Radar phase difference unmanned aerial vehicle height measurement method based on frequency spectrum refinement calibration

Abstract

The invention belongs to the field of unmanned aerial vehicle technologies and radars, and discloses a radar phase difference unmanned aerial vehicle height measurement method based on frequency spectrum refinement calibration, which comprises the following steps: receiving a radar echo carrier frequency signal, and forming a radar data cube by an intermediate frequency signal after frequency mixing processing; performing time-frequency domain conversion on the data cube distance dimension to obtain a radar one-dimensional distance image sequence; interference clutters are filtered based on a constant false alarm detection technology, and a maximum peak value in a range is searched to determine a distance threshold of an observation area; performing frequency spectrum refinement processing on a specific frequency band of the observation area by using Chirp-Z transformation, constructing a quadratic polynomial in combination with parabola fitting to perform spectral line calibration, and obtaining frequency information of the observation area after frequency spectrum refinement calibration; and extracting a phase sequence of an observation area distance threshold, unwrapping adjacent frame phases by using phase unwrapping, and determining height information of the unmanned aerial vehicle relative to the ground by combining frequency and phase information. The unmanned aerial vehicle can be assisted to quickly feed back the relative ground height, and the collision risk is avoided.

Description

Technical Field

[0001] The present invention relates to the field of UAV technology and radar, and in particular to a radar phase difference UAV altitude measurement method based on spectrum refinement calibration. Background Art

[0002] With the rapid development of computer science and UAV technology, improving the safety and stability of UAV operation has become the key to current research. UAVs are widely used in agricultural irrigation, power inspection, logistics and transportation due to their economic efficiency, high flexibility and high adaptability. During the take-off and landing process of UAVs, it is necessary to obtain the ground height information in real time to adjust the take-off and landing speed and achieve precise control.

[0003] At present, drone altitude measurement usually adopts: barometer measurement, RTK measurement and lidar measurement. However, the drone flight environment is complex and changeable. The barometer altitude measurement technology is affected by air pressure changes and airflow, and can only obtain an approximate value of the flight altitude. When using RTK technology to measure the altitude of a drone, it is necessary to ensure that the communication between the drone and the base station is stable, which is easily affected by satellite conditions and electromagnetic interference. Lidar has a shielding effect in rainy and foggy weather, has poor perception ability, and is not suitable for complex environments. Summary of the invention

[0004] The purpose of the present invention is to provide a radar phase difference UAV altitude measurement method based on spectrum refinement calibration, which can take into account the altitude measurement of UAVs in different complex environments. The method provides high-precision, high-resolution and high-adaptability detection at a relatively low economic cost, and assists UAVs to quickly feedback ground altitude and avoid collision risks.

[0005] In order to achieve the above object, the present invention adopts the following solution:

[0006] A radar phase difference UAV altitude measurement method based on spectrum refinement calibration includes the following steps:

[0007] Step M1: receiving radar echo carrier frequency signals, and mixing the intermediate frequency signals to form radar data cubes;

[0008] Step M2: Use discrete Fourier transform to transform the distance dimension of the data cube into time-frequency domain to obtain a radar one-dimensional range image sequence;

[0009] Step M3: Filter out ground interference clutter based on constant false alarm detection technology, traverse the spectrum peaks, and search for the maximum value of the peak within the range to determine the distance threshold of the observation area;

[0010] Step M4: Use Chirp-Z transform to perform spectrum refinement processing on a specific frequency band in the observation area, and construct a quadratic polynomial in combination with parabola fitting to perform spectrum line calibration, so as to obtain the frequency information of the observation area after spectrum refinement and calibration;

[0011] Step M5: Extract the phase sequence of the distance threshold of the observation area, use phase unwrapping to unfold the phases of adjacent frames, and combine the frequency and phase information to determine the relative height information of the UAV to the ground.

[0012] The intermediate frequency signal S after the mixing process in step M1 IF The model is:

[0013] s IF (t) = Acos(2π(μτt+f c τ)),τ≤t≤T

[0014] Where A is the modulation signal amplitude, f c is the starting carrier frequency, T is the FM period, B is the FM bandwidth, and the FM slope is μ=B / T.

