Radar high-speed maneuvering target adaptive detection method for suppressing blind speed sidelobe

Abstract

The invention discloses a radar high-speed maneuvering target adaptive detection method for suppressing a blind speed sidelobe, and the method comprises the steps: building a maneuvering target binary hypothesis test problem under a clutter background based on a high-speed maneuvering target radar echo data model; a two-step method is adopted to obtain GLRT test statistics of a binary hypothesis test problem, and detection of a strongest target in a scene is realized; the echo of the detected target is regarded as interference, and a maneuvering target binary hypothesis test problem containing interference terms is constructed; an iterative generalized matching subspace detector is deduced based on a two-step method, blind speed sidelobe suppression of a detected target is realized, and detection of a secondary strong target is realized; and when the output of the detector is smaller than a preset detection threshold, judging and detecting all moving targets in the scene. According to the iterative generalized matching subspace detector provided by the invention, uniform clutters can be effectively suppressed while coherent accumulation is carried out on a maneuvering target, blind speed sidelobe suppression of the target is realized, and the problems of false alarm and weak target leak detection caused by the blind speed sidelobe are effectively relieved.

Description

Technical Field

[0001] This invention belongs to the field of radar signal processing and moving target detection, and specifically relates to an adaptive radar target detection method for suppressing blind velocity sidelobes. Background Technology

[0002] Radar, with its all-weather, all-day capability and ability to simultaneously detect and track multiple targets, has become a crucial reconnaissance tool in modern information warfare. However, the presence of strong ground clutter and sea clutter can severely degrade radar target detection performance during radar detection missions. Adaptive radar target detection technology, with its excellent clutter suppression and target detection capabilities, is one of the important research directions in modern radar signal processing. Currently, radar target adaptive detectors, represented by the Generalized Likelihood Ratio Test (GLRT) and Adaptive Matched Filtering (AMF), can effectively suppress clutter while ensuring coherent accumulation of the target signal, significantly improving the target's output signal-to-clutter plus noise ratio (SCNR). However, classical adaptive processing methods typically require the target signal to be within the same range cell, which limits its effective accumulation time. For high-speed maneuvering target echoes exhibiting cross-range cell movement, classical adaptive processing algorithms struggle to achieve effective coherent accumulation and cannot obtain the desired accumulation gain. In recent years, scholars have proposed various long-term coherent accumulation algorithms to address the problem of targets easily moving across distances during long-term accumulation. Early researchers proposed the Keystone transform method, which uses a slow-time-dimensional scaling transform to decouple the linear coupling between fast and slow time, thereby eliminating linear movement of the target echo envelope. Later, a long-term coherent accumulation algorithm based on the Radon-Fourier transform (RFT) was proposed. This algorithm extracts the target signal along the target's trajectory based on the target's radial uniform motion characteristics and performs phase compensation to achieve coherent accumulation. The RFT can essentially be considered a generalized Doppler filter bank, exhibiting optimal detection performance for uniformly moving targets against a Gaussian white noise background. However, these long-term coherent accumulation algorithms primarily focus on improving the accumulation gain for targets moving across distances, with limited ability to suppress clutter. Subsequently, an adaptive RFT (ARFT) algorithm based on the maximum output SCNR principle was proposed, simultaneously achieving adaptive clutter suppression and long-term coherent accumulation for maneuvering targets. However, both the RFT and ARFT algorithms exhibit blind velocity sidelobe phenomena in their coherent accumulation results, further contributing to severe false alarms or missed detections of weak targets. Therefore, researching how to achieve adaptive detection of maneuvering targets in strong clutter backgrounds while effectively suppressing blind velocity sidelobe phenomena is of great significance for improving radar detection performance in multi-maneuvering target scenarios. Summary of the Invention

[0003] Purpose of the invention: The purpose of this invention is to provide an adaptive radar target detection method for suppressing blind velocity sidelobes, which is used to solve the problems of false alarms and missed detections of weak targets caused by blind velocity sidelobes of strong targets in multiple high-speed maneuvering target detection tasks under strong clutter background.

