Radar antenna signal processing method and system

Abstract

The invention relates to the technical field of signal processing, in particular to a radar antenna signal processing method and system, and the method comprises the following steps: collecting an antenna time domain echo extraction channel energy curve, calculating the initial phase attenuation curvature of a period length energy peak value, and building an interference characteristic parameter set; matching real-time echoes, judging homology, eliminating interference, extracting main echoes, extracting main peak position spacing through frequency spectrum transformation, calculating a judgment threshold, comparing, and judging and executing decoupling operation to obtain target frequency spectrum data. According to the method, by extracting Doppler channel coherent energy distribution and establishing a periodic interference characteristic parameter set, the intrinsic characteristics of periodic interference can be accurately described, rapid judgment and effective stripping of interference source homology are achieved, and by analyzing the Doppler center frequency change trend and generating a pulse delay compensation quantized value, the interference source homology can be quickly judged and effectively stripped. The stable target identification and frequency decoupling operation under the interference background is realized, and the target identification and frequency control level in the variable environment is obviously enhanced.

Description

Technical Field

[0001] This invention relates to the field of signal processing technology, and in particular to a radar antenna signal processing method and system. Background Technology

[0002] Signal processing technology involves the acquisition, transformation, analysis, and reconstruction of analog or digital signals to improve the effectiveness of information extraction and transmission. Core aspects include signal acquisition and conversion, filtering and noise reduction, feature extraction and recognition, time-frequency analysis, pattern recognition, and signal restoration. This technology is widely used in various engineering fields such as communications, radar, sonar, image recognition, biomedical engineering, and automatic control. Especially in radar systems, signal processing is a crucial step in target detection, tracking, and identification, playing a decisive role in system performance. Traditional radar antenna signal processing methods refer to preprocessing, detecting, and estimating parameters of echo signals in the radar receiver, primarily addressing needs such as suppressing clutter interference, enhancing target signals, and distinguishing multiple targets. Traditional methods employ matched filtering to achieve pulse compression and improve range resolution, use constant false alarm rate (CFAR) detection to threshold the signal for target detection, and combine Doppler filtering or beamforming techniques to improve angular or velocity resolution. In terms of multi-target analysis, traditional methods often rely on Fourier transform-like techniques to estimate the distance and angle of targets. However, their processing capabilities are significantly limited in the context of high clutter or low signal-to-noise ratio, making it difficult to meet the needs of high-precision detection in complex environments.

[0003] Traditional radar antenna signal processing methods mainly rely on matched filtering and constant false alarm rate (CFAR) detection. While these methods can improve range and velocity resolution to some extent, their effectiveness in target recognition is significantly limited in high clutter backgrounds and low signal-to-noise ratio (SNR) scenarios. This is especially true in environments with densely distributed multiple targets or periodically prominent interference sources. Traditional methods cannot effectively separate interference from the main echo, resulting in severe spectrum fusion and a chaotic distribution of the target echo's main peak, making accurate positioning and frequency identification difficult. Furthermore, due to insufficient spectrum decoupling capability, they cannot sensitively track frequency change trends in dynamic scenarios, limiting the accuracy of target energy focusing and echo consistency analysis. The overall processing flow lacks detailed modeling and iterative correction mechanisms for interference characteristics, leading to a decrease in target detection and frequency tracking accuracy and affecting stability and response capabilities in complex environments. Summary of the Invention

[0004] To address the technical problems existing in the prior art, this invention provides a radar antenna signal processing method, including the following steps: S1: Acquire the time-domain echo signal of the radar antenna, extract the coherent energy distribution curve of the Doppler channel, calculate the period length, peak energy amplitude, initial phase and attenuation curvature, and establish a set of periodic interference characteristic parameters. S2: Based on the set of periodic interference characteristic parameters, perform periodic back-projection matching operation on the real-time echo data, calculate the phase error value to determine the homogeneity of the interference source, remove the interference signal components, and obtain the main echo sequence after interference stripping. S3: Using the main echo sequence after interference stripping, extract the Doppler center frequency change trend, calculate the slope change rate value and interval fluctuation amplitude value, divide the velocity error prediction sub-interval, and generate pulse delay compensation quantization value. S4: Call the pulse delay compensation quantization value, perform time delay compensation operation on the sub-interval pulse, calculate the echo amplitude peak consistency value within the pulse group and the interference boundary contrast value between groups, and establish an optimized coherent pulse data group.

[0005] As a further aspect of the present invention, the periodic interference characteristic parameter set includes interference energy distribution type, periodic structure index, phase characteristic parameter, and energy attenuation law parameter; the main echo sequence includes target reflection component, interference residual intensity indicator, and time-domain waveform reconstruction index; the pulse delay compensation quantization value includes frequency drift correction value, velocity disturbance correction factor, and prediction error limit value within sub-interval; and the optimized coherent pulse data group includes time alignment consistency index, amplitude fluctuation suppression characteristic, and inter-group interference separation intensity indicator.

[0006] As a further aspect of the present invention, the specific steps of S1 are as follows: S101: Acquire radar antenna time-domain echo signal, perform frame structure identification, slice the pulse sequence in each frame of data, extract the signal peak index sequence based on the phase continuity and amplitude stability of a single pulse waveform, detect the periodic trend of the repetition interval based on the time interval sequence of adjacent pulses, and generate a set of period length values ​​by statistically classifying and frequency detecting the values ​​of the repetition interval. S102: Based on the set of period length values, locate the pulse index position where the period start point of the repetition frequency is located, call the complex amplitude data of the corresponding pulse in the original radar antenna time domain echo signal, aggregate the amplitude magnitude of the pulse within the period and perform envelope curve fitting to generate a group of energy peak amplitude values. S103: Call the initial phase information of the complex amplitude corresponding to the starting position of the cycle in the energy peak amplitude value group, and calculate the cosine phase difference sequence of the phase sequence at the starting point of each cycle in combination with the cycle length value. Retrieve the rate of change of the starting direction of the phase difference change trend curve within the cycle, and establish a set of periodic interference characteristic parameters by regression fitting the mean of the rate of change of the starting direction on the cycle.

[0007] As a further aspect of the present invention, the specific steps of S2 are as follows: S201: Based on the set of periodic interference characteristic parameters, construct a periodic waveform template sequence, call the continuous data segment with the same period length in the real-time echo data, perform point-by-point complex multiplication operation between the template waveform and the data segment, extract the real part to form a matching degree sequence, calculate the dot product ratio between the matching degree sequence at the period position and the template waveform, and obtain the periodic back projection matching factor column. S202: Call the periodic back projection matching factor column, combine the initial phase value and the attenuation curvature value, construct the desired phase sequence curve, normalize the amplitude based on the difference sequence between the complex phase sequence and the desired phase sequence at each period position in the echo data, calculate the mean square value of the normalized phase difference as the phase error value at each period position, and generate the phase error evaluation result. S203: Call the phase error evaluation result, compare the mean square error value with the set interference source same source determination threshold item by item. If the mean square error value is less than the interference source same source determination threshold, the position of the same source interference component is determined within the determination period, and the complex amplitude value in the real-time echo data is marked as the interference component. The amplitude is reduced to zero for the marked interference component to obtain the main echo sequence after interference removal.

