A target interference elimination method based on radar Doppler characteristics

Abstract

The application discloses a target interference elimination method based on radar Doppler characteristics, and relates to the technical field of radar signal processing, which comprises the following steps: collecting echo signals of a target area, then acquiring three-axis speed information and seat attitude angle information output by a seat inertial navigation module; performing frequency domain matching filtering distance compression and cross-pulse coherent accumulation on the echo signals, forming a range-Doppler joint spectrum matrix, then extracting energy proportion of spectrum response of each channel to perform spatial weighted fusion, then calculating Doppler spectrum entropy change, energy center migration and spectrum dispersion degree change to perform weighted fusion and conversion, generating an interference probability, and then identifying an abnormal interference area; the application enables the ejection seat height measurement radar to continuously output stable and reliable target distance results under high speed, large attitude change and multiple interference environment, and improves the reliability of target interference elimination and the radar ranging precision.

Description

Technical Field

[0001] This invention relates to the field of radar signal processing technology, and in particular to a target interference elimination method based on radar Doppler characteristics. Background Technology

[0002] With the increasing demands of modern aviation equipment for low-altitude rescue capabilities and intelligent sensing capabilities, ejection seat altimeter systems, as key sensing units in pilot active rescue systems, directly impact parachute deployment timing and rescue success rates through their ranging accuracy and environmental adaptability. Traditional ejection seat altimeter devices primarily employ radio ranging technology, transmitting electromagnetic waves to the ground and receiving reflected echoes from targets to determine the distance between the seat and the ground based on electromagnetic wave propagation time. With the development of microwave integration technology, digital signal processing technology, and high-performance embedded processors, existing altimeter radars are gradually evolving from single-distance measurement modes to multi-dimensional information fusion, including distance, Doppler, and angle. Among these, solutions based on frequency-modulated continuous wave radar, pulse compression radar, and coherent accumulation processing, by improving range resolution, enhancing weak echo detection capabilities, and utilizing Doppler frequency shift to acquire target motion state information, have become important development directions in the field of airborne ranging and target detection.

[0003] However, existing technologies directly identify interference and remove it by setting a certain feature value to zero based solely on whether it exceeds a fixed threshold. This not only easily leads to misjudgment, but also the threshold setting is highly dependent on prior knowledge of the scenario, lacks the ability to adapt to continuously changing interference scenarios, and lacks a dynamic constraint mechanism based on the platform's own motion state. It cannot predict the theoretical target Doppler features based on the real-time speed and posture changes of the seat, nor can it actively detect and probabilistically suppress unknown types of interference. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a target interference elimination method based on radar Doppler features, which solves the problems of existing technologies lacking adaptability to continuously changing interference scenarios and lacking dynamic constraint mechanisms based on the platform's own motion state.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a target interference elimination method based on radar Doppler features, which includes: acquiring echo signals from the target area, and then acquiring three-axis velocity information and seat attitude angle information output by the seat inertial navigation module;

[0008] After performing frequency domain matched filtering, range compression, and cross-pulse coherent accumulation on the echo signal to form a range-Doppler joint spectral matrix, the energy proportion of the spectral response of each channel is extracted and spatially weighted and fused. Then, the changes in Doppler spectral entropy, energy center migration, and spectral dispersion are calculated, weighted, fused, and transformed to generate interference probability, and then abnormal interference areas are identified.

[0009] Based on the triaxial velocity information and seat attitude angle information, the theoretical target Doppler frequency and theoretical Doppler range are obtained. The Doppler matching degree is calculated, and combined with the Doppler entropy mutation evaluation index, the target echo credibility is formed.

[0010] Based on the target echo confidence and interference probability, a target retention weight is constructed to selectively suppress the range-Doppler joint spectrum, resulting in a target enhancement spectrum. Based on the target enhancement spectrum, the target actual range conversion and target removal are performed to obtain the final ranging result.

[0011] As a preferred embodiment of the target interference elimination method based on radar Doppler features described in this invention, the method of acquiring echo signals from the target area refers to acquiring echo signals from the space area surrounding the ejection seat using four directional microstrip radars to obtain complex echo signals from each channel in the fast time domain.

[0012] As a preferred embodiment of the target interference elimination method based on radar Doppler features described in this invention, wherein: the re-acquisition of the three-axis velocity information and seat attitude angle information output by the seat inertial navigation module refers to the acquisition of the three-axis velocity information and seat attitude angle information of the seat in the navigation coordinate system by the inertial navigation module installed on the ejection seat;

[0013] The three-axis velocity information includes the navigation coordinate system. Axial velocity components, navigation coordinate system Axial velocity components, navigation coordinate system The axial velocity component and seat attitude angle information include seat pitch angle, seat roll angle, and seat yaw angle; the inertial navigation module includes a three-axis accelerometer and a three-axis gyroscope.

[0014] As a preferred embodiment of the target interference elimination method based on radar Doppler features described in this invention, the step of performing frequency-domain matched filtering range compression and cross-pulse coherent accumulation on the echo signal to form a range-Doppler joint spectral matrix is ​​as follows:

[0015] After obtaining the locally stored transmit reference signal as the matched filter template, the frequency domain matched filtering method is used to perform complex conjugate matching on the complex echo signals of each channel in the fast time domain through the matched filter template, so as to obtain the distance compression echo of the complex echo signals of each channel in the current pulse period after matched filtering.

[0016] The distance compression echoes corresponding to the same distance cell within multiple consecutive pulse cycles are arranged in slow time order to obtain a slow time series. The slow time series are used as columns of a matrix, and the time series in the slow time series are used as rows to form a coherent accumulation data matrix.