[0015] The radar data cube data in step M1 has a first dimension for storing all samples of each Chirp pulse, a second dimension for storing all Chirp pulses in a single Frame, and a third dimension corresponding to the antenna array channel.

[0016] The acquisition of the radar one-dimensional range image sequence in step M2 is to perform an N-point range-dimensional fast Fourier transform (FFT) on the discrete sequence of multiple Chirp pulses in each frame of data cube after sampling. After processing, the radar one-dimensional range image sequence X(k) is:

[0017]

[0018] The constant false alarm detection technology in step M3 is that each time the constant false alarm detection is performed, the unit under test (CUT) slides on the edge of the reference unit and the protection unit, and each time the sliding obtains the average power of the sampled signal level of the reference window along the front and back edges as the threshold. If the detection value CUT is greater than the threshold T, it is judged to be the target level, and if the detection value CUT is less than the threshold T, it is judged to be the noise level. The basic logic expression is:

[0019]

[0020] Among them, CUT is the unit under test, and T is the decision threshold.

[0021] The determination of the distance threshold of the observation area in step M3 is specifically as follows: after filtering out the interference clutter of the ground objects by constant false alarm detection, first obtain the peak point A of the ground target of the one-dimensional range image. peak , and then traverse the intermediate frequency signal SIF The spectrum peak value is then searched for the maximum amplitude value of the peak point within the observed terrain area (max(A peak ) to determine the ground relative radar range threshold and frequency.

[0022] In step M4, the Chirp-Z transform is used to refine the spectrum of the frequency band near the main frequency of the observation area, specifically: the ground target peak point max(A peak ) index is recorded as K, and the frequency values ​​of the spectrum lines K-1 and K+1 on both sides of the peak are calculated, recorded as f1 and f2 respectively, so as to determine the refined specific frequency band f1-f2 and the number of frequency refinement points M, and calculate the refined frequency value f of the observation area czt .

[0023] In step M4, a quadratic polynomial is constructed in combination with parabola fitting to perform spectral line calibration, and the specific steps are as follows:

[0024] Step 4.1: Refine the frequency value f according to Chirp-Z transform czt The quadratic polynomial model is constructed for the spectrum lines on the left and right sides, and a rectangular coordinate system is established with the real spectrum line as the coordinate origin. Combining the relationship between frequency and index, it can be known that the coordinates of the peak spectrum line are B(-β,X(B)), and the coordinates of the adjacent spectrum lines with an index difference of 1 are A(-(1+β),X(A)) and C((1-β),X(C));

[0025] Step 4.2: Construct a quadratic polynomial from the three spectral line coordinates:

[0026]

[0027] Among them, a and b are fitting parameters of the quadratic polynomial;

[0028] Step 4.3: Use the formula in step 4.2 to eliminate the parameters of the quadratic polynomial. The frequency resolution is Δf, and the frequency deviation δ is:

[0029]

[0030] The formula for extracting the phase sequence of the observation area distance threshold in step M5 is:

[0031]

[0032] Among them, A peak is the peak point of the ground target in the one-dimensional range image.

[0033] The determination of the relative height information of the drone to the ground in step M5 is specifically as follows:

[0034]

[0035] Among them, H is the radar height value, c is the speed of light, and λ is the wavelength of the transmitted signal.

[0036] Compared with the prior art, the present invention has the following advantages:

[0037] 1) The present invention can take into account the altitude measurement of UAVs in different complex environments with relatively low economic cost;

[0038] 2) The present invention provides high-precision and high-resolution UAV altitude detection, which can assist the UAV to quickly feedback the ground altitude and avoid the risk of collision. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 This is a flow chart of a radar phase difference UAV altitude measurement method based on spectrum refinement calibration described in the present invention;

[0040] Figure 2 This is a schematic diagram of a radar altimetry and positioning system according to an embodiment of the present invention;

