CN117849799B - A method for compensating residual motion errors of harmonic synthetic aperture radar - Google Patents

Abstract

The present invention discloses a method for compensating residual motion error of harmonic synthetic aperture radar, and the method comprises the following steps: step 1, constructing a harmonic synthetic aperture radar echo model; step 2, separating echo signal data; step 3, coarse focusing of SAR images; step 4, estimating residual error phase; and step 5, compensating error phase. The present invention uses residual phase error compensation of fundamental frequency image data to compensate harmonic SAR images, which fully utilizes the advantage of high signal-to-noise ratio of fundamental frequency images, has higher estimation accuracy than directly estimating phase error curves using harmonic images, and can effectively improve the focusing performance of harmonic images.

Description

A method for compensating residual motion errors of harmonic synthetic aperture radar

Technical Field

The invention belongs to the field of harmonic synthetic aperture radar, and in particular relates to a harmonic synthetic aperture radar residual motion error compensation method.

Background technique

Most of the scatterers in nature are linear scatterers. When traditional radar is used to illuminate natural scenes, there are a lot of clutter in the reflected waves. Harmonic radar uses the harmonic radiation characteristics of nonlinear targets for imaging and detection, effectively suppressing clutter and interference, and discovering stealth/hidden targets. Harmonic radar has the advantage of being able to filter linear scatterers in nature, and can effectively filter out the clutter generated by linear targets. Since the devices used in methods such as inflated false targets, strong interference, decoys, and anti-stealth do not have harmonic reflection characteristics, the above methods are only effective for traditional baseband radars and have no effect on harmonic radars. Harmonic radar can effectively filter out the above interference or camouflage. In traditional SAR images, the targets of interest are mixed with scatterers in the environment and are difficult to distinguish. Compared with synthetic aperture radars working at baseband, harmonic synthetic aperture radar (Harmonic Synthetic Aperture Radar: H-SAR) can effectively filter out interference from baseband, and H-SAR images will not be filled with various clutter scatterers like traditional SAR images. Since conventional natural scatterers do not have harmonic scattering characteristics, H-SAR images will be much cleaner and a large part of scatterers in nature can be automatically filtered out, highlighting the scatterers with harmonic characteristics of interest.

Although harmonic radar has strong advantages in anti-stealth, anti-false target and anti-interference in the field of target detection and imaging, it has a difficulty that is difficult to overcome, that is, the harmonic scattering energy of nonlinear targets is low, and the harmonic signal-to-noise ratio received by the receiver is low, resulting in the inability to accurately estimate the residual motion error by directly using harmonic data, which ultimately leads to the defocusing of H-SAR images. The above defects seriously affect the application of H-SAR. In practice, when nonlinear targets receive radar signal radiation, they reflect both the fundamental frequency signal and the harmonic signal, and the receiver can receive both the fundamental frequency signal and the harmonic signal at the same time. Since the motion error is generated by the platform movement, it has nothing to do with whether it works at the fundamental frequency or the harmonic. The harmonic signal-to-noise ratio is low, and it is difficult to directly use the harmonic signal to accurately estimate the residual motion error curve. Compared with the harmonic signal, the fundamental frequency signal has a higher signal-to-noise ratio, and a relatively accurate residual motion error curve can be extracted.

Summary of the invention

The present invention utilizes the baseband signal to estimate the residual motion error to compensate the harmonic signal, thereby improving the imaging performance of the harmonic signal.

