A method for stabilizing the transmit-receive time delay of a sar radar system

Abstract

The present application relates to SAR radar signal processing technical field, specifically to a kind of SAR radar system transmitting time delay stable method, comprising: acquisition radar operating frequency, transmitting pulse repetition frequency, system operating temperature and transmitter power amplifier operating voltage and other operating parameters, and system reference time delay value under nominal operating condition is obtained by table lookup method.The real-time compensation time delay value is obtained by using improved adaptive filter algorithm, combining frequency and pulse repetition frequency variation amount for iterative update, and real-time correction to reference time delay.In the SAR imaging range direction pulse compression stage, according to the time delay value, the radar echo sampling time sequence is adjusted, the compensated pulse compression data is generated, and the imaging processing is completed.The method can quickly track the time delay drift caused by operating condition change, accurately eliminate time delay deviation in key imaging link, effectively suppress the image distortion caused by time delay fluctuation, so that the system can output stable high-resolution SAR radar image under different operating environments.

Description

Technical Field

[0001] This invention relates to the field of SAR radar signal processing technology, and in particular to a method for stabilizing the transmit and receive delays of a SAR radar system. Background Technology

[0002] Existing SAR radar systems often use fixed compensation methods to control transmit and receive delays based on operating parameters such as temperature and high voltage, or employ conventional filtering algorithms for delay correction. The overall compensation logic is not linked to real-time changes in the radar's operating frequency and transmit pulse repetition frequency; both the acquisition and correction of the delay reference rely on static parameter matching. When system operating conditions fluctuate, delay drift cannot be quickly tracked, correction parameters are updated lag-wise, and a continuous deviation exists between the reference delay and the actual system delay.

[0003] Traditional latency compensation methods primarily focus on the data preprocessing stage, neglecting precise time-shift adjustments in the core imaging process of range pulse compression. Residual latency errors are directly carried over into subsequent imaging procedures, affecting pulse compression performance and range resolution. The combined latency disturbances caused by various environmental and operational parameter variations cannot be covered by a single fixed compensation strategy, leading to a continuous decline in transmit and receive latency stability during long-term system operation.

[0004] The existing technology system cannot fully meet the requirements of combining the radar operating frequency and the pulse repetition frequency variation, using an improved adaptive filter to iteratively correct the time delay in real time, and the stable control of dynamic time shift compensation of the sampling time sequence in the range pulse compression stage. It is difficult to guarantee the consistency and high resolution output of SAR imaging under all operating conditions. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a method for stabilizing the transmit and receive delays of a SAR radar system.

[0006] To achieve the above objectives, the present invention employs the following technical solution: a method for stabilizing the transmit and receive delays of a SAR radar system, comprising: Obtain the current operating parameter set of the SAR radar system, which includes the radar operating frequency, transmit pulse repetition frequency, system internal operating temperature, and transmitter power amplifier operating voltage. Based on the internal operating temperature of the system and the operating voltage of the transmitter power amplifier, the system reference delay value is obtained by looking up a table. The system reference delay value corresponds to the initial delay value under the nominal operating condition. Based on the current set of operating parameters, the system reference delay value is corrected in real time using an improved adaptive filter algorithm to generate a real-time compensated delay value. The improved adaptive filter algorithm iteratively updates the parameters based on the changes in the radar operating frequency and the transmit pulse repetition frequency. In the range pulse compression stage of SAR imaging processing, the sampling time sequence of the original radar echo data is time-shifted according to the real-time compensation delay value to generate pulse compressed data after delay compensation. SAR imaging is performed based on the time-delay compensated pulse compression data to generate the final high-resolution radar image.

[0007] As a further aspect of the present invention, based on the internal operating temperature of the system and the operating voltage of the transmitter power amplifier, the system reference delay value is obtained by a lookup table method, including: Establish a preset mapping relationship table between the system's internal operating temperature, the transmitter power amplifier's operating voltage value, and the system's reference time delay value. The preset mapping relationship table is constructed based on the measured system time delay calibration data under multiple combinations of temperature and high voltage values. The system collects real-time readings of the internal operating temperature sensor and the operating voltage monitoring value of the transmitter power amplifier of the SAR radar system. The internal operating temperature sensor readings include the temperature of the transmitter power amplifier heat sink and the temperature of the receiver local oscillator module. Using the transmitter power amplifier heat sink temperature, receiver local oscillator module temperature, and transmitter power amplifier operating voltage monitoring values ​​as joint inputs, interpolation is performed in the preset mapping table to obtain the interpolation system reference delay that is closest to the current operating condition. Temperature drift compensation is performed on the reference time delay of the interpolation system. The temperature drift compensation is calculated based on the historical variation trend of the temperature of the transmitter power amplifier heat sink and the temperature of the receiver local oscillator module. Output the system reference delay value after temperature drift compensation.

[0008] As a further aspect of the present invention, the improved adaptive filter algorithm iteratively updates parameters based on the changes in the radar operating frequency and the transmitted pulse repetition frequency, including: Initialize the filter weight vector and convergence step size factor of the improved adaptive filter algorithm. The filter weight vector is used to characterize the nonlinear mapping relationship between time delay and operating parameters. Within each radar operating pulse repetition cycle, the instantaneous radar operating frequency reading of the current pulse and the instantaneous value of the transmitted pulse repetition frequency are collected, and the frequency difference with the previous cycle is calculated to obtain the change in radar operating frequency and the change in transmitted pulse repetition frequency. The radar operating frequency change, the transmit pulse repetition frequency change, and the system internal operating temperature and transmitter power amplifier operating voltage value in the current operating parameter set are used together to form the adaptive filter input vector at the current moment. The system reference time delay value is used as a reference signal input to the improved adaptive filter algorithm; The predicted value of the real-time compensation delay value is calculated based on the adaptive filter input vector, the reference signal, and the current filter weight vector. The predicted value is compared with the time delay residual value calculated based on the measured echo to generate an error signal. The filter weight vector and convergence step size factor are adjusted using the error signal and a preset iterative update rule to complete one parameter iterative update.

[0009] As a further aspect of the present invention, the predicted value is compared with the time delay residual value calculated based on the measured echo to generate an error signal, including: While the SAR system is performing normal imaging operations, the leakage signal sampling sequence of the radar transmission channel and the raw echo data sampling sequence of the receiving channel are collected in real time. The leakage signal sampling sequence and the original echo data sampling sequence are cross-correlated to calculate the measured transmit and receive delay values. Read the theoretical time delay value under the current working condition without correction by the improved adaptive filter algorithm from the preset mapping table; The difference between the measured transmit and receive delay value and the theoretical delay value is calculated to obtain the delay residual value calculated based on the measured echo. The difference between the predicted value of the real-time compensation delay output by the improved adaptive filter algorithm and the measured value of the transmit / receive delay is calculated, and the difference is used as the error signal for updating the filter parameters.

