CN121325112B

CN121325112B - Radar dynamic anti-interference method and system based on interference source positioning

Info

Publication number: CN121325112B
Application number: CN202511891962.5A
Authority: CN
Inventors: 杨凯; 刘宝明; 张君; 高军超; 张宇; 张�杰
Original assignee: Galileo Tianjin Technology Co ltd
Current assignee: Galileo Tianjin Technology Co ltd
Priority date: 2025-12-16
Filing date: 2025-12-16
Publication date: 2026-02-27
Anticipated expiration: 2045-12-16
Also published as: CN121325112A

Abstract

The application relates to the technical field of information and discloses a radar dynamic anti-interference method and system based on interference source positioning. The method comprises the steps of obtaining current signal data and a historical signal sequence, determining an initial deviation value, extracting the historical deviation sequence according to the initial deviation value to obtain a deviation change trend vector, calculating trend slope and fluctuation amplitude, determining an adjustment trigger signal, determining a weight coefficient update increment according to the interaction effect of the extraction characteristic of the adjustment trigger signal, carrying out iterative optimization by combining convergence speed parameters to obtain an optimized convergence speed value, adjusting deviation compensation model parameters according to the convergence speed value, outputting a deviation compensation result when a compensation residual is lower than a preset threshold value, carrying out positioning calculation according to the deviation compensation result, and determining corrected positioning coordinates. According to the application, the model is dynamically optimized through the sliding window, gradient descent and other technologies, the problem of positioning deviation caused by complex environment signal interference is solved, and the positioning precision and response speed are improved.

Description

Radar dynamic anti-interference method and system based on interference source positioning

Technical Field

The application relates to the technical field of information, in particular to a radar dynamic anti-interference method and system based on interference source positioning.

Background

The radar is used as core sensing equipment in the fields of navigation, national defense, traffic and the like, and the positioning accuracy directly determines the system efficiency. However, in a complex environment, multipath effects, obstruction, non-line-of-sight propagation and dynamic interference source switching are liable to cause unpredictable fluctuation of the intensity, frequency and modulation mode of a radar receiving signal, so as to cause positioning deviation, seriously influence the reliable operation of the radar in the complex environment, and a high-efficiency dynamic anti-interference technology is needed to solve the core problem.

The prior radar anti-interference technology has obvious defects in coping with complex dynamic interference. Firstly, most technologies rely on a single static deviation compensation model, so that the space-time dynamic change of signal characteristics cannot be adapted, for example, when signal mutation is caused by reflection of high-rise buildings in urban canyons or deviation modes are changed due to attenuation of indoor scene walls, the static model is difficult to adjust a compensation strategy in real time, and the positioning accuracy is low. Secondly, the feature interaction processing capability is insufficient, namely the influence of interaction among signal strength, frequency and modulation modes on deviation is obvious, but the prior art lacks accurate analysis on the feature interaction, the updating of the weight coefficient is lagged, the dynamic contribution of each feature to the deviation can not be reflected in time, and the compensation residual error is aggravated. The model optimization and the real-time performance are difficult to balance, and although a sliding window or an iterative optimization algorithm is adopted in part of technologies, the window size is fixed, the signal change frequency cannot be matched, and when algorithms such as gradient descent and the like process high-dimensional characteristics, the problems of slow convergence or high calculation complexity are easy to occur, and the response delay during signal mutation is difficult to meet the dual requirements of radar on the real-time performance and low residual error. In addition, the prior art does not establish a cooperative mechanism of interference source positioning and anti-interference, only simply compensates deviation without associating interference source dynamics, and the anti-interference robustness is insufficient.

Aiming at the defects, the method dynamically adapts the interference change in a mode of signal data acquisition, deviation trend analysis, triggering dynamic adjustment, weight iterative optimization, model parameter adaptation and positioning coordinate correction, and improves the anti-interference capability and positioning accuracy of the radar.

Disclosure of Invention

The application relates to the technical field of information and discloses a radar dynamic anti-interference method and system based on interference source positioning. The model is dynamically optimized through the sliding window, gradient descent and other technologies, so that the problem of positioning deviation caused by complex environment signal interference is solved, and the positioning precision and response speed are improved.

In a first aspect, the present application provides a radar dynamic anti-interference method based on interference source positioning, the method comprising:

Step S101, acquiring current signal data and a historical signal sequence, extracting characteristic parameters of signal intensity, frequency and modulation mode, and calculating an initial deviation value according to the characteristic parameters of the signal intensity, frequency and modulation mode;

Step S102, extracting a historical deviation sequence from the initial deviation value based on a dynamic sliding window, and performing time sequence analysis on the historical deviation sequence to generate a deviation change trend vector;

step S103, calculating trend slope and fluctuation amplitude by using the deviation change trend vector, and determining an adjustment trigger signal according to the relationship between the trend slope and the fluctuation amplitude and a preset threshold value;

Step S104, extracting related feature interaction influence from the deviation change trend vector according to the adjustment trigger signal, determining the correlation weight among the features, and calculating a weight coefficient update increment according to the correlation weight;

Step S105, updating the convergence speed parameter according to the weight coefficient updating increment, and optimizing the updated convergence speed parameter through a gradient descent iteration optimization algorithm to obtain an optimized convergence speed value;

Step S106, adjusting parameters of the deviation compensation model according to the optimized convergence speed value, calculating a compensation residual error of the adjusted deviation compensation model, and outputting a deviation compensation result if the compensation residual error is lower than a preset residual error threshold;

And S107, adjusting the weight of the input parameters of the positioning algorithm according to the deviation compensation result, and carrying out weighted calculation on the adjusted input parameters through the weighted positioning algorithm to determine the corrected positioning coordinates.

Optionally, the step S101 includes:

Acquiring current interference signal data and a historical signal sequence in real time through radar signal receiving equipment, and calling the historical signal sequence in a preset time length matched with a current radar working mode from a pre-established signal database;

Calculating initial deviation values according to the characteristic parameters of the current interference signals and the historical signal sequences, wherein the initial deviation values comprise intensity deviation components, frequency deviation components and modulation mode deviation components of the interference signals and the historical signals respectively;

And determining a weight coefficient of the deviation component according to the variance of the historical signal sequence, obtaining an initial deviation value through weighted summation calculation, and storing the initial deviation value in association with a calculation time stamp, a radar working mode and an interference scene label.

Optionally, the step S102 includes:

a sliding window technology is applied, a historical deviation sequence is extracted from a prestored deviation database, and the window size of the sliding window technology is dynamically adjusted according to the signal change frequency;

Performing time sequence analysis on the historical deviation sequence to generate a deviation change trend vector;

and carrying out normalization processing on the deviation change trend vector to obtain a standardized deviation change trend vector.

Optionally, the step S103 includes:

Extracting a time domain linear regression slope component in the deviation change trend vector, and determining the time domain linear regression slope component as a trend slope in a time window;

Calculating the fluctuation amplitude of the time domain and frequency domain characteristic components of the deviation change trend vector by adopting a standard deviation algorithm;

And if the trend slope exceeds a preset slope threshold and the fluctuation amplitude meets an abnormality judgment condition, marking the trend slope as a mutation mode and generating an adjustment trigger signal.