The technical solution of the present invention is a method for compensating residual motion error of a harmonic synthetic aperture radar, the method comprising the following steps:

Step 1, constructing a harmonic synthetic aperture radar echo model; the harmonic synthetic aperture radar transmits a baseband signal, and after irradiating a nonlinear target, the echo signal reflected by the nonlinear target contains both the baseband signal and the harmonic signal. The echo signal received by the receiver contains N harmonic signals, of which the 1st harmonic signal is called the baseband signal;

Step 2, separation of echo signal data: Separating the received echo signal data by bandpass frequency domain filtering to obtain independent 1st to Nth harmonic echo signal data;

Step 3, coarse focusing of the SAR image; coarse focusing imaging is performed on the 1st to Nth harmonic signals, the motion trajectory error is fitted by combining the inertial navigation data and the radar parameters to generate motion compensation data, and SAR imaging processing is performed on the 1st to Nth harmonic signal data respectively to obtain a coarse focusing SAR image;

Step 4, residual error phase estimation: extracting the residual motion error phase, using the baseband signal SAR image to extract the error phase generated by the residual motion error;

Step 5, error phase compensation; the residual motion phase error coefficients of the 2nd to Nth harmonic signals are generated by the error phase generated by the residual motion error estimated by the baseband signal SAR image; the coarsely focused SAR image is Fourier transformed in azimuth, the coarsely focused SAR image is converted into the range-Doppler domain, and the range-Doppler data of the 2nd to Nth harmonic coarse focused SAR image is compensated with the generated residual motion error coefficients, and then the compensated data is inversely Fourier transformed in azimuth to obtain finely focused images of the 2nd to Nth harmonic signals. At this point, finely focused images of the 1st to Nth harmonics are all obtained.

Further, the specific implementation method of step 1 is that the radar transmits a baseband signal, and the transmitted baseband signal The waveform is represented as:

(1)

in, Indicates fast time, /> represents the distance dimension envelope,/> Indicates the center frequency of the transmitted signal, /> Indicates the modulation frequency of the transmitted baseband signal, /> represents an exponential function with a natural constant as the base, and j represents a complex number;

The baseband signal emitted by the radar hits the nonlinear target. The echo reflected by the nonlinear target contains both the baseband signal and the harmonic signal. The echo signal received by the receiver Expressed as:

(2)

in, Indicates slow time, /> Represents two-dimensional data, which forms a two-dimensional data matrix after digital collection. represents the slow time dimension signal envelope,/> Indicates the order of harmonics, /> represents the propagation speed of electromagnetic signals, Indicates the instantaneous slant range between the target and the radar;

Assume that the position coordinates of the nonlinear target in the range-azimuth plane are , based on the geometric relationship between the radar platform and the nonlinear target, the instantaneous slant range/> Expressed as:

(3)

in, Indicates the radar movement speed, /> represents the linear motion term coefficient, /> Represents the distance bending term coefficient.

Further, the specific implementation method of step 2 is that the echo signal received by the receiver contains both the fundamental frequency signal and the harmonic signal, each harmonic signal is mixed together in the time domain, each harmonic signal is separated from the collected echo signal before imaging processing, and each harmonic signal is imaged separately;

The harmonic signal components are separated by frequency domain filtering. First, the echo signal is Fourier transformed in the fast time dimension. At this time, each harmonic is separated in the spectrum. Bandpass filtering is used to extract each harmonic signal. The baseband signal of the nth harmonic is obtained after distance Fourier transform, bandpass filtering, distance inverse Fourier transform, and center frequency shifting. Expressed as:

(4)

Furthermore, the specific implementation method of step 3 is to combine the inertial navigation data with the radar parameters to fit the motion trajectory error and generate motion compensation data. Due to the limited accuracy of the inertial navigation data, a residual error component is introduced. The instantaneous slant range after inertial navigation data compensation is Expressed as:

(5)

in, is the residual motion error after motion compensation, /> is the instantaneous slant distance after motion compensation. At this time, /> By instantaneous slant distance/> and residual motion error/> It consists of two parts;

Residual motion error after inertial navigation data compensation Less than one range gate sampling length, that is, after range pulse pressure, motion compensation, range movement correction, and range curvature correction processing, the target energy can be concentrated within one range gate, and only the influence of residual motion error on azimuth focusing is considered;

The two-dimensional time domain data of the nth harmonic signal obtained after range pulse pressure, motion compensation, range movement correction, and range curvature correction It is expressed as:

(6)

in, is the bandwidth of the baseband signal, /> is the symplectic function;

At this time, the azimuth matching filter is performed on formula (6), and the obtained two-dimensional image must be defocused in the azimuth direction. The azimuth defocused image obtained by the nth harmonic signal after azimuth matching filtering is: .

