Arc-synthetic aperture FOD moving target false alarm elimination method

Abstract

The application discloses a circular arc synthetic aperture FOD moving target false alarm elimination method. The method performs moving target detection on pulse compression data of a distance unit of an imaged target, distinguishes moving targets from stationary targets, performs segmented moving target detection on the moving targets, and eliminates false alarm targets of micro-movement according to a speed variation trend of the targets, so that the problem that moving false alarm targets can be effectively distinguished in an imaging graph with only distance and angle is solved.

Description

Technical Field

[0001] This invention relates to a method for eliminating false alarms of moving targets using circular arc synthetic aperture (FOD), belonging to the field of radar monitoring. Background Technology

[0002] Foreign object debris (FOD) can easily be ingested by aircraft engines, causing engine failure or affecting normal aircraft operation. Examples of such debris include metal parts, gravel, machine tools, and plastic products. FOD is one of the main threats to airport runway safety. The detection of potential FOD has evolved from initial manual inspections to the use of radar for 24 / 7 runway scanning, accurately reporting the location of FOD relative to the radar, including distance and angle information.

[0003] Radars used for detecting foreign objects (FOOs) on airport runways employ Arc Synthetic Aperture Radar (Arc-SAR). This radar uses a small antenna as a single radiating element, moving this element along a line to receive echo signals from the same target at different locations. These echo signals are then correlated, demodulated, and compressed to synthesize an equivalent large antenna, achieving high azimuth resolution independent of range. The high resolution of Arc-SAR is crucial for pinpointing the location of FEOs on runways. However, Arc-SAR's fundamental principle relies on radar rotation to obtain target echo signals at different spatial locations. These echo signals are then coherently processed to obtain an angularly focused target spectrum. The synthesis theory assumes the target is completely stationary. When the target moves slightly, its equivalent phase center fluctuates, resulting in significant phase fluctuations in the echo. This leads to high angular sidelobes during synthesis, affecting the detection of angular targets within the same range unit and causing false alarms. The current method uses clutter maps to eliminate false alarms and distinguishes whether a target is present by comparing it with the background. This method requires extremely high phase stability. Summary of the Invention

[0004] This invention aims to provide a method for eliminating false alarms of moving targets using circular arc synthetic aperture (FOD). By detecting moving targets, compensating for the phase difference caused by the radar rotation speed using the known radar rotation speed, using segmented moving target detection, and statistically analyzing the trend of segmented speed changes, the method achieves the purpose of eliminating false alarms of moving targets.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] Step 1) Perform Fast Fourier Transform on the ADC (Analog-to-Digital Converter) data to obtain pulse compression data. Perform Fast Fourier Transform on the pulse compression data to obtain a two-dimensional spectral signal. After phase compensation of the two-dimensional spectral signal, perform Inverse Fast Fourier Transform to obtain the imaging result. Among them, the ADC data is the digital signal obtained by sampling the mixed radar signal and converting the sampled signal into a digital signal through analog-to-digital conversion.

[0007] Step 2): Perform constant false alarm rate (CFAR) detection on the imaging results. If the target is detected, obtain the distance unit and angle unit of the target.

[0008] Step 3) Perform a Fast Fourier Transform on the pulse compression data of the target's range unit to obtain the target's motion state; the process of performing a Fast Fourier Transform on the pulse compression data corresponding to the target's range unit is to perform MTD (Moving Target Detection), and find the maximum value of the obtained data. If the maximum value appears at the first point in the velocity dimension, it is a stationary target; if the maximum value appears at a non-first point in the velocity dimension, it is a moving target.

[0009] Step 4): Determine whether the target is a moving target based on the motion state. If the target is a moving target, adjust the radar rotation speed V. r The resulting phase difference is compensated by multiplying the pulse compression data by exp(-j*2*π*f). d If the target is stationary, proceed to step 5); otherwise, the process ends. c represents the speed of light, f0 represents the radar carrier frequency, λ represents the wavelength of the electromagnetic wave, and f d This indicates the Doppler translation caused by the radar's rotational speed;

[0010] Step 5) Divide the pulse pressure data corresponding to the distance unit of the moving target into segments. For each segment of data, execute step 3) to find the velocity point where the target is located. The velocity point where the target is located is found by finding the maximum value of the data after performing a fast Fourier transform. The point in the velocity dimension where the maximum value is located is the velocity point.

[0011] Step 6), calculate y i With Y i The sum of squared residuals between them is as follows:

[0012]

[0013] Where n is the total number of velocity points, 1≤i≤n, Y i Let y be the true velocity value of the velocity point where the target is located in the i-th segment. i =ai+b,

[0014] Step 7): If the sum of squared residuals is equal to 0, the target is considered stationary; otherwise, the target is moving and is judged as a false alarm.