[0004] Technical solution: The adaptive detection method for high-speed maneuvering targets by radar with suppression of blind sidelobes as described in this invention includes the following steps:

[0005] (1) Based on the radar echo data model of high-speed maneuvering targets, construct a binary hypothesis testing problem for maneuvering targets in clutter background;

[0006] (2) The GLRT test statistic of the binary hypothesis test problem in step (1) is obtained by using a two-step method to realize the detection of the strongest target in the scene;

[0007] (3) Treat the echoes of the detected targets as interference and construct a binary hypothesis testing problem for maneuvering targets that includes interference terms;

[0008] (4) Based on the two-step method, the iterative generalized matching subspace detector is derived to achieve blind velocity sidelobe suppression of the detected target and further achieve the detection of the secondary target;

[0009] (5) Iterate through steps (3) and (4) until the detector output is less than the preset detection threshold, and declare that all moving targets in the scene have been detected.

[0010] Furthermore, the implementation process of step (1) is as follows:

[0011] Assuming the radar transmits a linear frequency modulated (LFM) signal and that the target moves radially at a uniform velocity relative to the radar, the pulse-compressed echo signal is represented as:

[0012]

[0013] in, For slow time, The pulse repetition interval, M represents the number of pulses within a coherent processing interval. To save time, The initial slant distance of the target. The radial velocity of the target, At the speed of light, For the target's complex range, For wavelength, Let be the radar bandwidth; further expressing the above equation in discrete matrix form, that is:

[0014]

[0015] in, This represents the distance cell number, where N is the number of distance cells. For range resolution, Because the coherent accumulation time required to detect weak, high-speed maneuvering targets is relatively long, the target's RCM effect cannot be ignored; motion parameters Define for unknown parameters For search parameters The corresponding target echo matrix;

[0016] The adaptive detection problem for maneuvering targets is transformed into the following binary hypothesis testing problem:

[0017]

[0018] in, and These represent the main data and auxiliary data in the received data, respectively. and These represent the clutter data in the detection window and the clutter window, respectively. For the target echo matrix, The target range.

[0019] Furthermore, the implementation process of step (2) is as follows:

[0020] According to the Neyman-Person criterion, the optimal detector is the likelihood ratio test:

[0021]

[0022] According to the GLRT criterion, respectively using and If the maximum likelihood estimate replaces the unknown parameter in the above equation, then the GLRT test statistic is:

[0023]

[0024] The above equation is simplified using a two-step method, that is, first assuming the covariance matrix... Given the data, derive the GLRT test statistic for the main data, and then obtain the result based on the auxiliary data. exist and The maximum likelihood estimate under the assumptions is used to replace the theoretical value in GLRT, resulting in the GAMF detector as follows:

[0025]

[0026] Represent GAMF as It is considered as a subspace detector when the projection matrix is ​​the identity matrix;

[0027] The clutter covariance matrix is ​​obtained from auxiliary data. Maximum likelihood estimate :

[0028]

[0029] Since the target's motion parameters are unknown, it is necessary to discretize the slant range-velocity parameter space and perform a traversal search according to the parameter grid, i.e.:

[0030]

[0031] in, To find the minimum slope distance, To achieve the minimum search speed, and These are the slant range search interval and the velocity search interval, respectively. and These are the distance and velocity numbers being searched, respectively; if and only if the search parameters and the actual target motion parameters When consistent, The detector has the highest output gain; For clutter covariance matrix It has CFAR characteristics, directly relating the detector output to the false alarm rate. Corresponding detection threshold Comparison is used to determine whether a target exists.

[0032] Furthermore, the relationship between the detection threshold and the false alarm rate is as follows: If the detector output If the objective is true, then the objective is declared to exist; otherwise, the objective is declared to not exist.

[0033] Furthermore, the implementation process of step (3) is as follows:

[0034] Construct a binary hypothesis testing problem for maneuvering targets that includes interference terms. Treat the detected targets as spurious interference with known structures. Achieve interference cancellation by projecting the received data onto the orthogonal complement space of the interference subspace:

[0035]

[0036] in, For the reason Interference items corresponding to the target were detected. The unknown magnitude of the interference term.