Optionally, the step S104 includes:

analyzing and extracting an interference scene label, a time stamp, a trend slope and a fluctuation amplitude according to the adjustment trigger signal;

Matching scene data sets in the historical trend database according to the interference scene tags, and extracting characteristic interaction data from the scene data sets according to the time stamp, the trend slope and the fluctuation amplitude;

analyzing the association degree among all the features in the feature interaction data set by adopting a pearson correlation coefficient matrix, and determining the correlation weights of three groups of feature interactions of signal strength-frequency, signal strength-modulation mode and frequency-modulation mode;

And obtaining a basic weight coefficient of the current deviation compensation model, and calculating a weight coefficient update increment according to the basic weight coefficient, the correlation weight and the dynamic adjustment coefficient.

Optionally, the step S105 includes:

Acquiring current convergence speed parameters from a parameter cache region of a radar deviation compensation model, wherein the parameter cache region comprises the convergence speed parameters, an updating time stamp and corresponding interference scene labels;

The update increment of the weight coefficient is respectively weighted and summed with an intensity influence coefficient, a frequency influence coefficient and a modulation influence coefficient to calculate the update quantity of the convergence speed parameter, the update quantity of the convergence speed parameter is overlapped with the current convergence speed parameter, and the updated convergence speed parameter is obtained, wherein the intensity influence coefficient, the frequency influence coefficient and the modulation influence coefficient are determined by historical data statistics;

and adopting a gradient descent method to iteratively optimize the updated convergence speed parameter.

Optionally, the step S106 includes:

calling initial parameters of a current deviation compensation model, wherein the initial parameters comprise a deviation compensation coefficient and a compensation offset;

According to the optimized convergence speed value, the deviation compensation coefficient and the compensation offset are adjusted, so that the adjusted parameters are adapted to the convergence speed of the deviation compensation model;

Inputting characteristic parameters of a current signal into an adjusted deviation compensation model, outputting a real-time compensation value after moving average filtering treatment, retrieving an ideal compensation value matched with the current interference scene and signal characteristics, and taking the absolute value error of the real-time compensation value and the ideal compensation value as a compensation residual error;

If the compensation residual is lower than the residual threshold, outputting a deviation compensation result containing the filtered real-time compensation value, the new compensation coefficient, the new compensation offset, the time stamp and the scene tag, and if the compensation residual is higher than the residual threshold, returning to the step S105 to iterate and optimize the convergence speed parameter again.

Optionally, the step S107 includes:

The deviation compensation result is disassembled into an x-direction component and a y-direction component which correspond to the original positioning coordinates of the radar;

Adjusting input parameters of a positioning algorithm according to the x-direction component and the y-direction component, wherein the input parameters of the positioning algorithm comprise a distance estimated value corresponding to signal strength, a phase offset value corresponding to frequency and a positioning error correction value corresponding to a modulation mode;

And (3) adopting a weighted positioning method, multiplying each adjusted input parameter by the corresponding adjusted weight respectively, and summing the products to obtain the corrected coordinate component.

In a second aspect, the present application provides a radar dynamic anti-interference system based on interference source positioning, the system comprising:

The signal acquisition module is used for acquiring current signal data and a historical signal sequence, calculating and determining an initial deviation value, and guaranteeing the instantaneity and accuracy of data acquisition;

the deviation analysis module extracts a historical deviation sequence according to the initial deviation value, generates a deviation change trend vector through processing, and intuitively reflects the change characteristics of the deviation along with time;

the trigger judging module is used for calculating a trend slope and a fluctuation amplitude by adopting the deviation change trend vector, and immediately determining and generating an adjustment trigger signal if the slope exceeds a preset threshold value, and starting a subsequent compensation adjustment flow;

The weight optimization module extracts characteristic interaction effects according to the adjustment trigger signals, analyzes characteristic correlation to determine correlation weights, calculates weight coefficient update increments, and provides basis for model parameter update;

The convergence optimization module is used for acquiring a current convergence speed parameter, updating the parameter by combining with a weight coefficient updating increment, optimizing through an iterative optimization algorithm to obtain an optimized convergence speed value, and improving the response speed of the model;

The compensation output module is used for adjusting deviation compensation model parameters according to the optimized convergence speed value, calculating a compensation residual error, and outputting a deviation compensation result for subsequent positioning correction if the residual error is lower than a preset threshold value;

And the positioning correction module is used for integrating the deviation compensation result into a positioning calculation flow, adjusting the input parameters of a positioning algorithm, calculating and determining corrected positioning coordinates through the positioning algorithm, and realizing accurate positioning.

In a third aspect, the application provides a radar dynamic anti-interference device based on interference source positioning, which comprises a memory and a processor, wherein the memory stores a computer program which can be run on the processor, and the processor realizes the radar dynamic anti-interference method based on the interference source positioning when executing the computer program.

The application provides a radar dynamic anti-interference method and system based on interference source positioning, which are suitable for an anti-interference positioning scene of a radar in a complex environment, and can solve the problems of poor static model suitability, insufficient feature interaction processing, unbalanced model optimization and real-time performance and weak anti-interference robustness in the prior art, and compared with the prior art, the beneficial effects of the technical scheme of the application are at least as follows:

Firstly, an initial deviation value is determined by acquiring current signal data and a historical signal sequence, a historical deviation sequence is dynamically extracted by combining a sliding window technology, a deviation change trend vector is generated, the limitation of a traditional static deviation compensation model is broken through, analysis dimensionality can be adaptively adjusted according to signal change frequency, space-time dynamic change of signal characteristics in a complex environment is adapted, and the problem of sudden drop of positioning accuracy when a signal mutation or a deviation mode is changed is avoided.

Secondly, based on the characteristic interaction influence among the signal intensity, the frequency and the modulation mode of the adjustment triggering signal extraction, the update increment of the weight coefficient is determined through characteristic correlation analysis, the defect of the characteristic interaction processing capacity in the prior art is overcome, the dynamic contribution of each characteristic to the deviation can be reflected by the weight coefficient in real time, the compensation residual error caused by weight update hysteresis is reduced, and the accuracy of deviation compensation is improved.

Thirdly, combining the weight coefficient updating increment and the convergence speed parameter, adopting a gradient descent iterative optimization algorithm to obtain an optimized convergence speed value, ensuring scientificity of model parameter adjustment, and dynamically optimizing balance calculation complexity and convergence efficiency, avoiding the problem of slow convergence or response delay when the traditional algorithm processes high-dimensional characteristics, and meeting the dual requirements of radar on positioning instantaneity and low residual error.

Fourthly, the deviation compensation result is integrated into a positioning calculation flow, the corrected positioning coordinates are determined through a weighted positioning algorithm, a cooperative mechanism of positioning and anti-interference of the interference source is established, the deviation is not compensated only, but the dynamic characteristics of the associated interference source are optimized, the anti-interference robustness of the radar in complex scenes such as multi-interference source switching and multipath effect is enhanced, and the stable and reliable operation of the radar system is ensured.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings may be obtained based on these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flow chart of a radar dynamic anti-interference method based on interference source positioning of the present application;

FIG. 2 is a diagram of a radar dynamic anti-interference processing procedure based on interference source positioning according to the present application;

FIG. 3 is a schematic diagram of a radar dynamic anti-interference system based on interference source positioning according to the present application;

fig. 4 is a schematic structural diagram of a radar dynamic anti-interference device based on interference source positioning according to the present application.

