Robust detection method and device for radar extended target under weighted generalized inverse gaussian clutter

Abstract

This invention relates to the field of radar target detection technology, and provides a robust method and apparatus for radar extended target detection under weighted generalized inverse Gaussian clutter. The invention uses a clutter prior information vector and a moment estimation order vector to estimate the clutter parameter vector; based on the training sample matrix, the data matrix to be detected, and the estimated values ​​of the clutter parameter vector, it estimates the inverse expectation of the texture component; based on the training sample matrix, the data matrix to be detected, the signal steering vector, and the estimated values ​​of the inverse expectation of the texture component, it constructs a test statistic; and based on the test statistic, it determines the target state. This invention solves the problems of radar extended target detection failing to meet real-time processing requirements and exhibiting poor robustness.

Description

Technical Field

[0001] This invention relates to the field of radar target detection technology, and in particular to a robust method and apparatus for radar extended target detection under weighted generalized inverse Gaussian clutter. Background Technology

[0002] In radar observation, targets with large size or complex structures are called extended targets. Their size is relatively large compared to the radar wavelength, or they possess complex shapes and scattering characteristics. Such targets exhibit large scattering cross-sections or complex scattering features in radar images or echoes. During radar signal processing, the statistical characteristics of observed clutter are generally considered to conform to a Gaussian distribution, allowing for target detection of extended targets based on the central limit theorem.

[0003] With the rapid development of modern radar technology, the performance indicators of radar systems have been continuously optimized, especially the significant improvement in resolution. When the radar operates at high resolution, the environment in which the observed clutter is located will transform into a complex Gaussian clutter environment, and the number of scatterers contained in a single resolution cell will decrease sharply. This will disrupt the conditions for the application of the central limit theorem, resulting in the observed clutter exhibiting significant non-Gaussian statistical characteristics.

[0004] In recent years, extended target detection in non-Gaussian clutter environments has become a research hotspot in the field of radar signal processing. On the one hand, in such scenarios, existing technologies use traditional detectors to detect received signals, determining whether to classify the received signal as a target based on a detection threshold. However, traditional detectors need to recalculate the detection threshold based on clutter amplitude each time they detect data, which cannot meet the real-time processing requirements of radar and has poor practicality. On the other hand, in actual working environments, signal mismatch problems caused by radar antenna sidelobe interference, multipath propagation effects, and array calibration errors are becoming increasingly prominent. These factors cause the radar target echo to not match the preset target steering vector, severely restricting the target detection performance of the radar system and resulting in poor robustness of radar extended target detection, making it difficult to meet practical application requirements.

[0005] Therefore, overcoming the shortcomings of the existing technology is an urgent problem to be solved in this technical field. Summary of the Invention

[0006] The technical problem to be solved by this invention is to provide a robust method and apparatus for radar extended target detection under weighted generalized inverse Gaussian clutter. The purpose is to utilize the prior statistical characteristics of clutter to achieve joint optimization of target energy accumulation and clutter suppression, thereby improving the robustness of radar extended target detection. By dynamically estimating target and clutter parameters, the detection performance of radar for extended targets in a compound Gaussian clutter environment is significantly improved, solving the problems of radar's inability to meet the real-time processing requirements and poor detection robustness when detecting extended targets.

[0007] The present invention adopts the following technical solution:

[0008] In a first aspect, the present invention provides a robust radar target detection method under weighted generalized inverse Gaussian clutter, comprising:

[0009] The clutter parameter vector is estimated using the clutter prior information vector and the moment estimation order vector.

[0010] Based on the estimated values ​​of the training sample matrix, the data matrix to be detected, and the clutter parameter vector, estimate the inverse expectation of the texture components;

[0011] A test statistic is constructed based on the training sample matrix, the data matrix to be detected, the signal steering vector, and the estimated value of the inverse expectation of the texture components;

[0012] The target state is determined based on the test statistic.

