A segmented aperture imaging and positioning method for synthetic aperture radar on multi-rotor UAVs

Abstract

This invention belongs to the field of signal processing technology and provides a segmented aperture imaging and positioning method for multi-rotor UAV-borne synthetic aperture radar. The method includes: 1) acquiring target echoes from the UAV-borne synthetic aperture radar system; 2) estimating the motion state of the maneuvering platform based on the echo signals; 3) segmenting the echo signals according to the platform motion state; 4) performing motion compensation on each segment of the echo signals according to the platform motion state; 5) performing a two-dimensional Fourier transform on each compensated echo signal to obtain a two-dimensional spectrum, and using a series inversion method to decompose the two-dimensional spectrum to obtain phase filters for each segment; 6) multiplying each segment's two-dimensional spectrum by the corresponding phase filter, and then performing a two-dimensional inverse Fourier transform on the two-dimensional spectrum to obtain images of each segment; 7) performing geometric correction on each segment's images, and then stitching the images together to obtain a full-aperture imaging result; 8) stitching together the platform trajectories of each segment to obtain the complete platform trajectory coordinates.

Description

Technical Field

[0001] This invention belongs to the field of signal processing technology, specifically relating to a segmented aperture imaging and positioning method for multi-rotor unmanned aerial vehicle-borne synthetic aperture radar. Background Technology

[0002] With the widespread application of remote sensing technology in civilian applications such as mapping, surface deformation detection, and weather observation, the miniaturization of remote sensing equipment has become an inevitable trend. Since the 1990s, unmanned aerial vehicle (UAV) platforms have gradually become an important platform for synthetic aperture radar (SAR) due to their small size, portability, high flexibility, and ease of operation. Numerous scholars have conducted extensive research on UAV-based SAR systems from both system design and imaging algorithm perspectives. In recent years, with the development of UAV navigation and control technology, various models of multi-rotor UAVs have entered the market in large numbers, but they have not yet been widely used in the field of remote sensing. This is mainly because traditional airborne imaging algorithms for multi-rotor UAV-borne SAR have unstable imaging effects and are limited by the performance of their onboard inertial navigation systems. Therefore, researching imaging algorithms for multi-rotor UAV-borne SAR that do not rely on inertial navigation systems is of great significance.

[0003] Currently, airborne imaging algorithms used by research institutions both domestically and internationally primarily rely on data acquired by high-precision inertial navigation systems for motion compensation and imaging. Furthermore, because the servo systems carried by large platforms can compensate for changes in platform attitude angles in real time without considering these angle variations, they cannot address three key issues inherent in multi-rotor UAV-borne synthetic aperture radar (SAR) systems: 1) The trajectory of multi-rotor platforms is typically highly unstable due to environmental factors such as wind, potentially leading to severe shaking and sharp turns; 2) The unique flight principle of multi-rotors introduces high-frequency errors and incident angle errors caused by rotor vibration and fuselage tilting; 3) The combination of highly maneuverable trajectories and low-cost inertial navigation systems results in significantly poor motion state and attitude angle data used for motion compensation. Moreover, the weight limitations of UAVs prevent the carrying of high-precision inertial navigation equipment, thus reducing the positioning accuracy of the UAV navigation system. If the aircraft's flight trajectory could be deduced from the imaging results during flight, it could provide a basis for the navigation system, enabling more intelligent trajectory planning. Therefore, researching imaging and positioning algorithms for multi-rotor UAV-borne SAR that do not rely on inertial navigation equipment is of great significance. Summary of the Invention

[0004] This invention addresses the shortcomings and deficiencies of existing airborne imaging algorithms, aiming to provide a segmented aperture imaging (SAI) and positioning method for multi-rotor UAV-borne synthetic aperture radar. This method is applicable to synthetic aperture radar systems without inertial navigation equipment or with low-precision inertial navigation equipment, thereby improving the focusing effect and imaging efficiency of radar imaging.

[0005] This invention provides a segmented aperture imaging method for multi-rotor UAV-borne synthetic aperture radar, used for segmented aperture imaging based on the raw echo signal of synthetic aperture radar. It includes the following steps: Step 1, performing range pulse compression on the raw echo signal s(t,η) to obtain the range pulse compression signal si. RC (t,η), where t is the range fast time and η is the azimuth slow time, based on the range pulse compression signal s. RC Phase history of strong scattering points in (t,η) Estimated speed of computer-controlled moving platform and estimated beam center angle Step 2, based on the estimated speed The direction will be the distance from the pulse compression signal s RC (t,η) is divided into N segments and the corresponding segment pulse compression signal s for each segment. RC,i (t,η), i=1…N; Step 3, based on the estimated velocity and estimated beam center angle Calculate the segment pulse compression signal s corresponding to each segment. RC,i Phase compensation amount of (t,η) Segment pulse pressure signal s RC,i (t,η) multiplied by the motion error compensation filter N compensated signals s are obtained MC,i (t,η), imaginary number Step 4, for the compensated signal s MC,i Performing a two-dimensional Fourier transform on (t,η) yields N two-dimensional spectra s. MC,i (f,f d The series inversion method is used to analyze the two-dimensional spectrum s. MC,i (f,f d The azimuth compression filter H is constructed by decomposing the components. AC,i Where f is the frequency corresponding to the time t between the distances, f d The Doppler frequency corresponding to the azimuth slow time η; Step 5, for the two-dimensional spectrum s MC,i (f,f d Multidirectional compression filter H AC,i Then, a two-dimensional inverse Fourier transform is performed to obtain N imaging results s. IMG,i (t, η); Step 6, sequentially image the corresponding images s of adjacent segments. IMG,i (t,η) The strong focal points in the overlapping region are aligned with the dimensional envelope and coherently accumulated. The non-overlapping regions are then stitched together to obtain the final imaging result S. all .

[0006] The multi-rotor UAV-borne synthetic aperture radar segmented aperture imaging method provided by this invention may also have the following feature: In step 1, the range pulse compression signal s RC Phase history of strong scattering points in (t, η) The phase history of the strong scattering point was obtained by performing a second-order fitting. The speed of the computer-controlled moving platform is estimated based on the second-order coefficient β and the first-order coefficient α. And the estimated beam center angle Where t is the distance fast time, η is the azimuth slow time, and o(η) is the higher-order phase error. Here, λ is the constant phase term, and λ is the wavelength of the system's transmitted signal. This is an estimated value for the system reference distance.

[0007] The multi-rotor UAV-borne synthetic aperture radar segmented aperture imaging method provided by this invention may also have the following feature: wherein, in step 2, based on the estimated velocity... The direction will be the distance from the pulse compression signal s RC (t,η) is divided into N segments with the same velocity direction. Then, it is determined whether the length of each segment is less than the length of a synthetic aperture. If it is less, the segment is expanded to both sides to reach the length of a synthetic aperture, and finally, N segments of pulse compression signals s are obtained. RC,i (t,η), i=1…N.

[0008] The multi-rotor UAV-borne synthetic aperture radar segmented aperture imaging method provided by this invention may also have the following feature: wherein, in step 3, the phase compensation amount In the formula, R0 represents each segment The mean of θ0 is the angle of view of N estimated beam centers. The mean.