[0008] As a further aspect of the present invention, the specific steps of S3 are as follows: S301: Based on the main echo sequence after interference stripping, extract the Doppler spectrum corresponding to the pulse and locate the peak amplitude frequency point index, aggregate the Doppler center frequency sequence corresponding to the pulse, arrange them sequentially by the pulse time index, construct a complete time-frequency trajectory curve, and obtain the Doppler center frequency change trend sequence. S302: Call the frequency difference between adjacent pulses in the Doppler center frequency change trend sequence, perform first-order difference calculation and normalization on the difference sequence, construct the frequency change slope sequence, perform sliding window operation on the frequency change slope sequence, calculate the mean square error and range of the slope change values ​​within the sliding window as the change rate index and fluctuation amplitude index of the segment, and generate a group of slope change rate values ​​and fluctuation amplitude values. S303: Based on the slope change rate value and fluctuation amplitude value group, set the speed error critical value range, divide the frequency change trend segment into intervals, and if any index in the frequency change trend segment exceeds the speed error critical value range, mark the frequency change trend segment as a prediction sub-interval, and perform linear fitting on the frequency change trend in the prediction sub-interval to generate pulse delay compensation quantization value.

[0009] As a further aspect of the present invention, the specific steps of S4 are as follows: S401: Call the pulse delay compensation quantization value, perform delay adjustment on the pulses in the velocity error prediction sub-interval in the time domain, sum the original time index and the correction amount, and reconstruct the pulse arrangement order according to the corrected index to obtain the pulse sequence set after delay compensation; S402: Based on the pulse sequence set after delay compensation, extract the peak amplitude within the pulse group by dividing it into groups. Obtain the consistency factor within the group by calculating the difference between the peak and trough peak within the group and dividing it by the peak value. Contrast index is constructed based on the difference between the edge pulse amplitude of each group and the edge pulse amplitude of the adjacent group. The contrast index and consistency factor are normalized and jointly arranged to generate a set of amplitude consistency and contrast index. S403: Based on the amplitude consistency and contrast index set, score and filter all pulse data groups, retain pulse data groups whose joint index values ​​are in the upper quantile interval, and aggregate the selected pulse sequences in chronological order to establish optimized coherent pulse data groups.

[0010] As a further aspect of the present invention, the process of scoring and filtering all pulse data groups according to the amplitude consistency and contrast index set, and retaining pulse data groups whose joint index values ​​are located in the upper quantile interval, specifically involves: presetting consistency weight values ​​and contrast weight values, and limiting the sum of the consistency weight values ​​and the contrast weight values ​​to one. Extract the normalized intra-group consistency factor and the contrast index from the set of amplitude consistency and contrast indexes; perform a multiplication operation on the intra-group consistency factor and the consistency weight value to obtain the consistency component value, and perform a multiplication operation on the contrast index and the contrast weight value to obtain the contrast component value; The consistency sub-value and the contrast sub-value are summed to generate the score value corresponding to each pulse data group; Perform a descending sort operation on the score values ​​of the pulse data group, and calculate the total number of values ​​in the sorted score value sequence; Multiply the total number of values ​​by a preset preferred ratio coefficient, and round the result down to obtain the truncation position index value; The set of values ​​in the scoring sequence from the first index to the truncation position index is defined as the upper quantile interval; Pulse data sets whose score values ​​fall within the upper quantile interval are determined to be retained, while pulse data sets whose score values ​​do not fall within the upper quantile interval are discarded.

[0011] As a further aspect of the present invention, the method further includes step S5: S5: Based on the optimized coherent pulse data group, perform a spectrum transformation operation to extract the target echo main peak position value and spacing value. Combine the statistical distribution characteristics of the main peak spacing to calculate the dynamic spacing judgment threshold. Compare the main peak spacing value with the dynamic spacing judgment threshold. When the main peak spacing value is less than the dynamic spacing judgment threshold, perform a spectrum splitting and decoupling operation according to the main peak spacing weight. Calculate the energy focusing degree value and frequency offset value of the decoupled target. If the main peak spacing value is greater than the dynamic spacing judgment threshold, lock the target frequency coordinates and obtain the target spectrum decoupling positioning data. The target spectrum decoupling and positioning data includes the position of the main peak in the frequency domain, the identifier of the decoupling target frequency, and the quantification value of the degree of spectrum separation and focusing.

[0012] As a further aspect of the present invention, the specific steps of S5 are as follows: S501: Based on the optimized coherent pulse data group, perform Fourier transform operation, extract the amplitude response curve in the frequency domain, locate the frequency index corresponding to the amplitude as the main peak position, and calculate the frequency interval value between adjacent main peaks to obtain the target echo main peak position and spacing sequence. S502: Call the target echo main peak position and spacing sequence, perform histogram distribution statistics, extract the main mode interval of frequency spacing distribution, and calculate the weighted average of the median of the main mode interval and the mean of the whole sequence as the dynamic distance reference value. Combine the variance value of the distribution curve to perform exponential adjustment and generate the dynamic spacing judgment threshold. S503: Compare the spacing between adjacent main peaks with the dynamic spacing judgment threshold item by item. If the spacing between main peaks is less than the dynamic spacing judgment threshold, assign a spectrum decoupling weight according to the reciprocal of the corresponding spacing. Perform multi-channel splitting processing on the spectrum amplitude curve according to the spectrum decoupling weight, locate the decoupling target frequency band range respectively, extract the frequency coordinate value of the corresponding main peak and mark it as the target frequency position, and call the target frequency coordinate and the spectrum amplitude data corresponding to the two frequency points before and after to perform local extreme value search, lock the frequency band range where the target peak point is located, and obtain the target spectrum decoupling positioning data.