[0017] A discrete Fourier transform is performed on each column of the coherent accumulation data matrix to map the frequency changes caused by the target motion to the Doppler domain. After obtaining the range-Doppler two-dimensional spectral response values ​​corresponding to each channel, the spatial gain weights corresponding to each channel are calculated. Then, the range-Doppler two-dimensional spectral response values ​​are weighted and fused to obtain the range-Doppler joint spectrum, forming the range-Doppler joint spectrum matrix.

[0018] As a preferred embodiment of the target interference elimination method based on radar Doppler features described in this invention, the step of extracting the energy proportion of the spectral response of each channel and performing spatial weighted fusion, then calculating the change in Doppler spectral entropy, the energy center migration, and the change in spectral dispersion and performing weighted fusion and transformation to generate the interference probability, is as follows:

[0019] Each range cell is treated as an independent analysis object. For any range cell, the corresponding range-Doppler joint spectrum is extracted. Then, the range-Doppler joint spectrum within each range cell is normalized using energy probability normalization to obtain the normalized energy corresponding to each Doppler frequency in the range cell.

[0020] Based on normalized energy, the Doppler spectral entropy value corresponding to the distance cell is calculated using the information entropy method. Then, the Doppler spectral entropy values ​​are sorted in chronological order to form a Doppler spectral entropy value sequence, and the difference in Doppler spectral energy distribution between adjacent time points in the sequence is calculated.

[0021] Based on the energy probability, the Doppler energy center and energy center migration of the range cell are calculated. Then, based on the Doppler energy center and normalized energy, the Doppler spectrum dispersion of the range cell is calculated.

[0022] By using preset weights for Doppler spectrum entropy change, Doppler energy center migration, and Doppler spectrum dispersion, the entropy change, energy center migration, and Doppler spectrum dispersion are weighted and fused to obtain an evaluation value for unknown interference mutation.

[0023] The probability of interference is obtained by performing a probability transformation on the evaluation value of unknown disturbance mutation using an exponential function.

[0024] As a preferred embodiment of the target interference elimination method based on radar Doppler features described in this invention, the re-identification of abnormal interference regions refers to comparing the interference probability with a preset interference mutation discrimination threshold. If the interference probability is greater than the preset interference mutation discrimination threshold, it is considered that there is an abnormal change in the range-Doppler joint spectrum corresponding to the current range cell, and it is marked as an unknown interference candidate region. If the interference probability is less than or equal to the preset interference mutation discrimination threshold, it is considered that there is no abnormal change in the range-Doppler joint spectrum corresponding to the current range cell, and it is marked as a normal candidate region.

[0025] As a preferred embodiment of the target interference elimination method based on radar Doppler characteristics described in this invention, the following steps are taken: After obtaining the theoretical target Doppler frequency and theoretical Doppler range, the Doppler matching degree is calculated, specifically as follows:

[0026] The initial beam direction vector of each channel of the four-way directional microstrip radar in the antenna coordinate system is read. An attitude rotation matrix is ​​established based on the seat attitude angle information. The initial beam direction vector is then transformed to the navigation coordinate system using the attitude rotation matrix to obtain the unit vector of the actual detection direction of the radar. Multiplication is then performed in combination with the three-axis velocity information to obtain the radial motion velocity of the seat along the radar detection direction.

[0027] Based on the radar operating wavelength and radial velocity, the theoretical target Doppler frequency is determined, and the theoretical target Doppler range is constructed.

[0028] For any distance cell, the maximum value search method is used to traverse the theoretical target Doppler range, select the position of the maximum energy peak, obtain the candidate frequency of the moving target, and calculate the degree of Doppler matching by obtaining the absolute difference between the candidate frequency of the moving target and the Doppler frequency of the theoretical target through absolute value calculation.

[0029] As a preferred embodiment of the target interference elimination method based on radar Doppler features described in this invention, the method involves: combining the Doppler entropy mutation evaluation index to form a target echo credibility index to extract the interference probability; combining the Doppler matching degree to perform target credibility modeling; and calculating the target echo credibility of the range cell.

[0030] As a preferred embodiment of the target interference elimination method based on radar Doppler features described in this invention, the following steps are described: Based on the target echo confidence and interference probability, a target retention weight is constructed to selectively suppress the range-Doppler joint spectrum to obtain the target enhancement spectrum; based on the target echo confidence and interference probability, a target retention weight for the range unit is constructed, and the target retention weight is multiplied by the range-Doppler joint spectrum to obtain the target enhancement spectrum.

[0031] As a preferred embodiment of the target interference elimination method based on radar Doppler features described in this invention, the step of performing target actual range conversion and target elimination based on the target enhancement spectrum to obtain the final ranging result is as follows:

[0032] After performing modulo squaring on the enhanced spectrum of the target, the actual echo power is obtained. The range cell with the largest actual echo power is selected as the target range cell, and the target range cell is converted into the actual target range.

[0033] Based on the actual distance of the target within several detection cycles, the distance change between adjacent detection cycles is obtained, and the theoretical distance change is calculated based on the radial motion velocity.

[0034] The distance change is subtracted from the theoretical distance change and the absolute value is taken to obtain the distance change anomaly evaluation value. If the distance change anomaly evaluation value is less than the distance change allowable error threshold, the current ranging result is considered to conform to the seat motion law, and the current target distance output is maintained. If the distance change anomaly evaluation value is greater than or equal to the distance change allowable error threshold, the current ranging result is considered to have an abnormal jump, the current distance unit is removed, and the next effective target actual distance is searched again from the target enhancement spectrum.