[0013] Furthermore, the estimation of the clutter parameter vector using the clutter prior information vector and the moment estimation order vector includes:

[0014] The first set of equations is constructed using the clutter prior information vector and the moment estimation order vector.

[0015] Solving the first system of equations yields the parameter moment estimates;

[0016] The parameter moment estimate is used as the initial value of the loss function;

[0017] The loss value of the loss function is minimized, and when the loss value meets a preset condition, the estimated clutter parameter vector is obtained.

[0018] Furthermore, the first set of equations is as follows:

[0019] ;

[0020] in, , This represents the weighted generalized inverse Gaussian distribution. dimensional weight vector, , Let represent the right-scale parameter vector corresponding to the weighted generalized inverse Gaussian distribution. , Let represent the left-scale parameter vector corresponding to the weighted generalized inverse Gaussian distribution. , This represents the shape parameter vector corresponding to the weighted generalized inverse Gaussian distribution; Representing the prior information vector of clutter First-order sample moments; , represents the moment estimation order vector, where Indicates the weighted number; , represents the clutter prior information vector, where This represents the length of the clutter prior information vector. Indicates transpose; Indicates the order is The second type of modified Bessel function.

[0021] Furthermore, the expression for the loss function is:

[0022] ;

[0023] in, , Represents the length of the vector. , , The number of intervals representing the range of clutter amplitude. This represents the amplitude value corresponding to each interval.

[0024] Further, estimating the inverse expectation of the texture components based on the estimated values ​​of the training sample matrix, the data matrix to be detected, and the clutter parameter vector includes:

[0025] The covariance matrix is ​​estimated iteratively using the training sample matrix;

[0026] Using the estimated value of the covariance matrix, the data matrix to be detected, and the estimated value of the clutter parameter vector, the inverse expectation of the texture components is estimated.

[0027] Further, constructing the test statistic based on the training sample matrix, the data matrix to be detected, the signal steering vector, and the estimated value of the inverse expectation of the texture components includes:

[0028] A test statistic is constructed using the estimated value of the covariance matrix, the data matrix to be detected, the signal steering vector, and the estimated value of the inverse expectation of the texture components.

[0029] Furthermore, determining the target state based on the test statistic includes:

[0030] The detection threshold is determined based on the test statistic and the preset false alarm probability.

[0031] If the test statistic is greater than the detection threshold, the target state is that the target exists; otherwise, the target state is that the target does not exist.

[0032] Furthermore, the expression for the test statistic is:

[0033] ;

[0034] in, , Represents absolute value. Indicates the first One data point to be detected. This indicates the conjugate transpose. Indicates the first The estimated value of the reciprocal expectation of the texture components in the data to be detected. Represents the signal steering vector. This represents the iterative estimate of the covariance matrix;

[0035] The expression for the detection threshold is:

[0036] ;

[0037] in, , For the number of Monte Carlo simulations, To preset the false alarm probability, For rounding operations, For sequence Arrange from largest to smallest The maximum value, , This indicates the first data point of the test data containing only clutter components. The first experiment The estimated value of each data point to be detected. This indicates the first data point of the test data containing only clutter components. The estimated value of the inverse expectation of the texture components in this experiment. This indicates the first data point of the test data containing only clutter components. The estimated value of the covariance matrix for this experiment.

[0038] Secondly, the present invention also provides a robust radar target detection device under weighted generalized inverse Gaussian clutter, comprising:

[0039] At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the processor for performing the weighted generalized inverse Gaussian clutter radar extended target robust detection method described in the first aspect.

[0040] Thirdly, the present invention also provides a non-volatile computer storage medium storing computer-executable instructions, which are executed by one or more processors to perform the weighted generalized inverse Gaussian clutter robust target detection method for radar as described in the first aspect.

[0041] Fourthly, a computer program product containing instructions is provided that, when executed on a computer or processor, causes the computer or processor to perform a robust radar target detection method under weighted generalized inverse Gaussian clutter as described in the first aspect.