[0013] A radar antenna signal processing system, comprising: The echo acquisition module acquires the time-domain echo signal from the radar antenna, calls the sampling time point sequence, antenna scanning angle sequence, and received signal amplitude sequence, performs energy normalization processing on multiple Doppler channels in each signal segment, extracts the normalized energy distribution curve, and generates a set of periodic interference characteristic parameters. The interference stripping module matches the sampling time sequence of the real-time echo data with the corresponding echo phase data according to the set of periodic interference characteristic parameters. It calls the interference phase reference sequence and the real-time echo phase difference sequence, performs point-by-point difference calculation on the phase difference and determines whether it is within the phase error limit threshold, and obtains the main echo sequence after interference stripping. Based on the main echo sequence after interference stripping, the frequency trend analysis module extracts the Doppler center frequency value under the corresponding time series, calls the frequency time distribution curve and calculates the rate of change of derivative and the frequency change range of each segment in segments. It normalizes the rate of change of slope and fluctuation amplitude in continuous segments to obtain the pulse delay compensation quantization value. The delay compensation module calls the pulse delay compensation quantization value, performs time delay adjustment operation on the echo pulse in the sub-interval, obtains the start time of the sub-interval, the echo pulse sampling period and the center time of adjacent pulses, and calculates the difference between the center time of each group of pulses and the delayed time series to establish an optimized coherent pulse data group. The spectrum decoupling module performs a spectrum transformation operation on each optimized pulse sequence based on the optimized coherent pulse data group, extracts the spacing data between the main spectrum peak position sequence and adjacent peaks, calls the adjacent main peak position difference sequence and statistically analyzes the distribution range and frequency, and obtains the target spectrum decoupling positioning data.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, by extracting the coherent energy distribution of the Doppler channel and establishing a set of periodic interference characteristic parameters, the intrinsic characteristics of periodic interference can be accurately characterized, enabling rapid judgment and effective separation of the homogeneity of interference sources, ensuring the purity of the main echo information. By analyzing the trend of Doppler center frequency changes and generating pulse delay compensation quantization values, the consistency and boundary contrast between echo pulses are improved, enhancing the frequency tracking capability under dynamic targets. By comparing the spectral main peak spacing with the dynamic judgment threshold, combined with spectral splitting decoupling and energy focusing calculation, the spectral discrimination capability and positioning accuracy in multi-target dense scenes are improved, realizing stable target identification and frequency decoupling operations under interference background, significantly enhancing the target identification and frequency control level in variable environments. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of the steps of the present invention; Figure 2 This is a detailed schematic diagram of S1 of the present invention; Figure 3 This is a detailed schematic diagram of S2 of the present invention; Figure 4This is a detailed schematic diagram of S3 of the present invention; Figure 5 This is a detailed schematic diagram of S4 of the present invention; Figure 6 This is a detailed schematic diagram of S5 of the present invention; Figure 7 This is a system module diagram of the present invention. Detailed Implementation

[0017] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0018] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0019] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0020] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0021] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0022] Please see Figure 1 This invention provides a radar antenna signal processing method, comprising the following steps: S1: Acquire the time-domain echo signal of the radar antenna, extract the coherent energy distribution curve of the Doppler channel, calculate the period length, peak energy amplitude, initial phase and attenuation curvature, and establish a set of periodic interference characteristic parameters. S2: Based on the periodic interference characteristic parameter set, perform periodic back projection matching operation on the real-time echo data, calculate the phase error value to determine the homogeneity of the interference source, remove the interference signal components, and obtain the main echo sequence after interference removal. S3: Using the main echo sequence after interference removal, extract the trend of Doppler center frequency change, calculate the slope change rate and interval fluctuation amplitude, divide the velocity error prediction sub-interval, and generate pulse delay compensation quantization values. S4: Call the pulse delay compensation quantization value, perform time delay compensation operation on the pulse of the sub-interval, calculate the echo amplitude peak consistency value within the pulse group and the interference boundary contrast value between groups, and establish an optimized coherent pulse data group; S5: Based on the optimized coherent pulse data group, perform spectrum transformation operation to extract the target echo main peak position value and spacing value. Combine the statistical distribution characteristics of the main peak spacing to calculate the dynamic spacing judgment threshold. Compare the main peak spacing value with the dynamic spacing judgment threshold. When the main peak spacing value is less than the dynamic spacing judgment threshold, perform spectrum splitting and decoupling operation according to the main peak spacing weight. Calculate the energy focusing degree value and frequency offset value of the decoupled target. If the main peak spacing value is greater than the dynamic spacing judgment threshold, lock the target frequency coordinates and obtain the target spectrum decoupling positioning data. The set of periodic interference characteristic parameters includes interference energy distribution type, periodic structure index, phase characteristic parameters, and energy attenuation law parameters. The main echo sequence includes target reflection components, interference residual intensity indicators, and time-domain waveform reconstruction indicators. The pulse delay compensation quantization values ​​include frequency drift correction values, velocity disturbance correction factors, and prediction error limit values ​​within sub-intervals. The optimized coherent pulse data set includes time alignment consistency indicators, amplitude fluctuation suppression characteristics, and inter-group interference separation intensity indicators. The target spectrum decoupling and positioning data includes frequency domain main peak position, decoupling target frequency indicators, and quantization values ​​of spectrum separation and focusing degree.

[0023] Please see Figure 2 The specific steps of S1 are as follows: S101: Acquire radar antenna time-domain echo signal, perform frame structure identification, slice the pulse sequence in each frame of data, extract the signal peak index sequence based on the phase continuity and amplitude stability of a single pulse waveform, detect the periodic trend of the repetition interval based on the time interval sequence of adjacent pulses, and generate a set of period length values ​​by statistically classifying and frequency detecting the values ​​of the repetition interval. The analog intermediate frequency (IF) signal output from the radar receiver is transmitted to a high-precision analog-to-digital converter. The sampling frequency is set to 200 MHz to satisfy the Nyquist sampling theorem, obtaining a continuous discrete digital signal sequence. A preset Barker code sequence is used as the frame synchronization identifier. A sliding window correlation matching operation is performed on the digital signal stream. Once the correlation peak exceeds a set zero-eight times the normalized amplitude threshold, the start position of the data frame is determined, and a fixed-length time window of data is extracted. The background noise level of the extracted data frame is estimated, and a dynamic amplitude threshold value four times higher than the background noise level is set. Data points within the frame are traversed to identify signal segments whose amplitude values ​​continuously exceed the threshold value. Detailed waveform quality analysis is performed on each signal segment to check whether the instantaneous phase exhibits linear change characteristics and the amplitude packet... To determine if the network has a flat-top shape, noise interference with disordered phase or drastic amplitude fluctuations is eliminated. For valid pulse waveforms that pass quality verification, the sampling point index at the moment when the amplitude reaches its maximum value is accurately located. The peak indices of valid pulses are arranged in chronological order, and the time difference between the peak indices of two adjacent pulses is calculated one by one. The time difference is converted into a specific time interval value, and a statistical histogram covering the period range is constructed. The interval width of the histogram is set to one microsecond. The calculated time interval values ​​are mapped to the corresponding histogram intervals for counting and accumulation. For example, when the statistical results show that there are 500 pulse pairs in the interval from 19.5 microseconds to 20.5 microseconds, the interval where the count peak is located is identified as a candidate period. The central trend value of the interval values ​​in the interval is calculated using the weighted average method to generate a set of period length values.