[0010] As a further aspect of the present invention, cross-correlation processing is performed on the leakage signal sampling sequence and the original echo data sampling sequence to calculate the measured transmit / receive delay value, including: The leakage signal sampling sequence is subjected to bandpass filtering to extract the start edge features of the transmitted pulse, and the position of its feature point is recorded as the transmission time reference point; The original echo data sampling sequence is subjected to pulse compression processing to extract the peak position of the ground object scattered echo, and the peak position is recorded as the echo arrival time reference point; Calculate the position difference between the echo arrival time reference point and the transmission time reference point in the sampling sequence to obtain the difference in the number of sampling points; Based on the sampling rate of the analog-to-digital converter of the SAR radar system, the difference in the number of sampling points is converted into a time difference to obtain the preliminary transmit and receive delay; The preliminary transmit / receive delay calculated over multiple consecutive pulse repetition cycles is subjected to moving average filtering to suppress random noise interference, resulting in the final measured transmit / receive delay value.

[0011] As a further aspect of the present invention, time-shifting adjustment is performed on the sampling time sequence of the original radar echo data based on the real-time compensation delay value, including: Read the current block of raw radar echo data to be processed from the data buffer of the SAR radar system; Extract the original sampling time corresponding to each sampling point in the original radar echo data block to form an original sampling time sequence; Read the latest real-time compensation delay value from the output of the improved adaptive filter algorithm at the current moment; Subtract the real-time compensation delay value from each sampling time in the original sampling time sequence to generate the compensated sampling time sequence; Based on the compensated sampling time sequence, the echo amplitude sequence in the original radar echo data block is rearranged. The rearrangement is achieved by interpolation calculation of the time position of the sampling points to generate an echo data sequence aligned with the time axis.

[0012] As a further aspect of the present invention, the echo amplitude sequence in the original radar echo data block is rearranged based on the compensated sampling time sequence, including: Determine the time interval between adjacent sampling points in the compensated sampling time sequence; Based on the ratio of the time interval to the original sampling time interval, determine whether data interpolation is needed; When the ratio is not equal to one, the sinc interpolation algorithm is used to resample the echo amplitude sequence and calculate the corresponding echo amplitude on the new, uniformly distributed compensated sampling time sequence. When the ratio is equal to one, the echo amplitude sequence is directly shifted as a whole, and the number of shift points is determined by rounding down the quotient of the real-time compensation delay value and the sampling time interval. The echo amplitude sequence, after being resampled or shifted, is output in the order of the compensated sampling time sequence to form a time-axis aligned echo data sequence.

[0013] As a further aspect of the present invention, SAR imaging is performed based on the time-delay-compensated pulse compression data, including: Perform a range-direction Fast Fourier Transform on the time-axis aligned echo data sequence to convert the data to the range frequency domain; In the range frequency domain, the time-axis aligned echo data sequence is multiplied by the frequency domain form of the matched filter of the transmitted signal to complete range pulse compression and generate one-dimensional range image data. Azimuth resampling is performed on one-dimensional range image data of multiple consecutive pulses to correct for Doppler frequency changes caused by platform motion; Azimuth Fast Fourier Transform is performed on the data after azimuth resampling to convert the data to the two-dimensional frequency domain. In the two-dimensional frequency domain, based on the geometric model of SAR imaging, the range migration correction factor and the secondary range compression factor are calculated and applied; Perform a two-dimensional inverse fast Fourier transform on the two-dimensional frequency domain data after all corrections are completed, and convert the data back to the two-dimensional time domain to obtain the final SAR complex image data. The SAR complex image data is moduloed to generate the final high-resolution radar image.

[0014] As a further aspect of the present invention, the azimuth resampling process for one-dimensional range image data of multiple consecutive pulses includes: Extract one-dimensional range profile data from two adjacent pulses, calculate the cross-correlation function of the one-dimensional range profile data from the two adjacent pulses, and estimate the azimuth position offset caused by the change in platform velocity based on the peak position of the cross-correlation function. Based on the azimuth position offset, construct an azimuth resampling interpolation kernel function; Using the azimuth resampling interpolation kernel function, non-uniform interpolation is performed on the one-dimensional range profile data of the current pulse along the azimuth direction to align its azimuth sampling position with the sampling grid of the reference pulse; Iteratively process the one-dimensional range image data of all pulses to ensure that all data have a uniform equivalent sampling interval in the azimuth direction.

[0015] As a further aspect of the present invention, in the two-dimensional frequency domain, based on the geometric model of SAR imaging, the range migration correction factor and the secondary range compression factor are calculated and applied, including: After completing the azimuth-to-fast Fourier transform, a two-dimensional frequency domain echo data matrix is ​​obtained, wherein the dimensions of the two-dimensional frequency domain echo data matrix include the range frequency dimension and the azimuth frequency dimension. Based on the platform motion parameters and imaging geometry of the SAR radar system, a range migration equation is constructed, which describes the coupling relationship between the target slant range history in the range frequency dimension and the azimuth frequency dimension. Based on the range migration equation, the range migration correction factor corresponding to each range frequency unit and azimuth frequency unit is calculated in the two-dimensional frequency domain. From the echo data matrix in the two-dimensional frequency domain, based on the current range frequency cell index and azimuth frequency cell index, read the complex echo value corresponding to the corresponding cell; The complex echo value read is multiplied by the calculated range migration correction factor using a complex conjugate multiplication to eliminate the phase coupling described by the range migration equation, thus completing the range migration correction. After completing the range migration correction, the second-order range compression factor is calculated based on the stationary phase principle for the cross-coupling term of range frequency and azimuth frequency. From the two-dimensional frequency domain data matrix after range migration correction, read the complex echo value corresponding to the corresponding cell based on the current range frequency cell index and azimuth frequency cell index; The complex echo value read is multiplied by the calculated second-order range compression factor using a complex conjugate multiplication to eliminate the phase error caused by the cross-coupling term between the range frequency and the azimuth frequency, thus completing the second-order range compression.