[0037] Furthermore, the implementation process of step (4) is as follows:

[0038] The GLRT test statistic for the principal data in the binary hypothesis testing problem after detecting the first target is as follows:

[0039]

[0040] Assuming Maximum likelihood estimation as follows:

[0041]

[0042] Assuming and Maximum likelihood estimate , as follows:

[0043]

[0044]

[0045] in, , , , , ;Will and Assuming and Maximum likelihood estimate , and Substituting into GLRT, we further obtain:

[0046]

[0047] Defined in Weighted projection matrix under inner product:

[0048]

[0049] in, It is the identity matrix. exist Projecting echo data onto a matrix using a weighted inner product The orthogonal complement space of the subspace spanned by the column will Rewritten as:

[0050]

[0051] The above formula first projects the echo data onto the interference term. The orthogonal complement space of the subspace spanned by the columns is used to eliminate interference from the main lobe and blind velocity sidelobes of the detected target in the echo, thereby achieving target clutter suppression and long-term coherent accumulation. Finally, an traversal search of the parameter space is performed. The output is higher than the detection threshold. If so, then the existence of the second target is declared, and the estimated values ​​of its motion parameters are as follows:

[0052]

[0053] detector Further promotion, assuming The target has been detected, the first... The estimated motion parameters of the detected targets are: The binary hypothesis testing problem is modified to the following form:

[0054]

[0055] in, This represents the interference term from the i-th detected target. The unknown magnitude of this interference term; therefore, the GLRT test statistic for the principal data corresponding to the above equation is:

[0056]

[0057] in, Similarly, it can be deduced that for detecting the first... The detector output when there are 1 target Represented as:

[0058]

[0059] Wherein, projection matrix Represented as:

[0060]

[0061] in, = For those already detected Disturbance terms for each target, Project the echo data onto The orthogonal complement of the subspace spanned by the columns eliminates Interference from the main lobe and blind sidelobes of the detected target; similarly, if The output exceeded the detection threshold. Then the first The estimated motion parameters of the target are as follows:

[0062] .

[0063] The present invention discloses a storage medium storing a computer program, which, when executed by at least one processor, implements the steps of the adaptive detection method for high-speed maneuvering radar targets with blind sidelobes as described above.

[0064] An electronic device according to the present invention includes a memory and a processor, wherein:

[0065] Memory is used to store computer programs that can run on a processor;

[0066] The processor is configured to, while running the computer program, execute the steps of the adaptive detection method for high-speed maneuvering radar targets with blind sidelobes as described above.

[0067] Beneficial effects: Compared with the prior art, the beneficial effects of this invention are as follows: For the problem of detecting multiple high-speed maneuvering targets in a uniform clutter background, the Iterative Generalized Matched Subspace Detector (IGMSD) proposed in this invention achieves adaptive detection of multiple targets one by one from strong to weak. While adaptively suppressing clutter, it effectively suppresses the detected blind velocity sidelobes by projecting the echo data onto the orthogonal complement of the subspace of the detected target echo matrix, thus solving the problem of false alarms and missed detection of weak targets caused by blind velocity sidelobes. Furthermore, the detector proposed in this invention has a constant false alarm rate (CFAR) characteristic for the clutter covariance matrix, avoiding the signal-to-noise ratio loss caused by additional CFAR processing. Attached Figure Description

[0068] Figure 1 This is a flowchart of the present invention;

[0069] Figure 2 A schematic diagram of the pulse compression-received echo data structure for adaptive detection of maneuvering targets;

[0070] Figure 3 This refers to the multi-target echo data after pulse compression in the simulation experiment;

[0071] Figure 4 The detection result of the IGMSD algorithm proposed in this invention for target 1;

[0072] Figure 5 The results of the IGMSD algorithm proposed in this invention for suppressing blind velocity sidelobes of target 1 and detecting target 2 are shown.

[0073] Figure 6The results of blind velocity sidelobe suppression for target 1 and target 2 using the IGMSD algorithm proposed in this invention are shown.

[0074] Figure 7 The graphs show the target detection probability curves of the IGMSD algorithm proposed in this invention and the classic moving target detection algorithm. Detailed Implementation

[0075] The present invention will now be further described in detail with reference to the accompanying drawings and specific embodiments.