Detailed Description

The embodiment of the application provides a radar dynamic anti-interference method and system based on interference source positioning. The terms "first," "second," "third," "fourth" and the like in the description and in the claims and in the above drawings, if any, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments described herein may be implemented in other sequences than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or apparatus.

The application relates to the technical field of information and discloses a radar dynamic anti-interference method and system based on interference source positioning. The method comprises the steps of obtaining current signal data and a historical signal sequence, determining an initial deviation value, extracting the historical deviation sequence according to the initial deviation value to obtain a deviation change trend vector, calculating trend slope and fluctuation amplitude, determining an adjustment trigger signal if the trend slope exceeds a preset threshold, determining a weight coefficient update increment according to the interaction influence of the extraction characteristic of the adjustment trigger signal, carrying out iterative optimization by combining a convergence speed parameter to obtain an optimized convergence speed value, adjusting a deviation compensation model parameter according to the convergence speed value, outputting a deviation compensation result when a compensation residual is lower than the preset threshold, carrying out positioning calculation according to the deviation compensation result, and determining corrected positioning coordinates. According to the application, the model is dynamically optimized through the sliding window, gradient descent and other technologies, the problem of positioning deviation caused by complex environment signal interference is solved, and the positioning precision and response speed are improved.

For ease of understanding, a specific flow of an embodiment of the present application is described below, referring to fig. 1, and an embodiment of a radar dynamic anti-interference method based on interference source positioning in the embodiment of the present application includes:

Step S101, current signal data and a historical signal sequence are obtained, signal intensity, frequency and modulation mode characteristic parameters are extracted, and initial deviation values are calculated according to the signal intensity, frequency and modulation mode characteristic parameters.

In a specific embodiment, the process of executing step S101 may specifically include the following steps:

Specifically, a signal receiving device (such as a high-sensitivity antenna module, which is adapted to a radar working frequency band, such as an X-band and an S-band) carried by the radar system is started, an interference signal in a current environment is sampled in real time, the sampling frequency needs to be matched with the radar working frequency (such as a sampling rate of 100MHz commonly used in the X-band radar), the acquired current interference signal data comprises a time domain waveform (such as voltage amplitude data of a continuous time sequence) and a frequency spectrum characteristic (such as power distribution of the signal at different frequency points), and the method can be directly used for subsequent characteristic extraction. And simultaneously, a historical signal sequence matched with the current radar working mode (such as a search mode and a tracking mode) is called from a pre-established signal database, the signal database stores signal records under different past interference scenes (such as electromagnetic interference, multipath reflection interference and dynamic interference source switching scenes), each record comprises a time stamp, a signal type label (such as urban canyon multipath interference and airport electromagnetic interference) and corresponding signal strength, frequency and modulation mode parameters, and a time stamp filtering mechanism is used for selecting a historical signal sequence within the last 30 minutes to construct a comparison benchmark when the historical signal sequence is called, so that timeliness of historical data is ensured.

Secondly, calculating a received signal strength indication through a power detection module arranged in signal receiving equipment, quantifying the received signal strength indication into dBm units (the value range is usually-40 dBm to-100 dBm, such as the strength indication of a current interference signal is-68 dBm), applying a fast Fourier transform algorithm to the time domain waveform of the current interference signal data according to frequency, converting the time domain signal into a frequency domain signal, identifying and extracting a main frequency component in a frequency domain map through a peak detection algorithm, adopting a digital filtering technology to eliminate out-of-band noise in the extraction process, improving the resolution of frequency parameters through a cubic spline interpolation algorithm, and identifying a modulation mode (such as ASK, FSK, QPSK or FM and AM) adopted by the interference signal according to a modulation mode through a demodulation circuit of a radar system by combining a maximum likelihood estimation algorithm. Through the three extracted characteristic parameters, the problem of deviation evaluation distortion caused by incomplete characteristic extraction and insufficient precision is solved.

Further, a multidimensional difference quantization method is adopted to respectively calculate deviation components of three types of characteristic parameters, wherein the signal intensity deviation components are algebraic differences of the current signal intensity value and average signal intensity in the same interference scene in a historical signal sequence, if the current intensity is indicated as-68 dBm, the average intensity of the 'urban multipath interference' scene in the historical signal sequence is indicated as-72 dBm, the signal intensity deviation components are 4dBm, the frequency deviation components are calculated by adopting Euclidean distance algorithm, namely, the currently extracted main frequency is different from the frequency average value of the historical signal sequence, if the current main frequency is 10.03GHz, the frequency deviation components are 0.03GHz, the modulation mode deviation components are endowed with qualitative values according to the mode matching degree, if the current modulation mode is consistent with the modulation mode of the historical signal sequence, the components are 0, and if the current modulation mode is inconsistent with the modulation mode, the components are endowed with preset qualitative deviation values. Then, the weight coefficient of each deviation component is determined according to the statistical variance of the historical signal sequence, if the variance of the signal intensity in the historical data is larger, the weight of the signal intensity deviation component is set to 0.5, the weight of the frequency deviation component is set to 0.3, the weight of the modulation mode deviation component is set to 0.2, and the initial deviation value is obtained through calculation of a weighted summation formula (initial deviation value = signal intensity deviation component x intensity weight + frequency deviation component x frequency deviation component + modulation mode deviation component x modulation weight).

And finally, storing the calculated initial deviation value into a local cache of a radar system or an associated deviation database, wherein the storage format comprises a calculation time stamp (such as '2028-08-21-15:48:55'), a current radar working mode (such as 'tracking mode') and an interference scene tag (such as 'urban multipath interference'), ring buffer management data are adopted in the storage process, and the newly generated initial deviation value automatically covers expiration data exceeding the preset storage duration in the database, so that the storage efficiency is maintained. Meanwhile, an index optimization technology (such as a composite index based on an interference scene tag and a time stamp) is adopted in a database query link, so that the adjustment of a historical deviation sequence by a sliding window technology in a subsequent step is accelerated, and the problem of low subsequent processing efficiency caused by data dispersion and storage disorder is solved.

And S102, extracting a historical deviation sequence from the initial deviation value based on a dynamic sliding window, and carrying out time sequence analysis on the historical deviation sequence to generate a deviation change trend vector.