[0016] Compared with the prior art, the advantages and positive effects of the present invention are as follows: Based on the system's internal operating temperature and the transmitter power amplifier's operating voltage, a lookup table method is used to obtain the system's reference delay value. This allows for the rapid identification of the initial delay reference under nominal operating conditions, providing a stable and consistent initial basis for subsequent delay correction and reducing error fluctuations caused by the lack of reference correction. An improved adaptive filter algorithm is employed, combining changes in radar operating frequency and transmit pulse repetition frequency for iterative parameter updates. This enables real-time capture of delay drift caused by operating condition fluctuations, dynamically and accurately correcting the system's reference delay value. This generates a real-time compensated delay value highly adapted to current operating parameters, meeting the delay variation requirements under different operating conditions and mitigating the impact of parameter fluctuations on delay stability.

[0017] In the range-direction pulse compression stage of SAR imaging processing, the sampling time sequence of the original radar echo data is time-shifted based on the real-time compensation delay value. This directly eliminates transmit and receive delay errors in the core imaging stage, generating pulse-compressed data with sufficient delay compensation, thus preventing delay errors from propagating to subsequent imaging processes. SAR imaging based on delay-compensated pulse-compressed data can reduce image distortion caused by delay drift, maintain the high resolution of radar images, and ensure stable imaging performance even under fluctuating operating conditions. Imaging consistency under different operating environments is improved, and the problem of image quality degradation caused by delay fluctuations is effectively alleviated. Attached Figure Description

[0018] Figure 1 This is a state diagram of a method for stabilizing the transmit and receive delays of a SAR radar system according to the present invention. Figure 2 A flowchart for obtaining the system's baseline delay value; Figure 3 This is a flowchart illustrating the process of generating error signals. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0020] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0021] See Figure 1 A method for stabilizing the transmit and receive delays of a SAR radar system is presented, with the following overall implementation scheme: During operation, the SAR radar system acquires its current operating parameter set, which includes the radar operating frequency, transmit pulse repetition frequency, system internal operating temperature, and transmitter power amplifier operating voltage. Based on the system internal operating temperature and transmitter power amplifier operating voltage, a system reference delay value is obtained using a lookup table method. This value corresponds to the initial delay value under nominal operating conditions. Based on all parameters in the current operating parameter set, an improved adaptive filter algorithm is used to correct the aforementioned system reference delay value in real time, thereby generating a real-time compensated delay value. The core feature of this improved adaptive filter algorithm is that its parameter iterative update process is based on the changes in the radar operating frequency and transmit pulse repetition frequency. In the range-direction pulse compression stage of SAR imaging processing, based on the calculated real-time compensated delay value, the sampling time sequence of the original radar echo data is time-shifted to generate time-delay compensated pulse compressed data. Based on this batch of time-delay compensated pulse compressed data, SAR imaging processing is completed to generate the final high-resolution radar image.

[0022] In one embodiment of the present invention, during the process of obtaining the system reference delay value, refer to... Figure 2A pre-defined mapping table is established between the system's internal operating temperature, transmitter power amplifier operating voltage, and system reference delay value. This table is constructed based on system delay calibration data obtained from pre-measured measurements under multiple temperature and high voltage combinations. Real-time acquisition of the SAR radar system's internal operating temperature sensor readings and transmitter power amplifier operating voltage monitoring values ​​is performed. Specifically, the internal operating temperature sensor readings include the transmitter power amplifier heatsink temperature and the receiver local oscillator module temperature. Using the acquired transmitter power amplifier heatsink temperature, receiver local oscillator module temperature, and transmitter power amplifier operating voltage monitoring values ​​as joint inputs, interpolation is performed in the pre-defined mapping table to obtain the interpolated system reference delay closest to the current operating conditions. Temperature drift compensation is then applied to this interpolated system reference delay. The compensation amount is calculated based on the historical variation trends of the transmitter power amplifier heatsink temperature and receiver local oscillator module temperature, and the temperature drift-compensated system reference delay value is output. When applying the improved adaptive filter algorithm, the filter weight vector and convergence step size factor are first initialized. The filter weight vector is used to characterize the nonlinear mapping relationship between delay and operating parameters. Within each radar pulse repetition cycle, the instantaneous radar operating frequency reading of the current pulse and the instantaneous value of the transmitted pulse repetition frequency are acquired, and the frequency difference between them and the previous cycle is calculated to obtain the radar operating frequency change and the transmitted pulse repetition frequency change. These two frequency changes, along with the system internal operating temperature and transmitter power amplifier operating voltage values ​​from the current operating parameter set, constitute the adaptive filter input vector at the current moment. The previously obtained system reference time delay value is used as the reference signal input to the improved adaptive filter algorithm. Based on the current adaptive filter input vector, the reference signal, and the current filter weight vector, the predicted value of the real-time compensation time delay is calculated.

[0023] In practical implementation, a preset mapping relationship table is established between the system's internal operating temperature, the transmitter power amplifier's operating voltage, and the system's reference delay value. This preset mapping relationship table is constructed based on pre-measured system delay calibration data under multiple temperature and high voltage combinations. The system's internal operating temperature sensor readings and transmitter power amplifier operating voltage monitoring values ​​are collected in real time. The internal operating temperature sensor readings include the transmitter power amplifier heatsink temperature and the receiver local oscillator module temperature. Using the transmitter power amplifier heatsink temperature, receiver local oscillator module temperature, and transmitter power amplifier operating voltage monitoring values ​​as joint inputs, interpolation is performed in the preset mapping relationship table to obtain the interpolated system reference delay that is closest to the current operating condition. Temperature drift compensation is performed on the interpolated system reference delay. The temperature drift compensation is calculated based on the historical change trends of the transmitter power amplifier heatsink temperature and the receiver local oscillator module temperature, and the system reference delay value after temperature drift compensation is output. In some embodiments, the filter weight vector and convergence step size factor of the improved adaptive filter algorithm are initialized. The filter weight vector is used to characterize the nonlinear mapping relationship between time delay and operating parameters. Within each radar pulse repetition cycle, the instantaneous radar operating frequency reading of the current pulse and the instantaneous value of the transmit pulse repetition frequency are collected, and the frequency difference with the previous cycle is calculated to obtain the radar operating frequency change and the transmit pulse repetition frequency change. The radar operating frequency change, the transmit pulse repetition frequency change, and the system internal operating temperature and transmitter power amplifier operating voltage values ​​in the current operating parameter set are used together to form the adaptive filter input vector at the current moment. The system reference time delay value is used as the reference signal input to the improved adaptive filter algorithm. Based on the adaptive filter input vector, the reference signal, and the current filter weight vector, the predicted value of the real-time compensation time delay value is calculated. Optionally, the parameter iteration update of the improved adaptive filter algorithm follows the following rules: ,in: This represents the filter weight vector at the k-th iteration. This represents the convergence step size factor at the k-th iteration. This represents the error signal at the k-th iteration. This represents the adaptive filter input vector at the k-th iteration. Filter weight vector. The dimension of the adaptive filter input vector Consistent dimensions, convergence step size factor The radar is dynamically adjusted based on the norm of the change in radar operating frequency and the change in transmitted pulse repetition frequency. It can be understood that, in practical implementation, the error signal... The adaptive filter input vector is generated by comparing the predicted real-time compensation delay value with the delay residual value calculated based on the measured echo. The system consists of the radar operating frequency variation, the transmitted pulse repetition frequency variation, the system internal operating temperature, and the transmitter power amplifier operating voltage. The system internal operating temperature includes the transmitter power amplifier heatsink temperature and the receiver local oscillator module temperature. The filter weight vector... The update process is performed within each pulse repetition cycle to ensure that the real-time compensation delay value can track the dynamic changes in system delay.