[0076] like Figure 1 As shown, this invention proposes an adaptive radar target detection method for suppressing blind velocity sidelobes, and the specific implementation steps are as follows:

[0077] Step 1: Based on the radar echo data model of high-speed maneuvering targets, construct a binary hypothesis testing problem for moving targets across distance cells under uniform clutter background.

[0078] Assuming the radar transmit signal is a linear frequency modulated (LFM) signal, and that the target moves radially at a uniform velocity relative to the radar, the pulse-compressed echo signal can be expressed as:

[0079]

[0080] in, The initial slant distance of the target. The radial velocity of the target, At the speed of light, For the target's complex range, For wavelength, Let be the radar bandwidth. Further expressing the above equation in discrete matrix form, we get:

[0081]

[0082] in, The sampling interval is... For range resolution, The pulse repetition interval, Because weak, high-speed maneuvering targets have a large radial velocity relative to the radar and require a long coherent accumulation time, the range cell migration (RCM) amount far exceeds... The target's RCM effect cannot be ignored. Furthermore, motion parameters... Define for unknown parameters For search parameters The corresponding target echo matrix.

[0083] Figure 2 A schematic diagram of the pulse compression-received echo data structure for adaptive detection of maneuvering targets is shown, in which... For reference window length, The detection window length is defined as follows: The primary data and auxiliary data are defined as echo data within the detection window and reference window, respectively, and the detection window includes all range cells containing the moving target.

[0084] The adaptive detection problem for maneuvering targets is equivalent to the following binary hypothesis testing problem:

[0085]

[0086] in, and These represent the main data and auxiliary data in the received data, respectively. and These represent the clutter data in the detection window and the clutter window, respectively. For the target echo matrix, The target amplitude. The clutter vector within a single range cell. Assume that the mean is zero and the covariance matrix is... The complex Gaussian distribution vector is assumed, and it is assumed that the clutter vectors of each range cell are independently and identically distributed. Therefore, and The joint probability density can be expressed as:

[0087]

[0088] in, , , Represents the determinant of a matrix.

[0089] Step 2: Use a two-step method to obtain the GLRT test statistic for the binary hypothesis testing problem in Step 1. This enables adaptive detection of the strongest target in the scene.

[0090] According to the Neyman-Person criterion, the optimal detector is the likelihood ratio test:

[0091]

[0092] in, for Assuming and The joint probability density function, for Assuming The probability density function is given by the false alarm rate. The detection threshold is set below. Because... and Due to the unknown nature of the problem and the complexity of the numerator integral, the likelihood ratio test in the above formula cannot be practically applied. According to the GLRT criterion, respectively... and If the maximum likelihood estimate replaces the unknown parameter in the above equation, then the GLRT test statistic is:

[0093]

[0094] Since the numerator of the above equation contains multiple integrals, it is difficult to obtain an analytical solution for the GLRT test statistic. A two-step method can be used to simplify the equation: first, assume the covariance matrix... Given the data, derive the GLRT test statistic for the main data, and then obtain the result based on the auxiliary data. exist and The maximum likelihood estimate under the assumption is used to replace the theoretical value in GLRT, thus obtaining the AMF detector.

[0095] First, assume Given, then the main data The GLRT test statistic is:

[0096]

[0097] about The partial derivatives are:

[0098]

[0099] in, For conjugate operations, further let ,get Maximum likelihood estimate:

[0100]

[0101] Will The GLRT test statistic is substituted into the equation, and then expressed equivalently as:

[0102]

[0103] Since the above equation can be viewed as a generalized AMF detector for high-speed maneuvering targets across units, it is called a Generalized Adaptive Matched Filter (GAMF) detector. To maintain naming consistency with the iterative detector discussed below, GAMF will be re-expressed as follows: This refers to a special subspace detector whose projection matrix is ​​an identity matrix.

[0104] The second step is to obtain the clutter covariance matrix based on the auxiliary data. Maximum likelihood estimation :

[0105]

[0106] in, For the first Auxiliary data vectors for each reference unit.