[0024] S102: Based on the period length numerical set, locate the pulse index position where the period start point of the repetition frequency is located, call the complex amplitude data of the corresponding pulse in the original radar antenna time domain echo signal, aggregate the amplitude magnitude of the pulse within the period and perform envelope curve fitting to generate a group of energy peak amplitude values. The most frequently occurring principal period value is selected as the benchmark reference, for example, two thousand sampling points are selected as the principal period. Using the position of the first confirmed valid pulse in the data frame as the anchor point, the principal period value is used to extrapolate along the time axis to predict the theoretically correct positions of each pulse. A small tolerance search window with a width of five is set before and after each predicted position to cover time offsets caused by jitter or Doppler effects. In the original time-domain echo signal, the actual pulse response peak is searched within the search window. The complex sampling data corresponding to the actual pulse peak time is extracted, and the in-phase and quadrature components are obtained separately. The sum of the squares of the in-phase and quadrature components is then calculated... The instantaneous amplitude magnitude of each pulse is obtained by taking the square root. For example, for a sampling point with three in-phase components and four quadrature components, the amplitude magnitude is calculated to be five. The amplitude magnitudes are arranged into an amplitude sequence according to the order in which the pulses appear. The least squares method is applied to the amplitude sequence to perform polynomial curve fitting. By iteratively adjusting the coefficients of the fitting function, the fitting curve can smoothly pass through each discrete amplitude point and reflect the energy change profile during the antenna beam scanning process. Differential operation is performed on the envelope curve generated by the fitting to find the stationary point where the first derivative is zero and the second derivative is negative. The amplitude value corresponding to the stationary point represents the true signal strength after eliminating the influence of noise, and a set of energy peak amplitude values ​​is generated.

[0025] S103: Call the initial phase information of the complex amplitude corresponding to the starting position of the cycle in the energy peak amplitude value group, and calculate the cosine phase difference sequence of the phase sequence at the starting point of each cycle in combination with the cycle length value. Retrieve the rate of change of the starting direction of the phase difference change trend curve within the cycle, and establish a set of periodic interference characteristic parameters by regression fitting the mean of the rate of change of the starting direction on the cycle. For each locked periodic pulse, the precise initial phase angle is extracted using the in-phase and quadrature components of the complex data through a four-quadrant arctangent operation. The phase angles are arranged according to a time sequence, and the phase difference between the starting pulses of two adjacent periods is calculated. To eliminate the discontinuities caused by phase winding, a cosine trigonometric function transformation is performed on the calculated phase difference, mapping it to a continuous numerical range from -1 to +1. For example, when the phase difference is 60 degrees, the calculated cosine value is 0.5. A trend curve of the cosine phase difference over time is plotted, with a focus on capturing the trend curve at each modulation... For the first five data points at the beginning of the cycle, the slope of the numerical change at the starting point is calculated using the numerical difference method to quantify the instantaneous direction and rate of change of the phase difference at the beginning of the cycle. Data on the rate of change of the starting direction of multiple cycles are collected to construct a mean regression statistical model. The long-term mean parameter and the regression velocity parameter are estimated. For example, the calculated long-term mean is close to 0.02 and the regression velocity parameter is 0.5. By analyzing the fluctuation of the current rate of change around the long-term mean, the phase stability characteristics of the signal are quantified. The model parameters obtained through statistical regression analysis are integrated to establish a set of periodic interference characteristic parameters.

[0026] Please see Figure 3 The specific steps of S2 are as follows: S201: Based on the periodic interference feature parameter set, construct a periodic waveform template sequence, call the continuous data segment with the same period length in the real-time echo data, perform point-by-point complex multiplication operation between the template waveform and the data segment, extract the real part to form a matching degree sequence, calculate the dot product ratio between the matching degree sequence at the period position and the template waveform, and obtain the periodic back projection matching factor column. A standardized periodic waveform template sequence is constructed to accurately simulate the time-domain fluctuation characteristics of the interference source under ideal conditions. For example, a digital waveform containing sinusoidal oscillations with a specific exponentially decaying envelope is constructed. Real-time echo data from the radar receiver's buffer is called up, and a continuous data segment with a preset period length is extracted as the processing unit, such as a data block with a length of two thousand sampling points. The constructed template waveform and the extracted real-time data segment are then subjected to point-by-point complex multiplication, which involves multiplying the complex value of each point in the template with the complex value of the corresponding point in the data segment. The amplitude and phase are then processed using complex number arithmetic rules. The information is extracted from the real part of the complex sequence of the calculation result to form a matching degree sequence that reflects the similarity of the signal waveform. The dot product ratio between the matching degree sequence at the current period position and the original template waveform is calculated. This is done by multiplying the values ​​in the matching degree sequence with the corresponding values ​​in the template waveform, summing them, and dividing by the energy normalization factor of the template waveform. Through the projection calculation process, the projection components of the real-time signal in the interference feature space are quantified. For example, when the calculated dot product ratio is 0.95, it indicates that the current data segment contains strong interference signals, while a ratio of 0.1 indicates that it is mainly background noise. The periodic back-projection matching factor column is obtained.

[0027] S202: Call the periodic back projection matching factor column, combine the initial phase value and the attenuation curvature value to construct the desired phase sequence curve, perform amplitude normalization based on the difference sequence between the complex phase sequence and the desired phase sequence at each period position in the echo data, calculate the mean square value of the normalized phase difference as the phase error value at each period position, and generate the phase error evaluation result. Combining the initial phase value obtained during the interference analysis phase with the curvature value describing the signal energy attenuation trend, an interpolation algorithm is used to construct a theoretically desired phase sequence curve. This curve represents the ideal trajectory of phase change over time when the signal is confirmed as interference. The complex phase sequence, marked as a potential period position, is read from the echo data. The difference between the actual measured phase and the theoretically desired phase sequence is calculated to obtain the original phase error sequence. To eliminate the influence of signal strength fluctuations on error assessment, the phase difference sequence is normalized based on the instantaneous amplitude of the echo data. This involves using amplitude weights to weight and scale the phase deviation, and calculating the mean square value of the normalized phase difference sequence. Specifically, each normalized difference in the sequence is squared, and the sum of the squared values ​​is divided by the sequence length. For example, if the phase fluctuation of a certain period closely matches the desired curve, the accumulated mean square value is only 0.02 radians squared, while for incoherent clutter signals, the mean square value is as high as 3.5 radians squared, generating the phase error assessment result.

[0028] S203: Call the phase error evaluation result, compare the error mean square value with the set interference source same source determination threshold item by item. If the error mean square value is less than the interference source same source determination threshold, the location of the same source interference component is determined within the determination period, and the complex amplitude value in the real-time echo data is marked as the interference component. The amplitude is reduced to zero for the marked interference component to obtain the main echo sequence after interference removal. A threshold for determining the homology of interference sources is set based on statistical characteristics. The threshold is set with reference to the phase noise level in a clean environment. For example, the threshold is set to 0.15 to distinguish between coherent interference and random noise. The mean square value of the phase error calculated for each cycle is compared with the threshold. If the mean square value of the error for a certain cycle is strictly less than 0.15, it indicates that the signal at the location has a very high homology with the interference source model. The data within the determination period belongs to the location of the interference component. The index address corresponding to the real echo data is locked in the computer memory, the complex amplitude value at the location is extracted and marked as the interference component, and then the amplitude is zeroed. That is, the real and imaginary values ​​in the storage unit are forcibly reset to zero, which is equivalent to a precise digital cut-off of the interference pulse in the time domain. For the region where the error value is higher than the threshold, the original data is kept unchanged. By traversing the entire data stream and executing the above stripping logic, the interference signals with periodicity and coherence are effectively removed, and the main echo sequence after interference stripping is obtained.

