Airborne sar moving target imaging method based on doppler parameter estimation

Abstract

The application relates to an airborne SAR moving target imaging method based on Doppler parameter estimation, and aims to solve the problem that a traditional RD algorithm is not applicable to airborne SAR moving target imaging. The process is as follows: 1. the Doppler center frequency and the Doppler frequency modulation slope of a target are derived; 2. a distance frequency domain signal is obtained; 3. a distance compression function is constructed to obtain distance compressed echo data; 4. an azimuth compression function is constructed to obtain azimuth compressed data; 5. a particle swarm optimization algorithm is adopted to obtain the target Doppler center frequency and the Doppler frequency modulation slope corresponding to the minimum entropy value; 6. a distance migration correction function is constructed to obtain two-dimensional frequency domain data after distance migration correction; 7. a distance matching function is constructed to obtain distance compressed azimuth frequency domain data; 8. an azimuth matching function is constructed to obtain two-dimensional time domain data, and imaging is completed. The application is used in the field of airborne SAR moving target imaging.

Description

Technical Field

[0001] This invention relates to an airborne SAR moving target imaging method. Background Technology

[0002] During scene imaging, the airborne synthetic aperture radar (SAR) moves at a constant speed along its heading and transmits a linear frequency modulated (LFM) signal at a certain pulse repetition frequency. The received scene echo signal exhibits two-dimensional coupling in the range and azimuth directions, i.e., range migration phenomenon in the system response curve. Traditional RD or CS algorithms can effectively correct the range migration error of the echo signal and complete range and azimuth compression processing to achieve two-dimensional imaging. Typically, the scene targets studied by airborne SAR imaging technology are mostly stationary targets. However, when the target's motion state changes, such as moving ground targets and surface ships, the radial motion between the radar and the target becomes more complex, making it impossible to process the range migration and two-dimensional compression of the target echo signal using conventional RD or CS algorithms. Furthermore, moving targets in the scene are usually non-cooperative, and their motion state is unpredictable, making it impossible to accurately obtain parameters such as range delay and Doppler frequency. Therefore, new imaging algorithms are needed to solve the problem of moving target imaging in airborne SAR. Summary of the Invention

[0003] The purpose of this invention is to solve the problem that traditional RD algorithms are not suitable for airborne SAR moving target imaging, and to propose an airborne SAR moving target imaging method based on Doppler parameter estimation.

[0004] The specific process of the airborne SAR moving target imaging method based on Doppler parameter estimation is as follows:

[0005] Step 1: Assume the motion parameters V of the moving target in the scene. U With V V Given that a moving target is considered a stationary target, the radar transmitter moves along the X-axis in the direction V. T -V U Along the Y-axis -V V The speed of flight was used to deduce the synthetic oblique angle of the radar to the target, and further deduce the Doppler center frequency and Doppler modulation slope of the target.

[0006] Step 2: The radar samples the echoes of moving targets in the scene to obtain raw echo data. For raw echo data Perform a range-direction FFT to obtain the range frequency domain signal S1(f) r ,t m );

[0007] Where FFT stands for Fast Fourier Transform; For distance to time domain; f rFor the range-direction frequency; t m For azimuth to time domain; S1(f r ,t m () represents the data after the original data has undergone a distance-to-Fourier transform;

[0008] Step 3: Construct the distance-directed compression function H Range (f r ,t m ), will S1(f r ,t m Multiply by H Range (f r ,t m Then perform range-wise IFFT to obtain range-compressed echo data.

[0009] Step 4: Construct the azimuth compression function The distance-compressed echo data from step three Perform azimuth FFT and then multiply by Then perform azimuth-oriented IFFT to obtain azimuth-compressed data.

[0010] Where f a The azimuth frequency domain frequency;

[0011] Step 5: Compress the azimuth data from Step 4. Calculate the entropy of a two-dimensional image Using entropy as the cost function, the particle swarm optimization algorithm is employed to obtain the target Doppler center frequency f corresponding to the minimum entropy. dc and Doppler frequency modulation slope K a ;

[0012] Step 6: Use the Doppler center frequency f estimated in Step 5. dc Construct distance migration correction function H1(f) r ,t m ), and the distance frequency domain data S1(f r ,t m Multiply by H1(f) r ,t m Then, perform an azimuth-directed FFT to obtain the range-migration corrected two-dimensional frequency domain data S2(f). r ,f a );

[0013] Step 7: Process the two-dimensional frequency domain data S2(f) from Step 6. r ,f a Multiply by the distance matching function H2(f) r ,f aThen perform range-directed IFFT to obtain range-compressed azimuth frequency domain data.