[0041] Figure 3 This is a schematic diagram of the radar height measurement state of the unmanned aerial vehicle according to an embodiment of the present invention;

[0042] Figure 4 This is a flowchart of spectrum refinement calibration process according to an embodiment of the present invention;

[0043] Figure 5 This is a schematic diagram of spectral line calibration according to an embodiment of the present invention;

[0044] Figure 6 The following is a flowchart of the constant false alarm detection process according to an embodiment of the present invention. DETAILED DESCRIPTION

[0045] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0046] The height-measuring radar module described in this embodiment includes: a voltage-controlled oscillator, which is responsible for generating a tunable linear frequency-modulated signal; a power divider, which divides the linear frequency-modulated signal into a transmission signal and a reference signal, the transmission signal is radiated to the external environment through a transmission antenna, and the reference signal is guided to a mixer for mixing with an echo signal; a transceiver antenna, which is used to transmit the frequency-modulated signal and receive the echo signal; a mixer, which filters out the high-frequency component of the echo signal to reduce the signal frequency and stabilize it; a signal processor, which is responsible for analyzing and processing the signal to obtain the distance information of the observed target, wherein the distance information is used to indicate the relative height between the UAV and the ground in the detection area.

[0047] A radar phase difference UAV altitude measurement method based on spectrum refinement calibration, the specific implementation process is as follows:

[0048] Step M1: Receive the radar multi-input and output array carrier signal and perform mixing processing to form a radar data cube.

[0049] The radar is installed directly below the drone to measure the height. The intermediate frequency signal after the mixing process is a modulated signal transmitted by the height measuring radar to the observation area. After the signal is irradiated on the ground, it is reflected to form an echo signal. After the mixing process, the intermediate frequency signal is formed. IF The model is:

[0050] s IF (t) = Acos(2π(μτt+f c τ)),τ≤t≤T

[0051] Where A is the modulation signal amplitude, f c is the starting carrier frequency, T is the frequency modulation period, B is the frequency modulation bandwidth, the frequency modulation slope is μ=B / T, τ is the delay, and the radar data cube is formed after mixing processing.

[0052] The radar data cube data has a first dimension for storing all Sample samples of each Chirp pulse, a second dimension for storing all Chirp pulses in a single Frame, and a third dimension corresponding to the antenna array channel.

[0053] Step M2: Use discrete Fourier transform to convert the distance dimension of the data cube into time-frequency domain to obtain a one-dimensional radar range image sequence.

[0054] The acquisition of the radar one-dimensional range image sequence is to perform N-point range-dimensional fast Fourier transform (FFT) on the discrete sequence of multiple Chirp pulses in each frame of data cube after sampling. After processing, the radar one-dimensional range image sequence X(k) is:

[0055]

[0056] Step M3: Filter out ground interference clutter based on constant false alarm detection technology, traverse the spectrum peaks, and search for the maximum value of the peak in the range to determine the distance threshold of the observation area.

[0057] The constant false alarm detection technology is that each time the constant false alarm detection is performed, the unit under test (CUT) slides on the edge of the reference unit and the protection unit. Each time the sliding is performed, the average power of the sampled signal level of the reference window along the front and back edges is obtained as the threshold. If the detection value CUT is greater than the threshold T, it is judged to be the target level. If the detection value CUT is less than the threshold T, it is judged to be the noise level. The basic logic expression is:

[0058]

[0059] Among them, CUT is the unit under test, and T is the decision threshold.

[0060] The step of determining the distance threshold of the observation area is as follows: After filtering out the interference clutter of the ground objects by constant false alarm detection, first obtain the peak point A of the ground target of the one-dimensional range image. peak , and then traverse the intermediate frequency signal S IF The spectrum peak value is finally searched for the maximum amplitude of the peak point within the observed terrain area (max(A peak ) to determine the ground relative radar range threshold and frequency.

[0061] Step M4: Use Chirp-Z transform to perform spectrum refinement processing on the specific frequency band of the observation area, and construct a quadratic polynomial in combination with parabola fitting to perform spectrum line calibration, so as to obtain the frequency information of the observation area after spectrum refinement and calibration.