[0029] Please see Figure 4 The specific steps of S3 are as follows: S301: Based on the main echo sequence after interference removal, extract the Doppler spectrum corresponding to the pulse and locate the peak amplitude frequency index. Aggregate the Doppler center frequency sequence corresponding to the pulse, arrange them sequentially by pulse time index, construct a complete time-frequency trajectory curve, and obtain the Doppler center frequency change trend sequence. For each discrete pulse echo data segment, a Fast Fourier Transform (FFT) algorithm with 1024 or higher points is used to convert the time-domain signal into a frequency-domain signal. To reduce the impact of spectral leakage on detection accuracy, the data is windowed before the transformation, for example, using a Hamming window to weight the time-domain data. The magnitude of the complex data at each frequency point in the transformed spectrum is calculated, i.e., the square root of the sum of the squares of the real and imaginary parts. The amplitude data of the entire spectrum is traversed, and the frequency point with the largest amplitude is located by a numerical comparison algorithm, which is used as the peak index of the current pulse. The peak index is then calculated by dividing the set sampling rate by the variable frequency. The frequency resolution is obtained by changing the number of points. The peak index is multiplied by the frequency resolution to calculate the specific Doppler center frequency value. For example, when the sampling rate is 2 kHz and the number of transformation points is 1,000, if the peak is located at the 100th point, the corresponding Doppler frequency is 200 Hz. The processed pulse Doppler center frequency values ​​are stored in the memory buffer in the order of pulse transmission and reception. In order to eliminate frequency jitter caused by individual noise points, cubic spline interpolation or smoothing filtering algorithms are used to perform trajectory smoothing processing on the discrete frequency sequence to obtain the Doppler center frequency change trend sequence.

[0030] S302: Call the frequency difference between adjacent pulses in the Doppler center frequency change trend sequence, perform first-order difference calculation and normalization on the difference sequence, construct the frequency change slope sequence, perform sliding window operation on the frequency change slope sequence, calculate the mean square error and range of the slope change values ​​within the sliding window as the change rate index and fluctuation amplitude index of the segment, and generate the slope change rate value and fluctuation amplitude value group. A point-by-point difference operation is performed, which involves subtracting the frequency value of the previous pulse from the Doppler frequency value of the subsequent pulse to obtain the frequency change between adjacent pulses. The radar's pulse repetition interval is used as a normalization reference. The calculated frequency change is divided by the time interval to obtain the instantaneous slope of the frequency change. For example, when the frequency difference between adjacent pulses is 50 Hz and the pulse interval is 1 millisecond, the calculated frequency change slope is 50 kHz per second. A frequency change slope sequence containing the calculation results is constructed. A sliding window of length 10 is set, and the window moves gradually along the slope sequence, moving one data point at a time. Within each window... Within a given location, the statistical window covers ten slope values. The arithmetic mean of these values ​​is calculated, and the sum of the squares of the differences between each value and the mean is calculated. Dividing this sum by the number of values ​​yields the root mean square deviation, which serves as a rate indicator for measuring the stability of frequency changes. Simultaneously, the maximum and minimum slope values ​​within the window are identified, and their differences are calculated to obtain the range, which serves as a fluctuation amplitude indicator for measuring the severity of frequency fluctuations. For example, if the maximum slope within the window is 60 Hz and the minimum is 20 Hz, then the fluctuation amplitude is 40 Hz. These two statistical measures calculated for each window are paired and stored to generate a set of slope change rate values ​​and fluctuation amplitude values.

[0031] S303: Based on the slope change rate value and fluctuation amplitude value group, set the speed error critical value range, divide the frequency change trend segment into intervals, if any index in the frequency change trend segment exceeds the speed error critical value range, mark the frequency change trend segment as the prediction sub-interval, and perform linear fitting on the frequency change trend in the prediction sub-interval to generate pulse delay compensation quantization value. For the preset tolerance of target maneuver characteristics, a critical range of speed error values ​​is set, including the upper limit threshold of slope change rate and the upper limit threshold of fluctuation amplitude. For example, the mean square error threshold is set to 5 Hz squared per second, and the range threshold is 20 Hz per second. Each segment of the frequency change trend is scanned and evaluated in detail. The calculated actual speed value and fluctuation amplitude value are compared with the set thresholds respectively. If the calculated slope mean square error exceeds 5 or the slope range exceeds 20 in a certain segment, it indicates that the frequency change of the area exhibits abnormally violent fluctuations without a nonlinearity. The segment is determined to be an unstable prediction sub-interval. The original Doppler frequency data points in the sub-interval are extracted, and a univariate linear regression model is constructed using the least squares method. A straight line equation that can best fit the scatter points is found. The deviation between the theoretical frequency value and the actual measured frequency value corresponding to the fitted straight line at each time is calculated. Using the frequency modulation slope parameter of the radar signal, the frequency deviation value is converted into a time dimension delay through a linear proportional relationship. For example, the corresponding microsecond-level delay is calculated based on the frequency deviation of 50 Hz and the frequency modulation slope of 1 MHz per microsecond, and a pulse delay compensation quantization value is generated.

[0032] Please see Figure 5 The specific steps of S4 are as follows: S401: Call the pulse delay compensation quantization value, perform delay adjustment on the pulses in the velocity error prediction sub-interval in the time domain, sum the original time index and the correction amount, and reconstruct the pulse arrangement order according to the corrected index to obtain the pulse sequence set after delay compensation; For each pulse data point marked within the velocity error prediction sub-interval, the corresponding compensation time value is read. For example, if a pulse requires a 0.5 microsecond time advance compensation, the time compensation value is converted into an offset of the number of sampling points based on the radar's sampling rate parameter. Assuming the sampling rate is 10 MHz, the 0.5 microseconds are converted into an index displacement of five sampling points. Addition or subtraction is performed on the original pulse's timestamp index, and the original index value is added to the calculated sampling point offset to obtain the corrected new time index position. A temporary sorting container is created in memory, and the time-corrected pulse data and the new index identifier are placed into the container. Since the time adjustment causes slight timing discrepancies in some pulses, a quicksort algorithm is used to rearrange the pulses according to the new time index value to ensure that the pulses are strictly aligned in chronological order. The reconstructed pulse queue is traversed to check for index overlap. For overlapping positions, interpolation smoothing is used to maintain waveform continuity. The sorted data stream is then solidified to obtain the pulse sequence set after delay compensation.