[0035] The abnormal jumps include the sudden appearance of false targets at close range due to strong fuselage reflection, the generation of false range peaks due to mutual interference between the dual-seat radars, the random drift of range peaks due to transient spatial clutter, and the generation of incorrect range responses after beam shift caused by large attitude changes.

[0036] The beneficial effects of this invention are as follows: By analyzing the Doppler energy distribution within the range cell, an unknown interference detection mechanism based on changes in Doppler spectral entropy, energy center migration, and spectral dispersion is established. This eliminates the need for pre-establishing a fixed interference template to identify anomalous echoes such as strong fuselage reflections, dual-seat radar interference, and transient spatial clutter, thus improving the radar's adaptability to unknown dynamic interference. Furthermore, by introducing velocity and attitude information output from the inertial navigation module, a theoretical target Doppler range matching the real-time motion state of the seat is established, imbuing the target discrimination process with kinematic constraints and reducing the probability of false targets during high-speed descent, roll, and dive. Secondly, a soft mask suppression method based on target credibility and interference probability is adopted, continuously adjusting according to the probability relationship between targets and interference in different range-Doppler cells, improving range-keeping capability under weak target conditions. Therefore, this invention enables the ejection seat altimeter radar to continuously output stable and reliable target range results under high-speed, large-attitude-change, and multi-interference environments, and improves the reliability of target interference elimination and radar ranging accuracy. Attached Figure Description

[0037] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the 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.

[0038] Figure 1 This is a flowchart of the target interference elimination method based on radar Doppler features in Example 1.

[0039] Figure 2 This is a flowchart of the target enhancement spectrum obtained in Example 1.

[0040] Figure 3 This is a flowchart of the final ranging result obtained in Example 1. Detailed Implementation

[0041] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0042] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0043] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0044] Example 1, referring to Figures 1-3 This is the first embodiment of the present invention, which provides a target interference elimination method based on radar Doppler features, including the following steps:

[0045] S1. Collect echo signals from the target area, and then obtain the three-axis velocity information and seat attitude angle information output by the seat inertial navigation module.

[0046] Specifically, the echo signals of the space area around the ejection seat are collected by four directional microstrip radars to obtain the complex echo signals of each channel in the fast time domain.

[0047] Among them, the four-way directional microstrip radars correspond to different spatial detection directions and are used to obtain target reflection signals at different azimuth angles, so as to reduce the blind spot of single-direction detection caused by attitude changes during the high-speed movement of the seat.

[0048] It should be noted that the four-way directional microstrip radar needs to be installed on the mounting base on the back of the ejection seat or on both sides of the headrest. The beam directions of the four microstrip antennas are spatially symmetrically distributed along the pitch and yaw axes of the seat. Specifically, the first radar points to the upper left front of the seat, the second radar points to the upper right front of the seat, the third radar points to the lower left front of the seat, and the fourth radar points to the lower right front of the seat. The installation angle between each radar and the coordinate system of the seat body is pre-calibrated by an optical calibrator during seat assembly and fixed in the non-volatile storage unit of the radar signal processor. After the ejection seat is activated and enters high-speed motion, the four-way directional microstrip radar simultaneously radiates linear frequency modulated continuous wave or pulse modulated signal in their respective detection directions at a set pulse repetition frequency, and synchronously acquires the antenna received signals of each channel. The radar receiving front end amplifies, down-converts, and quadratures the radio frequency signal received by the antenna through low noise amplification, down-conversion, and quadrature demodulation to obtain analog zero intermediate frequency in-phase and quadrature signals. Then, it is sampled and quantized by an analog-to-digital converter at a set sampling frequency to obtain the complex echo signals of each channel in the fast time domain. The complex echo signals are composed of in-phase components as the real part and quadrature components as the imaginary part. The sampling frequency is set according to the radar transmit signal bandwidth to ensure that the length of the fast time dimension range resolution unit is not greater than half of the radar range resolution.

[0049] Furthermore, an inertial navigation module installed on the ejection seat collects the seat's three-axis velocity information and seat attitude angle information in the navigation coordinate system. The three-axis velocity information includes the navigation coordinate system... Axial velocity components, navigation coordinate system Axial velocity components, navigation coordinate system The axial velocity component and seat attitude angle information include seat pitch angle, seat roll angle, and seat yaw angle.

[0050] The inertial navigation module includes a three-axis accelerometer and a three-axis gyroscope.

[0051] S2. After performing frequency domain matched filtering, distance compression, and cross-pulse coherent accumulation on the echo signal to form a range-Doppler joint spectral matrix, the energy proportion of the spectral response of each channel is extracted and spatially weighted and fused. Then, the changes in Doppler spectral entropy, energy center migration, and spectral dispersion are calculated, weighted, fused, and transformed to generate interference probability, and then abnormal interference areas are identified.

[0052] S2.1 After obtaining the locally stored transmit reference signal as the matched filter template, the frequency domain matched filtering method is used to perform complex conjugate matching on the complex echo signals of each channel in the fast time domain through the matched filter template, so as to obtain the distance compression echo of the complex echo signals of each channel in the current pulse period after matched filtering.

[0053] Specifically, the matched filtering process involves: performing a Fast Fourier Transform (FFT) on the fast time-domain complex echo signal sequence acquired during the current pulse period to convert it to the frequency domain; simultaneously, inverting the digital sampling sequence of the corresponding transmit reference signal in the time domain, taking its complex conjugate, and then performing a FFT to obtain the matched filter frequency domain transfer function; then performing a complex multiplication operation on the frequency domain data of the echo signal and the matched filter frequency domain transfer function at each frequency point; and performing an inverse FFT on the multiplication result to restore it to the time domain, thereby obtaining the distance-compressed echo of the channel within the current pulse period.