[0024] In practical implementation, the interpolation query of the preset mapping table adopts a bilinear interpolation method. It takes the transmitter power amplifier heatsink temperature, receiver local oscillator module temperature, and transmitter power amplifier operating voltage monitoring values ​​as inputs, and performs linear interpolation calculations in a three-dimensional data grid. The temperature drift compensation calculation relies on the historical sequence of temperature sensor readings, and predicts the additional impact of temperature changes on time delay through first-order difference. The improved adaptive filter algorithm's convergence step size factor... The adjustment strategy is based on the energy normalization of the input vector to stabilize the convergence performance of the filter. It can be understood that the acquisition of the system reference delay value and the operation of the improved adaptive filter algorithm are processed in parallel. The system reference delay value provides a reference signal for the filter, while the adaptive filter input vector incorporates information on changes in real-time operating parameters. The system reference delay value obtained through a lookup table is used as a reference, and after correction by the adaptive filter, the final real-time compensation delay value is generated for compensation.

[0025] In one embodiment of the present invention, when generating the error signal during the iteration process of the improved adaptive filter algorithm, see [reference needed]. Figure 3While performing normal imaging, the SAR system simultaneously acquires leakage signal sampling sequences from the radar's transmitting channel and raw echo data sampling sequences from the receiving channel. Cross-correlation processing is performed on the acquired leakage signal sampling sequences and raw echo data sampling sequences to calculate the measured transmit / receive delay values. The theoretical delay value, corrected by the unmodified adaptive filter algorithm under the current operating condition, is retrieved from a preset mapping table. The difference between the measured transmit / receive delay values ​​and the theoretical delay values ​​is calculated to obtain the delay residual value calculated based on the measured echo. The difference between the predicted real-time compensation delay value output by the improved adaptive filter algorithm and the measured transmit / receive delay values ​​is calculated and used as the error signal for updating the filter parameters. During the cross-correlation processing of the leakage signal sampling sequences and raw echo data sampling sequences, bandpass filtering is applied to the leakage signal sampling sequences to extract the start-edge features of the transmitted pulse, and the positions of these feature points are recorded as transmission time reference points. The original echo data sampling sequence is pulse-compressed to extract the peak position of the ground object scattered echoes, and this peak position is recorded as the echo arrival time reference point. The position difference between the echo arrival time reference point and the transmission time reference point in the sampling sequence is calculated to obtain the sampling point difference. According to the sampling rate of the analog-to-digital converter of the SAR radar system, the sampling point difference is converted into a time difference to obtain the preliminary transmit / receive delay. The preliminary transmit / receive delay calculated within multiple consecutive pulse repetition periods is subjected to moving average filtering to suppress random noise interference, thereby obtaining the final measured transmit / receive delay value.

[0026] In practical implementation, while the SAR system performs normal imaging operations, it simultaneously acquires the leakage signal sampling sequence from the radar transmission channel and the original echo data sampling sequence from the receiving channel in real time. The leakage signal sampling sequence and the original echo data sampling sequence are cross-correlated to calculate the measured transmit / receive delay values. From a preset mapping table, the theoretical delay value corrected by the unmodified adaptive filter algorithm under the current operating conditions is read. The difference between the measured transmit / receive delay value and the theoretical delay value is calculated to obtain the delay residual value calculated based on the measured echo. The difference between the predicted value of the real-time compensation delay value output by the improved adaptive filter algorithm and the measured transmit / receive delay value is calculated, and this difference is used as the error signal for updating the filter parameters. In some embodiments, the leakage signal sampling sequence is bandpass filtered to extract the start edge features of the transmitted pulse. The feature point positions of the start edge features are recorded as transmission time reference points. The original echo data sampling sequence is pulse compressed to extract the peak positions of the ground object scattered echoes. The peak positions of the ground object scattered echoes are recorded as echo arrival time reference points. The position difference between the echo arrival time reference point and the transmission time reference point in the sampling sequence is calculated to obtain the sampling point difference. According to the sampling rate of the analog-to-digital converter of the SAR radar system, the sampling point difference is converted into a time difference to obtain the preliminary transmit and receive delay. Optionally, the time delay residual value calculated based on the measured echo can be used... Defined by the following formula: ,in: This represents the measured transmit / receive delay value. This represents the theoretical time delay value, read from a preset mapping table, under the current operating condition and corrected by the unmodified adaptive filter algorithm. Error signal. Then the predicted value of the real-time compensation delay value Compared with the measured transmit and receive delay values The difference constitutes, i.e. It's understandable that the time delay residual value... This reflects the deviation between the actual system delay and the calibrated theoretical value, while the error signal... The parameters are directly used to drive the improved adaptive filter algorithm, ensuring real-time compensation of the predicted delay values. Converging to the measured transmit / receive delay values .

[0027] In practical implementation, the start edge characteristic of the transmitted pulse is determined by detecting the threshold point of the envelope of the leaked signal sampling sequence. The peak position of the ground object scattered echo is obtained by finding the maximum amplitude point after pulse compression matched filtering of the original echo data sampling sequence and the reference transmitted signal copy. When calculating the difference in the number of sampling points, both the transmission time reference point and the echo arrival time reference point are index numbers of the discrete sampling sequence. The difference multiplied by the sampling time interval is the preliminary transmit / receive delay. Optionally, the preliminary transmit / receive delay calculated within multiple consecutive pulse repetition cycles is subjected to moving average filtering. The window length of the moving average is set according to the system pulse repetition frequency and the expected delay fluctuation bandwidth. The delay values ​​within the window are arithmetically averaged to suppress random noise interference, and the final measured transmit / receive delay value is output. It can be understood that through moving average filtering, the random fluctuation of the measured transmit / receive delay value is reduced, making the generated error signal more stable. The process is smoother and more stable, which helps improve the convergence and steady-state performance of the adaptive filter algorithm. In some embodiments, cross-correlation processing is performed in a digital signal processor, while pulse compression and moving average filtering are both implemented in a pipelined manner within a field-programmable gate array (FPGA) logic.