[0033] S402: Based on the pulse sequence set after delay compensation, extract the peak amplitude within the pulse group by dividing it into groups. Obtain the consistency factor within the group by calculating the difference between the peak and trough peak within the group and dividing it by the peak value. Contrast index is constructed based on the difference between the edge pulse amplitude of each group and the edge pulse amplitude of the adjacent group. The contrast index and consistency factor are normalized and jointly arranged to generate a set of amplitude consistency and contrast index. A fixed grouping step size is set, for example, dividing the sequence into data units, each containing twenty consecutive pulses. For each group, the amplitude values ​​of the pulses within the group are iterated, and the maximum peak amplitude and minimum trough amplitude within the group are compared and extracted. For example, if the maximum amplitude in a group is five volts and the minimum amplitude is four volts, the difference between the two is calculated to obtain a fluctuation range of one volt. The fluctuation difference is divided by the maximum peak amplitude to calculate a consistency factor of 0.2 within the group. This factor quantifies the dispersion of pulse amplitudes within the group, locates the edge position of each pulse group, and extracts the current group. The amplitude value of the last pulse is compared with the amplitude value of the first pulse of the next adjacent group. The absolute difference between the amplitudes of these two edge pulses is calculated. For example, if the amplitude at the end of the previous group is 4.5 volts and the amplitude at the beginning of the next group is 4.6 volts, the difference is 0.1 volts. This difference is defined as the contrast index. To eliminate the influence of dimensions, the consistency factor and contrast index distribution of the statistical groups are analyzed. The maximum and minimum value normalization method is used to map the calculation results to the interval between zero and one. The normalized consistency values ​​and contrast values ​​are combined according to the corresponding group index to generate the amplitude consistency and contrast index set.

[0034] S403: Based on the amplitude consistency and contrast index set, all pulse data groups are scored and screened, and pulse data groups with joint index values ​​in the upper quantile interval are retained. The selected pulse sequences are aggregated in chronological order to establish optimized coherent pulse data groups. A weighted scoring model is constructed to comprehensively evaluate the quality of each pulse data group. The scoring logic is set to pursue high amplitude stability and smooth transition between groups. The normalized consistency factor is inverted as the stability score, and the contrast index is inverted as the coherence score. For example, if the consistency factor of a group is 0.1, the stability score is 0.9, and the contrast is 0.05, the coherence score is 0.95. The two scores are added together to obtain the total score of 1.85. The total scores of the pulse groups are sorted and statistically analyzed from high to low, and the upper quantile interval is defined as the screening criterion. For example, the pulse groups with the highest total scores are retained, and the pulse groups with scores below the quantile line are regarded as substandard data and are removed. The selected high-scoring pulse groups are extracted from the original dataset and seamlessly spliced ​​and aggregated according to their chronological order on the original time axis to establish optimized coherent pulse data groups.

[0035] Please see Figure 6 The specific steps of S5 are as follows: S501: Based on the optimized coherent pulse data group, perform Fourier transform operation, extract the amplitude response curve in the spectral domain, locate the frequency index corresponding to the amplitude as the main peak position, and calculate the frequency interval value between adjacent main peaks to obtain the target echo main peak position and spacing sequence. To effectively suppress sidelobe interference during spectral analysis, a weighted window function is applied to the data sequence, such as a Blackman window for smooth weighting. A high-precision Fast Fourier Transform (FFT) algorithm is then used, with 4096 transform points set to ensure sufficient frequency resolution for detecting minute frequency offsets. The discrete-time signal is converted into a complex-frequency sequence. The real and imaginary parts of each frequency point are squared and summed, and the sum is then square-rooted. Point-by-point, an amplitude response curve reflecting the signal energy distribution characteristics is constructed in the spectral domain. A detection threshold based on the statistical characteristics of background noise is set on the amplitude curve, and the entire frequency band is traversed. The range is determined by finding local maxima where the amplitude not only exceeds the threshold but is also significantly higher than the amplitudes of the adjacent frequency points. These local maxima are identified as the main peak positions of the target echo, and the corresponding frequency axis indices are recorded. For example, in a single scan, three strong signal points with frequency indices of 500, 650, and 800 are locked. The main peaks are arranged in ascending order of frequency, and the actual frequency differences between adjacent main peak frequency indices are calculated sequentially. For example, if the first interval is calculated to be 150 Hz, the second interval is also 150 Hz. The determined main peak frequency position information is associated with and stored with the calculated adjacent interval values ​​to obtain the target echo main peak position and spacing sequence.

[0036] S502: Call the target echo main peak position and spacing sequence, perform histogram distribution statistics, extract the main mode interval of the frequency spacing distribution, and calculate the weighted average of the median of the main mode interval and the mean of the whole sequence as the dynamic distance reference value. Combine the variance value of the distribution curve to perform exponential adjustment and generate the dynamic spacing judgment threshold. The statistical interval width of the histogram is set to 5 Hz to cover the frequency distribution range. Frequency interval values ​​in the sequence are traversed, mapped one by one, and accumulated into the corresponding histogram statistical bins. After statistical analysis, the bin containing the largest number of samples is identified as the principal mode interval. For example, if analysis reveals that most interval values ​​are concentrated in the 150 Hz to 155 Hz interval, the center frequency value of the principal mode interval is extracted as the representative mode. Simultaneously, the interval values ​​in the sequence are summed and divided by the total number of values ​​to calculate the arithmetic mean. A weighted fusion strategy is used, assigning a high weight to the mode (e.g., 0.8) and a low weight to the mean (e.g., 0.2). A basic dynamic distance reference value is obtained through weighted calculation. The dispersion of the interval values ​​relative to the reference value, i.e., the variance, is calculated. An exponential decay adjustment logic based on variance is constructed. When the variance is large, it indicates an unstable distribution, and a small decay coefficient is automatically generated. This decay coefficient is used to correct the basic reference value downwards, for example, tightening the 152 Hz reference value to 130 Hz as a strict standard for distinguishing independent targets from aliased targets, generating a dynamic distance judgment threshold.

[0037] S503: Compare the spacing between adjacent main peaks with the dynamic spacing judgment threshold item by item. If the spacing between main peaks is less than the dynamic spacing judgment threshold, assign a spectrum decoupling weight according to the reciprocal of the corresponding spacing. Perform multi-channel splitting processing on the spectrum amplitude curve according to the spectrum decoupling weight, locate the decoupling target frequency band range respectively, extract the frequency coordinate value of the corresponding main peak and mark it as the target frequency position, and call the target frequency coordinate and the spectrum amplitude data corresponding to the two frequency points before and after to perform local extreme value search, lock the frequency band range where the target peak point is located, and obtain the target spectrum decoupling positioning data. During the spectrum scanning process, each recorded frequency interval value is checked one by one. If the frequency difference between two adjacent main peaks is found to be less than the set threshold of 130 Hz, for example, the actual measured interval is only 40 Hz, the system determines that there is a spectrum aliasing or strong coupling between the two main peaks, and immediately starts the decoupling process. The reciprocal of the actual interval value is calculated and mapped to the spectrum decoupling weight. The smaller the interval, the greater the separation weight. Based on the weight, the parameters of the multi-channel filter are configured, and multi-channel splitting processing is performed on the original spectrum amplitude curve. In each independent channel, the energy of adjacent interference peaks is suppressed by bandpass filtering. The amplitude maxima are searched again in the decoupled independent frequency bands to locate the center frequency coordinates of the separated target and mark them as the confirmed target frequency position. In order to improve the positioning accuracy, the center frequency coordinates and the amplitude data of the two adjacent frequency points are read, and a local three-point parabolic interpolation operation is performed. The extreme points of the fitted curve are used to estimate the true simulated peak frequency, lock the precise frequency band range where the target energy is most concentrated, and obtain the target spectrum decoupling positioning data.