[0009] The multi-rotor UAV-borne synthetic aperture radar segmented aperture imaging method provided by this invention may also have the following feature: wherein, in step 4, the azimuth compression filter In the formula, f is the frequency corresponding to the distance and time t. d f is the Doppler frequency corresponding to the azimuth slow time η. c Let c be the carrier frequency of the system's transmitted signal, and c be the speed of light.

[0010] The multi-rotor UAV-borne synthetic aperture radar segmented aperture imaging method provided by this invention may also have the following feature: wherein, in step 6, the imaging results s corresponding to adjacent segments are first processed. IMG,i Geometric correction is performed on (t,η) to obtain the corrected imaging result. Then the corrected imaging results Rotation The corrected imaging results were obtained with the slant range perpendicular to the trajectory of the moving platform. Then, the corrected imaging results corresponding to adjacent segments are processed sequentially. The strong focal points in the overlapping region are aligned with the distance-dimensional envelope and coherently accumulated. The non-overlapping regions are then stitched together to obtain the final imaging result S. all .

[0011] The multi-rotor UAV-borne synthetic aperture radar segmented aperture imaging method provided by this invention may also have the following feature: wherein, step 6 processes the imaging result s IMG,i Geometric correction is performed in the following sub-steps: Step 6-1, adjust the imaging result s IMG,i Perform an azimuth-to-Fourier transform on (t,η) to obtain the range-Doppler domain map s. IMG,i (t,f d Step 6-2: Based on the Fourier transform characteristics and the target spatial geometry, construct the expression for the tilt correction filter to correct image tilt: In the formula, l represents the distance-axis scale of a typical building; step 6-3, the distance Doppler domain map s IMG,i (t,f d Multiplied by the tilt correction filter H GC-1 The tilt-corrected frequency domain plot was obtained. Step 6-4, adjust the tilt-corrected frequency domain plot. Perform an inverse Fourier transform in the azimuth direction to obtain the tilt-corrected time-domain plot. Step 6-5: Based on the target spatial geometry, obtain the expression for the stretching / compression factor: Step 6-6: Substitute the stretching / compression factor expression into the tilt-corrected time-domain plot. From this, the deformed time-domain diagram was obtained. Steps 6-7: Analyze the deformed time-domain plot. Perform a distance-to-Fourier transform to obtain the deformed frequency domain map. Steps 6-8: Based on the Fourier transform characteristics and the target spatial geometry, construct the expression for the two-position correction filter to correct image translation: Steps 6-9: Obtain the deformed frequency domain diagram. Multiplied by position correction filter H GC-3 The geometrically corrected frequency domain plot was obtained. Steps 6-10: Perform geometrically corrected frequency domain plots... Perform an inverse Fourier transform to the range to obtain the corrected imaging results for each segment after geometric correction.

[0012] The multi-rotor UAV-borne synthetic aperture radar segmented aperture imaging method provided by this invention may also have the following feature: wherein, in step 1, the velocity is estimated. and estimated beam center angle Calculate using the following steps: Step 1-1, based on the Doppler modulation slope K... a and Doppler center f dc The definition of K is obtained. a with f dc The calculation formula: In the formula, (·) represents the derivative of the positional slow time η; Steps 1-2, the space formed by the original echo signals is called the signal space, and the range pulse compression signal s RC Phase history of strong scattering points in (t,η) A second-order fit yields the expression for this strong scattering point in the signal space: In the formula, β is the coefficient of the second-order term, and α is the coefficient of the first-order term. The phase term is constant, and o(η) is the higher-order phase error; Steps 1-3, the phase history is... Substitute the expressions into K in step 1-1 respectively a The calculation formula and f dc From the calculation formula, K in the signal space is obtained. a with f dc The expression: Steps 1-4: The space formed by the actual positions of the target and the maneuvering platform is called the target space. Based on the spatial geometric relationship, the phase history in the target space is obtained. The expression: In the formula, v represents the velocity of the maneuvering platform, θ represents the angle of the beam center caused by the motion of the maneuvering platform, R0 represents the closest distance between the target and the maneuvering platform, and λ is the wavelength of the system's transmitted signal; Steps 1-5, the phase history in the target space is... Substitute the expressions into K in step 1-1 respectively a The calculation formula and f dc From the calculation formula, K in the target space is obtained. a with f dc The expression: Steps 1-6: Compare K in the signal space respectively. a The expression and K in the target space a The expression for f in the signal space dc The expression and f in the target space dc The right side of the expression yields the estimated velocity, estimated beam center angle, and estimated range:

[0013] The multi-rotor UAV-borne synthetic aperture radar segmented aperture imaging method provided by this invention may also have the following feature: wherein, in step 4, a series inversion method is used to analyze the two-dimensional spectrum s MC,i (f,f d The azimuth compression filter expression H is obtained by decomposition. AC,i The process is carried out in the following sub-steps: Step 4-2-1, obtain the two-dimensional spectrum s corresponding to the segment according to the stationary phase point method. MC,i (f,f d The expression for ) is: In the formula, f c The carrier frequency of the system's transmitted signal is represented by c, and the speed of light is represented by c. Step 4-2-2: The series inversion method is used to analyze the two-dimensional spectrum s from step 4-2-1. MC，i (f,f d Decompose the expression of f to obtain the elimination of f and f d The two-dimensional spectrum expression of the coupling term: Step 4-2-3, based on the elimination of f and f in step 4-2-2 d The two-dimensional spectral expression of the coupling term is used to construct the expression for the ideal phase filter: Step 4-2-4, estimate the speed Estimated beam center angle Distance estimate Substituting into the expression for the ideal phase filter, we obtain the expression for the azimuth compression filter:

[0014] This invention also provides a segmented aperture imaging method for multi-rotor UAV-borne synthetic aperture radar, used to calculate the UAV flight trajectory based on the original echo signal of the UAV-borne synthetic aperture radar. Its features include the following steps: Step S1, performing range pulse compression on the original echo signal s(t,η) to obtain a range pulse compression signal si. RC (t, η), based on the distance pulse compression signal s RC Phase history of strong scattering points in (t, η) Estimated speed of computer-controlled moving platform and estimated beam center angle Step S2, based on the estimated speed The direction will be the distance from the pulse compression signal s RC (t,η) is divided into N segments and the corresponding segment pulse compression signal s for each segment. RC,i (t,η), i=1…N, and based on the estimated velocity and estimated beam center angle Calculate the platform trajectory coordinates corresponding to the i-th segment. k = 1…M, where M is the azimuth length of each segment. In the formula, θ in This represents the angle between the UAV-borne synthetic aperture radar beam and the ground normal; Step S3, in the adjacent regions of the i-th and (i-1)-th segments, extract three strong scattering points, where the coordinates of the strong scattering points in the i-th segment are respectively... The coordinates of the strong scattering points in the (i-1)th segment are respectively Step S4, based on the coordinates of the strong scattering point and Calculate the rotation matrix γ for the i-th and (i-1)-th segments. Step S5, set the platform trajectory coordinates of the i-th segment. Rotate to the coordinate system of the platform trajectory of the (i-1)th segment, aligning the coordinates of the platform trajectories of adjacent segments, i.e.: Step S6: Perform coherent accumulation of the platform trajectory coordinates in the overlapping areas of the i-th and (i-1)-th segments, and concatenate the platform trajectory coordinates in the non-overlapping areas to obtain the concatenated trajectory coordinates of the i-th and (i-1)-th segments [P]. x ,P y ,P z ], Step S7: Repeat steps S3-S6 until the coordinates of the spliced ​​trajectory of all adjacent segments are obtained. This yields the platform's final trajectory coordinates.