Furthermore, the specific implementation method of step 4 is to perform phase gradient autofocusing algorithm processing on the imaging result of the baseband echo signal to estimate the residual phase error;

Perform azimuth Fourier transform on the 1st to Nth images to obtain range-Doppler data expressed as , where /> is the Doppler frequency;

The baseband echo SAR image is used as the data to be processed, and the estimated value of the error phase gradient is obtained by using the weighted maximum likelihood criterion. for:

(7)

Where M is the number of selected range gate samples, Represents the range gate weighting coefficient, which is determined by the total energy of the range gate./> To obtain the complex phase function, /> To take the complex conjugate, /> and Indicates two adjacent Doppler frequencies in the range-Doppler domain;

Integrate the error phase gradient to obtain the error phase for:

(8).

Furthermore, the specific implementation method of step 5 is to use the error phase of the fundamental frequency echo SAR image estimated in step 4 to construct the error compensation coefficients of the 1st to Nth harmonic images, and the nth error compensation coefficient Expressed as:

(9)

Estimated error compensation coefficient With range-Doppler data/> Multiply and perform inverse Fourier transform in azimuth to obtain the image after self-focusing and the nth harmonic image after fine focusing/> Expressed as:

(10)

in, Indicates that inverse Fourier transform is performed along the azimuth direction; at this point, the finely focused images of the 1st to Nth harmonic signals are all obtained.

The beneficial effects of the present invention include:

The present invention uses residual phase error estimated by fundamental frequency image data to compensate harmonic SAR images, which fully utilizes the advantage of high signal-to-noise ratio of fundamental frequency images. Compared with directly using harmonic images to estimate phase error curves, it has higher estimation accuracy and can effectively improve the focusing performance of harmonic images.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a flow chart of a method for compensating residual motion errors of a harmonic synthetic aperture radar according to the present invention.

Detailed ways

The invention discloses a harmonic synthetic aperture radar residual motion error compensation method. The present invention adopts residual motion error estimated by fundamental frequency data to compensate harmonic SAR image. The process of the present invention is shown in FIG1. Step 1: harmonic synthetic aperture radar transmits fundamental frequency signal, and after irradiating nonlinear target, the echo signal reflected by nonlinear target contains fundamental frequency signal and harmonic signal at the same time. The echo signal received by the receiver contains N harmonics, wherein the 1st harmonic is called fundamental frequency; Step 2: echo data separation, the received echo data is separated by bandpass frequency domain filtering to obtain independent 1st to Nth harmonic echo data; Step 3: coarse focusing imaging is performed on the 1st to Nth harmonics, the motion trajectory error is fitted by combining inertial navigation data and radar parameters to generate motion compensation data, and SAR imaging processing is performed on the 1st to Nth harmonic data respectively. Due to the limited accuracy of inertial navigation data, the image obtained at this time is a coarse focusing image, the image focusing effect is poor, and the resolution and image effect of the image are difficult to meet the actual requirements; Step 4: extracting the residual motion error phase, using the higher signal-to-noise ratio characteristic of the fundamental frequency signal, and using the fundamental frequency SAR image to extract the error phase generated by the residual motion error. The baseband signal SAR image is segmented along the range-azimuth direction, and the baseband SAR image is segmented into several overlapping data blocks. The error phase curve of each data block is extracted by the self-focusing algorithm, and the obtained error curve is fitted to obtain the final residual phase error curve, thereby extracting the residual motion error curve; step 5, the residual motion error curve obtained by combining the baseband SAR image estimation and the harmonic parameters are used to generate the residual motion phase error coefficients of the second to Nth harmonics. The coarse focus image is Fourier transformed in the azimuth direction, and the coarse focus image is converted to the range-Doppler domain. The range-Doppler data of the coarse focus image of the second to Nth harmonics is compensated with the generated residual motion error coefficients, and then the compensated data is inversely Fourier transformed in the azimuth direction to obtain the fine focus images of the second to Nth harmonics. At this point, the fine focus images of the first to Nth harmonics are all obtained.