[0054] It should be noted that the waveform parameters of the transmitted reference signal are completely consistent with those of the actual transmitted signal of the radar, and the transmitted reference signal can be acquired through real-time coupling.

[0055] For example, in one implementation: a directional coupling branch is connected in series between the forward power output port of the radar transmitter and a circulator or directional coupler. The coupling end of this branch is connected to the auxiliary receiving channel of the radar receiver front end via an RF cable. At the start of each transmit pulse, the auxiliary receiving channel performs down-conversion and analog-to-digital conversion on the coupled transmit signal sample using the same sampling clock and sampling frequency as the main receiving channel, obtaining a digital sampling sequence that is strictly synchronized with the current transmit pulse. The signal processor accurately determines the start time of the coupling acquisition window through a synchronization trigger signal. The acquisition window length covers the entire transmit pulse width and reserves at least one additional sampling point corresponding to a range resolution unit to ensure that the transmit signal falls completely within the acquisition window. After acquisition, the digital sampling sequence is used as the transmit reference signal digital sampling sequence for the current pulse period or the current coherent processing interval.

[0056] S2.2 Arrange the distance compression echoes corresponding to the same distance unit within multiple consecutive pulse cycles in slow time order to obtain a slow time series. Use the slow time series as columns of a matrix and the time series in the slow time series as rows to form a coherent accumulation data matrix.

[0057] It should be noted that: each column of the coherent accumulation data matrix represents the phase history (slow time series) of a certain fixed range cell under several consecutive pulses, which retains all Doppler phase information caused by the target motion; each row of the coherent accumulation data matrix represents the fast time sampling snapshot of all range cells within a certain pulse repetition period.

[0058] S2.3. Perform a discrete Fourier transform on each column of the coherent accumulation data matrix to map the frequency changes caused by the target motion to the Doppler domain. After obtaining the range-Doppler two-dimensional spectral response values ​​corresponding to each channel, calculate the spatial gain weights corresponding to each channel, and then perform weighted fusion of the range-Doppler two-dimensional spectral response values ​​to obtain the range-Doppler joint spectrum, forming the range-Doppler joint spectrum matrix.

[0059] The spatial gain weight for each channel is calculated using the following expression:

[0060]

[0061] In the formula, Indicates the first Spatial gain weight of the road channel This represents the modulo operation for complex numbers. Indicates the first Distance of the road passage - Doppler two-dimensional spectral response value, Indicates the total number of radar channels. Indicates the first Distance of the road passage - Doppler two-dimensional spectral response value, Represents small constants (e.g., );

[0062] The distance-Doppler joint spectrum is obtained, and its expression is:

[0063]

[0064] In the formula, This represents the distance-Doppler joint spectrum.

[0065] It should be noted that: spatial gain weight It reflects the proportion of signal energy that the current channel accounts for in the entire multi-channel receiving system. Channels with higher signal energy are assigned higher weights, thereby achieving an adaptive beamforming effect in space.

[0066] S2.4 Treat each range cell as an independent analysis object, extract the corresponding range-Doppler joint spectrum for any range cell, and then normalize each range-Doppler joint spectrum within the range cell using energy probability normalization to obtain the normalized energy corresponding to each Doppler frequency in the range cell.

[0067] It should be noted that: the range cell exists in the range-Doppler joint spectrum matrix, and each column of the range-Doppler joint spectrum matrix corresponds to a fixed range cell number; while the normalized energy is the probability distribution of the echo energy in the current range cell along the Doppler frequency dimension. The more concentrated the energy is in a few Doppler frequencies, the sharper the normalized energy distribution is, and vice versa.

[0068] S2.5. Based on normalized energy, the Doppler spectral entropy value corresponding to the distance cell is calculated using the information entropy method. Then, the Doppler spectral entropy values ​​are sorted according to time order to form a Doppler spectral entropy value sequence, and the difference in Doppler spectral energy distribution between adjacent time points in the sequence is calculated.

[0069] The Doppler spectral entropy value corresponding to the distance cell is calculated using the information entropy method, and the expression is as follows:

[0070]

[0071] In the formula, Distance unit In time The corresponding Doppler spectral entropy value, This indicates the number of Doppler frequencies contained in the current range cell (determined by counting the number of rows in the range-Doppler joint spectrum matrix). Distance unit In time No. Normalized energy of each Doppler frequency Represents the natural logarithm operation;

[0072] The difference in Doppler spectral energy distribution between adjacent time points in a sequence is calculated using the following expression:

[0073]

[0074] In the formula, Indicates time Relative to time The amount of entropy change, This represents the absolute value operation. Distance unit In time No. Normalized energy of each Doppler frequency Represents small constants (e.g., ).

[0075] It should be noted that the entropy change is calculated using KL divergence. In the calculation of entropy change, if the normalized energy value of a certain Doppler frequency position is zero at both the current and previous times, then the contribution of that term is directly set to zero, and no further division or logarithmic operations are performed.

[0076] S2.6. Based on the energy probability, calculate the Doppler energy center and energy center migration of the range cell, and then calculate the Doppler spectrum dispersion of the range cell based on the Doppler energy center and normalized energy.