[0015] The role and effect of invention

[0016] The multi-rotor UAV-borne synthetic aperture radar segmented aperture imaging and positioning method of the present invention employs a method of estimating platform motion parameters based on echo signal phase history and segmenting echo signals to accurately compensate for platform motion errors, achieving good imaging focusing effect and high imaging success rate. This improves the efficiency of data acquisition for multi-rotor UAV platforms, effectively enabling high-resolution imaging of multi-rotor UAV platform synthetic aperture radar systems. Furthermore, during imaging, the platform trajectory can be effectively estimated to obtain the platform's coordinate information relative to the measurement area, achieving the effect of platform positioning.

[0017] Compared with the prior art, the present invention has the following advantages:

[0018] Compared with traditional airborne synthetic aperture radar imaging algorithms, this invention takes into account the relationship between platform motion and signal phase, and can be applied to synthetic aperture radar imaging systems without inertial navigation equipment or with low-precision inertial navigation equipment for imaging.

[0019] Compared with traditional airborne synthetic aperture radar imaging algorithms, this invention takes into account the oblique view effect caused by the attitude angle change of the UAV platform, compensates for the coupled phase of range and azimuth, improves the image focusing effect, and is applicable to synthetic aperture radar imaging systems without antenna servo.

[0020] Compared with traditional airborne synthetic aperture radar imaging algorithms, this invention uses phase filter multiplication instead of interpolation, which improves the imaging speed of a single image in each segment.

[0021] Compared with traditional airborne synthetic aperture radar imaging algorithms, this invention uses a method of parallel imaging of each segment and then stitching them together to obtain a complete image, which improves the imaging speed of the complete image.

[0022] Compared with traditional airborne synthetic aperture radar imaging algorithms, this invention can estimate the platform's position while imaging, achieving the effect of platform localization and facilitating the establishment of an automated, intelligent, and integrated UAV detection system. Attached Figure Description

[0023] Figure 1 Flowcharts from embodiments of the present invention;

[0024] Figure 2 A schematic diagram of the target space and signal space of an airborne synthetic aperture radar system in an embodiment of the present invention;

[0025] Figure 3 A flowchart of the geometric correction process in an embodiment of the present invention;

[0026] Figure 4 A schematic diagram of trajectory estimation in an embodiment of the present invention;

[0027] Figure 5 Graphs showing the measured trajectory and attitude angle changes in embodiments of the present invention;

[0028] Figure 6 A schematic diagram illustrating the spatial relationship between the platform trajectory and the dot matrix target in an embodiment of the present invention;

[0029] Figure 7 Comparison of imaging results for each segment, full aperture imaging results, and imaging results of traditional imaging methods in the embodiments of the present invention;

[0030] Figure 8 The point spread function comparison results of the present invention and the conventional imaging method in the embodiments of the present invention;

[0031] Figure 9 Image results of conventional imaging methods, imaging results of the present method, and optical model comparison diagrams in the embodiments of the present invention;

[0032] Figure 10 The embodiment of the present invention provides a comparison of the platform estimated trajectory curve and the inertial navigation measurement trajectory curve, as well as an error curve diagram. Detailed Implementation

[0033] To make the technical means, creative features, objectives and effects of this invention easy to understand, the following describes in detail a multi-rotor UAV-borne synthetic aperture radar segmented aperture imaging and positioning method with reference to embodiments and accompanying drawings.

[0034] In this invention, after the UAV takes off, the synthetic aperture radar onboard the UAV transmits a carrier frequency of f through a transmitting antenna. c The linear frequency modulated signal is transmitted, scattered by the target, and then received by the radar through the receiving antenna. The received signal is the original echo signal s(t,η), where t is the range fast time and η is the azimuth slow time. The synthetic aperture length L... s =R×θ BW Where R is the reference distance, θ BW This refers to the azimuth beamwidth.

[0035] The present invention provides a segmented aperture imaging method for multi-rotor UAV-borne synthetic aperture radar, used for segmented aperture imaging based on the raw echo signal of UAV-borne synthetic aperture radar, comprising the following steps:

[0036] Step 1: Perform range pulse compression on the original echo signal s(t,η) to obtain the range pulse compression signal s RC (t,η), where t is the range fast time and η is the azimuth slow time, based on the range pulse compression signal s. RC Phase history of strong scattering points in (t,η) Estimated speed of computer-controlled moving platform and estimated beam center angle

[0037] Step 2, based on the estimated speed The direction will be the distance from the pulse compression signal s RC (t,η) is divided into N segments and the corresponding segment pulse compression signal s for each segment. RC,i (t,η), i=1…N;

[0038] Step 3, based on the estimated speed and estimated beam center angle Calculate the segment pulse compression signal s corresponding to each segment. RC,i Phase compensation amount of (t,η) Segment pulse pressure signal s RC,i (t,η) multiplied by the motion error compensation filter N compensated signals s are obtained MC,i(t,η), imaginary number

[0039] Step 4, for the compensated signal s MC,i Performing a two-dimensional Fourier transform on (t,η) yields N two-dimensional spectra s. MC,i (f,f d The series inversion method is used to analyze the two-dimensional spectrum s. MC,i (f,f d The azimuth compression filter H is constructed by decomposing the components. AC,i Where f is the frequency corresponding to the time t between the distances, f d The Doppler frequency corresponding to the azimuth slow time η;

[0040] Step 5, for the two-dimensional spectrum s MC,i (f,f d Multidirectional compression filter H AC,i Then, a two-dimensional inverse Fourier transform is performed to obtain N imaging results s. IMG,i (t,η);

[0041] Step 6: Sequentially process the imaging results s corresponding to adjacent segments. IMG,i (t,η) The strong focal points in the overlapping region are aligned with the dimensional envelope and coherently accumulated. The non-overlapping regions are then stitched together to obtain the final imaging result S. all .

[0042] In step 1, the distance pulse compression signal s RC Phase history of strong scattering points in (t,η) The phase history of the strong scattering point was obtained by performing a second-order fitting. The speed of the computer-controlled moving platform is estimated based on the second-order coefficient β and the first-order coefficient α. And the estimated beam center angle Where t is the distance fast time, η is the azimuth slow time, and o(η) is the higher-order phase error. Here, λ is the constant phase term, and λ is the wavelength of the system's transmitted signal. This is an estimated value for the system reference distance.

[0043] In step 2, based on the estimated speed The direction will be the distance from the pulse compression signal s RC (t,η) is divided into N segments with the same velocity direction. Then, it is determined whether the length of each segment is less than the length of a synthetic aperture. If it is less, the segment is expanded to both sides to reach the length of a synthetic aperture, and finally, N segments of pulse compression signals s are obtained. RC,i (t,η), i=1…N.

[0044] In step 3, the phase compensation amount In the formula, R0 represents each segment The mean of θ0 is the angle of view of N estimated beam centers. The mean.

[0045] In step 4, the azimuth compression filter In the formula, f is the frequency corresponding to the distance and time t. d f is the Doppler frequency corresponding to the azimuth slow time η. c Let c be the carrier frequency of the system's transmitted signal, and c be the speed of light.