[0062] The Chirp-Z transform is used to refine the spectrum of the frequency band near the main frequency of the observation area, specifically: screening the ground target peak point max(A peak ) index is recorded as K, and the frequency values ​​of the spectrum lines K-1 and K+1 on both sides of the peak are calculated, recorded as f1 and f2 respectively, so as to determine the refined specific frequency band f1-f2 and the number of frequency refinement points M, and calculate the refined frequency value f of the observation area czt .

[0063] The method of combining parabola fitting to construct a quadratic polynomial for spectral line calibration comprises the following specific steps:

[0064] 1) According to the Chirp-Z transform, the frequency value f is refined czt The quadratic polynomial model is constructed for the spectrum lines on the left and right sides, and a rectangular coordinate system is established with the real spectrum line as the coordinate origin. Combining the relationship between frequency and index, it can be known that the coordinates of the peak spectrum line are B(-β,X(B)), and the coordinates of the adjacent spectrum lines with an index difference of 1 are A(-(1+β),X(A)) and C((1-β),X(C));

[0065] 2) Construct a quadratic polynomial from the coordinates of the three spectral lines:

[0066]

[0067] Among them, a and b are fitting parameters of the quadratic polynomial;

[0068] 3) Use the formula in step 4.2 to eliminate the parameters of the quadratic polynomial. The frequency resolution is Δf, and the frequency deviation δ is:

[0069]

[0070] Step M5: Extract the phase sequence of the distance threshold of the observation area, use phase unwrapping to unfold the phases of adjacent frames, and combine the frequency and phase information to determine the relative height information of the UAV to the ground.

[0071] The determination of the relative height information of the drone to the ground is specifically as follows:

[0072]

[0073] Among them, A peak is the peak point of the ground target in the one-dimensional range image.

[0074] Since the obtained phase value is always between [-π,π], phase ambiguity is very likely to occur. Whenever the phase difference between consecutive frames is greater than or less than ±π, the phase is unwrapped by subtracting 2π to obtain the unfolded phase.

[0075] The determination of the relative height information of the drone to the ground is specifically as follows:

[0076]

[0077] Among them, H is the altitude value measured by the radar, c is the signal propagation speed, and λ is the wavelength of the transmitted signal.

[0078] This embodiment introduces in detail the specific steps of measuring the height of a drone using a radar sensor. The one-dimensional range image is obtained by performing Fourier transform on the mixed signal, and clutter is filtered out using constant false alarm technology to determine the target distance threshold. The spectrum is refined and calibrated using Chirp-z transform, and the phase unwrapping technology is combined to accurately measure the height of the drone.

Claims

1. A radar phase difference UAV altitude measurement method based on spectrum refinement calibration, comprising the following steps: Step M1: receiving radar echo carrier frequency signals, and mixing the intermediate frequency signals to form radar data cubes; Step M2: Use discrete Fourier transform to transform the distance dimension of the data cube into time-frequency domain to obtain a radar one-dimensional range image sequence; Step M3: Filter out ground interference clutter based on constant false alarm detection technology, traverse the spectrum peaks, and search for the maximum value of the peak within the range to determine the distance threshold of the observation area; Step M4: Use Chirp-Z transform to perform spectrum refinement processing on a specific frequency band in the observation area, and construct a quadratic polynomial in combination with parabola fitting to perform spectrum line calibration, so as to obtain the frequency information of the observation area after spectrum refinement and calibration; Step M5: Extract the phase sequence of the distance threshold of the observation area, use phase unwrapping to unfold the phases of adjacent frames, and combine the frequency and phase information to determine the relative height information of the UAV to the ground.

2. The radar phase difference UAV altitude measurement method based on spectrum refinement calibration according to claim 1 is characterized in that: The intermediate frequency signal S after the mixing process in step M1 IF The model is: s IF (t)=Acos(2π(μτt+f c τ)),τ≤t≤T Where A is the modulation signal amplitude, f c is the starting carrier frequency, T is the FM period, B is the FM bandwidth, and the FM slope is μ=B / T.