[0107] Since the target's motion parameters are unknown, it is necessary to discretize the slant range-velocity parameter space and perform a traversal search according to the parameter grid, i.e.:

[0108]

[0109] in, To find the minimum slope distance, To achieve the minimum search speed, and These are the slant range search interval and the velocity search interval, respectively. and These are the distance and speed numbers being searched, respectively. The search parameters are valid if and only if... and the actual target motion parameters When consistent, the output gain of the GAMF detector is highest. Furthermore, For clutter covariance matrix It has CFAR characteristics, so the detector output can be directly compared with the false alarm rate. Given detection threshold The comparison determines whether a target exists. The relationship between the detection threshold and the false alarm rate is as follows: If the detector output If the objective is true, then the objective is declared to exist; otherwise, the objective is declared to not exist.

[0110] Step 3: Start iteratively performing adaptive detection of the remaining targets and blind velocity sidelobe suppression of the detected targets. Treat the echoes of the detected targets as interference and construct a binary hypothesis testing problem for maneuvering targets that includes interference terms.

[0111] Since the main lobe and blind velocity sidelobes of a detected target coexist, suppressing the blind velocity sidelobes can be achieved by treating the detected target as interference and suppressing it. Based on this idea, a binary hypothesis testing problem can be constructed as follows: the detected target is treated as a spurious target with a known structure, and interference is eliminated by projecting the received data onto the orthogonal complement space of the interference subspace.

[0112]

[0113] in, This is the pulse compression echo matrix of the detected target. The unknown magnitude of the interference term.

[0114] Step 4: For the first In the next iteration, the iterative generalized matching subspace detector is derived based on a two-step method. The detector projects the echo data onto the orthogonal complement of the subspace of the detected target echo matrix, thus achieving the detection of the target echo. Blind velocity sidelobe suppression of the first target and further achieving the second Adaptive detection of individual targets.

[0115] The above equation is a binary hypothesis testing problem (i.e. The GLRT test statistic for the principal data at time ( ) is:

[0116]

[0117] Assuming Maximum Likelihood Estimation (MLE) as follows:

[0118]

[0119] exist Assuming that, let The complex logarithm of the molecule is the objective function. :

[0120]

[0121] beg about and Taking the partial derivatives and setting them to zero, we get:

[0122]

[0123] get and The MLE is as follows:

[0124]

[0125]

[0126] in, , , , , .Will , and Substituting into GLRT, we further obtain:

[0127]

[0128] Defined in Weighted projection matrix under inner product:

[0129]

[0130] in, It is the identity matrix. exist Projecting echo data onto a matrix using a weighted inner product The orthogonal complement of the subspace spanned by the columns can be represented as... Rewritten as:

[0131]

[0132] The above formula can be understood as first projecting the echo data onto the interference term. The orthogonal complement space of the subspace spanned by the columns eliminates interference from the main lobe and blind velocity sidelobes of the detected target in the echo, thereby achieving target clutter suppression and long-term coherent accumulation. Finally, an traversal search of the parameter space is performed. The output is higher than the detection threshold. If so, then the existence of the second target is declared, and the estimated values ​​of its motion parameters are as follows:

[0133]

[0134] detector Further promotion, assuming The target has been detected, the first... The estimated motion parameters of the detected targets are: The binary hypothesis testing problem is modified to the following form:

[0135]

[0136] in, This represents the interference term from the i-th detected target. This represents the unknown magnitude of the interference term. Therefore, the GLRT test statistic for the principal data corresponding to the above equation is:

[0137]

[0138] in, Similarly, for detecting the first... The detector output when there are 1 target It can be represented as:

[0139]

[0140] Wherein, projection matrix It can be represented as:

[0141]

[0142] in, = For those already detected Disturbance terms for each target, Project the echo data onto The orthogonal complement of the subspace spanned by the columns eliminates Interference from the main lobe and blind sidelobes of the detected target. Similarly, if The output exceeded the detection threshold. Then the verdict will be pronounced. There exists a target, and the estimated motion parameters corresponding to that target are as follows:

[0143]

[0144] Step 5: Iterate through steps 3 and 4 until the detector output is less than the preset detection threshold. That is, to declare that all moving targets in the scene have been detected.

[0145] The present invention also provides a storage medium storing a computer program, which, when executed by at least one processor, implements the steps of the adaptive detection method for high-speed maneuvering radar targets with suppressed blind sidelobes as described above.