[0042] Fifthly, the present invention also provides a robust detection system for extended radar targets under weighted generalized inverse Gaussian clutter, including a robust detection device for extended radar targets under weighted generalized inverse Gaussian clutter as described in the second aspect, and using the robust detection method for extended radar targets under weighted generalized inverse Gaussian clutter as described in the first aspect to complete the interaction of the robust detection device for extended radar targets under weighted generalized inverse Gaussian clutter as described in the second aspect.

[0043] Unlike existing technologies, the present invention has at least the following beneficial effects:

[0044] This invention constructs a test statistic by estimating the clutter parameter vector and the inverse expectation of the texture components after receiving the signal. Utilizing the prior statistical properties of clutter, it achieves joint optimization of target energy accumulation and clutter suppression, solving the signal mismatch problem and improving the robustness of radar extended target detection. It enables the test statistic to be directly obtained after the received signal is processed by the detector, eliminating the need to recalculate the detection threshold based on clutter amplitude each time, as is done with traditional detectors. The detector efficiently extracts extended target features through parallel processing, achieving integrated clutter suppression, signal accumulation, and target detection, meeting the real-time processing requirements of radar and demonstrating strong practicality. Attached Figure Description

[0045] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments of the present invention will be briefly described below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0046] Figure 1This is a flowchart illustrating a robust radar target detection method under weighted generalized inverse Gaussian clutter provided in an embodiment of the present invention.

[0047] Figure 2 This is a schematic diagram illustrating the principle of a robust radar target detection method under weighted generalized inverse Gaussian clutter provided in an embodiment of the present invention.

[0048] Figure 3 This is a flowchart illustrating step 10 provided in an embodiment of the present invention;

[0049] Figure 4 This is a flowchart illustrating step 20 provided in an embodiment of the present invention;

[0050] Figure 5 This is a flowchart illustrating step 40 provided in an embodiment of the present invention;

[0051] Figure 6 This is a comparison chart of the detection probability of the method of this embodiment of the invention and existing methods under different signal-to-noise ratios, provided by an embodiment of the invention.

[0052] Figure 7 This is a structural framework diagram of a radar extended target robust detection system under weighted generalized inverse Gaussian clutter provided in an embodiment of the present invention;

[0053] Figure 8 This is a schematic diagram of the architecture of a radar extended target robust detection device under weighted generalized inverse Gaussian clutter provided in an embodiment of the present invention. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0055] Unless the context otherwise requires, throughout the specification and claims, the term "comprising" is interpreted as openly inclusive, meaning "including, but not limited to." In the description of the specification, terms such as "one embodiment," "some embodiments," "exemplary embodiment," "example," "specific example," or "some examples" are intended to indicate that a particular feature, structure, material, or characteristic associated with that embodiment or example is included in at least one embodiment or example of this disclosure. The illustrative representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics mentioned may be included in any suitable manner in any one or more embodiments or examples; that is, although they may be incorporated into embodiments or examples using the above terms for reasons such as order and position, it does not limit them to be incorporated in combination by a single embodiment or example.

[0056] In the description of this invention, it should be understood that the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this disclosure and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this disclosure.

[0057] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of embodiments of this disclosure, unless otherwise stated, "a plurality of" means two or more. Furthermore, for example, the description may use the prefix "A" or "B" to describe the same type of nouns as two independent entities. In this case, the corresponding features defined with "A" and "B" are used only to distinguish between similar entities and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features.

[0058] In describing some embodiments, the terms "coupled," "coupled," and "connected," and their derivative expressions, may be used. For example, the term "connected" may be used in describing some embodiments to indicate that two or more components have direct physical or electrical contact with each other. Similarly, the term "coupled" may be used in describing some embodiments to indicate that two or more components have direct physical or electrical contact. However, the terms "connected" or "coupled" may also refer to two or more components that do not have direct contact with each other but still cooperate or interact with each other, such as "optical coupling," "wireless connection," etc. The embodiments disclosed herein are not necessarily limited to the scope of this invention.