The specific technical solution steps are as follows:

Step 1. Harmonic radar echo model construction;

The radar transmits a linear frequency modulation signal, the transmission signal The waveform can be expressed as:

(1)

in, Indicates fast time, /> represents the distance dimension envelope,/> Indicates the center frequency of the transmitted signal, /> Indicates the modulation frequency of the transmitted signal, /> represents an exponential function with a natural constant as the base, and j represents a complex number;

The radar transmits a baseband signal, which is then irradiated onto a nonlinear target. The echo reflected by the nonlinear target contains both the baseband signal and the harmonic signal. The echo signal received by the receiver is It can be expressed as:

(2)

in, Indicates slow time, /> Represents two-dimensional data, which forms a two-dimensional data matrix after digital collection. represents the slow time dimension signal envelope,/> Indicates the order of harmonics, /> represents the propagation speed of electromagnetic signals, Indicates the instantaneous slant range between the target and the radar.

Assume that the position coordinates of the target in the range-azimuth plane are , based on the geometric relationship between the radar platform and the target, the instantaneous slant range/> It can be expressed as:

(3)

in, Indicates the radar movement speed, /> represents the linear motion term coefficient, /> Represents the distance bending term coefficient.

Step 2. Echo data separation;

The echo received by the receiver contains both the fundamental frequency signal and the harmonic signal. The harmonics are mixed together in the time domain, and the echo signal cannot be imaged directly. Therefore, before imaging, it is necessary to separate the harmonics from the collected echo signal and perform imaging on each harmonic separately.

The present invention uses frequency domain filtering to separate harmonic signal components. First, the collected signal is Fourier transformed in the fast time dimension. At this time, each harmonic is separated in the spectrum, and each harmonic can be extracted by bandpass filtering. The baseband signal of the nth harmonic obtained after distance Fourier transform, bandpass filtering, distance inverse Fourier transform, and center frequency shifting It can be expressed as:

(4)

Step 3. Coarse focusing of SAR image;

Combine the inertial navigation data with the radar parameters to fit the motion trajectory error and generate motion compensation data. Due to the limited accuracy of the inertial navigation data, a residual error component is introduced. The instantaneous slant distance after inertial navigation data compensation It can be expressed as:

(5)

in, is the residual motion error after motion compensation, /> is the instantaneous slant distance after motion compensation. At this time, /> By the motion-free slope distance/> and residual motion error/> It consists of two parts.

Assume that the residual motion error after inertial navigation data compensation is It is less than one range gate sampling length, that is, after range pulse pressure, motion compensation, range movement correction, and range curvature correction processing, the target energy can be concentrated within one range gate, and only the influence of residual motion error on azimuth focusing is considered.

The two-dimensional time domain data of the nth harmonic obtained after range pulse pressure, motion compensation, range movement correction, and range curvature correction It can be expressed as:

(6)

in, is the bandwidth of the baseband signal, /> is the Symgine function.

At this time, the azimuth matching filter is performed on formula (6), and the obtained two-dimensional image must be defocused in the azimuth direction. The azimuth defocused image obtained by azimuth matching filtering of the nth harmonic is: .

Step 4. Residual error phase estimation;

Since the second to Nth harmonics are weaker than the fundamental frequency echo energy, the present invention implements a phase gradient autofocusing algorithm to process the imaging result of the fundamental frequency echo signal to estimate the residual phase error.

Perform azimuth Fourier transform on the 1st to Nth images to obtain range-Doppler data expressed as , where /> is the Doppler frequency.