[0015] According to embodiments of the present invention, the present invention can be further optimized, and the optimized technical solution is as follows:

[0016] In one preferred embodiment, in step 5), the number of segments is a power of 2. Choosing a power of 2 can speed up processing time and processing speed when performing a Fast Fourier Transform;

[0017] Based on the same concept, the present invention also provides an electronic device, the electronic device including a memory and one or more processors; the memory stores one or more programs, which, when executed by the one or more processors, cause the one or more processors to perform the steps of the method described above.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: The method proposed in this invention performs moving target detection on the pulse compression data of the distance unit where the target is located after imaging, distinguishes between moving targets and stationary targets, and then performs segmented moving target detection on the moving targets. Based on the velocity dimension change trend of the target, false alarm targets with slight movement can be eliminated. Moving false alarm targets can be effectively distinguished in the imaging image with only distance and angle, thus improving the accuracy of moving false alarm target recognition. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the Arc-SAR radar installation;

[0020] Figure 2 This is an analytical diagram of the Arc-SAR radar installation;

[0021] Figure 3 This is a top-down view of the Arc-SAR radar under rotation.

[0022] Figure 4 This is a target velocity dimension curve diagram in an embodiment of the present invention;

[0023] Figure 5 This is a trend diagram of the target velocity point change in an embodiment of the present invention;

[0024] Figure 6 This is a trend diagram of the velocity point change of a stationary target in another embodiment of the present invention;

[0025] Figure 7 This is a comparison chart of the actual and predicted values ​​of the velocity of a stationary target in another embodiment of the present invention;

[0026] Figure 8This is a trend diagram of the velocity point change of a moving target in another embodiment of the present invention;

[0027] Figure 9 This is a comparison chart of the actual and predicted values ​​of the velocity of a moving target in another embodiment of the present invention;

[0028] Figure 10 This is a flowchart of the moving target false alarm determination process of the present invention;

[0029] in, Figure 2 In the diagram, S, A, and B are the antenna phase centers (the intersections of the projections of the line connecting OP at different angles on the rotation plane and the rotation circle), ω is the angular velocity of the platform (antenna) rotation, L is the arm length, O is the rotation center (center of the circle), P(r0,τ0) is the point target in the scene, r0 is the distance between target P and the rotation center, τ0 is the azimuth position, H is the height difference between the radar and the target, and β is the vector... The angle between the plane of rotation and the plane of rotation. R(η) is the slant range of the target, where η = τ n +t', for frequency modulated continuous wave (FMCW) SAR radar, τ n t' represents slow time, and t' represents fast time. Figure 4 In the figure, (a) is the velocity curve of a stationary target, and (b) is the velocity curve of a moving target. Figure 5 In the figure, (a) shows the trend of velocity points of a stationary target, and (b) shows the trend of velocity points of a moving target. MTD: Moving targets detection. Detailed Implementation

[0030] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0031] Arc-SAR radar antennas are mounted on a fixed robotic arm. The relative motion between the robotic arm and the target is achieved by rotating the robotic arm, thereby obtaining Doppler information. Figure 1 and Figure 2 The geometric model of Arc-SAR and its imaging detection principle are given:

[0032] Where S, A, and B are the antenna phase centers (the intersections of the projections of the line connecting OP at different angles on the rotation plane and the rotation circle), ω is the angular velocity of the platform (antenna) rotation, L is the arm length, O is the rotation center (center of the circle), P(r0,τ0) is the point target in the scene, r0 is the distance between target P and the rotation center, τ0 is the azimuth position, H is the height difference between the radar and the target, and β is the vector... The angle between the plane of rotation and the plane of rotation. R(η): The slant range of the target, where η = τ n +t', for frequency modulated continuous wave (FMCW) SAR radar, τ n t' represents slow time, and t' represents fast time.

[0033] For any point target P(r0,τ0), the straight-line distance to the radar is:

[0034]

[0035] Usually L << r0. Applying Taylor expansion to equation (1) and retaining the first-order terms, we obtain equation (2). Therefore, equation (1) can be approximated as:

[0036] R(η)≈r0-Lcosω(η-τ0)cosβ (2)

[0037] Where θ=ω(η-τ0), for FMCW SAR, the slant range R(η) contains two parts: one is the target range information; the other is the migration of range cells (mainly affected by azimuth). Therefore, traditional range-Doppler focusing cannot be directly used in FMCW SAR.