[0038] Please see Figure 7 A radar antenna signal processing system, comprising: The echo acquisition module acquires the time-domain echo signal from the radar antenna, calls the sampling time point sequence, antenna scanning angle sequence, and received signal amplitude sequence, performs energy normalization processing on multiple Doppler channels in each signal segment, extracts the normalized energy distribution curve, and generates a set of periodic interference characteristic parameters. The interference stripping module matches the sampling time sequence of real-time echo data with the corresponding echo phase data based on the periodic interference characteristic parameter set. It calls the interference phase reference sequence and the real-time echo phase difference sequence, performs point-by-point difference calculation on the phase difference, and determines whether it is within the phase error limit threshold to obtain the main echo sequence after interference stripping. The frequency trend analysis module extracts the Doppler center frequency value under the corresponding time series based on the main echo sequence after interference stripping, calls the frequency time distribution curve and calculates the rate of change of derivative and the frequency change range of each segment in segments. It normalizes the rate of change of slope and fluctuation amplitude in continuous segments to obtain the pulse delay compensation quantization value. The delay compensation module calls the pulse delay compensation quantization value, performs time delay adjustment operation on the echo pulses in the sub-interval, obtains the start time of the sub-interval, the echo pulse sampling period and the center time of adjacent pulses, and calculates the difference between the center time of each group of pulses and the delayed time series to establish an optimized coherent pulse data group. The spectrum decoupling module performs spectrum transformation on each optimized pulse sequence based on the optimized coherent pulse data group, extracts the spacing data between the main spectrum peak position sequence and adjacent peaks, calls the adjacent main peak position difference sequence and statistically analyzes the distribution range and frequency, and obtains the target spectrum decoupling positioning data.

[0039] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A radar antenna signal processing method, characterized in that, Includes the following steps: S1: Acquire the time-domain echo signal of the radar antenna, extract the coherent energy distribution curve of the Doppler channel, calculate the period length, peak energy amplitude, initial phase and attenuation curvature, and establish a set of periodic interference characteristic parameters. S2: Based on the set of periodic interference characteristic parameters, perform periodic back-projection matching operation on the real-time echo data, calculate the phase error value to determine the homogeneity of the interference source, remove the interference signal components, and obtain the main echo sequence after interference stripping. S3: Using the main echo sequence after interference stripping, extract the Doppler center frequency change trend, calculate the slope change rate value and interval fluctuation amplitude value, divide the velocity error prediction sub-interval, and generate pulse delay compensation quantization value. S4: Call the pulse delay compensation quantization value, perform time delay compensation operation on the sub-interval pulse, calculate the echo amplitude peak consistency value within the pulse group and the interference boundary contrast value between groups, and establish an optimized coherent pulse data group.

2. The radar antenna signal processing method according to claim 1, characterized in that, The set of periodic interference characteristic parameters includes interference energy distribution type, periodic structure index, phase characteristic parameters, and energy attenuation law parameters. The main echo sequence includes target reflection component, interference residual intensity indicator, and time-domain waveform reconstruction index. The pulse delay compensation quantization value includes frequency drift correction value, velocity disturbance correction factor, and prediction error limit value within sub-interval. The optimized coherent pulse data group includes time alignment consistency index, amplitude fluctuation suppression characteristics, and inter-group interference separation intensity indicator.

3. The radar antenna signal processing method according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Acquire radar antenna time-domain echo signal, perform frame structure identification, slice the pulse sequence in each frame of data, extract the signal peak index sequence based on the phase continuity and amplitude stability of a single pulse waveform, detect the periodic trend of the repetition interval based on the time interval sequence of adjacent pulses, and generate a set of period length values ​​by statistically classifying and frequency detecting the values ​​of the repetition interval. S102: Based on the set of period length values, locate the pulse index position where the period start point of the repetition frequency is located, call the complex amplitude data of the corresponding pulse in the original radar antenna time domain echo signal, aggregate the amplitude magnitude of the pulse within the period and perform envelope curve fitting to generate a group of energy peak amplitude values. S103: Call the initial phase information of the complex amplitude corresponding to the starting position of the cycle in the energy peak amplitude value group, and calculate the cosine phase difference sequence of the phase sequence at the starting point of each cycle in combination with the cycle length value. Retrieve the rate of change of the starting direction of the phase difference change trend curve within the cycle, and establish a set of periodic interference characteristic parameters by regression fitting the mean of the rate of change of the starting direction on the cycle.

4. The radar antenna signal processing method according to claim 3, characterized in that, The specific steps of S2 are as follows: S201: Based on the set of periodic interference characteristic parameters, construct a periodic waveform template sequence, call the continuous data segment with the same period length in the real-time echo data, perform point-by-point complex multiplication operation between the template waveform and the data segment, extract the real part to form a matching degree sequence, calculate the dot product ratio between the matching degree sequence at the period position and the template waveform, and obtain the periodic back projection matching factor column. S202: Call the periodic back projection matching factor column, combine the initial phase value and the attenuation curvature value, construct the desired phase sequence curve, normalize the amplitude based on the difference sequence between the complex phase sequence and the desired phase sequence at each period position in the echo data, calculate the mean square value of the normalized phase difference as the phase error value at each period position, and generate the phase error evaluation result. S203: Call the phase error evaluation result, compare the mean square error value with the set interference source same source determination threshold item by item. If the mean square error value is less than the interference source same source determination threshold, the position of the same source interference component is determined within the determination period, and the complex amplitude value in the real-time echo data is marked as the interference component. The amplitude is reduced to zero for the marked interference component to obtain the main echo sequence after interference removal.

5. The radar antenna signal processing method according to claim 4, characterized in that, The specific steps for S3 are as follows: S301: Based on the main echo sequence after interference stripping, extract the Doppler spectrum corresponding to the pulse and locate the peak amplitude frequency point index, aggregate the Doppler center frequency sequence corresponding to the pulse, arrange them sequentially by the pulse time index, construct a complete time-frequency trajectory curve, and obtain the Doppler center frequency change trend sequence. S302: Call the frequency difference between adjacent pulses in the Doppler center frequency change trend sequence, perform first-order difference calculation and normalization on the difference sequence, construct the frequency change slope sequence, perform sliding window operation on the frequency change slope sequence, calculate the mean square error and range of the slope change values ​​within the sliding window as the change rate index and fluctuation amplitude index of the segment, and generate a group of slope change rate values ​​and fluctuation amplitude values. S303: Based on the slope change rate value and fluctuation amplitude value group, set the speed error critical value range, divide the frequency change trend segment into intervals, and if any index in the frequency change trend segment exceeds the speed error critical value range, mark the frequency change trend segment as a prediction sub-interval, and perform linear fitting on the frequency change trend in the prediction sub-interval to generate pulse delay compensation quantization value.