[0077] The expression for calculating the Doppler energy center of the distance cell is as follows:

[0078]

[0079] In the formula, Distance unit In time Doppler energy center Indicates the first One Doppler frequency;

[0080] The energy center migration amount is expressed as:

[0081]

[0082] In the formula, Indicates time Relative to time The amount of energy center migration, Distance unit In time Doppler energy center This represents the absolute value operation;

[0083] The expression for calculating the Doppler spectrum dispersion of a range cell is:

[0084]

[0085] In the formula, Distance unit In time The degree of dispersion of the Doppler spectrum.

[0086] It should be noted that the degree of dispersion is mathematically equivalent to the second moment of the normalized energy distribution relative to the Doppler energy center. Its dimension is the square of the frequency unit. The larger the value, the wider the range of echo energy distribution in the Doppler frequency dimension and the lower the degree of energy concentration. Conversely, the smaller the value, the more concentrated the energy is near the Doppler energy center and the more compact the spectral structure.

[0087] S2.7. By using preset weights for Doppler spectrum entropy change, Doppler energy center migration, and Doppler spectrum dispersion, the entropy change, energy center migration, and Doppler spectrum dispersion are weighted and fused to obtain an evaluation value for unknown interference mutation.

[0088] The weighted fusion of entropy change, energy center migration, and Doppler spectrum dispersion is expressed as follows:

[0089]

[0090] In the formula, Distance unit In time Evaluation value of unknown interference mutations The weights represent the changes in Doppler spectral entropy. Indicates the Doppler energy center migration weights. The weights represent the degree of dispersion of the Doppler spectrum. Distance unit In time The degree of dispersion of the Doppler spectrum.

[0091] It should be noted that: before performing weighted fusion, all parameters should be normalized; the Doppler spectral entropy change weight is used to control the contribution of the Doppler spectral entropy change to the fusion result, the Doppler energy center migration weight is used to control the contribution of the energy center migration to the fusion result, and the Doppler spectrum dispersion weight is used to control the contribution of the Doppler spectrum dispersion change to the fusion result. All three weight values ​​are positive real numbers, and their sum is one.

[0092] For example, in one exemplary approach: if the interference to be suppressed mainly manifests as abrupt changes in spectral shape, a relatively large Doppler spectral entropy change weight value can be selected; if the interference mainly manifests as overall frequency drift, a relatively large Doppler energy center migration weight value can be selected; and if the interference mainly manifests as spectral broadening or splitting, a relatively large Doppler spectral dispersion weight value can be selected. As an exemplary range of values, the Doppler spectral entropy change weight can take any value between 0.3 and 0.5, the Doppler energy center migration weight can take any value between 0.2 and 0.4, and the Doppler spectral dispersion weight can take any value between 0.2 and 0.4.

[0093] Preferably, the present invention can make the weight of Doppler spectral entropy change 0.4, the weight of Doppler energy center migration 0.3, and the weight of Doppler spectrum dispersion 0.3. Therefore, it can maintain a balanced response to each feature quantity under various typical interference scenarios and avoid a single feature quantity from overdoing the evaluation value.

[0094] S2.8. Use an exponential function to perform probability transformation on the evaluation value of unknown interference mutations to obtain the interference probability.

[0095] Among them, the probability transformation of the evaluation value of unknown disturbance mutation is performed using an exponential function, and the expression is as follows:

[0096]

[0097] In the formula, Distance unit In time The probability of interference, This represents an exponential function.

[0098] It should be noted that: when the evaluation value of the unknown interference mutation approaches positive infinity, the interference probability approaches one, indicating that a highly significant mutation has occurred in the current range-Doppler joint spectrum, which is highly likely to originate from unknown interference; when the evaluation value of the unknown interference mutation approaches negative infinity, the interference probability approaches zero, indicating that the Doppler spectrum has almost no change, which is highly likely to originate from the target echo; when the evaluation value of the unknown interference mutation is equal to zero, the interference probability is equal to 0.5, indicating that it is impossible to determine whether interference exists.

[0099] Furthermore, the interference probability is compared with a preset interference mutation discrimination threshold. If the interference probability is greater than the preset interference mutation discrimination threshold, it is considered that there is an abnormal change in the range-Doppler joint spectrum corresponding to the current range cell, and it is marked as an unknown interference candidate region. If the interference probability is less than or equal to the preset interference mutation discrimination threshold, it is considered that there is no abnormal change in the range-Doppler joint spectrum corresponding to the current range cell, and it is marked as a normal candidate region.

[0100] It should be noted that the preset interference mutation discrimination threshold can be calculated in a standard test scenario where the ejection seat is stationary or in uniform linear motion and there are no interference sources around. The interference probability of each distance unit is calculated, the mean and standard deviation of the interference probability of each distance unit at each time are statistically analyzed, and then the mean is added to three times the standard deviation to generate the initial interference mutation discrimination threshold.

[0101] S3. Based on the three-axis velocity information and seat attitude angle information, obtain the theoretical target Doppler frequency and theoretical Doppler range, calculate the Doppler matching degree, and combine it with the Doppler entropy mutation evaluation index to form the target echo credibility.

[0102] S3.1 Read the initial beam direction vector of the antenna coordinate system of each channel of the four-way directional microstrip radar, establish the attitude rotation matrix according to the seat attitude angle information, and use the attitude rotation matrix to transform the initial beam direction vector to the navigation coordinate system to obtain the unit vector of the actual detection direction of the radar. Combine the three-axis velocity information to perform multiplication operation to obtain the radial motion velocity of the seat along the radar detection direction.