[0059] In the description of this invention, the expression “A and / or B” (where A and B are used to formally represent specific features) will be used. The corresponding expression includes the following three combinations: only A, only B, and a combination of A and B.

[0060] As used in this invention, “about,” “approximately,” or “approximately” includes the stated value and the average value within an acceptable range of deviation from a particular value, wherein the acceptable range of deviation is determined by a person skilled in the art taking into account the measurement under discussion and the error associated with the measurement of the particular quantity (i.e., the limitations of the measurement system).

[0061] Furthermore, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0062] Example 1:

[0063] To solve the above problems, such as Figure 1 As shown, this embodiment of the invention provides a robust radar target detection method under weighted generalized inverse Gaussian clutter, including:

[0064] Step 10: Estimate the clutter parameter vector using the clutter prior information vector and the moment estimation order vector.

[0065] The following describes the robust radar target detection method under weighted generalized inverse Gaussian clutter according to an embodiment of the present invention, based on specific examples:

[0066] use 3D matrix Represents the data matrix to be detected. Number of system channels To expand the number of distance cells occupied by the target (i.e., the target expansion dimension).

[0067] This invention uses hypothesis testing as a statistical method to determine whether the data to be detected contains the target signal. In hypothesis testing... Below, data to be detected Contains only clutter components In the hypothesis testing framework of radar signal processing, "the data to be detected contains only clutter components" means that, under the no-target assumption, the received signal data does not contain target echo signals, but only background clutter; this clutter may originate from various non-target signal sources such as ground clutter, sea clutter, and rain clutter. In this case, the null hypothesis... This indicates that the data to be detected contains only clutter, while the alternative hypothesis... This indicates that the data contains both clutter and target signals. By performing statistical tests on the data, it can be determined whether to accept the null hypothesis (i.e., determine that no target exists) or reject the null hypothesis (i.e., determine that a target exists).

[0068] Assuming clutter components Each column is independent and identically distributed, let the i-th column be denoted as . Listed as , It follows a pattern with a mean of zero and a covariance matrix of... The complex Gaussian distribution is denoted as . ,in, It is the first Data to be detected The clutter texture, corresponding to the probability density function is: . , representing the weighted generalized inverse Gaussian distribution. A weight vector of dimension; , represents the right-scale parameter vector corresponding to the weighted generalized inverse Gaussian distribution; , represents the left-scale parameter vector corresponding to the weighted generalized inverse Gaussian distribution; , represents the shape parameter vector corresponding to the weighted generalized inverse Gaussian distribution, and ,in Indicates the weighted number. This indicates transpose.

[0069] In hypothesis testing Down Includes clutter components and signal components , signal component Can be written as ,in, dimensional vector Represents the signal steering vector. dimensional vector Indicates signal amplitude. This indicates the transpose. In a real-world environment, the covariance matrix of clutter... It is usually unknown. For estimation Using training samples, assuming that there exists There are training samples, denoted as _n_. , . Contains only clutter components Therefore, the problem to be detected can be represented by the following binary hypothesis test:

[0070]

[0071] in, , .

[0072] Before step 10, the following matrix also needs to be constructed:

[0073] Signal steering vector (Dimension is) ), data matrix to be detected (Dimension is) Training sample matrix (Dimension is) ) and clutter prior information vector (Dimension is) ),in, , , indicating the first One data point to be detected. , , indicating the first One training sample data, This represents prior clutter information data arranged from smallest to largest. Number of system channels Indicates the target's expanded dimension. Indicates the number of training samples. This represents the length of the clutter prior information vector. This indicates transpose.

[0074] Step 20: Estimate the inverse expectation of the texture components based on the training sample matrix, the data matrix to be detected, and the estimated values ​​of the clutter parameter vector.

[0075] In step 20, the inverse expectation of the texture component is an estimate of the inverse expectation of the texture component. This process will be explained below.