In a specific embodiment, the process of executing step S102 may specifically include the following steps:

Specifically, when a sliding window technology is adopted and a historical deviation sequence of a latest time period is extracted from a prestored deviation database, a time interval is set forward by taking the current time corresponding to an initial deviation value as a reference, and a sliding window is started. The initial size of the sliding window is set to 10 seconds of data points, the number of data points in the window is dynamically adjusted according to the radar signal sampling rate, and if the radar sampling rate is 100MHz (i.e. 100×10 ⁶ signal data points are acquired per second), the 10 seconds window contains 1000 data points. The window boundary processing adopts a mirror image extension method to avoid data interception, the extension length is set to be 5% of the window size (such as 10 seconds window extension 0.5 seconds data point), and the boundary data deletion is prevented from causing incomplete sequence. Calculating the change rate of the signal frequency, if the change rate exceeds 5%/second (for example, the signal frequency suddenly changes from 10GHz to 10.5GHz and the change rate within 1 second is 5%), the window is reduced to 5 second data points, and if the frequency is stable (the change rate is lower than 0.5%/second), the window is expanded to 20 second data points, the adjustment step size is limited by the minimum response time of the system to 1 millisecond, and the window adjustment is ensured not to lag behind the signal change. Meanwhile, the data processing adopts a double-buffer mechanism, the front buffer stores the original deviation sequence called from the deviation database in real time, and the rear buffer temporarily stores the sequence segment to be processed, so that the data reading and writing conflict is avoided. In addition, the dynamic window adjusting algorithm also monitors the change of the signal spectrum entropy value, if the entropy value mutation exceeds a preset threshold value, window reconfiguration is immediately triggered, and the window size and the entropy value change rate are in negative correlation (the higher the entropy value change rate is, the smaller the window is), so that the problems that the fixed window cannot match the signal change frequency, the data redundancy is caused during high-frequency interference, the information is insufficient during low-frequency stabilization can be solved, the smooth distortion of the historical deviation sequence characteristics caused by the traditional fixed window is avoided, and the extracted sequence can accurately reflect the transient and steady state characteristics of deviation.

And secondly, extracting multi-scale characteristics of a historical deviation sequence extracted by a sliding window, firstly calculating the instantaneous variation of adjacent deviation values through a first-order difference algorithm (for example, the first-order difference result of the sequence [2.009,2.015,2.021,2.018,2.025] is [0.006,0.006, -0.003,0.007 ]), then calculating the variation acceleration through a second-order difference algorithm (for example, the second-order difference based on the first-order difference result is [0, -0.009,0.01 ]), and then adopting a linear regression algorithm to fit the original historical deviation sequence to obtain regression coefficients (for example, the slope is 0.004/second and the intercept is 2.005) reflecting the deviation time evolution trend, wherein the coefficients form time domain basic components of a deviation variation trend vector. Meanwhile, fourier transform is applied to the historical deviation sequence, periodic characteristics (such as the period of identifying the interference signal is 0.5 seconds) in the sequence are detected, and the amplitude (such as 0.005) of the main frequency component is normalized and then used as the frequency domain characteristics of the trend vector. The dimension of the deviation change trend vector is fixed to be 6 dimensions, the front three dimensions correspond to time domain features (respectively, the sequence mean value, the linear regression slope and the second-order differential curvature, such as [2.015,0.004,0.01 ]), and the rear three dimensions store frequency domain features (respectively, the normalized amplitudes of the first three main frequency components, such as [0.8,0.15,0.05 ]). The feature extraction stage adopts a parallel computing architecture, a time domain analysis (differential and linear regression) and a frequency domain analysis (Fourier transformation) are distributed to be independent operation units, the processing efficiency is improved, the vector generation algorithm adopts fixed point number operation optimization, the time domain feature computing precision is kept at 0.001 level, and the frequency domain feature resolution reaches 0.1Hz. By fusing time domain and frequency domain features, the problem of delayed response of simple time domain analysis to periodic interference is solved, so that the deviation change trend vector can reflect the instantaneous change speed and direction of deviation and capture the frequency features of periodic interference.

Finally, the standard deviation change trend vector is obtained by carrying out normalization processing on the deviation change trend vector, wherein the normalization processing adopts a maximum and minimum method, and the value range of each dimension characteristic is firstly counted from a training data set (comprising deviation trend vector samples under different working modes and different interference scenes) of a radar system, wherein the value range is that the time domain mean value is 1.8-2.2, the linear regression slope is-0.01-0.01/s, the second-order differential curvature is-0.02-0.02, and the normalization amplitude of three main frequency components in the frequency domain is 0-1. The normalized value of each dimension, such as the variation trend vector [2.015,0.004,0.01,0.8,0.15,0.05], is calculated according to the formula "normalized component= (original component-minimum value of dimension)/(maximum value of dimension-minimum value of dimension)", and normalized by the time domain mean value isLinear regression slope normalized toThe second order difference curvature is normalized to be%And (3) after the frequency domain three components are normalized, maintaining the original value, and finally obtaining a normalized deviation change trend vector [0.5375,0.7,0.75,0.8,0.15,0.05]. The normalization module is internally provided with a parameter automatic calibration function, the calibration period is set to be 1 hour, historical trend vector data of the last 24 hours are extracted from the deviation database in each calibration, the value range of each dimension is counted again, the normalization parameters are updated, and the timeliness of the parameters is ensured. Meanwhile, the normalization processing eliminates the dimension difference of deviation features under different radar working modes (such as a search mode and a tracking mode), so that the standard deviation change trend vector has cross-scene comparability, and the problem of unbalanced weight of the subsequent threshold comparison under different modes is solved.

And step S103, calculating trend slope and fluctuation amplitude by using the deviation change trend vector, and determining an adjustment trigger signal according to the relationship between the trend slope and the fluctuation amplitude and a preset threshold value.

In a specific embodiment, the process of executing step S103 may specifically include the following steps:

Specifically, when calculating the trend slope in the time window, a time-domain linear regression slope component in the bias variation trend vector needs to be extracted, and the component is derived from the least square fitting result of the historical bias sequence. For example, the extracted historical deviation sequence is [2.009,2.015,2.021,2.018,2.025] (corresponding to a 5 second sliding window, the time stamps are t ₁ to t ₅ in sequence), a linear regression fitting is used to obtain a fitting straight line equation y=0.004t+2.005 (t is in seconds, y is a deviation value), the slope of the straight line is 0.004/second, the value of the slope directly reflects the change speed of the deviation value with time in the time window, namely, the slope is regular deviation and is in an ascending trend, the negative is in a descending trend, the absolute value of the slope is larger, and the deviation change is more intense. The calculation of the trend slope is defined in strict correspondence with the slope value of the deviation change trend vector, and the time window is consistent with the sliding window in the step S102 (for example, 5 seconds), so that the consistency of the data time dimension is ensured, the trend judgment distortion caused by the mismatch of the windows is avoided, and the processing solves the problem that the prior art lacks quantification of the deviation change speed and cannot accurately identify the signal mutation.

Further, when calculating the fluctuation width of the deviation variation trend vector [0.5375,0.7,0.75,0.8,0.15,0.05], the standard deviation is calculated to be 0.294, and the standard deviation value is the fluctuation width, and the smaller the fluctuation width is, the more stable the deviation variation is, and conversely, the more severe the fluctuation is. The corresponding relation between the differential sequence and the fluctuation amplitude ensures that the fluctuation amplitude can accurately quantify the discrete degree of deviation change, and solves the problems that the deviation fluctuation is lack of effective evaluation and noise is easy to be misjudged as signal mutation in the prior art.