[0103] It should be noted that the initial beam direction vector was obtained through joint calibration using a vector network analyzer and optical alignment equipment when the radar was mounted on the seat. This vector is a three-dimensional unit column vector, and its three components represent the radar beam in the antenna coordinate system. axis, shaft and The direction cosine on the axis; while using the attitude rotation matrix to transform the initial beam direction vector to the navigation coordinate system is to perform matrix-vector multiplication operation between the attitude rotation matrix and the initial beam direction vector of the current channel. Specifically, the three rows of the attitude rotation matrix are multiplied by the three components of the initial beam direction vector one by one and then accumulated. The three results are sequentially formed into a new three-dimensional column vector. This vector is the actual detection direction unit vector in the navigation coordinate system after the initial beam direction vector of the radar in the antenna coordinate system is rotated by the seat attitude. Its three components represent the direction cosines of the actual beam pointing in the east, north and sky directions in the navigation coordinate system, respectively.

[0104] S3.2. Based on the radar operating wavelength and radial velocity, determine the theoretical target Doppler frequency and construct the theoretical target Doppler range.

[0105] The theoretical target Doppler frequency is determined by the following expression:

[0106]

[0107] In the formula, Indicates time The theoretical target Doppler frequency, Indicates the first Road channel antenna in time radial velocity, Indicates the radar operating wavelength. This represents the two-way Doppler coefficient generated by the radar transmit and receive paths.

[0108] The theoretical target Doppler range is constructed as follows:

[0109]

[0110] In the formula, Indicates time The corresponding theoretical target Doppler range, Indicates Doppler prediction error;

[0111] Regarding Doppler prediction error First, the attitude change between adjacent time intervals needs to be calculated. After obtaining the three-axis angular velocity of the seat, the Doppler frequency in the normal distance-Doppler joint spectrum is extracted. Then, the absolute difference between the theoretical target Doppler frequency and the actual Doppler frequency is calculated. The mean and standard deviation of the absolute difference are then statistically analyzed, and combined with... Principle settings.

[0112] It should be noted that: The design reason is that the radar signal undergoes two-way propagation in the transmission and reception paths, and the Doppler effect is accumulated twice; when the seat is stationary or the radial movement speed is zero, the theoretical target Doppler frequency is zero Hertz, which means that the echo signal does not have the Doppler frequency shift introduced by the seat movement.

[0113] S3.3 For any distance cell, the maximum value search method is used to traverse the theoretical target Doppler range, select the position of the maximum energy peak, obtain the candidate frequency of the moving target, and calculate the Doppler matching degree after obtaining the absolute difference between the candidate frequency of the moving target and the theoretical target Doppler frequency through absolute value operation.

[0114] The expression for calculating the Doppler matching degree is as follows:

[0115]

[0116] In the formula, Distance unit In time Doppler matching degree, Distance unit In time The absolute difference in frequency.

[0117] It should be noted that if multiple energy values ​​are the same and all are maximum values, the signal processor takes the average of the Doppler frequencies corresponding to these maximum value positions as the final candidate frequency.

[0118] Furthermore, the interference probability is extracted. Combined with Doppler matching degree The target confidence model is established, and the target echo confidence of the range cell is calculated.

[0119] The expression for calculating the target echo confidence level of the range cell is as follows:

[0120]

[0121] In the formula, Distance unit In time The credibility of the target echo.

[0122] It should be noted that: This is the complement of the interference probability, which represents the probability estimate that the current range cell is not interfered with at the current moment. The target echo confidence is the overall confidence level that the echo signal of the current range cell originates from the real target. The higher the Doppler matching degree, the better the echo frequency matches the seat kinematics prediction. The lower the interference probability, the more stable the echo Doppler spectrum is between adjacent moments and the fewer abrupt changes.

[0123] S4. Based on the target echo confidence and interference probability, construct the target retention weight to selectively suppress the range-Doppler joint spectrum to obtain the target enhancement spectrum. Based on the target enhancement spectrum, perform target actual range conversion and target removal to obtain the final ranging result.

[0124] Specifically, based on the target echo reliability and interference probability The target retention weights of the range cells are constructed, and the target retention weights are multiplied by the range-Doppler joint spectrum to obtain the target enhancement spectrum.

[0125] The target retention weights for constructing distance cells are generated based on a proportional relationship, and the specific expression is as follows:

[0126]

[0127] In the formula, Distance unit In time The target retains the weight.

[0128] It should be noted that: target retains weight. The calculation has automatic tendency capability and can cover three cases; for example, if the target has extremely high credibility and the probability of interference is extremely low (such as...). , ):

[0129]

[0130] This naturally resulted in a weight of 0.9, very close to 1, achieving the goal of maximizing retention;

[0131] When the target's credibility is extremely low and the probability of interference is extremely high (e.g.) , ):

[0132]

[0133] This naturally reduced the weight to 0.1, which is very close to 0, thus achieving strong interference suppression;

[0134] When the probability of the target and the probability of interference are extremely disparate (e.g.) , ):

[0135]

[0136] When the two are completely equal (e.g.) , ):

[0137]

[0138] Therefore, the target retains the weight. The computational design resulted in a continuous and smooth... Type mapper.

[0139] S4.1 After performing modulo squaring on the target enhancement spectrum, the actual echo power is obtained. The range cell with the largest actual echo power is selected as the target range cell, and the target range cell is converted into the actual target range.

[0140] It should be noted that if two or more range cells have the same maximum power value and are all global maximum values, the signal processor takes the range cell with the smaller index as the target range cell; and the radar range resolution can be generated by dividing the speed of electromagnetic waves in a vacuum by twice the bandwidth of the transmitted signal.