[0028] In one embodiment of the present invention, when adjusting the sampling time sequence of the original radar echo data according to the real-time compensation delay value, the current radar original echo data block to be processed is read from the data buffer of the SAR radar system. The original sampling time corresponding to each sampling point in this radar original echo data block is extracted to form the original sampling time sequence. The latest real-time compensation delay value is read from the output of the improved adaptive filter algorithm at the current time. The real-time compensation delay value is subtracted from each sampling time in the original sampling time sequence to generate the compensated sampling time sequence. Based on this compensated sampling time sequence, the echo amplitude sequence in the radar original echo data block is rearranged. During the rearrangement process, the time interval between adjacent sampling points in the compensated sampling time sequence is determined. Based on the ratio of this time interval to the original sampling time interval, it is determined whether data interpolation is required. When the ratio is not equal to one, the sinc interpolation algorithm is used to resample the echo amplitude sequence, and the corresponding echo amplitude on the new, uniformly distributed compensated sampling time sequence is calculated. When the ratio equals one, the echo amplitude sequence is directly shifted as a whole. The number of shift points is determined by rounding down the quotient of the real-time compensation delay value and the sampling time interval. The echo amplitude sequence after resampling or overall shifting is then output according to the order of the compensated sampling time sequence to form a time-axis aligned echo data sequence.

[0029] In practice, the current raw radar echo data block to be processed is read from the data buffer of the SAR radar system. The original sampling time corresponding to each sampling point in the raw radar echo data block is extracted to form the original sampling time sequence. The latest real-time compensation delay value is read from the output of the improved adaptive filter algorithm at the current time. The latest real-time compensation delay value is subtracted from each sampling time in the original sampling time sequence to generate the compensated sampling time sequence. Based on the compensated sampling time sequence, the echo amplitude sequence in the raw radar echo data block is rearranged. In some embodiments, the time interval between adjacent sampling points in the compensated sampling time sequence is determined. Based on the ratio of the time interval to the original sampling time interval, it is determined whether data interpolation is required. When the ratio of the time interval to the original sampling time interval is not equal to one, the sinc interpolation algorithm is used to resample the echo amplitude sequence and calculate the corresponding echo amplitude on the new, uniformly distributed compensated sampling time sequence. When the ratio of the time interval to the original sampling time interval is equal to one, the echo amplitude sequence is directly shifted as a whole. The number of shifted points is determined by rounding down the quotient of the real-time compensation delay value and the original sampling time interval. The echo amplitude sequence after resampling or overall shifting is output according to the order of the compensated sampling time sequence to form a time-axis aligned echo data sequence.

[0030] In practical implementation, the original sampling time sequence consists of equally spaced discrete time points. The compensated sampling time sequence, however, may become a non-uniform sequence due to the subtraction of a real-time compensation delay value that is not an integer multiple of the sampling interval. The sinc interpolation algorithm is applied precisely to recover the signal values ​​in the uniformly compensated sampling time sequence from the non-uniform original sampling points. Optionally, the sinc interpolation calculation is described by the following formula: ,in: Indicates the sampling time after compensation. The echo amplitude calculated above, Indicates the time of the original sampling. The echo amplitude collected above, This represents the system's original sampling time interval. It is an interpolation kernel function, summing index. Coverage A finite number of original sampling points centered on the center, This is the half-order of the interpolation filter. It can be understood that when the real-time compensation delay value is an integer multiple of the original sampling time interval, the ratio of the time interval to the original sampling time interval is equal to one. In this case, the compensated sampling time sequence is still uniform, and it only has a fixed offset of an integer number of sampling points from the original sequence. Overall translation is the most efficient processing method. When the real-time compensation delay value is not an integer multiple of the original sampling time interval, resampling is required using the sinc interpolation formula mentioned above to accurately achieve the time shift adjustment of the fractional sampling interval.

[0031] In some embodiments, the judgment and interpolation process can be accelerated by querying a pre-defined rule table, which associates different real-time compensation delay value ranges with corresponding processing modes. Optionally, referring to Table 1, a simplified processing rule table is shown, illustrating the different operations taken for different real-time compensation delay values: Table 1: Time Shift Adjustment Processing Rules , In practice, the overall translation operation is achieved by modifying the read pointer address of the data buffer, while the sinc interpolation resampling operation is completed in the digital signal processor through convolution operations. It can be understood that regardless of the processing method used, the ultimate goal is to generate a new echo amplitude sequence whose time axis is perfectly aligned with the ideal timing sequence. This time-aligned echo data sequence will be directly used for subsequent range-direction pulse compression processing to eliminate the phase error introduced by system transmit and receive delay fluctuations.

[0032] In one embodiment of the present invention, when performing SAR imaging based on time-delay compensated pulse compression data, a range-oriented Fast Fourier Transform is performed on the time-axis aligned echo data sequence to convert the data to the range frequency domain. In the range frequency domain, the time-axis aligned echo data sequence is multiplied by the frequency domain form of the matched filter of the transmitted signal to complete range-oriented pulse compression and generate one-dimensional range image data. Azimuth resampling processing is performed on the one-dimensional range image data of multiple consecutive pulses to correct for Doppler frequency variations caused by platform motion. In the azimuth resampling processing, one-dimensional range image data of two adjacent pulses are extracted, the cross-correlation function of these two one-dimensional range image data is calculated, and the azimuth position offset caused by platform velocity changes is estimated based on the peak position of the cross-correlation function. An azimuth resampling interpolation kernel function is constructed based on the estimated azimuth position offset. Using the constructed azimuth resampling interpolation kernel function, non-uniform interpolation is performed on the one-dimensional range image data of the current pulse along the azimuth direction to align its azimuth sampling position with the sampling grid of the reference pulse. Iterative processing of all pulses' one-dimensional range image data ensures all data have a uniform equivalent sampling interval in the azimuth direction. An azimuth-resampled data is then subjected to an azimuth-oriented Fast Fourier Transform (FFT) to convert the data to the two-dimensional frequency domain. In the two-dimensional frequency domain, based on the geometric model of SAR imaging, a range migration correction factor and a secondary range compression factor are calculated and applied. A two-dimensional inverse FFT is then performed on the corrected two-dimensional frequency domain data to convert the data back to the two-dimensional time domain, yielding the final SAR complex image data. This SAR complex image data is then moduloed to generate the final high-resolution radar image.