[0014] Step 8: Process the azimuth frequency domain data from Step 7. Multiply by the orientation matching function Then, an azimuth-oriented IFFT is performed to obtain two-dimensional time-domain data. Imaging complete.

[0015] The beneficial effects of this invention are as follows:

[0016] When a target in the scene moves, the radial motion of both the radar and the target undergoes additional changes caused by the target's motion. This makes the Doppler center frequency and Doppler modulation slope more complex, rendering the original RD algorithm ineffective. To effectively image moving targets, it is necessary to accurately estimate the Doppler parameters of the echo signal and construct a range migration correction function and a two-dimensional matched compression function based on these parameters. Therefore, this invention proposes an airborne SAR moving target imaging method based on Doppler parameter estimation.

[0017] In target motion scene models, such as the parameter model presented in this invention, traditional airborne SAR range-Doppler algorithms cannot effectively compensate for range migration errors of moving targets, and azimuth compression results in significant broadening, making precise focusing difficult. In contrast, the airborne SAR moving target imaging method based on Doppler parameter estimation can effectively correct for range migration errors of moving targets in the scene while maintaining azimuth compression, ensuring clear imaging of moving targets. In summary, the airborne SAR moving target imaging method based on Doppler parameter estimation outperforms the traditional airborne SAR range-Doppler algorithm in terms of range migration error compensation and azimuth compression imaging performance in target motion scene models. Attached Figure Description

[0018] Figure 1 This is a geometric model diagram of the moving target imaging scene used in this invention;

[0019] Figure 2 A comparison chart of distance migration correction using the traditional RD algorithm;

[0020] Figure 3 This is a directional compression map based on the traditional RD algorithm.

[0021] Figure 4 This is a comparison chart of distance migration correction algorithms of this invention;

[0022] Figure 5 This is a location compression diagram of the algorithm of this invention. Detailed Implementation

[0023] Specific implementation method one: Combining Figure 1This embodiment describes the specific process of the airborne SAR moving target imaging method based on Doppler parameter estimation as follows:

[0024] Step 1: Assume the motion parameters V of the moving target in the scene. U With V V Given that a moving target is considered a stationary target, the radar transmitter moves along the X-axis in the direction V. T -V U Along the Y-axis -V V The speed of flight was used to deduce the synthetic oblique angle of the radar to the target, and further deduce the Doppler center frequency and Doppler modulation slope of the target.

[0025] Step 2: The radar samples the echoes of moving targets in the scene to obtain raw echo data. For raw echo data Perform a range-direction FFT to obtain the range frequency domain signal S1(f) r ,t m The radar is mounted on the aircraft.

[0026] Where FFT stands for Fast Fourier Transform; For distance to time domain; f r For the range-direction frequency; t m For azimuth to time domain; S1(f r ,t m () represents the data after the original data has undergone a distance-to-Fourier transform;

[0027] Step 3: Construct the distance-directed compression function H Range (f r ,t m ), will S1(f r ,t m Multiply by H Range (f r ,t m Then perform range-wise IFFT to obtain range-compressed echo data.

[0028] Step 4: Construct the azimuth compression function The range-compressed echo data from step three Perform azimuth FFT and then multiply by Then perform azimuth-oriented IFFT to obtain azimuth-compressed data.

[0029] Where f a The azimuth frequency domain frequency;

[0030] Step 5: Compress the azimuth data from Step 4. Calculate the entropy of a two-dimensional image Using entropy as the cost function, the particle swarm optimization (PSO) algorithm is employed to obtain the target Doppler center frequency f corresponding to the minimum entropy. dc and Doppler frequency modulation slope K a ;

[0031] Step 6: Use the Doppler center frequency f estimated in Step 5. dc Construct distance migration correction function H1(f) r ,t m ), and the distance frequency domain data S1(f r ,t m Multiply by H1(f) r ,t m Then, perform an azimuth-directed FFT to obtain the range-migration corrected two-dimensional frequency domain data S2(f). r ,f a );

[0032] Step 7: Process the two-dimensional frequency domain data S2(f) from Step 6. r ,f a Multiply by the distance matching function H2(f) r ,f a Then perform range-directed IFFT to obtain range-compressed azimuth frequency domain data.