[0141] S4.2. Based on the actual distance of the target within several detection cycles, obtain the distance change between adjacent detection cycles, and calculate the theoretical distance change based on the radial motion velocity.

[0142] The theoretical distance change is calculated by multiplying the radial motion velocity by the detection period of the four-way directional microstrip radar.

[0143] S4.3. Subtract the theoretical distance change from the distance change and take the absolute value to obtain the distance change anomaly evaluation value. If the distance change anomaly evaluation value is less than the distance change allowable error threshold, the current ranging result is considered to conform to the seat motion law, and the current target distance output is maintained. If the distance change anomaly evaluation value is greater than or equal to the distance change allowable error threshold, the current ranging result is considered to have an abnormal jump, the current distance unit is removed, and the next effective target actual distance is searched again from the target enhancement spectrum.

[0144] Among these abnormal jumps are: strong fuselage reflection causing the sudden appearance of false targets at close range; mutual interference between the dual-seat radars causing false range peaks; transient spatial clutter causing random drift of range peaks; and large attitude changes causing beam shifts and resulting in erroneous range responses.

[0145] It should be noted that the allowable error threshold for distance variation can be set according to the radar range resolution;

[0146] For example, in one implementation: the permissible error threshold for distance variation is 0.5 meters. 0.5 meters is about one-third of the radar range resolution, which is sufficiently lenient to accommodate random fluctuations in ranging and inertial navigation velocity measurement noise during normal movement, avoiding the misjudgment and rejection of valid targets due to measurement noise; secondly, 0.5 meters is also sufficiently strict to effectively distinguish between continuous changes in the distance of normal targets and distance jumps caused by false targets when the ejection seat moves at a typical speed. Most false range peaks caused by fuselage reflection, multi-radar interference, or clutter usually manifest as distance changes greater than one meter within adjacent detection cycles.

[0147] In summary, by analyzing the Doppler energy distribution within the range cell, an unknown interference detection mechanism based on changes in Doppler spectral entropy, energy center migration, and spectral dispersion is established. This eliminates the need for pre-establishing a fixed interference template to identify anomalous echoes such as strong fuselage reflections, dual-seat radar interference, and transient spatial clutter, thus improving the radar's adaptability to unknown dynamic interference. Furthermore, by incorporating velocity and attitude information output from the inertial navigation module, a theoretical target Doppler range matching the real-time motion state of the seat is established, imbuing the target discrimination process with kinematic constraints and reducing the probability of false targets during high-speed descent, roll, and dive. Secondly, a soft mask suppression method based on target credibility and interference probability is employed, continuously adjusting according to the probability relationship between targets and interference in different range-Doppler cells, improving range-keeping capability under weak target conditions. Therefore, this invention enables the ejection seat altimeter radar to continuously output stable and reliable target range results under high-speed, large attitude change, and multi-interference environments, and improves the reliability of target interference elimination and radar ranging accuracy.

[0148] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A target interference elimination method based on radar Doppler features, characterized in that: include, The echo signal of the target area is collected, and then the three-axis velocity information and seat attitude angle information output by the seat inertial navigation module are obtained. After performing frequency domain matched filtering, range compression, and cross-pulse coherent accumulation on the echo signal to form a range-Doppler joint spectral matrix, the energy proportion of the spectral response of each channel is extracted and spatially weighted and fused. Then, the changes in Doppler spectral entropy, energy center migration, and spectral dispersion are calculated, weighted, fused, and transformed to generate interference probability, and then abnormal interference areas are identified. Based on the triaxial velocity information and seat attitude angle information, the theoretical target Doppler frequency and theoretical Doppler range are obtained. The Doppler matching degree is calculated, and combined with the Doppler entropy mutation evaluation index, the target echo credibility is formed. Based on the target echo confidence and interference probability, a target retention weight is constructed to selectively suppress the range-Doppler joint spectrum, resulting in a target enhancement spectrum. Based on the target enhancement spectrum, the target actual range conversion and target removal are performed to obtain the final ranging result.

2. The target interference elimination method based on radar Doppler features as described in claim 1, characterized in that: The acquisition of echo signals from the target area refers to the acquisition of echo signals from the space surrounding the ejection seat by four directional microstrip radars, resulting in complex echo signals from each channel in the fast time domain.

3. The target interference elimination method based on radar Doppler features as described in claim 1, characterized in that: The acquisition of the three-axis velocity information and seat attitude angle information output by the seat inertial navigation module refers to the acquisition of the three-axis velocity information and seat attitude angle information of the seat in the navigation coordinate system by the inertial navigation module installed on the ejection seat. The three-axis velocity information includes the navigation coordinate system. Axial velocity components, navigation coordinate system Axial velocity components, navigation coordinate system The axial velocity component and seat attitude angle information include seat pitch angle, seat roll angle, and seat yaw angle; the inertial navigation module includes a three-axis accelerometer and a three-axis gyroscope.

4. The target interference elimination method based on radar Doppler features as described in claim 1, characterized in that: The echo signal is subjected to frequency-domain matched filtering, range compression, and cross-pulse coherent accumulation to form a range-Doppler joint spectral matrix, as detailed below: After obtaining the locally stored transmit reference signal as the matched filter template, the frequency domain matched filtering method is used to perform complex conjugate matching on the complex echo signals of each channel in the fast time domain through the matched filter template, so as to obtain the distance compression echo of the complex echo signals of each channel in the current pulse period after matched filtering. The distance compression echoes corresponding to the same distance cell within multiple consecutive pulse cycles are arranged in slow time order to obtain a slow time series. The slow time series are used as columns of a matrix, and the time series in the slow time series are used as rows to form a coherent accumulation data matrix. A discrete Fourier transform is performed on each column of the coherent accumulation data matrix to map the frequency changes caused by the target motion to the Doppler domain. After obtaining the range-Doppler two-dimensional spectral response values ​​corresponding to each channel, the spatial gain weights corresponding to each channel are calculated. Then, the range-Doppler two-dimensional spectral response values ​​are weighted and fused to obtain the range-Doppler joint spectrum, forming the range-Doppler joint spectrum matrix.

