes: if the initial peak index is not unique, adjusting the segmentation parameters of the interval window to re-divide the interval window.

7. The radar ranging method based on frequency domain interpolation according to claim 1, characterized in that, Determining the narrow frequency band based on the initial peak index and frequency resolution includes: The half-width of the bandwidth is determined based on the frequency resolution and the bandwidth coefficient. The center frequency point is determined based on the initial peak index and the frequency resolution. The narrow frequency band is obtained by dividing the center frequency point into frequency band ranges based on the half-width of the frequency band.

8. The radar ranging method based on frequency domain interpolation according to claim 7, characterized in that, A second spectrum is obtained by frequency domain interpolation through linear frequency-modulated convolution within the narrow frequency band, including: Within the narrow frequency band, taking the frequency point corresponding to the initial peak index as the center, the frequency band range is determined according to the frequency band coefficient, and a local spectrum vector is constructed. A linear frequency-modulated convolution kernel is constructed based on the index of the local spectrum vector; The second spectrum is obtained by multiplying the linear frequency modulated convolution kernel with the local spectrum vector.

9. The radar ranging method based on frequency domain interpolation according to claim 8, characterized in that, Extract the target frequency corresponding to the maximum spectral value from the second spectrum, input the target frequency into the frequency-distance relationship expression, and obtain the detection range, including: Determine the target index value corresponding to the maximum spectral value in the second spectrum, and map the target index value to the corresponding frequency value on the frequency axis to obtain the target frequency; The target frequency is input into the frequency-distance relationship expression to obtain the detection distance.

10. A radar ranging device based on frequency domain interpolation, characterized in that, The device includes: The data transformation module is used to acquire the time-domain data of radar ranging and transform the time-domain data to generate a first spectrum. Detect the candidate peak index set in the first spectrum, and determine the corresponding initial peak index based on the candidate peak index set, wherein the candidate peak index set consists of the index values ​​corresponding to the maxima in the first spectrum; A narrow frequency band is determined based on the initial peak index and frequency resolution, and frequency domain interpolation is performed within the narrow frequency band by linear frequency modulation convolution to obtain a second spectrum, wherein the frequency resolution is determined based on the sampling frequency and the number of sampling points in the first spectrum. The distance determination module is used to extract the target frequency corresponding to the maximum spectral value from the second spectrum, input the target frequency into the frequency-distance relationship expression, and obtain the detection distance.

Images