[0046] In step 6, the imaging results s corresponding to adjacent segments are first processed. IMG,i Geometric correction is performed on (t,η) to obtain the corrected imaging result. Then the corrected imaging results Rotation The corrected imaging results were obtained with the slant range perpendicular to the trajectory of the moving platform. Then, the corrected imaging results corresponding to adjacent segments are processed sequentially. The strong focal points in the overlapping region are aligned with the distance-dimensional envelope and coherently accumulated. The non-overlapping regions are then stitched together to obtain the final imaging result S. all .

[0047] This invention also provides a segmented aperture imaging method for multi-rotor UAV-borne synthetic aperture radar, used to calculate the UAV flight trajectory based on the original echo signal of the UAV-borne synthetic aperture radar. Its features include the following steps: Step S1, performing range pulse compression on the original echo signal s(t,η) to obtain a range pulse compression signal si. RC (t, η), based on the distance pulse compression signal s RC Phase history of strong scattering points in (t,η) Estimated speed of computer-controlled moving platform and estimated beam center angle

[0048] Step S2, based on the estimated speed The direction will be the distance from the pulse compression signal s RC (t,η) is divided into N segments and the corresponding segment pulse compression signal s for each segment. RC,i (t,η), i=1…N, and based on the estimated velocity and estimated beam center angle Calculate the platform trajectory coordinates corresponding to the i-th segment. k = 1…M, where M is the azimuth length of each segment.

[0049]

[0050]

[0051]

[0052] In the formula, θ in This indicates the angle between the UAV-borne synthetic aperture radar beam and the ground normal.

[0053] Step S3: In the adjacent regions of the i-th and (i-1)-th segments, extract three strong scattering points. The coordinates of the strong scattering points in the i-th segment are as follows: The coordinates of the strong scattering points in the (i-1)th segment are respectively

[0054] Step S4, based on the coordinates of the strong scattering point and Calculate the rotation matrix γ for the i-th and (i-1)-th segments.

[0055]

[0056] Step S5, set the platform trajectory coordinates of the i-th segment. Rotate to the coordinate system of the platform trajectory of the (i-1)th segment, aligning the coordinates of the platform trajectories of adjacent segments, i.e.:

[0057]

[0058] Step S6: Perform coherent accumulation of the platform trajectory coordinates in the overlapping areas of the i-th and (i-1)-th segments, and concatenate the platform trajectory coordinates in the non-overlapping areas to obtain the concatenated trajectory coordinates of the i-th and (i-1)-th segments [P]. x ,P y ,P z ],

[0059]

[0060] Step S7, repeat steps S3-S6 until the coordinates of the spliced ​​trajectory of all adjacent segments are obtained [P] x ,P y ,P z This allows us to obtain the platform's final trajectory coordinates.

[0061] <Example>

[0062] In this embodiment, the mobile platform refers to a multi-rotor unmanned aerial vehicle (UAV), the airborne synthetic aperture radar is a Ku-band linear frequency modulated continuous wave radar, and the target refers to the ground area that needs to be imaged using synthetic aperture technology.

[0063] like Figure 1As shown, the segmented aperture imaging method of synthetic aperture radar on a multi-rotor UAV includes the following steps:

[0064] Step 1: After the mobile platform takes off, the airborne synthetic aperture radar transmits a carrier frequency of f through the transmitting antenna. c The transmitted linear frequency modulated (LFM) signal is scattered by the target and received by the radar through the receiving antenna. The received signal is the original echo signal s(t,η), where t is the range fast time and η is the azimuth slow time. Range pulse compression is applied to the original echo signal to obtain the pulse-compressed signal s. RC (t,η), based on the pulse compression signal s RC Phase history of strong scattering points in (t,η) Estimated speed of computer-controlled moving platform and estimated beam center angle The specific steps are as follows:

[0065] Step 1-1, based on the Doppler modulation slope K a and Doppler center f dc The definition of K is obtained. a with f dc The calculation formula:

[0066]

[0067]

[0068] In the formula, This represents the derivative of the azimuth time η, where η is the azimuth time. The phase history of a certain strong scattering point;

[0069] Steps 1-2, refer to Figure 2 The space formed by the original echo signals is called the signal space. Range pulse compression is performed on the original echo signals in the signal space to obtain the pulse-compressed signal s. RC (t,η), where t is the distance-time interval for the pulse compression signal s. RC Phase history of strong scattering points in (t,η) A second-order fit yields the expression for this strong scattering point in the signal space:

[0070]

[0071] In the formula, β is the coefficient of the second-order term in the azimuth slow time, α is the coefficient of the first-order term in the azimuth slow time, and o(η) is the higher-order phase error. The phase term is a constant.

[0072] Steps 1-3, convert the above formula <3> Substitution formula <1> With formula <2> From this, K in the signal space is obtained.a with f dc The expression:

[0073]

[0074]

[0075] Steps 1-4: The space formed by the actual positions of the target and the mobile platform is called the target space. Based on the geometric relationships within the target space and the law of cosines, the expression for the distance R between the target and the platform is obtained:

[0076]

[0077] In the formula, v represents the speed of the maneuvering platform, θ represents the angle of the beam center caused by the movement of the maneuvering platform, and R0 represents the closest distance between the target and the maneuvering platform;

[0078] The above formula <6> Taylor expansion at η = 0, retaining only the second-order terms, yields the expression for R:

[0079]

[0080] Based on the phase calculation formula and definition, the phase history in the target space is obtained. The expression:

[0081]

[0082] In the formula, λ is the wavelength of the system's transmitted signal;

[0083] Steps 1-5, convert the above formula <8> Substitution formula <1> With formula <2> In this process, K in the target space is obtained. a with f dc The expression:

[0084]

[0085]

[0086] Steps 1-6: Compare the above formulas respectively. <4> With formula <9> ,Mode <5> With formula <10> From the right side of the equation, we obtain the expressions for the estimated velocity, the estimated beam center angle, and the estimated range, respectively:

[0087]

[0088]

[0089]

[0090] Step 2, based on the estimated speed The direction will compress the pulse signal s RC (t, η) is divided into N segments and the corresponding segment compression signal s for each segment. RC,i (t,η), i=1…N, as follows:

[0091] Based on the estimated speed The positive and negative directions initially compress the pulse signal s RC (t, η) are divided into segments with the same velocity direction, and then the length of each segment is checked to see if it is less than the length of a composite aperture L. s =R×θ BW If the value is less than the required value, the segment is expanded to both sides to achieve a combined aperture length; otherwise, no processing is performed. Finally, the pulse compression signal s... RC (t,η) is divided into N segments, and the echo signal corresponding to each segment is represented as s. RC,i (t,η), i=1…N.

