Synthetic aperture radar doppler parameter estimation method based on inertial navigation data and image feedback

Abstract

This invention discloses a synthetic aperture radar (SAR) Doppler parameter estimation method based on inertial navigation (INS) data and image feedback, comprising: S1: performing initial coarse estimation of radar slant angle, Doppler center frequency, and ambiguity number using INS data; S2: performing coarse focusing imaging of the echo based on the coarsely estimated parameters to generate an initial SAR image; S3: estimating the baseband Doppler center frequency from the coarsely focused image using the energy centroid method; S4: combining the baseband frequency and the coarsely estimated ambiguity number to calculate the accurate overall Doppler center frequency and inversely derive the accurate slant angle. This invention, through an iterative feedback mechanism of coarse estimation-coarse imaging-fine estimation, combines the geometric advantages of INS data with image domain data, effectively improving the accuracy and reliability of Doppler parameter estimation. Furthermore, it does not rely on the assumption of a uniform scene, exhibits robust algorithm performance, and has strong engineering practicality, laying a crucial foundation for subsequent fine focusing processing of SAR images.

Description

Technical Field

[0001] This invention relates to the field of radar signal processing technology, and specifically to a method for estimating Doppler parameters in synthetic aperture radar imaging. Background Technology

[0002] Synthetic Aperture Radar (SAR) can provide high-quality SAR imagery in all weather and at all times, and is widely used in military and civilian fields. The Doppler center frequency is a core parameter in SAR signal processing; it reflects the geometric relationship between the radar beam center pointing and the platform's direction of motion, and its accuracy directly affects the azimuth positioning and focusing quality of the image. Furthermore, incorrect estimation of the Doppler center frequency ambiguity number can cause severe azimuth ambiguity.

[0003] Existing methods for estimating the Doppler center frequency can be mainly divided into three categories: First, geometric calculation methods based on high-precision inertial navigation systems (INS), which directly utilize platform speed and oblique angle to solve the problem, are costly and limited by the errors of the inertial navigation system itself; second, correlation methods or energy equalization methods based on echo data, which perform well in uniform scenes, but the estimation accuracy decreases in complex scenes or under low signal-to-noise ratio conditions; and third, estimation through parameter search, which involves large computational load and is prone to getting trapped in local optima.

[0004] In summary, existing methods have limitations in terms of estimation accuracy, computational efficiency, and scene adaptability, especially lacking effective means to use imaging results as feedback to correct parameters. Therefore, how to estimate Doppler parameters in a low-cost and high-accuracy manner and solve the problem of ambiguity number determination is a pressing technical challenge in the field of SAR imaging. Summary of the Invention

[0005] To address the issues of insufficient accuracy, computational complexity, or reliance on scene features in traditional Doppler parameter estimation methods, this invention proposes a synthetic aperture radar (SAR) Doppler parameter estimation method based on inertial navigation (INS) data and image feedback. This method combines the geometric computational advantages of INS data with the data advantages of the image domain. Through an iterative feedback mechanism of coarse estimation-coarse imaging-fine estimation, it effectively improves the accuracy and reliability of Doppler parameter estimation. Furthermore, it does not rely on the assumption of a uniform scene, exhibits robustness, and demonstrates strong engineering practicality, laying a crucial foundation for subsequent SAR image fine focusing processing.

[0006] This invention discloses a method for estimating Doppler parameters of synthetic aperture radar based on inertial navigation data and image feedback, comprising the following steps:

[0007] S1: Receive radar echo data and corresponding inertial navigation data (INS data), and use the INS data to make an initial coarse estimate of the radar's slant angle, Doppler center frequency, and ambiguity number.

[0008] S2: Based on the coarsely estimated parameters, coarse focusing imaging processing is performed on the radar echo to generate an initial synthetic aperture radar image.

[0009] S3: Extract the baseband Doppler center frequency directly from the coarse focused image or estimate it using the energy centroid method;

[0010] S4: Combining the baseband Doppler center frequency and the initial coarse estimate of the ambiguity number, the accurate overall Doppler center frequency is calculated, and the accurate oblique angle is further derived.

[0011] Preferably, in step S1, it is assumed that the transmitted signal emitted by the SAR radar is a linear frequency modulated (LFM) signal:

[0012]

[0013] In the formula, It is the waveform used in the transmitted signal emitted by the radar. Represents the natural exponential function. It is the imaginary unit. It is the width of the transmitted signal pulse emitted by the radar. It is a fast time variable of distance, representing the time history of the electromagnetic wave's round-trip propagation during the transmission-to-reception cycle of a single pulse. It is the frequency modulation slope of the transmitted signal. It is the carrier frequency of the transmitted signal.