The baseband echo SAR image is used as the data to be processed, and the estimated value of the error phase gradient is obtained by using the weighted maximum likelihood criterion. for:

(7)

Where M is the number of selected range gate samples, Represents the range gate weighting coefficient, which is determined by the total energy of the range gate./> To obtain the complex phase function, /> To take the complex conjugate, /> and Represents two adjacent Doppler frequency points in the range-Doppler domain.

By integrating the error phase gradient, we can obtain the error phase for:

(8)

In practical applications, since the motion residual error is spatially variable, it is necessary to divide the image after coarse focusing into two-dimensional blocks in the range and azimuth directions, estimate the error phase curve of each image data block, and use the estimated error phase curve to compensate each image data block separately.

Step 5. Error phase compensation;

The error phase of the fundamental frequency image estimated in step 4 is used to construct the error compensation coefficients of the 1st to Nth harmonic images. The nth error compensation coefficient It can be expressed as:

(9)

Estimated error compensation coefficient With range-Doppler data/> Multiply and perform inverse Fourier transform in azimuth to obtain the image after self-focusing and the nth harmonic image after fine focusing./> It can be expressed as:

(10)

in, Indicates that the inverse Fourier transform is performed along the azimuth direction. At this point, the fine-focus images of the 1st to Nth harmonics are all obtained.

It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims (6)