Finally, when judging whether the trend slope exceeds a preset threshold and generating an adjustment trigger signal, the preset threshold is required to be determined based on historical interference scene data statistics of the radar system, and different interference scenes correspond to different thresholds, wherein the preset threshold is set to be 0.005/second when the deviation mutation is frequent due to the fact that signals are easy to be reflected by buildings in urban multipath interference scenes, and the preset threshold is set to be 0.008/second when the signals are relatively stable in airport electromagnetic interference scenes. Comparing the calculated trend slope with a preset threshold value of a corresponding scene, if the trend slope exceeds the threshold value, marking that the current interference signal is in a sudden change mode, then generating an adjustment trigger signal, wherein the signal is a digital level signal, the high level (logic 1) is used for representing the trigger, the low level (logic 0) is used for representing the non-trigger, the signal needs to carry a time stamp, a current interference scene label (such as 'urban multipath interference'), trend slope and fluctuation amplitude data, and transmitting the data to the next step through an internal bus of a radar system for starting a subsequent deviation compensation adjustment flow. If the trend slope does not exceed the threshold, the abrupt mode is not marked, and an adjustment trigger signal is not generated, and the system maintains the current deviation compensation model parameters. The threshold judgment mechanism solves the problem that the static model cannot adjust the compensation strategy in real time when the signal suddenly changes, ensures that the system can start the compensation process in time by quickly generating the adjustment trigger signal, avoids the sudden drop of positioning accuracy, and meanwhile, the fluctuation amplitude can be used as an auxiliary judgment basis, if the trend slope slightly exceeds the threshold but the fluctuation amplitude is extremely small, the accidental noise can be judged, the adjustment is not triggered, and the judgment accuracy is further improved.

And step S104, extracting related feature interaction influence from the deviation change trend vector according to the adjustment trigger signal, determining the correlation weight among the features, and calculating a weight coefficient update increment according to the correlation weight.

In a specific embodiment, the process of executing step S104 may specifically include the following steps:

Specifically, when relevant characteristic interaction effects are extracted from the historical trend database according to the adjustment trigger signal, the adjustment trigger signal carries key retrieval information, including calculation time stamps (such as '2025-08-21:15:48:55'), interference scene labels (such as 'airport electromagnetic interference'), trend slopes (0.004/s) and fluctuation amplitudes (such as 0.0060), and the information is used as a composite condition for database retrieval. The historical trend database stores characteristic data under various interference scenes in the past operation of the radar, and each record comprises a real-time sampling value of signal intensity, frequency and modulation mode and an interactive correlation record among the three types of characteristics. In the extraction process, the interference scene label is matched with a scene data set in a database, records in 1 hour nearest to the current moment are screened according to time stamps, characteristic interaction data of similar interference change trend is further screened according to trend slope and fluctuation amplitude range, for example, 20 groups of signal intensity-frequency-modulation mode interaction records conforming to the slope and amplitude range are extracted from 1 hour data of an airport electromagnetic interference scene, and the extracted characteristic interaction influence is guaranteed to be highly matched with the current interference state.

Further, when the correlation weight among the features is determined by analyzing the feature interaction influence, the extracted 20 groups of feature interaction data are taken as input, and feature correlation analysis is carried out aiming at the interaction among the signal intensity, the frequency and the modulation mode. The invention calculates the linear correlation degree between features by using the pearson correlation coefficient. For example, when calculating the correlation coefficient of signal intensity and frequency, firstly, the average value of signal intensity and average value of frequency in 20 groups of data are counted, and the result is passed through the formula(WhereinIs a correlation coefficient, x is the signal strength,Average value of signal intensity, y is frequency,For frequency average) a correlation coefficient is calculated. When the correlation coefficientThe method comprises the steps of determining the correlation coefficient of the signal intensity and the modulation mode, and obtaining the correlation coefficient by quantizing the modulation mode when calculating the correlation coefficient of the signal intensity and the modulation modeThe method shows that the signal intensity and the modulation mode are weakly related, the influence of the switching of the modulation mode on the intensity is small, and the correlation coefficient of the frequency and the modulation mode is calculatedThe frequency and modulation scheme are shown to be moderately correlated, and partial modulation scheme switching is accompanied by frequency minor fluctuations. The correlation weight is distributed according to the absolute value of the correlation coefficient, and the larger the absolute value of the correlation coefficient is, the more the influence of corresponding characteristic interaction on deviation is obvious, so that the correlation weight of signal strength-frequency interaction is set to 0.55, the signal strength-modulation mode interaction is set to 0.2, and the frequency-modulation mode interaction is set to 0.25. The process solves the problem that the contribution degree of characteristic interaction to deviation cannot be quantified and the weight distribution is subjective in the prior art.

The method comprises the steps of according to the correlation weight, calculating a weight coefficient update increment, specifically, calling a basic weight coefficient (such as signal intensity basic weight 0.4, frequency weight 0.3 and modulation mode weight 0.3) of a current radar deviation compensation model, and calculating the weight coefficient update increment by adopting a weighted average method, wherein the specific formula is that the weight coefficient update increment= (signal intensity-frequency interaction weight x signal intensity basic weight x signal intensity-modulation mode interaction weight x modulation mode interaction weight + frequency-modulation mode interaction weight x frequency basic weight) x dynamic adjustment coefficient. The dynamic adjustment coefficient is determined according to the fluctuation amplitude, the larger the fluctuation is, the larger the adjustment coefficient is, the increment can be ensured to match the intensity of the interference change, and the weight coefficient update increment is (0.55×0.4+0.2×0.3+0.25×0.3) ×0.12= 0.0426, assuming that the adjustment coefficient is set to 0.12 when the fluctuation amplitude is 0.294. Further distributing the update increment to each characteristic weight, namely, signal intensity weight increment=0.55x 0.0426 (0.0234), frequency weight increment=0.25x 0.0426 (0.0107), modulation mode weight increment=0.2x 0.0426 (0.0085), and finally obtaining a weight coefficient update increment vector [0.0234, 0.0107 and 0.0085]. The calculated result is stored in a system weight buffer area and is associated with a time stamp for adjusting a trigger signal and an interference scene label, so that the problem that the dynamic contribution of characteristic interaction to deviation cannot be reflected in real time due to the fact that the weight coefficient is updated after the time stamp is updated is solved.

Step 105, updating the convergence speed parameter according to the weight coefficient updating increment, and optimizing the updated convergence speed parameter through a gradient descent iteration optimization algorithm to obtain an optimized convergence speed value.

In a specific embodiment, the process of executing step S105 may specifically include the following steps:

Specifically, when the current convergence speed parameter is obtained, the parameter is called from a parameter buffer area of the radar deviation compensation model, the parameter buffer area stores real-time parameters according to the radar working mode (such as tracking mode and searching mode), and each record comprises the convergence speed parameter, an update time stamp and a corresponding interference scene label (such as urban multipath interference and airport electromagnetic interference). The convergence speed parameter reflects the response speed of the deviation compensation model, the larger the value is, the faster the response of the model to deviation change is, for example, in a tracking mode, the interference change caused by the movement of a target needs to be quickly adapted, the initial value of the current convergence speed parameter is set to be 0.3, in a searching mode, the signal coverage is wide, the interference is relatively stable, and the initial value of the parameter is set to be 0.2. When the parameter is called, the parameter is accurately matched through the current radar working mode label, and meanwhile, the parameter updating time stamp is verified, so that the timeliness of the parameter is ensured.