[0014] When the transmitted signal encounters the target point and is reflected back to the radar, the two-dimensional echo signal received by the radar can be represented as:

[0015]

[0016] It is the echo signal received by the radar. It is a slow-time variable of azimuth, used to mark the continuous moments when the radar platform transmits and receives each pulse along its flight path. The azimuth aperture time, in its physical essence, is the observation time window corresponding to the formation of an effective synthetic antenna array. This indicates the distance from the radar antenna phase center to the point target. It is the speed of electromagnetic wave propagation, and its value is .

[0017] because The resulting delayed echo phase change is

[0018]

[0019] This phase term is called the Doppler phase. It is the wavelength of the transmitted signal. The first derivative of the Doppler phase with respect to time is defined as the Doppler center frequency. .

[0020]

[0021] In the formula, This is a rough estimate of the Doppler center frequency. Indicates to Conducting research on slow time Differentiating, This is the radar slant angle read from inertial navigation data. A rough estimate of the Doppler center frequency can be obtained from this formula.

[0022] Based on the coarse estimate of the Doppler center frequency and the pulse repetition frequency of the system, a coarse estimate of the Doppler ambiguity number is calculated.

[0023]

[0024] symbol This indicates the integer division operation.

[0025] Preferably, in step S2, based on the coarsely estimated oblique angle and coarsely estimated Doppler center frequency obtained in step S1, motion compensation and preliminary imaging processing are performed on the original echo data to generate a coarsely focused synthetic aperture radar complex image.

[0026] The purpose of this step is to obtain an intermediate image with basic azimuth energy focusing, which can be used for subsequent parameter analysis. Its imaging accuracy requirement is lower than the final fine image, but it must ensure that the degree of image defocus caused by parameter errors in step S2 is within the tolerance of the parameter estimation method in the subsequent step S3.

[0027] Specifically, coarse focusing imaging can be achieved using mature synthetic aperture radar imaging algorithms in this field, such as, but not limited to, range Doppler algorithm, linear frequency modulation scaling algorithm, or back projection algorithm.

[0028] Taking the range-Doppler algorithm as an example, range-matched filtering is performed on the raw echo data to improve range resolution. Based on the coarsely estimated angle of view obtained in step S1, range curvature (RCMC) in the echo data is calculated and compensated. Since the parameters are coarsely estimated, this step allows for the use of lower interpolation accuracy or a simplified model to balance processing speed and correction effect. An azimuth-matched filter is constructed with the coarsely estimated Doppler center frequency as the Doppler center and the Doppler modulation frequency coarsely estimated based on the platform velocity as the parameter. This filter performs azimuth processing on the data to complete focusing and obtain a coarsely focused complex image.

[0029] Preferably, in step S3, in order to estimate the baseband Doppler center frequency, it can be done in the following ways, but not limited to:

[0030] (1) Direct estimation method: Based on the center position of the coarse focused complex image, the baseband Doppler center frequency corresponding to the image is directly estimated;

[0031] (2) Energy centroid method: Calculate the energy centroid of the azimuth spectrum of the coarse focused complex image and take its corresponding frequency as the baseband Doppler center frequency. This method is robust and suitable for distributed scenarios.

[0032] Preferably, in step S4, the specific formula for calculating the precise overall Doppler center frequency is as follows:

[0033]

[0034] In the formula, It is the Doppler center frequency obtained in step S3. To accurately determine the overall Doppler center frequency.

[0035] Combined with known platform speed and radar wavelength The precise radar beam angle can be obtained by inverse solving. It is used for subsequent focusing imaging.

[0036] This invention aims to overcome the shortcomings of existing technologies and provide a highly accurate and robust Doppler parameter estimation method. This method combines the geometric computational advantages of inertial navigation data with image domain data, achieving automatic correction of initial parameters through a feedback loop of coarse estimation followed by coarse imaging and fine estimation.

[0037] The core improvement of this invention lies in:

[0038] 1. A preliminary image is formed using coarse parameters that may contain errors. Although the image is not optimal, it contains information that reflects the true phase and energy.

[0039] 2. Abandoning the traditional approach of relying solely on echo signals or pure geometric models, this paper proposes a method to directly extract the baseband Doppler center frequency from the energy distribution of the coarse-focused image itself.