[0033] Step 8: Process the azimuth frequency domain data from Step 7. Multiply by the orientation matching function Then, an azimuth-oriented IFFT is performed to obtain two-dimensional time-domain data. Imaging complete.

[0034] Specific Implementation Method Two: This implementation method differs from Specific Implementation Method One in that, in step one, it is assumed that the motion parameter V of the moving target in the scene is... U With V V Given that a moving target is considered a stationary target, the radar transmitter moves along the X-axis in the direction V. T -V U Along the Y-axis -V V The target's speed of flight was used to derive the radar's synthetic oblique angle of view, and further, the target's Doppler center frequency and Doppler modulation slope were derived; the specific process is as follows:

[0035] Taking the direction of radar platform motion as the positive X-axis, the Y-axis perpendicular to the X-axis, and the Z-axis perpendicular to the XOY plane; assuming the motion parameter V of the moving target in the scene... U With V V Given that a moving target is considered a stationary target, the radar transmitter moves along the X-axis in the direction V. T -V U Along the Y-axis -VV The speed of flight is used to deduce the synthetic oblique angle of the radar on the target as follows:

[0036]

[0037] Where θ T It is the original oblique angle of the radar transmitter, θ T1 It is the radar transmitter synthesized velocity V T The angle between ′ and the X-axis, α T It is the downward view of the radar transmitter; V T and V U These represent the velocities of the radar and the target along the X-axis, respectively; V V The target's velocity along the Y-axis; the radar transmitter's synthesized velocity.

[0038] Let there be a certain time t m The slant range between the radar and the point target P is

[0039]

[0040] Where (x) T0 ,y T0 ,H)(x P0 ,y P0 ,z P0 ) represent the starting positions of the radar platform and the target, respectively; R c H is the slant range at which the radar reaches the target at the initial moment; H is the radar platform height.

[0041] From this, we can deduce that the target's Doppler center frequency and Doppler modulation slope are respectively...

[0042]

[0043]

[0044] Where λ is the wavelength of the LFM signal; For slow time t m The value is taken when = 0.

[0045] The other steps and parameters are the same as in Specific Implementation Method 1.

[0046] Specific Implementation Method Three: This implementation method differs from Specific Implementation Method One or Two in that step three involves constructing the distance compression function H. Range (f r ,t m The expression is:

[0047]

[0048] Where IFFT stands for Inverse Fast Fourier Transform; j is an imaginary number, j 2 =-1; c is the speed of light, c = 3 × 10 8 m / s; f c For LFM signal carrier frequency; K r The frequency modulation slope of the LFM signal; This is the echo data after range compression.

[0049] Other steps and parameters are the same as in specific implementation method one or two.

[0050] Specific Implementation Method Four: This implementation method differs from Specific Implementation Methods One to Three in that step four involves constructing an azimuth compression function. The expression is:

[0051]

[0052] The other steps and parameters are the same as those in one of the specific implementation methods one to three.

[0053] Specific Implementation Method Five: This implementation method differs from Specific Implementation Methods One to Four in that the azimuth compression data from Step Four is processed in Step Five. Calculate the entropy of a two-dimensional image Using entropy as the cost function, the particle swarm optimization (PSO) algorithm is employed to obtain the target Doppler center frequency f corresponding to the minimum entropy. dc and Doppler frequency modulation slope K a The specific process is as follows:

[0054] The expressions for calculating entropy and the Particle Swarm Optimization (PSO) algorithm are as follows:

[0055]

[0056]

[0057] Where S is an intermediate variable. Σ(·) is the summation symbol; |·| is the absolute value symbol; x and v represent the particle's position and velocity, respectively; k represents the number of iterations; i represents the current particle index; ω is the inertia factor, usually taking values ​​in [0, 1, 0, 9]; c1 and c2 are learning factors, taking values ​​near 2; rand() is a random number in the range [0, 1]. and gbest k These are the historical optimal solutions for the i-th particle in the k-th iteration and the historical optimal solutions for the particle swarm in the k-th iteration, respectively.

[0058] The other steps and parameters are the same as those in one of the specific implementation methods one to four.