[0076] Step 30: Construct a test statistic based on the training sample matrix, the data matrix to be detected, the signal steering vector, and the estimated value of the inverse expectation of the texture component.

[0077] The process of constructing the test statistic will be explained below.

[0078] Step 40: Determine the target state based on the test statistic.

[0079] This invention constructs a test statistic by estimating the clutter parameter vector and the inverse expectation of the texture components after receiving the signal. Utilizing the prior statistical properties of clutter, it achieves joint optimization of target energy accumulation and clutter suppression, solving the signal mismatch problem and improving the robustness of radar extended target detection. It enables the test statistic to be directly obtained after the received signal is processed by the detector, eliminating the need to recalculate the detection threshold based on clutter amplitude each time, as is done with traditional detectors. The detector efficiently extracts extended target features through parallel processing, achieving integrated clutter suppression, signal accumulation, and target detection, meeting the real-time processing requirements of radar and demonstrating strong practicality.

[0080] like Figure 2 The diagram shown is a schematic representation of the robust radar target detection method under weighted generalized inverse Gaussian clutter according to an embodiment of the present invention. The following is a detailed description of the robust radar target detection method under weighted generalized inverse Gaussian clutter according to an embodiment of the present invention:

[0081] The composite Gaussian model is a statistical modeling framework for describing non-Gaussian clutter. This model decomposes clutter into a product of texture and speckle components. The choice of texture component distribution directly affects model accuracy. In traditional research, models such as the K-distribution, inverse Gamma distribution, and generalized inverse Gaussian distribution are widely used for texture component modeling; however, these models still have limitations in describing complex clutter environments. This invention's embodiment fits the clutter distribution based on a weighted generalized inverse Gaussian distribution model, effectively improving the accuracy of clutter characteristic descriptions in complex environments.

[0082] In one embodiment, such as Figure 3 As shown, step 10 includes:

[0083] Step 101: Construct the first set of equations using the clutter prior information vector and the moment estimation order vector.

[0084] The first set of equations is a set of equations for estimating the initial moments of clutter parameter vectors. In one embodiment, the first set of equations is:

[0085] ;

[0086] in, , This represents the weighted generalized inverse Gaussian distribution. dimensional weight vector, , Let represent the right-scale parameter vector corresponding to the weighted generalized inverse Gaussian distribution. , Let represent the left-scale parameter vector corresponding to the weighted generalized inverse Gaussian distribution. , This represents the shape parameter vector corresponding to the weighted generalized inverse Gaussian distribution; Represents the gamma function; Representing the prior information vector of clutter First-order sample moments; , represents the moment estimation order vector, where Indicates the weighted number; , represents the clutter prior information vector, where This represents the length of the clutter prior information vector. Indicates transpose; Indicates the order is The second type of modified Bessel function is expressed as: .

[0087] Step 102: Solve the first set of equations to obtain the parameter moment estimate.

[0088] The specific method for solving the first system of equations can be selected by those skilled in the art based on the specific application scenario; in one optional embodiment, the system of equations is solved using the trust region dogleg method, specifically by calling the fsolve function in MATLAB software, and the resulting parameter moment estimate is denoted as... ,in, , , , .

[0089] Step 103: Determine the parameter moment estimate as the initial value of the loss function.

[0090] The loss function is selected by those skilled in the art based on the specific application scenario; in one embodiment, the expression of the loss function is:

[0091] ;

[0092] in, , Represents the length of the vector. , ;in, express Medium to large and less than or equal to A vector composed of data, This represents the length of the vector. In this embodiment of the invention, the clutter amplitude range is... Evenly divided, The number of intervals representing the range of clutter amplitude, i.e., the number of evenly divided intervals. This represents the amplitude value corresponding to each interval. .

[0093] Step 104: Minimize the loss value of the loss function. When the loss value meets the preset conditions, the estimated clutter parameter vector is obtained.