[0146] The present invention also provides an electronic device, including a memory and a processor, wherein: the memory is used to store a computer program that can run on the processor; the processor is used to execute, when running the computer program, the steps of the adaptive detection method for high-speed maneuvering radar targets with suppressed blind sidelobes as described above.

[0147] The invention is verified through multi-objective simulation experiments. The specific simulation system parameters are shown in Table 1.

[0148] Table 1 Radar System Parameters

[0149]

[0150] Two moving targets with different intensities were set in the simulation scenario, with radial velocities of 40 m / s and 88 m / s, respectively. The scene clutter was set to uniform Gaussian correlated clutter. The SCNRs after target pulse compression were 2.5 dB and -9.5 dB, respectively. The multi-target echo data after pulse compression are as follows: Figure 3As shown. The method of this invention is used to perform adaptive long-term coherent accumulation and blind velocity sidelobe suppression on multi-target echoes, and the results are as follows. Figures 4 to 6 As shown, the outputs of IGMSD0 and IGMSD1 detected two moving targets in sequence, and the outputs of both IGMSD1 and IGMSD2 effectively suppressed the blind velocity sidelobes of the detected targets. The output of IGMSD2 did not exceed the detection threshold, so it was determined that all targets had been detected and the iteration was terminated without any false alarms.

[0151] To further verify the effectiveness of the method of this invention, the target detection probabilities of the proposed IGMSD algorithm and the classic MTD, RFT, and ARFT algorithms were statistically analyzed through 200 Monte Carlo experiments. The false alarm rate of the detector was set to... The detection probability curves of different algorithms are shown in the figure. Figure 7 As shown, the method of the present invention has a higher detection probability for high-speed maneuvering targets in cluttered backgrounds.

[0152] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A radar adaptive detection method for high-speed maneuvering targets with suppressed blind sidelobes, characterized in that, Includes the following steps: (1) Based on the radar echo data model of high-speed maneuvering targets, construct a binary hypothesis testing problem for maneuvering targets in clutter background; (2) The GLRT test statistic of the binary hypothesis test problem in step (1) is obtained by using a two-step method to realize the detection of the strongest target in the scene; (3) Treat the echoes of the detected targets as interference and construct a binary hypothesis testing problem for maneuvering targets that includes interference terms; (4) Based on the two-step method, the iterative generalized matching subspace detector is derived to achieve blind velocity sidelobe suppression of the detected target and further achieve the detection of the secondary target; (5) Iterate through steps (3) and (4) until the detector output is less than the preset detection threshold, and declare that all moving targets in the scene have been detected.

2. The adaptive detection method for high-speed maneuvering radar targets with suppression of blind velocity sidelobes according to claim 1, characterized in that, The implementation process of step (1) is as follows: Assuming the radar transmits a linear frequency modulated (LFM) signal and that the target moves radially at a uniform velocity relative to the radar, the pulse-compressed echo signal is represented as: in, For slow time, The pulse repetition interval, M represents the number of pulses within a coherent processing interval. To save time, The initial slant distance of the target. The radial velocity of the target, At the speed of light, For the target's complex range, For wavelength, Let be the radar bandwidth; further expressing the above equation in discrete matrix form, that is: in, This represents the distance cell number, where N is the number of distance cells. For range resolution, Because the coherent accumulation time required to detect weak, high-speed maneuvering targets is relatively long, the target's RCM effect cannot be ignored; motion parameters Define for unknown parameters For search parameters The corresponding target echo matrix; The adaptive detection problem for maneuvering targets is transformed into the following binary hypothesis testing problem: in, and These represent the main data and auxiliary data in the received data, respectively. and These represent the clutter data in the detection window and the clutter window, respectively. For the target echo matrix, The target range.