When updating the convergence rate parameter according to the weight coefficient update increment, the parameter update amount is calculated by using the weight coefficient update increment (such as the weight increment of the vector [0.0234, 0.0107, 0.0085] corresponding to the signal intensity, the frequency and the modulation mode) generated in the step S104 as input and adopting a weighted summation formula, wherein the parameter update amount= (the signal intensity weight increment x the intensity influence coefficient + the frequency weight increment x the frequency influence coefficient + the modulation mode weight increment x the modulation influence coefficient), the influence coefficient determines that the influence of the signal intensity on the response rate of the model is most obvious according to the historical data statistics, the intensity influence coefficient is set to be 0.5, the frequency is set to be 0.3, and the modulation mode influence is smaller and is set to be 0.2. Substitution weight coefficient update increment calculation, parameter update amount= (0.0234×0.5+0.0107×0.3+0.0085×0.2) = 0.01661. And then the updated quantity is overlapped with the current convergence rate parameter to obtain an updated convergence rate parameter, namely 0.3+0.01661= 0.31661 in a tracking mode and 0.2+0.01661= 0.21661 in a searching mode. In the updating process, the maximum value and the minimum value of the parameters are required to be limited to be not more than 0.8 and not less than 0.1, and the phenomenon that the model responds too quickly to oscillate or too slowly to produce hysteresis due to abnormal parameters is avoided.

The updated convergence rate parameter is optimized by a gradient descent method, the learning rate is set to 0.01 when the gradient descent parameter is initialized, the upper limit of iteration times is set to 100 times, and the convergence threshold is set to 0.001. The iteration process takes the sum of squares of deviation compensation residual errors as a loss function, and the loss function value is obtained by calculating the difference between the output of the current model and the target value in real time. And calculating the gradient of the loss function to the convergence speed parameter every iteration, wherein the gradient value is obtained by a numerical differentiation method, and the differentiation step length is set to be 0.001. The parameter updating is carried out along the opposite direction of the gradient, and the updating step length is determined by the product of the learning rate and the gradient value. The iteration termination condition is that the loss function value is lower than 0.001 or the maximum iteration number is reached, and the obtained convergence speed parameter value is the convergence speed value after optimization. The gradient descent iterative optimization solves the problems of slow convergence, high calculation complexity and response delay in signal mutation of the traditional iterative algorithm, and ensures that the convergence speed value after optimization can be quickly adapted to the current interference change.

And S106, adjusting parameters of the deviation compensation model according to the optimized convergence speed value, calculating a compensation residual error of the adjusted deviation compensation model, and outputting a deviation compensation result if the compensation residual error is lower than a preset residual error threshold value.

In a specific embodiment, the process of executing step S106 may specifically include the following steps:

Specifically, when the parameters of the deviation compensation model are adjusted according to the optimized convergence speed value, the current deviation compensation coefficient and the compensation offset-the deviation compensation coefficient are firstly adjusted from the parameter storage unit of the deviation compensation model to be used for quantifying the compensation weight of the signal characteristic to the deviation (such as the deviation compensation amount corresponding to each 1dBm change of the signal strength), the compensation offset is used for correcting the system deviation caused by the fixed interference, the initial values of the two parameters are required to be matched with the radar working mode, for example, in the tracking mode, the initial deviation compensation coefficient is 1.2 (high response requirement), the compensation offset is 0.05 (small fixed offset), and in the searching mode, the initial coefficient is 1.0 (stable requirement), and the offset is 0.03. The optimized convergence speed value (0.31661, tracking mode obtained in step S105) is used as an adjustment basis, an association logic of parameters and the convergence speed value is established, wherein the deviation compensation coefficient adopts positive correlation adjustment, the formula is 'new compensation coefficient=original compensation coefficient× (1+optimized convergence speed value×0.1)', the new compensation coefficient=1.2× (1+0.31661×0.1) ≡ 1.238 is obtained by substituting data, the faster the convergence speed is ensured, the larger the coefficient is to enhance the compensation sensitivity, the compensation offset adopts incremental adjustment, the formula is 'new compensation offset=original compensation offset+optimized convergence speed value×0.02', and the new compensation offset=0.05+0.31661×0.02≡ 0.0563 is obtained by substituting data, so that the offset is synchronously adapted to the disturbance dynamic change along with the convergence speed.

Further, when calculating the compensation residual error of the offset compensation model after adjustment, characteristic parameters (signal intensity-68 dBm, frequency 10.03GHz, modulation scheme FM) of the current interference signal are taken as input, the offset compensation model after adjustment (including new compensation coefficient 1.238 and new compensation offset 0.0563) is substituted, and a compensation value is output through model operation, namely "compensation value= (signal intensity offset component x compensation coefficient) + (frequency offset component x compensation coefficient x 0.8) + (frequency component weight 0.8) +compensation offset" (frequency offset component weight 0.8 is derived from historical characteristic correlation analysis), and if the signal intensity offset component is 4dBm and the frequency offset component is 0.03GHz, the compensation value= (4 x 1.238) + (0.03 x 1.238 x 0.8) +0.0563 +.5.038). And then determining a compensation target value, calling the compensation target value from a recorded historical trend database, selecting an ideal compensation value (such as 5.036) matched with the current interference scene (such as 'urban multipath interference') and the signal characteristics, calculating according to 'compensation residual= |adjusted model output compensation value-compensation target value|', and substituting the compensation residual= |5.038-5.036|=0.002 into data. In the calculation process, a digital filtering technology (such as moving average filtering, window 3 data points) is adopted to eliminate output fluctuation caused by interference signal noise, ensure residual calculation accuracy, and solve the problem that residual calculation is interfered by noise and cannot truly reflect compensation effect.

Finally, if the compensation residual is lower than a preset threshold, when the deviation compensation result is output, the preset threshold needs to be set by combining a radar working mode and an interference scene, wherein in a tracking mode, the threshold is set to be 0.005 due to high-precision positioning, and in a searching mode, the threshold is set to be 0.01 due to wide signal coverage range. Comparing the calculated compensation residual error 0.002 with the tracking mode threshold value 0.005, judging that the residual error is lower than the threshold value, and then generating a deviation compensation result, wherein the result comprises a compensation value 5.038 output by the adjusted model, a new deviation compensation coefficient 1.238, a new compensation offset 0.0563, a calculation timestamp (such as '2025-08-21:15:48:55') and an interference scene label 'urban multipath interference', storing the calculation timestamp and the interference scene label 'urban multipath interference' in a system compensation result buffer area, and transmitting the calculation result to the next process through an internal data bus. If the compensation residual is higher than the threshold, returning to the step S105 to perform iterative optimization again until the residual reaches the standard. The threshold value judgment and result output mechanism solves the problem that the compensation result is put into use without verification, so that positioning deviation is accumulated, and ensures that the output deviation compensation result can provide reliable support for subsequent positioning correction.