[0038] like Figure 3 As shown, for any θ, since the turntable has an angular velocity ω and the arm length L, the antenna end velocity is v = ωL, and the radial velocity of the antenna relative to the target point is v. r =vsinγ, where γ is the angle between vectors AB and AW, the coordinates of point A are (Lcosθ, Lsinθ), the coordinates of point W are (r0, 0), the line passing through the origin has vector AB as (vsinθ, vcosθ) and vector AW as (r0-Lcosθ, -Lsinθ), and the projection of vector AB onto vector AW is:

[0039]

[0040] Since point W and the plane of revolution are not on the same plane, therefore, v t =v r ·cosβ.

[0041] For Arc-SAR transmitting FMCW, imaging can be achieved by removing the residual video phase (RVP). After linear frequency modulated continuous wave mixing and filtering:

[0042]

[0043] s b (t) represents the radar-mixed signal, s T (t) represents the radar transmitted signal. For radar received signals, f0 is the carrier frequency, ψ is the amplitude, and K is the slope. Let t be the time delay, c be the speed of light, and v be the speed of light. t R represents the target's speed, and R represents the target's distance relative to the radar.

[0044] The quadratic term of the delay τ is negligible. Substitution formula (4):

[0045]

[0046] Substituting equations (6) and (7) into equation (5) reveals the relationship between phase and radar rotation:

[0047]

[0048] To compensate for the phase difference caused by range, velocity, and range migration (range changes due to changes in azimuth angle during imaging) in equation (6), the phase calculated for each range and each velocity under the ArcSAR model is stored in a file. During imaging, the phase is compensated to eliminate the spatial phase difference introduced by the ArcSAR model. The imaging result can be obtained by performing an inverse FFT on the phase-compensated data.

[0049] The embodiments of the present invention include the following steps:

[0050] Step 1: Perform FFT on the ADC data acquired by the radar to obtain pulse compression data, perform FFT on the pulse compression data to obtain a two-dimensional spectrum signal, and perform inverse FFT on the two-dimensional spectrum signal after phase compensation to obtain the imaging result; where ADC data is a digital signal obtained by sampling the mixed radar signal and converting it from analog to digital.

[0051] Step two: Perform constant false alarm rate (CFAR) detection on the imaging results. Once the target is detected, obtain the distance and angle units where the target is located. One sampling point is one distance unit; assuming a distance resolution of 0.1m, one distance unit is 0.1m. One pulse signal is one angle unit; assuming an angle resolution of 0.1°, one angle unit is 0.1°.

[0052] Step 3: Perform FFT on the pulse compression data corresponding to the range cell where the target is located to obtain the target's motion state [Reference: Tang Yao, Wang Wei, Zhang Yan, eds. Analysis and Research on MTD Processing of LFMCW Radar [J]. Firepower and Command Control, 2014, 39(11):67-71]. If the target is stationary, the target is at zero velocity; if the target is moving, the target is not at zero velocity. Figure 4 The curves in the diagram represent the magnitude of each velocity dimension. The highest point represents the target's velocity. For a target at zero velocity, the highest point is at the first velocity point; for a target not at zero velocity, the highest point is not at the first velocity point. For example, there are 256 points in the velocity dimension, each representing a velocity point; just as the x-axis in a Cartesian coordinate system has 256 values, each representing a point.

[0053] Step 4: If a moving target appears in Step 3, compensate for the phase difference caused by the radar's rotation speed to cancel it out. The radar's rotation speed is determined by the system application and is assumed to be uniform. Let's assume the radar's rotation speed is V. r The V r The resulting phase difference is as follows:

[0054]

[0055] in, c represents the speed of light; f0 represents the radar carrier frequency; λ represents the wavelength of the electromagnetic wave; f d The Doppler shift caused by the radar's rotation speed is represented; the phase difference caused by the radar's rotation is calculated, and this phase difference is compensated by multiplying the pulse compression data by exp(-j*2*π*f). d *t), to cancel out the phase caused by the radar rotation and eliminate the influence of the radar's own rotation.

[0056] Step 5: After eliminating the phase difference caused by the radar rotation speed, extract all pulse compression data of the target's range cell. Divide all pulse compression data of the target's range cell into small segments (e.g., 16 pulses per segment, 32 pulses per segment, 64 pulses per segment, etc.). The size of the data segments is allocated according to the total number of pulses emitted by the radar. Larger pulse counts require larger data segments, and smaller pulse counts require smaller data segments. Data selection is done using powers of 2 (selecting powers of 2 can speed up processing time and speed when performing Fast Fourier Transform). Then, perform Step 3 on the segmented data to find the target's velocity point and store it as Y. iThe velocity point index is stored as i, and this process continues until the last segment of moving target detection is completed. For example, when performing the first segment of moving target detection, if the target velocity point is at point 3, then velocity point index i = 1 and velocity point Y1 = 3. When performing the second segment of moving target detection, if the target velocity point is at point 5, then velocity point index i = 2 and velocity point Y2 = 5. Segmented moving target detection is used to obtain the trend of target velocity changes. If the target is moving, and its velocity is unknown and fluctuates, especially for objects with non-linear velocity changes like grass, moving target detection will identify non-linear velocity, allowing for the removal of the moving target.