[0094] The specific methods for setting the conditions and minimizing the loss function are to be selected by those skilled in the art based on the specific application scenario, and are not limited here.

[0095] In an alternative embodiment, the parameter moment estimation is performed. As a loss function The initial value is used to minimize the loss function using the Nelder-Mead algorithm, specifically by calling the fminsearch function in MATLAB. The resulting solution is denoted as... ,in, , , , .

[0096] The final clutter parameter vector is represented as follows: Dimension is .in, , representing the texture distribution corresponding to dimensional weight vector , representing the right scale parameter vector corresponding to the texture distribution; , representing the left-scale parameter vector corresponding to the texture distribution; This represents the shape parameter vector corresponding to the texture distribution, and , The weighted number is represented as the clutter order vector. Dimension is The numerical values ​​are selected by those skilled in the art based on the specific application scenario, and are not limited here.

[0097] like Figure 4 As shown, step 20 includes:

[0098] Step 201: Iteratively estimate the covariance matrix using the training sample matrix.

[0099] Wherein, the covariance matrix is ​​expressed as Its iterative estimate is expressed as The dimensions are all superscript Indicates an estimate.

[0100] In one embodiment, the expression for the iterative estimate of the covariance matrix is:

[0101]

[0102] in, , , Number of system channels Represents the trace of a matrix. Indicates the maximum number of iterations; Iteration The estimated value of the covariance matrix after the second iteration. , , for The initial values ​​of the intermediate matrix and the covariance matrix are: , This indicates the conjugate transpose. Indicates training samples, This represents a preset value indicating the number of training samples.

[0103] Step 202: Using the estimated value of the covariance matrix, the data matrix to be detected, and the estimated value of the clutter parameter vector, estimate the inverse expectation of the texture components.

[0104] Wherein, the inverse expectation of the texture component is expressed as dimensional vector The corresponding estimated value is expressed as ,in, , , indicating the first An estimate of the inverse expectation of the texture component in the data to be detected. In one embodiment, the estimate of the inverse expectation of the clutter texture component in the data to be detected is:

[0105] ;

[0106] in, , , indicating calculation Process parameters.

[0107] In one embodiment, step 30 includes:

[0108] A test statistic is constructed using the estimated value of the covariance matrix, the data matrix to be detected, the signal steering vector, and the estimated value of the inverse expectation of the texture components.

[0109] In one embodiment, the expression for the test statistic is:

[0110] ;

[0111] in, , Represents absolute value. Indicates the first One data point to be detected. This indicates the conjugate transpose. Indicates the first The estimated value of the reciprocal expectation of the texture components in the data to be detected. Represents the signal steering vector. This represents the iterative estimate of the covariance matrix.

[0112] After determining the test statistic, such as Figure 5 As shown, step 40 includes:

[0113] Step 401: Determine the detection threshold according to the test statistic and the preset false alarm probability.

[0114] The preset false alarm probability, or the preset value of the false alarm probability, is selected by those skilled in the art based on the specific application scenario. The false alarm probability refers to the probability that a radar system mistakenly identifies a target as present when there is no target. The higher the detection threshold, the lower the false alarm probability. The method of this embodiment can achieve constant false alarm rate (CFAR) detection.

[0115] In one embodiment, the expression for the detection threshold is:

[0116] ;

[0117] in, , For the number of Monte Carlo simulations, To preset the false alarm probability, For rounding operations, For sequence Arrange from largest to smallest The maximum value, , This indicates the first data point of the test data containing only clutter components. The first experiment The estimated value of each data point to be detected. This indicates the first data point of the test data containing only clutter components. The estimated value of the inverse expectation of the texture components in this experiment. This indicates the first data point of the test data containing only clutter components. The estimated value of the covariance matrix for this experiment.

[0118] Step 402: If the test statistic is greater than the detection threshold, the target state is that the target exists; otherwise, the target state is that the target does not exist.