[0092] Step 3, based on the estimated speed and estimated beam center angle Calculate the segment compression signal s corresponding to each segment. RC,i Phase compensation amount of (t,η) For segment compression signal s RC,i (t,η) multiplied by the motion error compensation filter N compensated signals s are obtained MC,i (t,η), imaginary number To achieve compression of each segment of the signal s RC,i Motion compensation is performed on (t, η). Specifically, the following sub-steps are followed:

[0093] Step 3-1: Calculate the effect of the slant distance change δR caused by the change in the motion state of each platform segment on the phase. The expression is:

[0094]

[0095] In the formula, v0 is The mean, θ0 is The mean;

[0096] Step 3-2: Construct the segment compression signal s corresponding to each segment. RC，i The expression for the phase compensation amount of (t, η):

[0097]

[0098] Step 3-3, construct the expression for the motion error compensation filter:

[0099]

[0100] Where j represents an imaginary number,

[0101] Steps 3-4, compress the segment signal s RC,i (t,η) multiplied by the formula above <14> The motion error compensation filter shown yields the compensated signal s. MC,i (t,η) completes motion compensation for each segment of the compressed signal.

[0102] Step 4, for the compensated signal s MC,i Performing a two-dimensional Fourier transform on (t,η) yields N two-dimensional spectra s. MC,i (f,f d The series inversion method is used to analyze the two-dimensional spectrum s. MC,i (f,f d The azimuth compression filter H is constructed by decomposing the components. AC,i Where f is the frequency corresponding to the time t between the distances, f d This is the Doppler frequency corresponding to the azimuth slow time η. The specific steps are as follows:

[0103] Step 4-1, for the compensated signal s MC,i Performing a two-dimensional Fourier transform on (t,η) yields the corresponding two-dimensional spectrum s for each segment. MC,i (f,f d ), where f is the frequency corresponding to the time t between distances, f d The Doppler frequency corresponding to the azimuth slow time η;

[0104] Step 4-2-1: Obtain the two-dimensional spectrum s of each segment using the stationary phase point method. MC，i (f,f d The expression for ) is:

[0105]

[0106] Among them, f c The carrier frequency of the system's transmitted signal is represented by , and c represents the speed of light.

[0107] Step 4-2-2: Use the series inversion method to perform the above equation. <17> Decomposition yields the elimination of f and f d The two-dimensional spectrum expression of the coupling term:

[0108]

[0109]

[0110] Step 4-2-3, according to the above formula <15> Construct the expression for the ideal phase filter:

[0111]

[0112] Step 4-2-4, estimate the speed Estimated beam center angle Distance estimate Substituting into the above formula <19> From this, we obtain the expression for the phase filter:

[0113]

[0114] Step 5, for the two-dimensional spectrum s MC，i (f,f d Multidirectional compression filter H AC，i Then, a two-dimensional inverse Fourier transform is performed to obtain N imaging results s. IMG，i (t, η).

[0115] Step 6, refer to Figure 3 The imaging results s corresponding to adjacent segments are processed sequentially. IMG,i (t,η) The strong focal points in the overlapping region are aligned with the dimensional envelope and coherently accumulated. The non-overlapping regions are then stitched together to obtain the final imaging result S. all The specific steps are as follows:

[0116] Step 6-1, analyze the imaging result s IMG,i Perform an azimuth-to-Fourier transform on (t, η) to obtain the range-Doppler domain map s. IMG,i (t,f d );

[0117] Step 6-2: Based on the Fourier transform characteristics and the target spatial geometry, construct the expression for the tilt correction filter to correct image tilt:

[0118]

[0119] In the formula, l is the distance dimension of a typical building;

[0120] Step 6-3, map the distance to the Doppler domain s IMG,i (t,f d Multiplied by the tilt correction filter H GC-1 The tilt-corrected frequency domain plot was obtained.

[0121] Step 6-4, adjust the tilt-corrected frequency domain plot. Perform an inverse Fourier transform in the azimuth direction to obtain the tilt-corrected time-domain plot.

[0122] Step 6-5: Based on the target spatial geometry, obtain the expression for the stretching / compression factor:

[0123]

[0124] Step 6-6, convert the above formula <22> Substitute the tilted corrected time domain diagram From this, the deformed time-domain diagram was obtained.

[0125] Steps 6-7: Analyze the deformed time-domain plot. Perform a distance-to-Fourier transform to obtain the deformed frequency domain map.

[0126] Steps 6-8: Based on the Fourier transform characteristics and the target spatial geometry, construct the expression for the two-position correction filter to correct image translation:

[0127]

[0128] Steps 6-9: Obtain the deformed frequency domain diagram. Multiply by the above formula <23> Position correction filter H in GC-3 The geometrically corrected frequency domain plot was obtained.

[0129] Steps 6-10: Perform geometrically corrected frequency domain plots... Perform an inverse Fourier transform to the distance to obtain the geometrically corrected time-domain graphs of each segment.

[0130] Steps 6-11: Transform the geometrically corrected time-domain plot... Rotate counterclockwise The degree is used to obtain the time-domain image to be stitched, where the slant range is perpendicular to the trajectory of the maneuvering platform.

[0131] Steps 6-12: sequentially stitch adjacent time-domain diagrams to be concatenated. Alignment of strong focal points in overlapping regions with the distance dimension envelope;

[0132] Steps 6-13 involve coherent accumulation of overlapping regions and sequential connection of non-overlapping regions to complete sub-aperture stitching and obtain the full-aperture imaging result S. all .

[0133] The positioning component of the segmented aperture imaging and positioning method for synthetic aperture radar on rotor-borne UAVs includes the following steps:

[0134] Step S1: Perform distance pulse compression on the original echo signal s(t,η) to obtain the pulse compressed signal s RC (t,η), according to the pulse compression signal s RC Phase history of strong scattering points in (t,η) Estimated speed of computer-controlled moving platform and estimated beam center angle

[0135] This step is exactly the same as step 1 in the imaging method section, and will not be repeated here.

[0136] Step S2, based on the estimated speed The direction will compress the pulse signal s RC (t,η) is divided into N segments and the corresponding segment compression signal s for each segment. RC,i (t,η), i=1…N. And based on the estimated beam center angle... Estimate the platform trajectory coordinates corresponding to the i-th segment. k = 1…M, where M is the azimuth length of each segment.

[0137]

[0138]

[0139]

[0140] Where, θ in This indicates the angle between the UAV-borne synthetic aperture radar beam and the ground normal.

[0141] Step S3: In the adjacent regions of the i-th and (i-1)-th segments, extract three strong scattering points. The coordinates of the strong scattering points in the i-th segment are as follows: The coordinates of the strong scattering points in the (i-1)th segment are respectively

[0142] Step S4, based on the coordinates of the strong scattering point and Calculate the rotation matrix of the i-th and (i-1)-th segments.

[0143]

[0144] Step S5, set the platform trajectory coordinates of the i-th segment. Rotate to the coordinate system of the platform trajectory of the (i-1)th segment, so that the coordinates of the platform trajectory of adjacent segments are aligned.

[0145]

[0146] Step S6: Perform coherent accumulation of the platform trajectory coordinates in the overlapping regions of the i-th and (i-1)-th segments, and concatenate the platform trajectory coordinates in the non-overlapping regions to obtain the concatenated trajectory coordinates of the i-th and (i-1)-th segments [P]. x ,P y ,P z ].

[0147]

[0148] Step S7, repeat steps S3 to S6 N-1 times until the coordinates of the spliced ​​trajectory of all adjacent segments are obtained [P]. x ,P y ,P z This allows us to obtain the platform's final trajectory coordinates.

[0149] The advantages of this method are further illustrated below with specific comparative verification results.