[0033] In practical implementation, SAR imaging is performed based on pulse compression data with time delay compensation. A range-oriented Fast Fourier Transform (FFT) is applied to the time-axis aligned echo data sequence to convert the data to the range frequency domain. In the range frequency domain, the time-axis aligned echo data sequence is multiplied by the frequency domain form of the matched filter of the transmitted signal to complete range-oriented pulse compression, generating one-dimensional range image data. Optionally, the frequency domain form of the matched filter of the transmitted signal is the conjugate Fourier transform of the time-domain waveform of the transmitted signal. The multiplication operation is performed point-by-point in the digital signal processor, and the generated one-dimensional range image data is arranged into a two-dimensional data matrix according to the slow-time pulse order. Azimuth resampling is performed on one-dimensional range image data of multiple consecutive pulses to correct for Doppler frequency variations caused by platform motion. One-dimensional range image data of two adjacent pulses are extracted, and the cross-correlation function of the one-dimensional range image data of two adjacent pulses is calculated. Based on the peak position of the cross-correlation function, the azimuth position offset caused by platform velocity variation is estimated. Based on the estimated azimuth position offset, an azimuth resampling interpolation kernel function is constructed. Using the constructed azimuth resampling interpolation kernel function, non-uniform interpolation is performed on the one-dimensional range image data of the current pulse along the azimuth direction to align its azimuth sampling position with the sampling grid of the reference pulse. The one-dimensional range image data of all pulses are iteratively processed to ensure that all data have a uniform equivalent sampling interval in the azimuth direction.

[0034] In some embodiments, the selection of the azimuth resampling interpolation kernel function is based on the characteristics of the motion error. Referring to Table 2, an implementation method for selecting different interpolation kernel functions according to the interval to which the peak position offset of the cross-correlation function belongs is shown: Table 2: Correspondence between Azimuth Position Offset and Interpolation Kernel Function , In practical implementation, the data after azimuth resampling is subjected to an azimuth-oriented Fast Fourier Transform (FFT) to convert the data to a two-dimensional frequency domain. In the two-dimensional frequency domain, based on the geometric model of SAR imaging, the range migration correction factor and the secondary range compression factor are calculated and applied. The range migration correction factor is used to compensate for the range curvature caused by the relative motion between the radar and the target. The calculation formula is expressed as follows: ,in: Indicates distance frequency and azimuth frequency The distance migration correction factor phase at that location, It is the imaginary unit. It is the radar operating wavelength. It is a distance frequency and azimuth frequency The range migration offset is jointly determined. From the echo data matrix in the two-dimensional frequency domain, based on the current range frequency cell index and azimuth frequency cell index, the complex echo value corresponding to the corresponding cell is read. The read complex echo value is multiplied by the calculated range migration correction factor using complex conjugate multiplication to eliminate the phase coupling described by the range migration equation, thus completing the range migration correction. A two-dimensional inverse fast Fourier transform is performed on the two-dimensional frequency domain data after all corrections are completed, converting the data back to the two-dimensional time domain to obtain the final SAR complex image data. The modulus of the SAR complex image data is then taken to generate the final high-resolution radar image. It can be understood that the azimuth resampling process corrects the azimuth sampling non-uniformity caused by the non-ideal motion of the platform, while the subsequent range migration correction accurately compensates for the phase introduced by range curvature in the two-dimensional frequency domain. The well-focused complex image data is obtained through the two-dimensional inverse fast Fourier transform, and the modulus operation generates an intensity image that can be interpreted.

[0035] In some embodiments, when calculating the cross-correlation function, range gate data from the strong scattering point set in the one-dimensional range profile of adjacent pulses are selected for calculation to improve estimation accuracy. When constructing the azimuth resampling interpolation kernel function, both the linear interpolation kernel function and the sinc interpolation kernel function are pre-stored in the processor memory for later use. Optionally, the azimuth resampling process is iterative. The one-dimensional range profile data of the first pulse is used as a reference grid, and the data of each subsequent pulse is sequentially cross-correlated and interpolated with the data aligned with the previous pulse until all pulses have been processed. It can be understood that through this pulse-by-pulse correction method, even if the platform has non-uniform motion, the generated final data is equivalent to uniform sampling in the azimuth direction, which meets the data uniformity requirements of subsequent azimuth Fourier transform processing, thereby ensuring the focusing quality of the final SAR image in the azimuth direction.

[0036] In one embodiment of the present invention, when calculating and applying the range migration correction factor and the second-order range compression factor in the two-dimensional frequency domain based on the geometric model of SAR imaging, after completing the azimuth-directed fast Fourier transform, a two-dimensional frequency domain echo data matrix is ​​obtained. This matrix includes both range and azimuth frequency dimensions. Based on the platform motion parameters and imaging geometry of the SAR radar system, a range migration equation is constructed, which describes the coupling relationship between the target slant range history and the range and azimuth frequency dimensions. Based on the constructed range migration equation, the range migration correction factor corresponding to each range and azimuth frequency unit is calculated in the two-dimensional frequency domain. From the echo data matrix in the two-dimensional frequency domain, the complex echo value corresponding to the corresponding unit is read according to the current range and azimuth frequency unit indices. The read complex echo value is multiplied by the calculated range migration correction factor using a complex conjugate multiplication to eliminate the phase coupling described by the range migration equation, thus completing the range migration correction. After completing the range migration correction, the second-order range compression factor is calculated based on the stationary phase principle for the cross-coupling term of the range and azimuth frequencies. From the two-dimensional frequency domain data matrix after range migration correction, the complex echo values ​​corresponding to the current range frequency cell index and azimuth frequency cell index are read. The read complex echo values ​​are then multiplied by the calculated quadratic range compression factor using complex conjugate multiplication to eliminate the phase error caused by the cross-coupling term between the range frequency and azimuth frequency, thus completing the quadratic range compression.

[0037] In practice, after completing the azimuth-to-fast Fourier transform, a two-dimensional frequency domain echo data matrix is ​​obtained. The dimensions of the two-dimensional frequency domain echo data matrix include the range frequency dimension and the azimuth frequency dimension. Based on the platform motion parameters and imaging geometry of the SAR radar system, a range migration equation is constructed. The range migration equation describes the coupling relationship between the target slant range history and the range frequency dimension. Based on the constructed range migration equation, the range migration correction factor corresponding to each range frequency unit and azimuth frequency unit is calculated in the two-dimensional frequency domain. From the two-dimensional frequency domain echo data matrix, according to the current range frequency unit index and azimuth frequency unit index, the complex echo value corresponding to the corresponding unit is read. The read complex echo value is multiplied by the calculated range migration correction factor by complex conjugate multiplication to eliminate the phase coupling described by the range migration equation and complete the range migration correction. In some embodiments, the calculation of the range migration correction factor depends on the specific values ​​of the radar platform velocity, wavelength, reference slant range, and range and azimuth frequencies. The complex conjugate multiplication operation of the complex echo value and the range migration correction factor is performed independently and in parallel on each range-azimuth frequency unit.