1. A method for compensating residual motion error of harmonic synthetic aperture radar, characterized in that the method comprises the following steps: Step 1, constructing a harmonic synthetic aperture radar echo model; the harmonic synthetic aperture radar transmits a baseband signal, and after irradiating a nonlinear target, the echo signal reflected by the nonlinear target contains both the baseband signal and the harmonic signal. The echo signal received by the receiver contains N harmonic signals, of which the 1st harmonic signal is called the baseband signal; Step 2, separation of echo signal data: Separating the received echo signal data by bandpass frequency domain filtering to obtain independent 1st to Nth harmonic signal data; Step 3, coarse focusing of the SAR image; coarse focusing imaging is performed on the 1st to Nth harmonic signals, the motion trajectory error is fitted by combining the inertial navigation data and the radar parameters to generate motion compensation data, and SAR imaging processing is performed on the 1st to Nth harmonic signal data respectively to obtain a coarse focusing SAR image; Step 4, residual error phase estimation: extracting the residual motion error phase, using the baseband signal SAR image to extract the error phase generated by the residual motion error; Step 5, error phase compensation; the residual motion phase error coefficients of the 2nd to Nth harmonic signals are generated by the error phase generated by the residual motion error estimated by the baseband signal SAR image; the coarsely focused SAR image is Fourier transformed in azimuth, the coarsely focused SAR image is converted into the range-Doppler domain, and the range-Doppler data of the 2nd to Nth harmonic coarse focused SAR image is compensated with the generated residual motion error coefficients, and then the compensated data is inversely Fourier transformed in azimuth to obtain finely focused images of the 2nd to Nth harmonic signals. At this point, finely focused images of the 1st to Nth harmonics are all obtained. 2. The method according to claim 1 is characterized in that the specific implementation method of step 1 is that the harmonic synthetic aperture radar transmits a baseband signal, and the transmitted baseband signal The waveform is represented as: (1) in, Indicates fast time, /> represents the distance dimension envelope,/> Indicates the center frequency of the transmitted signal, /> Indicates the modulation frequency of the transmitted baseband signal, /> represents an exponential function with a natural constant as the base, and j represents a complex number; The fundamental frequency signal emitted by the harmonic synthetic aperture radar hits the nonlinear target. The echo signal reflected by the nonlinear target contains both the fundamental frequency signal and the harmonic signal. The echo signal received by the receiver Expressed as: (2) in, Indicates slow time, /> Represents two-dimensional data, which forms a two-dimensional data matrix after digital collection, /> represents the slow time dimension signal envelope,/> Indicates the order of harmonics, /> Represents the propagation speed of electromagnetic signals, /> Indicates the instantaneous slant range between the target and the radar; Assume that the position coordinates of the nonlinear target in the range-azimuth plane are , based on the geometric relationship between the radar and the nonlinear target, the instantaneous slant range/> Expressed as: (3) in, Indicates the radar movement speed, /> represents the linear motion term coefficient, /> Represents the distance bending term coefficient. 3. The method according to claim 2 is characterized in that the specific implementation method of step 2 is that the echo signal received by the receiver contains both the fundamental frequency signal and the harmonic signal, the harmonic signals of each order are mixed together in the time domain, and the harmonic signals of each order are separated from the collected echo signal before imaging processing, and imaging processing is performed on each harmonic signal respectively; The harmonic signal components are separated by frequency domain filtering. First, the echo signal is Fourier transformed in the fast time dimension. At this time, each harmonic is separated in the spectrum. Bandpass filtering is used to extract each harmonic signal. The baseband signal of the nth harmonic is obtained after distance Fourier transform, bandpass filtering, distance inverse Fourier transform, and center frequency shifting. Expressed as: (4). 4. The method according to claim 3 is characterized in that the specific implementation method of step 3 is to combine the inertial navigation data with the radar parameters to fit the motion trajectory error to generate motion compensation data. Due to the limited accuracy of the inertial navigation data, a residual error component is introduced to obtain the instantaneous slant range after the inertial navigation data compensation. Expressed as: (5) in, is the residual motion error after motion compensation, /> is the instantaneous slant distance after motion compensation. At this time, /> By instantaneous slant distance/> and residual motion error/> It consists of two parts; Residual motion error after inertial navigation data compensation Less than one range gate sampling length, that is, after range pulse pressure, motion compensation, range movement correction, and range curvature correction processing, the target energy can be concentrated within one range gate, and only the influence of residual motion error on azimuth focusing is considered; The two-dimensional time domain data of the nth harmonic signal obtained after range pulse pressure, motion compensation, range movement correction, and range curvature correction It is expressed as: (6) in, is the bandwidth of the baseband signal, /> is the symplectic function; At this time, the azimuth matching filter is performed on formula (6), and the obtained two-dimensional image must be defocused in the azimuth direction. The azimuth defocused image obtained by the nth harmonic signal after azimuth matching filtering is: , that is, the coarsely focused SAR image of the 1st to Nth harmonic signals. 5. The method according to claim 4, characterized in that the specific implementation method of step 4 is to perform phase gradient autofocusing algorithm processing on the imaging result of the baseband echo signal to estimate the residual phase error; The coarsely focused SAR images of the 1st to Nth harmonic signals are transformed in azimuth direction to obtain the range-Doppler data expressed as , where /> is the Doppler frequency; The coarsely focused SAR image of the baseband signal is used as the data to be processed, and the estimated value of the error phase gradient is obtained using the weighted maximum likelihood criterion. for: (7) Where M is the number of selected range gate samples, Represents the range gate weighting coefficient, which is determined by the total energy of the range gate./> To obtain the complex phase function, /> To take the complex conjugate, /> With/> Indicates two adjacent Doppler frequencies in the range-Doppler domain; Estimate of the error phase gradient Integrate to obtain the residual error phase for: (8). 6. The method according to claim 5 is characterized in that the specific implementation method of step 5 is to construct the error compensation coefficients of the SAR images of the first to Nth harmonic signals using the error phase of the baseband signal SAR image estimated in step 4, and the nth error compensation coefficient Expressed as: (9) Estimated error compensation coefficient With range-Doppler data/> Multiply and perform inverse Fourier transform in azimuth to obtain the image after self-focusing and the nth harmonic image after fine focusing/> Expressed as: (10) in, Indicates that inverse Fourier transform is performed along the azimuth direction; at this point, the finely focused images of the 1st to Nth harmonic signals are all obtained.