[0150] 1. Simulation conditions:

[0151] As listed in Table 1:

[0152] Table 1 Simulation Parameters

[0153]

[0154]

[0155] 2. Simulation and Experiment

[0156] Verification 1: Under the conditions in Table 1, the original echo signal s of the dot matrix target obtained from the simulated and measured trajectory. r (t,η), the measured trajectory comes from the measured data collected by the inertial navigation system in a certain experiment, and its trajectory is as follows: Figure 5 As shown in (a) and (b). Among them, Figure 5 (a) shows the change curves of attitude angles of the maneuvering platform during flight recorded by the inertial navigation system. The three graphs from top to bottom are the change curves of roll angle, pitch angle and yaw angle relative to the platform movement. The horizontal axis in the graphs represents the path of the maneuvering platform from -200m to 200m. Figure 5 (b) shows the curves of the motion trajectory of the maneuvering platform recorded by the inertial navigation system in three dimensions. The three figures from top to bottom are the curves of the change in the azimuth dimension, the distance dimension, and the altitude dimension, respectively.

[0157] The dot matrix targets are 7×7 point targets with a range and azimuth interval of 25m. The spatial relationship between the dot matrix targets and the platform trajectory is as follows: Figure 6 As shown. Figure 6 The coordinate system shown, with X, Y, and Z corresponding to the azimuth, distance, and height dimensions respectively, is represented by R in the figure. ref This indicates that the system is at the current incident angle θ in The reference distance is θ, where H represents the average altitude of the maneuvering platform in flight. a θ represents the beamwidth in the azimuth dimension of the system. bw The beamwidth represents the system's distance dimension, S represents the width of the area under test, v represents the speed of the moving platform, and L represents the beamwidth of the system's distance dimension. sThis represents the length of a synthetic aperture. The lattice on the right side of the image represents the positional relationship of the simulated target, with the lattice center located at R on the left side of the image. ref Intersection with the Y-axis. The length and width of the dot matrix are each 7 points, with a dot spacing of 25m. The values ​​of other parameters are shown in Table 1.

[0158] For s r (t,η) distance pulse compression yields s RD (t,η), for s RD (t,η) Imaging comparisons were performed using the SAI algorithm and traditional imaging methods, respectively, and the results are as follows: Figure 7 , Figure 8 As shown, the path is segmented into three segments (segment 1, segment 2, and segment 3) by the SAI algorithm proposed in this paper. Figure 7 In the figures (a), (b), and (c), respectively, the imaging results after geometric correction of segments 1, 2, and 3 in this embodiment are shown. Since each segment can only image a part of the area, the imaging results of each segment only contain a part of the target area. In addition, at the edge of the azimuth dimension, the outline of the point target can be seen to be diffused. Figure 7 In this embodiment, (d) represents the full-aperture imaging result after stitching together each segment. Figure 7 In the diagram, (e) represents the imaging result obtained using the traditional imaging method. (Comparison) Figure 7 (d) and (e) show that the sharpness of the focal point outlines is different and the brightness is different, with the difference being most obvious in the points enclosed in the box. Figure 8 The results show a comparison of the point spread function between this embodiment and traditional imaging methods. Figure 8The upper figure shows a comparison of the point spread function curves of the third row of point targets in segments 1, 2, and 3. The lower figure shows a comparison of the point spread function curves of the third row of point targets in the SAI method proposed in this paper and the traditional imaging method. Analyzing the main lobe broadening at the -3dB position of each peak in the point spread function curves of segments 1, 2, and 3, the SAI method, and the traditional imaging method, we can obtain the extent of resolution expansion relative to the theoretical value. According to the statistical results in Table 2A), the smaller the main lobe broadening coefficient, the better the image focusing effect. Comparing the amplitudes of the main lobe and side lobes at each peak of these five curves, as shown in Table 2B, a larger main-to-side lobe ratio indicates a higher image signal-to-noise ratio, i.e., better image quality. Comparing the peak amplitude loss of these five curves allows us to observe the impact of incomplete aperture on the image. Although the peak amplitude loss is larger at the edges in each segment, the peak amplitude loss is significantly reduced in the final full-aperture image of SAI due to the complementary relationship between the segments. After measuring the point spread function curve index of a certain row of the imaging result, the amount of information contained in the complete image is judged by the image entropy. The higher the image focusing effect, the more information it contains, and the smaller the image entropy. The final full-aperture image of the SAI method has the smallest entropy. Comparing the main lobe broadening system, main lobe-to-side lobe ratio, peak amplitude loss, and image entropy, the SAI method proposed in this paper has higher resolution, greater signal-to-noise ratio, and peak amplitude loss comparable to the original method, while containing more information. Therefore, the SAI proposed in this paper has excellent imaging performance and solves the practical problems encountered by UAV-borne synthetic aperture radar.

[0159] Table 2 Comparison of Image Quality Indicators

[0160] A) Comparison of main lobe expansion coefficients

[0161]

[0162] B) Main lobe to side lobe ratio (dB)

[0163]

[0164]

[0165] C) Peak amplitude loss (dB)

[0166]

[0167] D) Image entropy

[0168]

[0169] Verification 2: Change the platform speed to 10 m / s and the platform flight altitude to 350 m among the conditions in Table 1, which are the test conditions for the actual flight test. The test platform is a multi-rotor UAV of model KWT-65, carrying a Ku-band small FM continuous wave synthetic aperture radar. Using this platform to fly about 100 flights, with a flight distance of about 1 km for each flight, multiple groups of echo data of a certain ground area are collected. After imaging multiple groups of echo data using the traditional method and the method of the present invention, the comparison of the focusing effects is as Figure 9 shown. Among them, Figure 9 (a) in it is the imaging effect of the traditional imaging method; Figure 9 (b) in it is the imaging effect of the method of this embodiment; Figure 9 (c) in it is the optical model corresponding to the imaging area.

[0170] Figure 9 It shows that the method of this embodiment has a better focusing effect than the traditional imaging method in terms of focusing effect. According to the optical image of the待测 target, it can be seen that the target presents a "工" shape. The "工" character contour of the imaging result using the method of the present invention is clearer than that using the traditional method. Taking the left side of the "工" shaped building in the image as an example, when the radar irradiates the glass, it will penetrate the glass and enter the room interior, and cannot return the echo to the radar. Therefore, the imaging result will show black, while there will be obvious reflections at the window edges due to the presence of right angles and metals, forming bright spots. The imaging result corresponding to the method of the present invention can clearly observe the floor intervals through the black areas between the bright spots, and the bright spots in the imaging result corresponding to the traditional method are all blurred into a mass, and it is impossible to clearly observe the glass between the floors. There is a road at the upper left of the "工" shaped building. Using the method of the present invention, the edge of the road can be clearly seen, the contour of the road edge obtained by the traditional method is not very clear, and there are extra longitudinal stripes. And the imaging effect of the garden contour around the "工" shaped building is also clearer using the method of the present invention. The overall signal-to-noise ratio of the image is also better than the traditional method. In addition, through multiple / multi-location flight tests, it is verified that this method has a good focusing effect on imaging 90% of the echo data, while the images with good focusing effect of the traditional method are about 35%. For images of the same scale (12500×8192) on the same computer platform, the imaging time of this method is about 320 s, and the imaging time of the traditional imaging method is about 5200 s, and the speed is increased by about 16 times.