[0059] Specific Implementation Method Six: This implementation method differs from Specific Implementation Methods One to Five in that step six uses the Doppler center frequency f estimated in step five. dc Construct distance migration correction function H1(f) r ,t m The expression is:

[0060]

[0061] Where H1 is the distance migration correction function.

[0062] The other steps and parameters are the same as those in one of the specific implementation methods one to five.

[0063] Specific Implementation Method Seven: This implementation method differs from Specific Implementation Methods One through Six in that the distance matching function H2(f) in step seven... r ,f a The expression is:

[0064]

[0065] Where f am It is the maximum value in the directional frequency domain; H2(f r ,f a ) is the distance matching function; It is S2(f r ,f a The azimuth frequency domain data after distance compression.

[0066] The other steps and parameters are the same as those in one of the specific implementation methods one to six.

[0067] Specific Implementation Method Eight: This implementation method differs from Specific Implementation Methods One through Seven in that the directional matching function in step eight... The expression is:

[0068]

[0069] f dc To obtain the target Doppler center frequency f corresponding to the minimum entropy value in step five. dc ;

[0070] in This is a directional matching function; for Two-dimensional time-domain data after azimuth compression.

[0071] The other steps and parameters are the same as those in any of the specific implementation methods one to seven.

[0072] The beneficial effects of the present invention are verified using the following embodiments:

[0073] Example 1:

[0074] The airborne SAR moving target imaging method based on Doppler parameter estimation in this embodiment is prepared according to the following steps:

[0075] Here is a parametric model:

[0076] Table 1. Basic parameter values ​​of the airborne SAR moving target imaging method based on Doppler parameter estimation.

[0077]

[0078]

[0079] Simulation results of moving targets are as follows Figure 2 , 3 4, 5; Figure 2 A comparison chart of distance migration correction using the traditional RD algorithm; Figure 3 This is a directional compression map based on the traditional RD algorithm. Figure 4 This is a comparison chart of distance migration correction algorithms of this invention; Figure 5 This is a location compression diagram of the algorithm of this invention. From Figure 2 and Figure 4 The comparison results clearly show that the distance migration correction effect of the method of the present invention is better than that of the traditional method. From Figure 3 and Figure 5 The comparison results clearly show that the azimuth focusing effect of the method of the present invention is better than that of the traditional method.

[0080] This invention may have other embodiments. Without departing from the spirit and essence of this invention, those skilled in the art can make various corresponding changes and modifications according to this invention, but these corresponding changes and modifications should all fall within the protection scope of the appended claims.