[0040] 3. The precise baseband Doppler center frequency extracted from the image is combined with the ambiguity number obtained from the inertial navigation data to finally output a higher quality image.

[0041] Beneficial effects: Through the above technical solution, the present invention has achieved a significant improvement in the accuracy of Doppler parameter estimation, reduced the dependence on scene uniformity, enhanced the overall robustness of the system, and the process is clear and easy to implement in engineering. Attached Figure Description

[0042] Figure 1 This is a schematic diagram of the geometric relationship of Doppler parameters based on INS data;

[0043] Figure 2 This is a flowchart illustrating the implementation of a synthetic aperture radar Doppler parameter estimation method based on inertial navigation data and image feedback.

[0044] Figure 3 This is a block diagram of the processing system for implementing the present invention;

[0045] Figure 4 This is a schematic diagram of Doppler blur.

[0046] Figure 5 It is a point target image directly imaged using only inertial navigation data;

[0047] Figure 6 It is an upsampled profile image directly imaged using only inertial navigation data;

[0048] Figure 7 It is an azimuth profile image directly imaged using only inertial navigation data;

[0049] Figure 8 It is a range profile image directly imaged using only inertial navigation data;

[0050] Figure 9 It is a point target image image captured using this method;

[0051] Figure 10 This is an upsampled profile image obtained using this method;

[0052] Figure 11 This is an azimuth profile image obtained using this method;

[0053] Figure 12 This is a range profile image obtained using this method. Detailed Implementation

[0054] The method for estimating Doppler parameters of synthetic aperture radar based on inertial navigation data and image feedback, proposed in this invention, will be described in detail below with reference to the accompanying drawings. Without loss of generality, a schematic diagram of the geometric relationship of Doppler parameters based on INS data (e.g.) is provided here. Figure 1 Taking the example shown below, the specific implementation steps of this method are described as follows: Figure 2 and Figure 3 As shown.

[0055] S1: Assume the transmitted signal from the SAR radar is a linear frequency modulated (LFM) signal:

[0056]

[0057] In the formula, It is the waveform used in the transmitted signal emitted by the radar. Represents the natural exponential function. It is the imaginary unit. It is the width of the transmitted signal pulse emitted by the radar. It is a fast time variable of distance, representing the time history of the electromagnetic wave's round-trip propagation during the transmission-to-reception cycle of a single pulse. It is the frequency modulation slope of the transmitted signal. It is the carrier frequency of the transmitted signal.

[0058] When the transmitted signal encounters the target point and is reflected back to the radar, the two-dimensional echo signal received by the radar can be represented as:

[0059]

[0060] It is the echo signal received by the radar. t is a slow-time variable for azimuth, used to mark the consecutive moments when the radar platform transmits and receives each pulse along its flight path. The azimuth aperture time, in its physical essence, is the observation time window corresponding to the formation of an effective synthetic antenna array. This indicates the distance from the radar antenna phase center to the point target. It is the speed of electromagnetic wave propagation, and its value is .

[0061] because The resulting delayed echo phase change is

[0062]

[0063] This phase term is called the Doppler phase. It is the wavelength of the transmitted signal. The first derivative of the Doppler phase with respect to time is defined as the Doppler center frequency. .

[0064]

[0065] In the formula, This is a rough estimate of the Doppler center frequency. Indicates to Conducting research on slow time Differentiating, This is the radar slant angle read from inertial navigation data. The Doppler center frequency can be roughly estimated using this formula.

[0066] Based on the coarse estimate of the Doppler center frequency and the pulse repetition frequency of the system, a coarse estimate of the Doppler ambiguity number is calculated.

[0067]

[0068] symbol This indicates the integer division operation.

[0069] S2: Based on the oblique angle, baseband Doppler center frequency and ambiguity number coarsely estimated in S1, motion compensation and preliminary imaging processing are performed on the original echo data to generate a coarsely focused synthetic aperture radar complex image.

[0070] Specifically, coarse focusing imaging can be achieved using mature synthetic aperture radar imaging algorithms in this field, such as, but not limited to, simplified versions of range Doppler algorithms, linear frequency modulation scaling algorithms, or back projection algorithms.

[0071] Taking the distance-Doppler algorithm as an example, the specific steps are as follows:

[0072] Perform a range-to-fast Fourier transform (FFT) on each pulse of the original echo data to convert it to the range-frequency domain-azimuth-time domain. Multiplying by the frequency domain response of the distance-directed matched filter completes the matched filtering operation.