[0171] Verification 3: Process the data of a certain actual test under the conditions of Verification 2. While obtaining the image, the trajectory of the platform can be estimated. The comparison of the trajectory estimated using the method of the present invention and the trajectory collected by the inertial navigation device is as Figure 10 shown. Among them, Figure 10(a) shows the comparison between the trajectory estimated by the method of the present invention and the trajectory collected by the inertial navigation device from top to bottom in the azimuth dimension, the distance dimension, and the height dimension. The blue curve represents the measurement result of the inertial navigation device, and the red curve represents the estimation result of the method of the present invention. It can be seen that the two basically overlap in the azimuth dimension. The high sampling rate of the multi-rotor platform ensures a small error in the azimuth dimension. There is a slight deviation between the two in the distance dimension. The blue curve will deviate at the edge of the red curve, especially at the three peaks with large fluctuations. In the height dimension, the deviation between the blue curve and the red curve is larger. Since the method of the present invention is based on two-dimensional image estimation, some errors will accumulate in the third dimension, namely the height dimension. Figure 9 (b) is the estimation error curve of the method in this embodiment. The curve shows that the deviations in the azimuth dimension, range dimension, and altitude dimension are approximately at the centimeter level, centimeter level, and decimeter level, respectively, reaching a level close to that of inertial navigation.

[0172] The role and effect of the embodiments

[0173] The segmented aperture imaging method of multi-rotor UAV-borne synthetic aperture radar provided in this embodiment mainly includes: 1) acquiring target echoes from the UAV-borne synthetic aperture radar system; 2) estimating the motion state of the maneuvering platform based on the echo signals; 3) segmenting the echo signals according to the platform motion state; 4) performing motion compensation on each segment of the echo signals according to the platform motion state; 5) performing a two-dimensional Fourier transform on each compensated echo signal to obtain a two-dimensional spectrum, and using a series inversion method to decompose the two-dimensional spectrum to obtain phase filters for each segment; 6) multiplying each segment's two-dimensional spectrum by the corresponding phase filter, and then performing a two-dimensional inverse Fourier transform on the two-dimensional spectrum to obtain images for each segment; 7) performing geometric correction on each segment's images, and then stitching the images together to obtain a full aperture imaging result; 8) stitching together the platform trajectories of each segment to obtain the complete platform trajectory coordinates. This embodiment employs a method based on echo signal phase history estimation of platform motion parameters and echo signal segmentation to accurately compensate for platform motion errors, achieving good imaging focusing effect and high imaging success rate. This improves the efficiency of data acquisition for multi-rotor UAV platforms and can effectively perform high-resolution imaging of synthetic aperture radar systems for multi-rotor UAV platforms. Simultaneously, this method can calculate the three-dimensional coordinates of the platform trajectory during imaging, achieving platform positioning. This can be applied to UAV navigation, providing possibilities for future intelligent and integrated detection systems.

[0174] Compared with existing technologies, it has the following advantages:

[0175] Compared with traditional airborne synthetic aperture radar imaging algorithms, this algorithm takes into account the relationship between platform motion and signal phase, and can be applied to synthetic aperture radar imaging systems without inertial navigation equipment or with low-precision inertial navigation equipment.

[0176] Compared with traditional airborne synthetic aperture radar imaging algorithms, this algorithm takes into account the squinting effect caused by the attitude angle change of the UAV platform, compensates for the coupled phase of range and azimuth, improves the image focusing effect, and is applicable to synthetic aperture radar imaging systems without antenna servo.

[0177] Compared with traditional airborne synthetic aperture radar imaging algorithms, this method uses phase filter multiplication instead of interpolation, which improves the imaging speed of a single image in each segment.

[0178] Compared with traditional airborne synthetic aperture radar imaging algorithms, this method uses parallel imaging of each segment and then stitches them together to obtain a complete image, which improves the imaging speed of the complete image.

[0179] Furthermore, experimental verification shows that the segmented aperture imaging (SAI) method for multi-rotor UAV-borne synthetic aperture radar proposed in this embodiment has high imaging resolution and fast computation speed. At the same time, it can estimate the platform trajectory, reducing the hardware requirements of the algorithm. This indicates that the present invention can be effectively applied to synthetic aperture radar imaging systems for small mobile platforms.

[0180] This embodiment derives in detail the correspondence between platform motion parameters and echo phase, and demonstrates that the second-order expansion of the two-dimensional spectrum under oblique viewing can achieve motion compensation and imaging without relying on inertial navigation equipment. Simulations compared the point spread function curves of each segment, the complete image, and traditional imaging algorithms using the method of this embodiment. Field measurements verified and compared the imaging results of the method of this invention with those of traditional algorithms, proving that this embodiment can effectively perform high-resolution imaging of synthetic aperture radar systems on multi-rotor UAV platforms.

[0181] The above embodiments are preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention.

Claims

1. A segmented aperture imaging method for a multi-rotor unmanned aerial vehicle (UAV)-borne synthetic aperture radar, used for segmented aperture imaging based on the raw echo signal of the UAV-borne synthetic aperture radar, characterized in that, The method comprises the following steps: Step 1, performing range pulse compression on the original echo signal to obtain a range pulse compressed signal where is the range fast time, is the azimuth slow time, according to the range pulse compressed phase history of the strong scattering point estimated velocity of the computer moving platform and estimated beam center squint angle ; Step 2, based on the estimated speed The direction of the distance pulse compression signal Segmentation Each segment and the segment pulse compression signal corresponding to each segment. ; Step 3, according to the estimated velocity and the estimated beam center squint angle calculate the phase compensation amount of the segment pulse pressure signal corresponding to each segment , multiply the segment pulse pressure signal by a motion error compensation filter , to obtain a compensated signal , imaginary ;​ Step 4, for the compensated signal Perform a two-dimensional Fourier transform to obtain Two-dimensional spectrum The two-dimensional spectrum is analyzed using a series inversion method. Decompose and construct an azimuth compression filter. ,in, To keep pace with the distance and time The corresponding frequency, To slow down the direction The corresponding Doppler frequency; Step 5, applying the two-dimensional spectrum to the azimuthal compression filter and then performing a two-dimensional inverse Fourier transform to obtain an imaging result ; Step 6, sequentially corresponding to the imaging results of adjacent segments The strong focusing point in the overlapping area is aligned with the distance envelope and accumulated in phase, and the non-overlapping area is spliced to obtain the final imaging result , In the step 3, the phase compensation amount , wherein, is the mean value of each segment , is the mean value of estimated beam center squint angles . In step 4, the azimuth compression filter where, is the range fast time corresponding frequency, is the azimuth slow time corresponding Doppler frequency, is the carrier frequency of the system transmitted signal, is the speed of light, In step 4, the two-dimensional spectrum is inverted using a series inversion method The process of decomposing to obtain the azimuth compression filter expression is performed in the following sub-steps: Step 4-2-1, the two-dimensional spectrum corresponding to the segment is obtained according to the stationary point method expression: ， wherein denotes the carrier frequency of the system transmission signal, denotes the speed of light; Step 4-2-2 - The two-dimensional spectrum expression in Step 4-2-1 is decomposed using the method of series inversion to obtain a two-dimensional spectrum expression that eliminates the coupling terms of and ​​ ； Step 4-2-3, the elimination according to the description in Step 4-2-2 and The two-dimensional spectral expression of the coupling term builds an ideal phase filter expression: ； Step 4 - 2 - 4, the estimated velocity , the estimated beam center squint angle , the distance estimate into the ideal phase filter expression, to obtain the azimuth compression filter expression: 。 2. The multi-copter UAV-borne synthetic aperture radar sub-aperture imaging method according to claim 1, characterized in that: wherein, In step 1, the distance pulse compression signal The phase history of the strong scattering point described in the text The phase history of the strong scattering point was obtained by performing a second-order fitting. According to the coefficients of the second-order terms and first-order term coefficients Calculate the estimated speed of the motorized platform. and the estimated beam center angle ,in, For the sake of speed, For location, slow time, For higher-order phase errors, For constant phase term, The wavelength of the system's transmitted signal. This is an estimated value for the system reference distance.