6. The radar antenna signal processing method according to claim 5, characterized in that, The specific steps of S4 are as follows: S401: Call the pulse delay compensation quantization value, perform delay adjustment on the pulses in the velocity error prediction sub-interval in the time domain, sum the original time index and the correction amount, and reconstruct the pulse arrangement order according to the corrected index to obtain the pulse sequence set after delay compensation; S402: Based on the pulse sequence set after delay compensation, extract the peak amplitude within the pulse group by dividing it into groups. Obtain the consistency factor within the group by calculating the difference between the peak and trough peak within the group and dividing it by the peak value. Contrast index is constructed based on the difference between the edge pulse amplitude of each group and the edge pulse amplitude of the adjacent group. The contrast index and consistency factor are normalized and jointly arranged to generate a set of amplitude consistency and contrast index. S403: Based on the amplitude consistency and contrast index set, score and filter all pulse data groups, retain pulse data groups whose joint index values ​​are in the upper quantile interval, and aggregate the selected pulse sequences in chronological order to establish optimized coherent pulse data groups.

7. The radar antenna signal processing method according to claim 6, characterized in that, The process of scoring and filtering all pulse data groups based on the amplitude consistency and contrast index set, and retaining pulse data groups whose joint index values ​​are located in the upper quantile interval, specifically involves: presetting consistency weight values ​​and contrast weight values, and limiting the sum of the consistency weight values ​​and the contrast weight values ​​to one. Extract the normalized intra-group consistency factor and the contrast index from the set of amplitude consistency and contrast indexes; perform a multiplication operation on the intra-group consistency factor and the consistency weight value to obtain the consistency component value, and perform a multiplication operation on the contrast index and the contrast weight value to obtain the contrast component value; The consistency sub-value and the contrast sub-value are summed to generate the score value corresponding to each pulse data group; Perform a descending sort operation on the score values ​​of the pulse data group, and calculate the total number of values ​​in the sorted score value sequence; Multiply the total number of values ​​by a preset preferred ratio coefficient, and round the result down to obtain the truncation position index value; The set of values ​​in the scoring sequence from the first index to the truncation position index is defined as the upper quantile interval; Pulse data sets whose score values ​​fall within the upper quantile interval are determined to be retained, while pulse data sets whose score values ​​do not fall within the upper quantile interval are discarded.

8. The radar antenna signal processing method according to claim 1, characterized in that, The method further includes step S5: S5: Based on the optimized coherent pulse data group, perform a spectrum transformation operation to extract the target echo main peak position value and spacing value. Combine the statistical distribution characteristics of the main peak spacing to calculate the dynamic spacing judgment threshold. Compare the main peak spacing value with the dynamic spacing judgment threshold. When the main peak spacing value is less than the dynamic spacing judgment threshold, perform a spectrum splitting and decoupling operation according to the main peak spacing weight. Calculate the energy focusing degree value and frequency offset value of the decoupled target. If the main peak spacing value is greater than the dynamic spacing judgment threshold, lock the target frequency coordinates and obtain the target spectrum decoupling positioning data. The target spectrum decoupling and positioning data includes the position of the main peak in the frequency domain, the identifier of the decoupling target frequency, and the quantification value of the degree of spectrum separation and focusing.

9. The radar antenna signal processing method according to claim 8, characterized in that, The specific steps of S5 are as follows: S501: Based on the optimized coherent pulse data group, perform Fourier transform operation, extract the amplitude response curve in the frequency domain, locate the frequency index corresponding to the amplitude as the main peak position, and calculate the frequency interval value between adjacent main peaks to obtain the target echo main peak position and spacing sequence. S502: Call the target echo main peak position and spacing sequence, perform histogram distribution statistics, extract the main mode interval of frequency spacing distribution, and calculate the weighted average of the median of the main mode interval and the mean of the whole sequence as the dynamic distance reference value. Combine the variance value of the distribution curve to perform exponential adjustment and generate the dynamic spacing judgment threshold. S503: Compare the spacing between adjacent main peaks with the dynamic spacing judgment threshold item by item. If the spacing between main peaks is less than the dynamic spacing judgment threshold, assign a spectrum decoupling weight according to the reciprocal of the corresponding spacing. Perform multi-channel splitting processing on the spectrum amplitude curve according to the spectrum decoupling weight, locate the decoupling target frequency band range respectively, extract the frequency coordinate value of the corresponding main peak and mark it as the target frequency position, and call the target frequency coordinate and the spectrum amplitude data corresponding to the two frequency points before and after to perform local extreme value search, lock the frequency band range where the target peak point is located, and obtain the target spectrum decoupling positioning data.

10. A radar antenna signal processing system, characterized in that, The system is used to implement the radar antenna signal processing method according to any one of claims 1-9, the system comprising: The echo acquisition module acquires the time-domain echo signal from the radar antenna, calls the sampling time point sequence, antenna scanning angle sequence, and received signal amplitude sequence, performs energy normalization processing on multiple Doppler channels in each signal segment, extracts the normalized energy distribution curve, and generates a set of periodic interference characteristic parameters. The interference stripping module matches the sampling time sequence of the real-time echo data with the corresponding echo phase data according to the set of periodic interference characteristic parameters. It calls the interference phase reference sequence and the real-time echo phase difference sequence, performs point-by-point difference calculation on the phase difference and determines whether it is within the phase error limit threshold, and obtains the main echo sequence after interference stripping. Based on the main echo sequence after interference stripping, the frequency trend analysis module extracts the Doppler center frequency value under the corresponding time series, calls the frequency time distribution curve and calculates the rate of change of derivative and the frequency change range of each segment in segments. It normalizes the rate of change of slope and fluctuation amplitude in continuous segments to obtain the pulse delay compensation quantization value. The delay compensation module calls the pulse delay compensation quantization value, performs time delay adjustment operation on the echo pulse in the sub-interval, obtains the start time of the sub-interval, the echo pulse sampling period and the center time of adjacent pulses, and calculates the difference between the center time of each group of pulses and the delayed time series to establish an optimized coherent pulse data group. The spectrum decoupling module performs a spectrum transformation operation on each optimized pulse sequence based on the optimized coherent pulse data group, extracts the spacing data between the main spectrum peak position sequence and adjacent peaks, calls the adjacent main peak position difference sequence and statistically analyzes the distribution range and frequency, and obtains the target spectrum decoupling positioning data.

Images