[0073]

[0074] After matched filtering, perform range-directed inverse fast Fourier transform (IFFT) and azimuth-directed fast Fourier transform to the range-Doppler domain. Then, distance migration correction (RCMC) is performed. Since the parameters are coarsely estimated, this step allows for the use of lower interpolation accuracy or a simplified model to balance processing speed with correction effectiveness.

[0075] When the aircraft travels at speed When flying in the positive azimuth direction, the instantaneous distance between the radar and the target can be expressed by the following formula:

[0076]

[0077] in, This represents the shortest instantaneous distance between the radar and the target, after transforming the data to the range-Doppler domain. hour, The corresponding Fourier transform is , This refers to range migration (RCM) in the range-Doppler domain. The essence of RCM is to measure the range migration of different point targets in the range-Doppler domain as a function of azimuth frequency. Corrected to closest distance .

[0078] Distance-Doppler domain after RCMC completion In the middle, the data is multiplied by an azimuth matched filter to perform azimuth processing, thereby completing the focusing and obtaining a coarsely focused complex image.

[0079] S3: To estimate the baseband Doppler center frequency, the following methods can be used, but are not limited to:

[0080] (1) Direct estimation method: Based on the center position of the coarse focused complex image, the baseband Doppler center frequency corresponding to the image is directly estimated;

[0081] (2) Energy centroid method: Calculate the energy centroid of the azimuth spectrum of the coarse-focused complex image, and use its corresponding frequency as the baseband Doppler center frequency. This method is robust and suitable for distributed scenarios. The specific steps are as follows:

[0082] After range compression of the echo data, a Fast Fourier Transform (FFT) is performed on the azimuth time series (i.e., the slow-time-dimensional signal) within each range gate to transform it into the Doppler domain. Then, for each range gate, the modulus square of its azimuth spectrum is calculated to obtain the power spectral density of that range gate. :

[0083]

[0084] in For Doppler frequency, This is the distance gate index.

[0085] To suppress noise and improve the robustness of the estimation, the power spectra of all range gates within the selected data processing block are superimposed and averaged to obtain a high signal-to-noise ratio average Doppler power spectrum. :

[0086]

[0087] in, The average number of distance gates participating.

[0088] Through calculation The first moment can be used to obtain the centroid, i.e., the estimated baseband Doppler center. :

[0089]

[0090] S4: Doppler center frequency obtained in step S3 And the fuzzy number roughly estimated in step S1 The specific formula for the precise overall Doppler center frequency is as follows:

[0091]

[0092] In the formula, To accurately determine the overall Doppler center frequency. Figure 4 A schematic diagram of Doppler blur is shown.

[0093] Combined with known platform speed and radar wavelength The precise radar beam angle can be obtained by inverse solving. The formula is as follows:

[0094]

[0095] Output high-precision Doppler parameters and Used for subsequent focusing imaging.

[0096] To verify the effectiveness of this invention, a point target simulation experiment was conducted. The radar platform was set to move at a constant speed along a straight line, with a true platform speed of 110 m / s and a true oblique angle of 30°. Based on geometric relationships, the true Doppler center frequency was calculated to be 3667 Hz, and the ambiguity number was 4.

[0097] In the simulation, an error was introduced into the oblique angle used in step S1 to simulate the error in the acquired inertial navigation data. Specifically, the initial oblique angle setting had a deviation of 3°, resulting in 33°. Therefore, the estimated Doppler center frequency in step S1 was 3994Hz, and the ambiguity number was 4. Other simulation parameters are shown in Table 1, and the coarse imaging results are as follows: Figure 5 As shown, the upsampling profile, azimuth profile, and range profile are respectively as follows: Figure 6 , Figure 7 and Figure 8 As shown.

[0098] Table 1 Experimental parameters

[0099]

[0100] Using the INS data containing errors mentioned above, perform the following two operations respectively:

[0101] 1. The obtained coarsely estimated oblique angle of 33° is directly used for imaging processing.

[0102] 2. Perform the entire process described in claim 1: First, use the coarsely estimated parameters to perform coarse focusing imaging, then estimate the baseband Doppler center frequency as -327.8Hz from the coarsely focused image, and finally combine the coarsely estimated ambiguity number to synthesize the accurate Doppler center frequency as 3672.2Hz and the accurate oblique angle as 30.05°.