In a specific embodiment, the process of executing step S107 may specifically include the following steps:

Specifically, when the deviation compensation result is integrated into the positioning calculation flow, the deviation compensation result needs to include the adjusted compensation value, the deviation compensation coefficient, the compensation offset, and the associated interference scene tag (such as urban multipath interference and airport electromagnetic interference) and the calculation time stamp. The positioning calculation process originally comprises an original coordinate generation step, at this time, a coordinate correction step based on a deviation compensation result is needed to be inserted, the compensation value is disassembled into components corresponding to the original positioning coordinates (such as an original x-direction compensation component corresponding to the x-coordinate and an original y-direction compensation component corresponding to the y-coordinate), so as to ensure that the compensation data corresponds to the coordinate components one by one, for example, when the original x-coordinate is 50 meters, if the x-direction compensation component in the compensation value is 0.0965 meters, the original coordinate components are directly associated for subsequent correction, and correction errors caused by data dislocation are avoided.

Further, when the input parameters of the positioning algorithm are adjusted according to the deviation compensation result, the input parameters of the positioning algorithm comprise a distance estimated value corresponding to the signal strength, a phase offset value corresponding to the frequency and a positioning error correction value corresponding to the modulation mode, and initial weights of the parameters need to be adjusted by combining with a weight coefficient update increment in the deviation compensation result. For example, if the signal strength weight increment is 0.0234, the frequency weight increment is 0.0107, the modulation scheme weight increment is 0.0085, the signal strength initial weight is 0.5, the frequency initial weight is 0.3, and the modulation scheme initial weight is 0.2 in the positioning algorithm, the adjusted signal strength weight is 0.5+0.0234=0.5234, the frequency weight is 0.3+0.0107=0.3107, and the modulation scheme weight is 0.2+0.0085= 0.2085 in the deviation compensation result. In the adjustment process, the weight change of each input parameter is directly associated with the corresponding weight increment in the deviation compensation result, so that the input parameter weight can reflect the dynamic contribution of each signal characteristic to positioning under the current interference in real time, and the problem of compensation residual errors caused by insufficient characteristic interaction processing capacity and delayed updating of weight coefficients in the prior art is solved.

Finally, when calculating the corrected positioning coordinates by a weighted positioning method, the adjusted input parameters (such as the signal strength distance estimated value, the frequency phase offset value and the modulation mode error corrected value after weight adjustment) are multiplied by the corresponding adjusted weights respectively, and then the products are summed to obtain corrected coordinate components. Taking x-direction coordinate calculation as an example, if the adjusted signal strength distance estimated value is 30m, multiplying by adjusted weight 0.5234 to obtain 15.702, the x-component corresponding to the frequency phase offset is 15m, multiplying by adjusted weight 0.3107 to obtain 4.0065, the x-direction error correction value corresponding to the modulation mode is 5m, multiplying by adjusted weight 0.2085 to obtain 1.0425, and summing the three to obtain the x-direction corrected coordinate, namely 15.702+4.0065+4.0065= 20.751 m. The y-direction coordinate calculation logic is the same, if the adjusted signal strength distance estimated value is 20 meters, multiplied by 0.5234 to obtain 10.468, the y-component corresponding to the frequency phase offset is 10 meters, multiplied by 0.3107 to obtain 3.107, the y-direction error correction value corresponding to the modulation mode is 3 meters, multiplied by 0.2085 to obtain 0.6255, and summed by 10.468+3.107+0.6255= 14.2005 meters, namely the y-direction corrected coordinate is obtained. In the weighting calculation process, the input parameters, the adjusted weights and the coordinate components form a clear data corresponding relation, the algorithm gives priority to the signal characteristics which are obviously affected under the current interference through weight distribution, the problems that a static model cannot adapt to the dynamic change of a complex environment signal and the positioning accuracy is suddenly reduced are solved, meanwhile, the cooperation of the positioning and anti-interference of an interference source is realized by associating with an interference scene label, and the multipath effect and the anti-interference robustness under a dynamic interference source switching scene are enhanced.

The foregoing describes a radar dynamic anti-interference method based on interference source positioning in the embodiment of the present application, and the following describes a radar dynamic anti-interference processing process 200 based on interference source positioning in the embodiment of the present application, referring to fig. 2, the processing process 200 is configured by seven processes including a signal acquisition process 201, a deviation analysis process 202, a slope and threshold value judgment process 203, a weight optimization process 204, a convergence optimization process 205, a compensation output process 206 and a positioning correction output process 207, and cooperatively implements signal processing and positioning correction functions, and the specific processes are as follows:

In the signal acquisition process 201, the signal receiving device samples the current interference signal in real time according to the radar working frequency, and at the same time, dynamically matches and retrieves the historical sequence according to the signal frequency from a database storing the historical signal of the same working mode for approximately 30 minutes, extracts the characteristics of signal strength, frequency (through fast fourier transform+peak detection extraction), modulation mode (demodulation circuit+maximum likelihood estimation identification) and the like after sampling, and completes the deviation calculation through multi-dimensional difference quantization (the strength deviation is the difference between the current mean value and the historical mean value, the frequency deviation is qualitatively assigned by Euclidean distance and the modulation mode deviation, and the weighted summation is determined according to the historical variance).

The bias analysis process 202 starts a sliding window of initial 10 seconds (including 1000 data points), expands to 20 seconds according to the signal frequency change rate (more than 5%/second is 5 seconds and less than 0.5%/second, and the double buffering mechanism guarantees no conflict in reading and writing), performs frequency domain analysis (fourier transform to extract frequency domain components) and time domain analysis (first-order/second-order differential calculation change amount is fitted with acceleration and linear regression to obtain time domain components), obtains regression slope, main amplitude and the like, normalizes by a maximum-minimum method (based on training data set statistical range and 1 hour calibration parameter), and generates a 6-dimensional (3 time domain+3 frequency domain) trend vector.

The slope and threshold judging process 203 extracts the time domain linear regression slope based on the trend vector, calculates the vector standard deviation to analyze the deviation, judges whether the slope exceeds the preset scene threshold (city multipath interference 0.005/s, airport electromagnetic interference 0.008/s), if yes, generates the high level trigger signal with the time stamp and scene label, if not, further judges whether the fluctuation amplitude exceeds the threshold, if yes, generates the adjustment trigger signal, and if not, re-executes the deviation analysis flow.

The weight optimization process 204 is driven by adjusting the trigger signal, extracting 20 groups of similar interaction data from the history base according to the time stamp, scene tag and slope/amplitude of the trigger signal, calculating the relevance of every two features (such as intensity-frequency, weight 0.55 when R epsilon (0.6,1)) through pearson correlation coefficient and normalizing to obtain the correlation weight, calculating the weight increment according to the' basic weight+ (interaction weight multiplied by basic weight) ×dynamic adjustment coefficient (according to the fluctuation amplitude, such as amplitude 0.294), and outputting to the convergence optimization process.