[0038] After completing range migration correction, a second-order range compression factor is calculated based on the stationary phase principle for the cross-coupling term between range and azimuth frequencies. This second-order range compression factor is used to compensate for the higher-order phase caused by the coupling between range and azimuth frequencies. Calculated using the following formula: ,in: Indicates distance frequency and azimuth frequency The quadratic distance compression factor at that location, It is the imaginary unit. It is distance frequency. It depends on the azimuth frequency The changing equivalent frequency modulation (FM). From the two-dimensional frequency domain data matrix after range migration correction, based on the current range frequency cell index and azimuth frequency cell index, the complex echo value corresponding to the corresponding cell is read. The read complex echo value is then multiplied by the calculated second-order range compression factor using complex conjugate multiplication to eliminate the phase error caused by the cross-coupling term between the range frequency and azimuth frequency, thus completing the second-order range compression. The second-order range compression factor can be understood as... The secondary phase term, which mainly acts on the range frequency, has a compensation amount that varies with the azimuth frequency. In high-resolution or large-angle imaging scenarios, the compensation of this coupled phase term is crucial for obtaining good range-oriented focusing.

[0039] In some embodiments, equivalent frequency modulation The analytical expression is determined by the radar system parameters and imaging geometry, and its calculation can be expressed as: ,in: It is the frequency modulation frequency for transmitting linear frequency modulation signals. It is the radar operating wavelength. It is the effective speed of the radar along the heading. This refers to the azimuth frequency. Optionally, the application of the range migration correction factor and the second-order range compression factor can be performed sequentially in a single loop. That is, each cell of the two-dimensional frequency domain data matrix is ​​first multiplied by the conjugate of the range migration correction factor, and then immediately multiplied by the conjugate of the second-order range compression factor, thereby reducing repeated read / write operations on the data matrix. In practice, the complex conjugate multiplication operation means multiplying the complex echo value by the conjugate of the correction factor (whose modulus is always 1 and contains only phase information). The result is that only the phase of the complex echo value is changed without changing its amplitude, thus accurately removing unwanted coupled phase terms and creating conditions for the subsequent two-dimensional inverse Fourier transform to generate a high-quality focused image.

[0040] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for stabilizing the transmit and receive delays of a SAR radar system, characterized in that, The method includes: Obtain the current operating parameter set of the SAR radar system, which includes the radar operating frequency, transmit pulse repetition frequency, system internal operating temperature, and transmitter power amplifier operating voltage. Based on the internal operating temperature of the system and the operating voltage of the transmitter power amplifier, the system reference delay value is obtained by looking up a table. The system reference delay value corresponds to the initial delay value under the nominal operating condition. Based on the current set of operating parameters, the system reference delay value is corrected in real time using an improved adaptive filter algorithm to generate a real-time compensated delay value. The improved adaptive filter algorithm iteratively updates the parameters based on the changes in the radar operating frequency and the transmit pulse repetition frequency. In the range pulse compression stage of SAR imaging processing, the sampling time sequence of the original radar echo data is time-shifted according to the real-time compensation delay value to generate pulse compressed data after delay compensation. SAR imaging is performed based on the time-delay compensated pulse compression data to generate the final high-resolution radar image.

2. The method for stabilizing the transmit and receive delays of a SAR radar system according to claim 1, characterized in that, Based on the system's internal operating temperature and the transmitter power amplifier's operating voltage, the system's reference time delay value is obtained using a lookup table method, including: Establish a preset mapping relationship table between the system's internal operating temperature, the transmitter power amplifier's operating voltage value, and the system's reference time delay value. The preset mapping relationship table is constructed based on the measured system time delay calibration data under multiple combinations of temperature and high voltage values. The system collects real-time readings of the internal operating temperature sensor and the operating voltage monitoring value of the transmitter power amplifier of the SAR radar system. The internal operating temperature sensor readings include the temperature of the transmitter power amplifier heat sink and the temperature of the receiver local oscillator module. Using the transmitter power amplifier heat sink temperature, receiver local oscillator module temperature, and transmitter power amplifier operating voltage monitoring values ​​as joint inputs, interpolation is performed in the preset mapping table to obtain the interpolation system reference delay that is closest to the current operating condition. Temperature drift compensation is performed on the reference time delay of the interpolation system. The temperature drift compensation is calculated based on the historical variation trend of the temperature of the transmitter power amplifier heat sink and the temperature of the receiver local oscillator module. Output the system reference delay value after temperature drift compensation.

3. The method for stabilizing the transmit and receive delays of a SAR radar system according to claim 1, characterized in that, The improved adaptive filter algorithm iteratively updates parameters based on the changes in the radar operating frequency and the transmitted pulse repetition frequency, including: Initialize the filter weight vector and convergence step size factor of the improved adaptive filter algorithm. The filter weight vector is used to characterize the nonlinear mapping relationship between time delay and operating parameters. Within each radar operating pulse repetition cycle, the instantaneous radar operating frequency reading of the current pulse and the instantaneous value of the transmitted pulse repetition frequency are collected, and the frequency difference with the previous cycle is calculated to obtain the change in radar operating frequency and the change in transmitted pulse repetition frequency. The radar operating frequency change, the transmit pulse repetition frequency change, and the system internal operating temperature and transmitter power amplifier operating voltage value in the current operating parameter set are used together to form the adaptive filter input vector at the current moment. The system reference time delay value is used as a reference signal input to the improved adaptive filter algorithm; The predicted value of the real-time compensation delay value is calculated based on the adaptive filter input vector, the reference signal, and the current filter weight vector. The predicted value is compared with the time delay residual value calculated based on the measured echo to generate an error signal. The filter weight vector and convergence step size factor are adjusted using the error signal and a preset iterative update rule to complete one parameter iterative update.

4. The method for stabilizing the transmit and receive delays of a SAR radar system according to claim 3, characterized in that, The predicted value is compared with the time delay residual value calculated based on the measured echo to generate an error signal, including: While the SAR system is performing normal imaging operations, the leakage signal sampling sequence of the radar transmission channel and the raw echo data sampling sequence of the receiving channel are collected in real time. The leakage signal sampling sequence and the original echo data sampling sequence are cross-correlated to calculate the measured transmit and receive delay values. Read the theoretical time delay value under the current working condition without correction by the improved adaptive filter algorithm from the preset mapping table; The difference between the measured transmit and receive delay value and the theoretical delay value is calculated to obtain the delay residual value calculated based on the measured echo. The difference between the predicted value of the real-time compensation delay output by the improved adaptive filter algorithm and the measured value of the transmit / receive delay is calculated, and the difference is used as the error signal for updating the filter parameters.