3. The adaptive detection method for high-speed maneuvering radar targets with suppression of blind sidelobes according to claim 1, characterized in that, The implementation process of step (2) is as follows: According to the Neyman-Person criterion, the optimal detector is the likelihood ratio test: According to the GLRT criterion, respectively using and If the maximum likelihood estimate replaces the unknown parameter in the above equation, then the GLRT test statistic is: The above equation is simplified using a two-step method, that is, first assuming the covariance matrix... Given the data, derive the GLRT test statistic for the main data, and then obtain the result based on the auxiliary data. exist and The maximum likelihood estimate under the assumptions is used to replace the theoretical value in GLRT, resulting in the GAMF detector as follows: Represent GAMF as It is considered as a subspace detector when the projection matrix is ​​the identity matrix; The clutter covariance matrix is ​​obtained from auxiliary data. Maximum likelihood estimate : Since the target's motion parameters are unknown, it is necessary to discretize the slant range-velocity parameter space and perform a traversal search according to the parameter grid, i.e.: in, To find the minimum slope distance, To achieve the minimum search speed, and These are the slant range search interval and the velocity search interval, respectively. and These are the distance and velocity numbers being searched, respectively; if and only if the search parameters and the actual target motion parameters When consistent, The detector has the highest output gain; For clutter covariance matrix It has CFAR characteristics, directly relating the detector output to the false alarm rate. Corresponding detection threshold Comparison is used to determine whether a target exists.

4. The adaptive radar target detection method for suppressing blind sidelobes according to claim 3, characterized in that, The relationship between the detection threshold and the false alarm rate is as follows: If the detector output If the objective is true, then the objective is declared to exist; otherwise, the objective is declared to not exist.

5. The adaptive detection method for high-speed maneuvering radar targets with suppression of blind sidelobes according to claim 1, characterized in that, The implementation process of step (3) is as follows: Construct a binary hypothesis testing problem for maneuvering targets that includes interference terms. Treat the detected targets as spurious interference with known structures. Achieve interference cancellation by projecting the received data onto the orthogonal complement space of the interference subspace: in, For the reason Interference items corresponding to the target were detected. The unknown magnitude of the interference term.

6. The adaptive detection method for high-speed maneuvering radar targets with suppression of blind sidelobes according to claim 1, characterized in that, The implementation process of step (4) is as follows: The GLRT test statistic for the principal data in the binary hypothesis testing problem after detecting the first target is as follows: Assuming Maximum likelihood estimation as follows: Assuming and Maximum likelihood estimate , as follows: in, , , , , ;Will and Assuming and Maximum likelihood estimate , and Substituting into GLRT, we further obtain: Defined in Weighted projection matrix under inner product: in, It is the identity matrix. exist Projecting echo data onto a matrix using a weighted inner product The orthogonal complement space of the subspace spanned by the column will Rewritten as: The above formula first projects the echo data onto the interference term. The orthogonal complement space of the subspace spanned by the columns is used to eliminate interference from the main lobe and blind velocity sidelobes of the detected target in the echo, thereby achieving target clutter suppression and long-term coherent accumulation. Finally, an traversal search of the parameter space is performed. The output is higher than the detection threshold. If so, then the existence of the second target is declared, and the estimated values ​​of its motion parameters are as follows: detector Further promotion, assuming The target has been detected, the first... The estimated motion parameters of the detected targets are: The binary hypothesis testing problem is modified to the following form: in, This represents the interference term from the i-th detected target. The unknown magnitude of this interference term; therefore, the GLRT test statistic for the principal data corresponding to the above equation is: in, Similarly, it can be deduced that for detecting the first... The detector output when there are 1 target Represented as: Wherein, projection matrix Represented as: in, = For those already detected Disturbance terms for each target, Project the echo data onto The orthogonal complement of the subspace spanned by the columns eliminates Interference from the main lobe and blind sidelobes of the detected target; similarly, if The output exceeded the detection threshold. Then the first The estimated motion parameters of the target are as follows: 。 7. A storage medium, characterized in that, The storage medium stores a computer program that, when executed by at least one processor, implements the steps of the adaptive detection method for high-speed maneuvering radar targets with suppressed blind sidelobes as described in any one of claims 1 to 6.

8. An electronic device, characterized in that, Includes memory and processor, wherein: Memory is used to store computer programs that can run on a processor; A processor, configured to, while running the computer program, perform the steps of the adaptive detection method for high-speed maneuvering radar targets with suppressed blind sidelobes as described in any one of claims 1 to 6.

Images