3. The radar phase difference UAV altitude measurement method based on spectrum refinement calibration according to claim 1 is characterized in that: The radar data cube data in step M1 has a first dimension for storing all samples of each Chirp pulse, a second dimension for storing all Chirp pulses in a single Frame, and a third dimension corresponding to the antenna array channel.

4. The radar phase difference UAV altitude measurement method based on spectrum refinement calibration according to claim 1 is characterized in that: The acquisition of the radar one-dimensional range image sequence in step M2 is to perform an N-point range-dimensional fast Fourier transform (FFT) on the discrete sequence of multiple Chirp pulses in each frame of data cube after sampling. After processing, the radar one-dimensional range image sequence X(k) is:

5. The radar phase difference UAV altitude measurement method based on spectrum refinement calibration according to claim 1 is characterized in that: The constant false alarm detection technology in step M3 is that each time the constant false alarm detection is performed, the unit under test (CUT) slides on the edge of the reference unit and the protection unit. Each time the sliding is performed, the average power of the sampled signal level of the reference window along the front and back edges is obtained as the threshold. If the detection value CUT is greater than the threshold T, it is judged to be the target level. If the detection value CUT is less than the threshold T, it is judged to be the noise level. The basic logic expression is: Among them, CUT is the unit under test, and T is the decision threshold.

6. The radar phase difference UAV altitude measurement method based on spectrum refinement calibration according to claim 1 is characterized in that: The determination of the distance threshold of the observation area in step M3 is specifically carried out as follows: after filtering out the interference clutter of the ground objects by constant false alarm detection, first obtain the peak point A of the ground target of the one-dimensional range image. peak , and then traverse the intermediate frequency signal S IF The spectrum peak value is finally searched for the maximum amplitude of the peak point within the observed terrain area (max(A peak ) to determine the ground relative radar range threshold and frequency.

7. The radar phase difference UAV altitude measurement method based on spectrum refinement calibration according to claim 1 is characterized in that: In step M4, the Chirp-Z transform is used to refine the spectrum of the frequency band near the main frequency of the observation area, specifically: the ground target peak point max(A peak ) index is recorded as K, and the frequency values ​​of the spectrum lines K-1 and K+1 on both sides of the peak are calculated, recorded as f1 and f2 respectively, so as to determine the refined specific frequency band f1-f2 and the number of frequency refinement points M, and calculate the refined frequency value f of the observation area czt .

8. The radar phase difference UAV altitude measurement method based on spectrum refinement calibration according to claim 1 is characterized in that: In step M4, a quadratic polynomial is constructed in combination with parabola fitting to perform spectral line calibration, and the specific steps are as follows: Step 4.1: Refine the frequency value f according to Chirp-Z transform czt The quadratic polynomial model is constructed for the spectrum lines on the left and right sides, and a rectangular coordinate system is established with the real spectrum line as the coordinate origin. Combining the relationship between frequency and index, it can be known that the coordinates of the peak spectrum line are B(-β,X(B)), and the coordinates of the adjacent spectrum lines with an index difference of 1 are A(-(1+β),X(A)) and C((1-β),X(C)); Step 4.2: Construct a quadratic polynomial from the three spectral line coordinates: Among them, a and b are fitting parameters of the quadratic polynomial; Step 4.3: Use the formula in step 4.2 to eliminate the parameters of the quadratic polynomial. The frequency resolution is Δf, and the frequency deviation δ is:

9. The radar phase difference UAV altitude measurement method based on spectrum refinement calibration according to claim 1 is characterized in that: The formula for extracting the phase sequence of the observation area distance threshold in step M5 is: Among them, A peak is the peak point of the ground target in the one-dimensional range image.

10. The radar phase difference UAV altitude measurement method based on spectrum refinement calibration according to claim 1 is characterized in that: The determination of the relative height information of the drone to the ground in step M5 is specifically as follows: Among them, H is the radar height value, c is the speed of light, and λ is the wavelength of the transmitted signal.

Images