5. The target interference elimination method based on radar Doppler features as described in claim 1, characterized in that: The process involves extracting the energy proportion of each channel's spectral response, performing spatial weighted fusion, and then calculating the changes in Doppler spectral entropy, energy center migration, and spectral dispersion, followed by weighted fusion and transformation to generate the interference probability. The details are as follows: Each range cell is treated as an independent analysis object. For any range cell, the corresponding range-Doppler joint spectrum is extracted. Then, the range-Doppler joint spectrum within each range cell is normalized using energy probability normalization to obtain the normalized energy corresponding to each Doppler frequency in the range cell. Based on normalized energy, the Doppler spectral entropy value corresponding to the distance cell is calculated using the information entropy method. Then, the Doppler spectral entropy values ​​are sorted in chronological order to form a Doppler spectral entropy value sequence, and the difference in Doppler spectral energy distribution between adjacent time points in the sequence is calculated. Based on the energy probability, the Doppler energy center and energy center migration of the range cell are calculated. Then, based on the Doppler energy center and normalized energy, the Doppler spectrum dispersion of the range cell is calculated. By using preset weights for Doppler spectrum entropy change, Doppler energy center migration, and Doppler spectrum dispersion, the entropy change, energy center migration, and Doppler spectrum dispersion are weighted and fused to obtain an evaluation value for unknown interference mutation. The probability of interference is obtained by performing a probability transformation on the evaluation value of unknown disturbance mutation using an exponential function.

6. The target interference elimination method based on radar Doppler features as described in claim 1, characterized in that: The re-identification of abnormal interference regions refers to comparing the interference probability with a preset interference mutation discrimination threshold. If the interference probability is greater than the preset interference mutation discrimination threshold, it is considered that there is an abnormal change in the range-Doppler joint spectrum corresponding to the current range unit, and it is marked as an unknown interference candidate region. If the interference probability is less than or equal to the preset interference mutation discrimination threshold, it is considered that there is no abnormal change in the range-Doppler joint spectrum corresponding to the current range unit, and it is marked as a normal candidate region.

7. The target interference elimination method based on radar Doppler features as described in claim 1, characterized in that: After obtaining the theoretical target Doppler frequency and the theoretical Doppler range, the Doppler matching degree is calculated as follows: The initial beam direction vector of each channel of the four-way directional microstrip radar in the antenna coordinate system is read. An attitude rotation matrix is ​​established based on the seat attitude angle information. The initial beam direction vector is then transformed to the navigation coordinate system using the attitude rotation matrix to obtain the unit vector of the actual detection direction of the radar. Multiplication is then performed in combination with the three-axis velocity information to obtain the radial motion velocity of the seat along the radar detection direction. Based on the radar operating wavelength and radial velocity, the theoretical target Doppler frequency is determined, and the theoretical target Doppler range is constructed. For any distance cell, the maximum value search method is used to traverse the theoretical target Doppler range, select the position of the maximum energy peak, obtain the candidate frequency of the moving target, and calculate the degree of Doppler matching by obtaining the absolute difference between the candidate frequency of the moving target and the Doppler frequency of the theoretical target through absolute value calculation.

8. The target interference elimination method based on radar Doppler features as described in claim 1, characterized in that: The target echo credibility index is formed by combining the Doppler entropy mutation evaluation index to extract the interference probability. Combined with the Doppler matching degree, the target credibility model is established, and the target echo credibility of the range cell is calculated.

9. The target interference elimination method based on radar Doppler features as described in claim 1, characterized in that: The target enhancement spectrum is obtained by constructing target retention weights based on target echo confidence and interference probability, and selectively suppressing the range-Doppler joint spectrum.

10. The target interference elimination method based on radar Doppler features as described in claim 1, characterized in that: The target actual distance conversion and target removal are performed based on the target enhancement spectrum to obtain the final ranging result, as detailed below: After performing modulo squaring on the enhanced spectrum of the target, the actual echo power is obtained. The range cell with the largest actual echo power is selected as the target range cell, and the target range cell is converted into the actual target range. Based on the actual distance of the target within several detection cycles, the distance change between adjacent detection cycles is obtained, and the theoretical distance change is calculated based on the radial motion velocity. The distance change is subtracted from the theoretical distance change and the absolute value is taken to obtain the distance change anomaly evaluation value. If the distance change anomaly evaluation value is less than the distance change allowable error threshold, the current ranging result is considered to conform to the seat motion law, and the current target distance output is maintained. If the distance change anomaly evaluation value is greater than or equal to the distance change allowable error threshold, the current ranging result is considered to have an abnormal jump, the current distance unit is removed, and the next effective target actual distance is searched again from the target enhancement spectrum. The abnormal jumps include the sudden appearance of false targets at close range due to strong fuselage reflection, the generation of false range peaks due to mutual interference between the dual-seat radars, the random drift of range peaks due to transient spatial clutter, and the generation of incorrect range responses after beam shift caused by large attitude changes.

Images