The convergence optimization process 205 receives the weight update increment, retrieves the current convergence speed parameter from the parameter buffer (tracking mode 0.3, search mode 0.2, verification timestamp), superimposes the current parameter by "parameter update amount = intensity increment x 0.5+ frequency increment x 0.3+ modulation increment x 0.2" (limit 0.1-0.8), iterates the optimization by gradient descent method (learning rate 0.01, iterates 100 times, convergence threshold 0.001, residual square sum is loss function), outputs the optimized speed value if the convergence condition is satisfied, and otherwise continues the iteration.

The compensation output process 206 inputs the optimized convergence rate, adjusts the compensation model parameters (such as tracking mode compensation coefficient 1.2- & 1.238) according to the "compensation coefficient=original coefficient× (1+convergence rate×0.1), offset=original offset+convergence rate×0.02", calculates the residual error between the compensation value and the ideal value of the history base (such as 5.038-5.036=0.002, and noise-removed by moving average filtering), determines whether the residual error is lower than the scene threshold (tracking mode 0.005, search mode 0.01), if yes, outputs the compensation result with parameters, time stamp, and scene tag, otherwise returns to the convergence optimization process for readjustment.

The positioning correction output process 207 inputs the compensation result, disassembles the compensation result into x/y coordinate components, inserts the positioning process correction step, adjusts the parameter weight (such as intensity weight 0.5-0.5234) input by the positioning algorithm according to the weight increment, calculates by a weighted positioning method (adjusted parameter x corresponding weight summation, such as x direction 15.702+4.0065+1.0425= 20.751 m), outputs correction coordinates after fusing the coordinates, and completes the whole process from signal acquisition to positioning accurate correction.

The foregoing describes a radar dynamic anti-interference method based on interference source positioning in the embodiment of the present application, and the following describes a radar dynamic anti-interference system based on interference source positioning in the embodiment of the present application, referring to fig. 3, an embodiment of the radar dynamic anti-interference system based on interference source positioning in the embodiment of the present application includes:

the signal acquisition module 301 is configured to acquire current signal data and a historical signal sequence, calculate and determine an initial deviation value, and ensure real-time performance and accuracy of data acquisition.

The deviation analysis module 302 extracts a history deviation sequence according to the initial deviation value, generates a deviation change trend vector through processing, and visually reflects the change characteristics of the deviation along with time.

The trigger judging module 303 calculates the trend slope and the fluctuation amplitude by using the deviation change trend vector, and if the slope exceeds the preset threshold, immediately determines and generates an adjustment trigger signal, and starts the subsequent compensation adjustment process.

The weight optimization module 304 extracts the feature interaction effect according to the adjustment trigger signal, analyzes the feature correlation to determine the correlation weight, calculates the weight coefficient update increment, and provides a basis for model parameter update.

The convergence optimization module 305 acquires the current convergence speed parameter, updates the parameter by combining the weight coefficient update increment, optimizes the parameter through an iterative optimization algorithm to obtain an optimized convergence speed value, and improves the response speed of the model.

And the compensation output module 306 is used for adjusting the deviation compensation model parameters according to the optimized convergence speed value, calculating a compensation residual error, and outputting a deviation compensation result for subsequent positioning correction if the residual error is lower than a preset threshold value.

The positioning correction module 307 integrates the deviation compensation result into a positioning calculation flow, adjusts the input parameters of a positioning algorithm, calculates and determines corrected positioning coordinates through the positioning algorithm, and realizes accurate positioning.

Fig. 3 above describes the radar dynamic anti-interference system based on the positioning of the interference source in the embodiment of the present invention in detail from the perspective of the modularized functional entity, and the radar dynamic anti-interference device based on the positioning of the interference source in the embodiment of the present invention is described in detail from the perspective of the hardware processing.

Referring to fig. 4, in an embodiment of the present invention, a radar dynamic anti-interference device 400 based on interference source positioning is further provided, where the radar dynamic anti-interference device based on interference source positioning may be a server, and the internal structure of the radar dynamic anti-interference device may be as shown in fig. 4. The radar dynamic anti-jamming device 400 based on jamming source positioning includes a processor 402, a memory 403, a display 404, an input device 405, a network interface 406, and a database 407 connected by a system bus 401. Wherein the computer-designed processor 402 is used to provide computing and control capabilities. The memory 403 of the radar dynamic anti-jamming device based on jamming source positioning comprises a non-volatile storage medium 4031, an internal memory 4032. The nonvolatile storage medium 4031 stores an operating system and a computer program. The internal memory 4032 provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database 407 of the radar dynamic anti-interference device based on interference source positioning is used for storing the corresponding data in this embodiment. The network interface 406 of the radar dynamic anti-jamming device based on jamming source positioning is used for communication with an external terminal through a network connection. The computer program being executable by a processor to implement the method described above.

It will be appreciated by those skilled in the art that the structure shown in fig. 4 is merely a block diagram of a portion of the structure associated with the present inventive arrangements and does not constitute a limitation of the radar dynamic anti-interference device based on interference source positioning to which the present inventive arrangements are applied. It will be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, systems and units may refer to the corresponding processes in the foregoing method embodiments, which are not repeated herein.

The foregoing embodiments are merely for illustrating the technical solution of the present invention, but not for limiting the same, and although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those skilled in the art that modifications may be made to the technical solution described in the foregoing embodiments or equivalents may be substituted for parts of the technical features thereof, and that such modifications or substitutions do not depart from the spirit and scope of the technical solution of the embodiments of the present invention in essence.

Claims

1. A radar dynamic anti-interference method based on interference source positioning, characterized in that the method comprises:

Step S105, updating the increment and updating the convergence speed parameter according to the weight coefficient, and optimizing the updated convergence speed parameter through a gradient descent iteration optimization algorithm to obtain an optimized convergence speed value;

2. The radar dynamic anti-interference method based on interference source positioning according to claim 1, wherein the step S101 includes:

3. The method of radar dynamic anti-interference based on interference source positioning according to claim 2, wherein the step S102 comprises:

4. The method of claim 1, wherein in step S103, the method further comprises:

5. The method of radar dynamic anti-interference based on interference source positioning according to claim 3, wherein in step S104, the method comprises:

according to the interference scene label, matching a scene data set in a historical trend database, and according to the time stamp, the trend slope and the fluctuation amplitude, extracting characteristic interaction data from the scene data set;

6. The radar dynamic anti-interference method based on interference source positioning according to claim 1, wherein in step S105, the method comprises:

7. The radar dynamic anti-interference method based on interference source positioning according to claim 1, wherein the step S106 includes:

8. The method of radar dynamic anti-interference based on interference source localization according to claim 7, wherein the step S107 comprises:

9. A radar dynamic anti-interference system based on interference source positioning, for implementing the radar dynamic anti-interference method based on interference source positioning according to any one of claims 1-8, the radar dynamic anti-interference system based on interference source positioning comprising:

10. A radar dynamic anti-jamming device based on jamming source positioning, characterized by comprising a memory and a processor, the memory storing a computer program executable on the processor, the processor implementing the radar dynamic anti-jamming method based on jamming source positioning of any of claims 1 to 8 when executing the computer program.