5. The method for stabilizing the transmit and receive delays of a SAR radar system according to claim 4, characterized in that, The leakage signal sampling sequence and the original echo data sampling sequence are cross-correlated to calculate the measured transmit and receive delay values, including: The leakage signal sampling sequence is subjected to bandpass filtering to extract the start edge features of the transmitted pulse, and the position of its feature point is recorded as the transmission time reference point; The original echo data sampling sequence is subjected to pulse compression processing to extract the peak position of the ground object scattered echo, and the peak position is recorded as the echo arrival time reference point; Calculate the position difference between the echo arrival time reference point and the transmission time reference point in the sampling sequence to obtain the difference in the number of sampling points; Based on the sampling rate of the analog-to-digital converter of the SAR radar system, the difference in the number of sampling points is converted into a time difference to obtain the preliminary transmit and receive delay; The preliminary transmit / receive delay calculated over multiple consecutive pulse repetition cycles is subjected to moving average filtering to suppress random noise interference, resulting in the final measured transmit / receive delay value.

6. The method for stabilizing the transmit and receive delays of a SAR radar system according to claim 1, characterized in that, Based on the real-time compensation delay value, the sampling time sequence of the original radar echo data is time-shifted and adjusted, including: Read the current block of raw radar echo data to be processed from the data buffer of the SAR radar system; Extract the original sampling time corresponding to each sampling point in the original radar echo data block to form an original sampling time sequence; Read the latest real-time compensation delay value from the output of the improved adaptive filter algorithm at the current moment; Subtract the real-time compensation delay value from each sampling time in the original sampling time sequence to generate the compensated sampling time sequence; Based on the compensated sampling time sequence, the echo amplitude sequence in the original radar echo data block is rearranged. The rearrangement is achieved by interpolation calculation of the time position of the sampling points to generate an echo data sequence aligned with the time axis.

7. The method for stabilizing the transmit and receive delays of a SAR radar system according to claim 6, characterized in that, Based on the compensated sampling time sequence, the echo amplitude sequence in the original radar echo data block is rearranged, including: Determine the time interval between adjacent sampling points in the compensated sampling time sequence; Based on the ratio of the time interval to the original sampling time interval, determine whether data interpolation is needed; When the ratio is not equal to one, the sinc interpolation algorithm is used to resample the echo amplitude sequence and calculate the corresponding echo amplitude on the new, uniformly distributed compensated sampling time sequence. When the ratio is equal to one, the echo amplitude sequence is directly shifted as a whole, and the number of shift points is determined by rounding down the quotient of the real-time compensation delay value and the sampling time interval. The echo amplitude sequence, after being resampled or shifted, is output in the order of the compensated sampling time sequence to form a time-axis aligned echo data sequence.

8. The method for stabilizing the transmit and receive delays of a SAR radar system according to claim 1, characterized in that, SAR imaging based on the time-delay compensated pulse compression data includes: Perform a range-direction Fast Fourier Transform on the time-axis aligned echo data sequence to convert the data to the range frequency domain; In the range frequency domain, the time-axis aligned echo data sequence is multiplied by the frequency domain form of the matched filter of the transmitted signal to complete range pulse compression and generate one-dimensional range image data. Azimuth resampling is performed on one-dimensional range image data of multiple consecutive pulses to correct for Doppler frequency changes caused by platform motion; Azimuth Fast Fourier Transform is performed on the data after azimuth resampling to convert the data to the two-dimensional frequency domain. In the two-dimensional frequency domain, based on the geometric model of SAR imaging, the range migration correction factor and the secondary range compression factor are calculated and applied; Perform a two-dimensional inverse fast Fourier transform on the two-dimensional frequency domain data after all corrections are completed, and convert the data back to the two-dimensional time domain to obtain the final SAR complex image data. The SAR complex image data is moduloed to generate the final high-resolution radar image.

9. A method for stabilizing the transmit and receive delays of a SAR radar system according to claim 8, characterized in that, The azimuth resampling process for one-dimensional range image data of multiple consecutive pulses includes: Extract one-dimensional range profile data from two adjacent pulses, calculate the cross-correlation function of the one-dimensional range profile data from the two adjacent pulses, and estimate the azimuth position offset caused by the change in platform velocity based on the peak position of the cross-correlation function. Based on the azimuth position offset, construct an azimuth resampling interpolation kernel function; Using the azimuth resampling interpolation kernel function, non-uniform interpolation is performed on the one-dimensional range profile data of the current pulse along the azimuth direction to align its azimuth sampling position with the sampling grid of the reference pulse; Iteratively process the one-dimensional range image data of all pulses to ensure that all data have a uniform equivalent sampling interval in the azimuth direction.

10. A method for stabilizing the transmit and receive delays of a SAR radar system according to claim 8, characterized in that, In the two-dimensional frequency domain, based on the geometric model of SAR imaging, the range migration correction factor and the secondary range compression factor are calculated and applied, including: After completing the azimuth-to-fast Fourier transform, a two-dimensional frequency domain echo data matrix is ​​obtained, wherein the dimensions of the two-dimensional frequency domain echo data matrix include the range frequency dimension and the azimuth frequency dimension. Based on the platform motion parameters and imaging geometry of the SAR radar system, a range migration equation is constructed, which describes the coupling relationship between the target slant range history in the range frequency dimension and the azimuth frequency dimension. Based on the range migration equation, the range migration correction factor corresponding to each range frequency unit and azimuth frequency unit is calculated in the two-dimensional frequency domain. From the echo data matrix in the two-dimensional frequency domain, based on the current range frequency cell index and azimuth frequency cell index, read the complex echo value corresponding to the corresponding cell; The complex echo value read is multiplied by the calculated range migration correction factor using a complex conjugate multiplication to eliminate the phase coupling described by the range migration equation, thus completing the range migration correction. After completing the range migration correction, the second-order range compression factor is calculated based on the stationary phase principle for the cross-coupling term of range frequency and azimuth frequency. From the two-dimensional frequency domain data matrix after range migration correction, read the complex echo value corresponding to the corresponding cell based on the current range frequency cell index and azimuth frequency cell index; The complex echo value read is multiplied by the calculated second-order range compression factor using a complex conjugate multiplication to eliminate the phase error caused by the cross-coupling term between the range frequency and the azimuth frequency, thus completing the second-order range compression.

Images