[0057] Step six: Analyze the trend of the target speed points using a univariate linear regression model, y i =ai+b,y i The predicted values ​​(calculated using parameters a and b) are obtained using the least squares method. Y i For the true velocity value of the velocity point where the target is located in the i-th segment, calculate the sum of squared residuals between the predicted and true values:

[0058]

[0059] Where n is the total number of velocity points stored in step five;

[0060] If the calculated sum of squared residuals equals 0, the velocity change is considered linear; if the calculated sum of squared residuals does not equal 0, the velocity change is considered non-linear. A sum of squared residuals equal to 0 indicates that the predicted value is the same as the actual value. If the target is stationary, the velocity change is linear; if the target is moving, the velocity change is non-linear.

[0061] Step 7: Determine whether the change in velocity point is linear. This will determine whether the target is stationary or moving. If it is a moving target, it can be identified as a false alarm, and the detected target point will be removed.

[0062] The method of this invention has been verified in Arc-SAR radar for moving target false alarm removal, and it can effectively remove moving target false alarms. Figures 6-9 As shown. Figure 6 and Figure 8 In the left figure, x is the angular dimension, where a point represents an angular unit; y is the distance dimension, where a point represents a distance unit; and z is the signal strength (unit: dB). Figure 6 These are the angular and distance dimensions of a stationary, real target. Figure 8 It represents the angular dimension and distance dimension of the target grass. Figure 6 and Figure 8In the right image, the entire angular dimension corresponding to the distance cell of the target in the left image is extracted, totaling 4096 points. Steps two through six are performed on this data to obtain the right image. Considering computational complexity and time, it is divided into 32 segments, each with 128 angular cells. Moving target detection is performed on each segment. a and b are obtained using the least squares method. Figure 7 If the predicted value completely matches the actual value and the sum of squared residuals is 0, it indicates a linear change. Figure 8 The predicted value differs significantly from the actual value, with a residual sum of squares of 30230, indicating a nonlinearity.

[0063] Based on the above measured data, it can be seen that the method of this embodiment can be used for false alarm elimination of moving targets.

[0064] The above embodiments should be understood as being used only to illustrate the present invention more clearly, and not to limit the scope of the present invention. After reading the present invention, any modifications of the present embodiments by those skilled in the art will fall within the scope defined by the appended claims.

Claims

1. A method for eliminating false alarms of moving targets using circular arc synthetic aperture (FOD), characterized in that, Includes the following steps: Step 1: Perform Fast Fourier Transform on the ADC data to obtain pulse compression data, perform Fast Fourier Transform on the pulse compression data to obtain a two-dimensional spectral signal, and perform Inverse Fast Fourier Transform on the two-dimensional spectral signal after phase compensation to obtain the imaging result; wherein, the ADC data is a digital signal obtained by sampling the mixed radar signal and converting the sampled signal into a digital signal through analog-to-digital conversion. Step 2: Perform constant false alarm rate (CFAR) detection on the imaging results. If the target is detected, obtain the range unit and angle unit of the target. Step 3: Perform a Fast Fourier Transform on the pulse compression data corresponding to the distance unit of the target to obtain the motion state of the target; Step 4: Determine whether the target is a moving target based on the described motion state. If the target is a moving target, adjust the radar rotation speed V. r The resulting phase difference is compensated by multiplying the pulse compression data by exp(-j*2*π*f). d If *t), proceed to step 5; otherwise, if the target is a stationary target, end; where, c represents the speed of light, f0 represents the radar carrier frequency, λ represents the wavelength of the electromagnetic wave, and f d This indicates the Doppler translation caused by the radar's rotational speed; Step 5: Divide the pulse pressure data corresponding to the distance unit of the moving target into segments. For each segment of data, execute step 3 to find the velocity point of the moving target. Step 6, calculate y i With Y i The sum of squared residuals between them is as follows: Where n is the total number of velocity points, 1≤i≤n, Y i Let y be the true velocity value of the velocity point where the moving target is located in the i-th segment. i =ai+b, Step 7: If the sum of squared residuals equals 0, the target is considered stationary; otherwise, the target is in motion and is considered a false alarm.

2. The method for eliminating false alarms of moving targets using circular arc synthetic aperture (FOD) according to claim 1, characterized in that, In step 5, the number of segments is a power of 2.

3. An electronic device, characterized in that, The electronic device includes a memory and one or more processors; the memory stores one or more programs that, when executed by the one or more processors, cause the one or more processors to perform the steps of the method of claim 1.

Images