Claims

1. An airborne SAR moving target imaging method based on Doppler parameter estimation, characterized in that: The specific process of the method is as follows: Step 1: Assume the motion parameters V of the moving target in the scene. U With V V Given that a moving target is considered a stationary target, the radar transmitter moves along the X-axis in the direction V. T -V U Along the Y-axis -V V The speed of flight was used to deduce the synthetic oblique angle of the radar to the target, and further deduce the Doppler center frequency and Doppler modulation slope of the target. Step 2: The radar samples the echoes of moving targets in the scene to obtain raw echo data. For raw echo data Perform a range-direction FFT to obtain the range frequency domain signal S1(f) r ,t m ); Where FFT stands for Fast Fourier Transform; For distance to time domain; f r For the range-direction frequency; t m For azimuth to time domain; S1(f r ,t m () represents the data after the original data has undergone a distance-to-Fourier transform; Step 3: Construct the distance-directed compression function H Range (f r ,t m ), will S1(f r ,t m Multiply by H Range (f r ,t m Then perform range-wise IFFT to obtain range-compressed echo data. Step 4: Construct the azimuth compression function The distance-compressed echo data from step three Perform azimuth FFT and then multiply by Then perform azimuth-oriented IFFT to obtain azimuth-compressed data. Where f a The azimuth frequency domain frequency; Step 5: Compress the azimuth data from Step 4. Calculate the entropy of a two-dimensional image Using entropy as the cost function, the particle swarm optimization algorithm is employed to obtain the target Doppler center frequency f corresponding to the minimum entropy. dc and Doppler frequency modulation slope K a ; Step 6: Use the Doppler center frequency f estimated in Step 5. dc Construct distance migration correction function H1(f) r ,t m ), and the distance frequency domain data S1(f r ,t m Multiply by H1(f) r ,t m Then, perform an azimuth-directed FFT to obtain the range-migration corrected two-dimensional frequency domain data S2(f). r ,f a ); Step 7: Process the two-dimensional frequency domain data S2(f) from Step 6. r ,f a Multiply by the distance matching function H2(f) r ,f a Then perform range-directed IFFT to obtain range-compressed azimuth frequency domain data. Step 8: Process the azimuth frequency domain data from Step 7. Multiply by the orientation matching function Then, an azimuth-oriented IFFT is performed to obtain two-dimensional time-domain data. Complete imaging; In step one, it is assumed that the motion parameter V of the moving target in the scene is... U With V V Given that a moving target is considered a stationary target, the radar transmitter moves along the X-axis in the direction V. T -V U Along the Y-axis -V V The target's speed of flight was used to derive the radar's synthetic oblique angle of view, and further, the target's Doppler center frequency and Doppler modulation slope were derived; the specific process is as follows: Taking the direction of radar platform motion as the positive X-axis, the Y-axis perpendicular to the X-axis, and the Z-axis perpendicular to the XOY plane; assuming the motion parameter V of the moving target in the scene... U With V V Given that a moving target is considered a stationary target, the radar transmitter moves along the X-axis in the direction V. T -V U Along the Y-axis -V V The speed of flight is used to deduce the synthetic oblique angle of the radar on the target as follows: Where θ T It is the original oblique angle of the radar transmitter, θ T1 It is the radar transmitter synthesized velocity V T The angle between ′ and the X-axis, α T It is the downward view of the radar transmitter; V T and V U These represent the velocities of the radar and the target along the X-axis, respectively; V V The target's velocity along the Y-axis; the radar transmitter's synthesized velocity. Let there be a certain time t m The slant range between the radar and the point target P is Where (x) T0 ,y T0 ,H)(x P0 ,y P0 ,z P0 ) represent the starting positions of the radar platform and the target, respectively; R c H is the slant range at which the radar reaches the target at the initial moment; H is the radar platform height. From this, we can deduce that the target's Doppler center frequency and Doppler modulation slope are respectively... Where λ is the wavelength of the LFM signal; |t m =0 means at slow time t m The value is taken when = 0; In step three, the distance compression function H is constructed. Range (f r ,t m The expression is: Where IFFT stands for Inverse Fast Fourier Transform; j is an imaginary number, j 2 =-1; c is the speed of light, c = 3 × 10 8 m / s; f c For LFM signal carrier frequency; K r The frequency modulation slope of the LFM signal; This is the echo data after range compression; In step four, an azimuth compression function is constructed. The expression is:

2. The airborne SAR moving target imaging method based on Doppler parameter estimation according to claim 1, characterized in that: The fifth step involves compressing the azimuth data from the fourth step. Calculate the entropy of a two-dimensional image Using entropy as the cost function, the particle swarm optimization algorithm is employed to obtain the target Doppler center frequency f corresponding to the minimum entropy. dc and Doppler frequency modulation slope K a The specific process is as follows: The expressions for calculating entropy and particle swarm optimization algorithms are as follows: Where S is an intermediate variable. ∑(·) is the summation symbol; |·| is the absolute value symbol; x and v represent the particle's position and velocity, respectively; k represents the number of iterations; i represents the current particle index; ω is the inertia factor; c1 and c2 are learning factors; rand() is a random number in the range [0, 1]. and gbest k These are the historical optimal solutions for the i-th particle in the k-th iteration and the historical optimal solutions for the particle swarm in the k-th iteration, respectively.

3. The airborne SAR moving target imaging method based on Doppler parameter estimation according to claim 2, characterized in that: In step six, the Doppler center frequency f estimated in step five is used. dc Construct distance migration correction function H1(f) r ,t m The expression is: Where H1 is the distance migration correction function.

4. The airborne SAR moving target imaging method based on Doppler parameter estimation according to claim 3, characterized in that: In step seven, the distance matching function H2(f) r ,f a The expression is: Where f am It is the maximum value in the directional frequency domain; H2(f r ,f a ) is the distance matching function; It is S2(f r ,f a The azimuth frequency domain data after distance compression.

5. The airborne SAR moving target imaging method based on Doppler parameter estimation according to claim 4, characterized in that: The directional matching function in step eight The expression is: in This is a directional matching function; for Two-dimensional time-domain data after azimuth compression.

Images