3. The multi-copter UAV-borne synthetic aperture radar sub-aperture imaging method according to claim 1, characterized in that: wherein In step 2, based on the estimated speed The direction of the distance pulse compression signal Divided sequentially into those with the same velocity direction Each segment is divided into segments, and then it is determined whether the length of each segment is less than the length of a synthetic aperture. If it is less, the segment is extended to both sides to the length of the synthetic aperture, thus obtaining the segmented product. The pulse pressure signal of the segment .

4. The multi-copter UAV-borne synthetic aperture radar sub-aperture imaging method according to claim 1, characterized in that: wherein In step 6, the imaging results corresponding to adjacent segments are first processed. Geometric correction is performed to obtain the corrected imaging result. Then the corrected imaging results Rotation The corrected imaging result is obtained with a slant range perpendicular to the trajectory of the motorized platform. Then, the corrected imaging results corresponding to adjacent segments are processed sequentially. The strong focal points in the overlapping region are aligned with the distance-dimensional envelope and coherently accumulated. The non-overlapping regions are then stitched together to obtain the final imaging result. .

5. The multi-copter UAV SAR sub-aperture imaging method of claim 4, characterized in that: wherein the step 6 geometrically corrects the imaging result carrying out the geometric correction is performed in the following substeps: Step 6-1, performing an azimuth Fourier transform on the imaging result to obtain a range-Doppler domain plot ; Step 6-2, according to the Fourier transform characteristics and the target space geometry, an expression of a tilt correction filter for correcting image tilt is constructed: ， In the formula, is the distance dimension of a typical building; Step 6-3, multiply the range-Doppler domain plot by the slant correction filter ;​​ Step 6-4, azimuth Fourier transform of the tilt-corrected frequency domain map ;​ Step 6-5, according to the target space geometry, an expression of a stretch / compression factor is obtained: ； Step 6-6, substituting the stretch / compression factor expression into the slant-corrected time-domain map to obtain a deformed-corrected time-domain map ; Step 6-7, performing a distance-wise Fourier transform on the deformed corrected time domain map to obtain a deformed corrected frequency domain map Step 6-7, performing a distance-wise Fourier transform on the deformed corrected time domain map to obtain a deformed corrected frequency domain map ; Step 6-8, according to the Fourier transform characteristics and the target space geometry, an expression of a two-position correction filter for correcting image translation is constructed: ； Step 6-9, the geometry-corrected frequency domain image is obtained by multiplying the position-corrected frequency domain image by the position correction filter Step 6-8, the position-corrected frequency domain image is obtained by multiplying the frequency domain image by the position correction filter Step 6-9, the geometry-corrected frequency domain image is obtained by multiplying the position-corrected frequency domain image by the position correction filter ; Step 6-10, inverse Fourier transform of the distance direction of the geometrically corrected frequency domain image to obtain a geometrically corrected time domain image of each segment Step 6-10, inverse Fourier transform of the distance direction of the geometrically corrected frequency domain image to obtain a geometrically corrected time domain image of each segment ; Step 6-11, the time domain image is geometrically corrected clockwise rotation degrees, to get the slant range perpendicular to the trajectory of the mobile platform to be spliced time domain image ; Step 6-12, sequentially to adjacent said to be spliced time domain map The strong focusing point in the overlapping area is in the distance dimension envelope alignment; Step 6-13, coherent accumulation is performed on the overlapping areas, and the non-overlapping areas are sequentially connected to complete sub-aperture stitching to obtain full-aperture imaging results .

6. The multi-copter UAV-borne synthetic aperture radar sub-aperture imaging method according to claim 1, characterized in that: wherein In step 1, the estimated velocity and the estimated beam center squint angle are calculated as follows: Step 1-1, definition of Doppler frequency modulation slope and Doppler center and the calculation formula of and ​ ， ， In the formula, denotes the azimuthal slow time derivative Steps 1-2, the space spanned by the raw echo signals is called the signal space, and the distance pulse-echo signals the phase history of the strong scatterer a second order fit to obtain an expression of the strong scatterer in the signal space: ， wherein is a second order term coefficient, is a first order term coefficient, is a constant phase term, is a high order phase error; Step 1-3, the phase history of step 1-1 of step 1-2 and the calculation formula of in the signal space and expressions: ， ； Step 1-4, the space constituted by the target and the actual position of the mobile platform is called target space, and the phase history in the target space is obtained according to the spatial geometric relationship Expression: ； wherein represents the velocity of the moving platform, represents the beam center angle of inclination caused by the moving platform, represents the closest distance between the target and the moving platform, is the wavelength of the system transmitted signal; Step 1-5, the expression of the phase history in the target space in step 1-1 in step 1-2 and the calculation formula of in step 1-3, respectively, to obtain the expression of and in the target space ， ； Step 1 - 6, compare the expression in the signal space with the expression in the target space , the right side of the expression in the signal space with the right side of the expression in the target space , to get the estimated velocity, the estimated beam center squint angle, and the range estimate ， ， 。 7. A multi-copter UAV SAR sub-aperture positioning method for calculating a UAV flight trajectory from raw echo signals of a UAV SAR characterized in that, The method comprises the following steps: Step S1, performing range pulse compression on the original echo signal to obtain a range pulse compressed signal , calculating the estimated velocity of the computer moving platform and the estimated beam center squint angle from the phase history of the strong scattering points in the range pulse compressed signal ;​​​ Step S2, segmenting the range pulse compression signal according to the direction of the estimated velocity is the azimuth length of each segment.​​​​​​​​ ， ， ， In the formula, denotes the angle between the UAV-borne synthetic aperture radar beam and the ground normal. Step S3, in the first and the second adjacent regions in the segment, extract three strong scattering points, the coordinates of the strong scattering points in the first segment are respectively , the coordinates of the strong scattering points in the second segment are respectively ; Step S4, based on the coordinates of the strong scattering point and Calculate the first The and the first Rotation matrix of each segment , ； Step S5, the first The platform trajectory coordinates of each segment Rotate to the first In the coordinate system of the platform trajectory of each segment, the coordinates of the platform trajectory of adjacent segments are aligned, that is: ； Step S6, phase accumulation is performed on the platform track coordinates of the overlapping areas of the first and second segments, and the platform track coordinates of the non-overlapping areas are spliced to obtain spliced track coordinates of the first and second segments ,​​​​ ； Step S7, repeat steps S3 to S6 until the spliced track coordinates of all adjacent segments are obtained , thereby obtaining the final track coordinates of the platform .

Images