[0103] The parameter estimates output by the two methods are compared with the assumed true values, and the error comparison is shown in the table below:

[0104] Table 2 Comparison of Parameter Estimation Errors

[0105]

[0106] Figures 5-8 and Figures 9-12 The imaging results of point targets obtained using inertial navigation data and those obtained using the precise parameters corrected by this method are presented in comparison. These include point target images, upsampled profile maps, and their azimuth and range profiles. The comparison shows that when imaging is performed directly using inertial navigation data, the azimuth defocus is severe, resulting in poor focusing quality. However, after correction using the method of this invention, the point target is significantly focused in the azimuth direction, the imaging quality is significantly improved, and the estimation error accuracy is greatly enhanced.

[0107] The simulation results above demonstrate that the synthetic aperture radar Doppler parameter estimation method based on inertial navigation data and image feedback proposed in this invention can effectively sense and compensate for measurement deviations in inertial navigation data, significantly improve the estimation accuracy of key imaging parameters such as the Doppler center frequency, and thus obtain SAR images with better focus and higher quality. This method provides an effective technical approach to improve SAR imaging quality.

Claims

1. A method for estimating Doppler parameters of synthetic aperture radar based on inertial navigation data and image feedback, characterized in that, Includes the following steps: S1: Acquire radar echo data and corresponding inertial navigation data. Based on the platform velocity, oblique angle and radar system parameters in the inertial navigation data, perform an initial coarse estimate of the Doppler center frequency and its corresponding ambiguity number to obtain the coarse estimate of the Doppler center frequency and the coarse estimate of the ambiguity number. S2: Based on the coarsely estimated Doppler center frequency and coarsely estimated ambiguity number, coarse focusing imaging processing is performed on the radar echo data to generate a coarsely focused synthetic aperture radar complex image with preliminary azimuth focusing. S3: Perform azimuth spectrum analysis on the coarse focused complex image, and extract its baseband Doppler center frequency using the energy centroid method to obtain a finely estimated baseband Doppler center frequency; S4: Combine the precisely estimated baseband Doppler center frequency with the coarsely estimated ambiguity number in step S1 to calculate the accurate overall Doppler center frequency, and deduce the accurate radar beam angle based on the accurate overall Doppler center frequency. Output the accurate Doppler parameters for subsequent fine focusing imaging.

2. The synthetic aperture radar Doppler parameter estimation method based on inertial navigation data and image feedback according to claim 1, characterized in that, The specific implementation method of the initial coarse estimation in step S1 is as follows: based on the radar platform velocity provided by the inertial navigation data... and oblique angle Combined with radar wavelength Calculate a rough estimate of the Doppler center frequency. : ； Based on the system pulse repetition frequency Calculate the rough estimate of the fuzzy number : ； in, This indicates the floor function.

3. The synthetic aperture radar Doppler parameter estimation method based on inertial navigation data and image feedback according to claim 1, characterized in that, The coarse focusing imaging processing in step S2 uses one of the range Doppler algorithm, linear frequency modulation scaling algorithm, or back projection algorithm. It only needs to ensure that the energy in the azimuth direction of the image is basically focused to meet the requirements of subsequent baseband frequency estimation.

4. The synthetic aperture radar Doppler parameter estimation method based on inertial navigation data and image feedback according to claim 1, characterized in that, The energy centroid method in step S3 includes the following sub-steps: S31: Perform an azimuth-to-Fourier transform on the coarse-focused complex image to obtain Doppler domain data; S32: Calculate the power spectral density of each range gate, and perform incoherent superposition and averaging of the power spectra of all range gates to obtain the average Doppler power spectrum. : ； in, For distance to gates, For the first S33: Calculate the first moment of the average Doppler power spectrum to obtain the baseband Doppler center frequency. : ； in, For Doppler frequency sampling points, This represents the number of sampling points in the azimuth direction.

5. The synthetic aperture radar Doppler parameter estimation method based on inertial navigation data and image feedback according to claim 1, characterized in that, The calculation and inversion method for the precise parameters in step S4 is as follows: the baseband Doppler center frequency obtained in step S3 is used as the reference. The fuzzy number coarsely estimated in step S1 Synthesizing, the accurate global Doppler center frequency is obtained. : ； Based on the precise total Doppler center frequency Platform speed and radar wavelength The inversion yields the accurate oblique angle. : ； The and The precise Doppler parameters are output and used for subsequent synthetic aperture radar fine-focusing imaging processing.

Images