[0119] Finally, compare the test statistic with the detection threshold. If the test statistic is greater than the detection threshold, the target is determined to exist; otherwise, the target is determined to not exist.

[0120] The effects of the present invention will be further explained below with reference to simulation experiments.

[0121] like Figure 6 The diagram shows a comparison of the detection probabilities of the method of this invention and existing methods under different signal-to-noise ratios. To simplify calculations, let the weighting factor... Number of radar system channels Target expanded dimension Dimension of signal subspace Number of training samples The number of iterations for estimation is 4, and the signal steering vector... ,in, represents an imaginary number, The normalized Doppler frequency of the target is set to 0.5.

[0122] The clutter prior information is the measured data acquired by the radar system. The clutter data of the 13th range cell is selected as the clutter prior information vector. clutter prior information length The target is located in distance cells 13 to 14, and the training sample data is located in distance cells 5 to 12 and 15 to 22.

[0123] After signal mismatch occurs, the actual signal steering vector is Signal amplitude vector After being randomly generated, the values ​​are kept constant, and the initial values ​​for clutter parameter moment estimation are set to... , , , The moment estimation order vector is set as The maximum number of iterations for both the fsolve and fminsearch functions is set to 100,000.

[0124] In the simulation experiment, the actual clutter covariance matrix can be obtained from... This is obtained by following step 201. The false alarm probability is set to... The signal-to-noise ratio is defined as follows: The signal-to-noise ratio was set to 30 dB, and the mismatch was defined as follows: .

[0125] Assume the mismatch is In order to obtain In the Doppler domain Uniform sampling of 10,000 points , For each Generate candidate steering vectors For each Calculate its with Mismatch between Choose to satisfy of As the true normalized Doppler frequency, and to determine the final true steering vector as .

[0126] The expression corresponding to the existing technical method used for comparison is as follows:

[0127] ;in, , .

[0128] from Figure 6 As can be seen, the method of this embodiment of the invention has a performance improvement of about 0.1 compared with the two-step generalized likelihood ratio test detector (GLRT) in the prior art when the mismatch is 0.6. That is, under the premise that the mismatch remains unchanged, the detection probability is improved by 0.1 when the signal-to-noise ratio is 30dB.

[0129] like Figure 7 As shown, this embodiment of the invention also provides a robust radar target detection system under weighted generalized inverse Gaussian clutter, comprising:

[0130] Data construction module: used to construct the signal matrix, the data matrix to be detected, the training sample matrix, and the clutter prior information vector.

[0131] Clutter parameter estimation module: Used to estimate clutter parameter vectors using clutter prior information vectors.

[0132] Covariance matrix estimation module: Used to estimate the covariance matrix using training samples.

[0133] Texture estimation algorithm module: used to calculate the estimated value of the inverse expectation of clutter texture components using the joint probability density function of the data to be detected.

[0134] Test statistic calculation module: Used to construct test statistics using the data to be tested, signal steering vector, clutter covariance matrix, and the inverse expectation of clutter texture components.

[0135] Detection threshold determination module: used to determine the detection threshold based on the false alarm probability.

[0136] The target decision module compares the test statistic and the detection threshold to determine whether the target exists. If the test statistic is greater than the detection threshold, the target is determined to exist; otherwise, the target is determined not to exist.

[0137] Example 2:

[0138] like Figure 8 The diagram shown is a schematic representation of an architecture of a robust radar target detection device under weighted generalized inverse Gaussian clutter according to an embodiment of the present invention. This robust radar target detection device under weighted generalized inverse Gaussian clutter includes one or more processors 21 and a memory 22. Figure 8 Take a processor 21 as an example.

[0139] Processor 21 and memory 22 can be connected via a bus or other means. Figure 8 Taking the example of a connection between China and Israel via a bus.

[0140] The memory 22, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs and non-volatile computer-executable programs, such as the robust detection method for extended radar targets under weighted generalized inverse Gaussian clutter in this embodiment. The processor 21 executes the robust detection method for extended radar targets under weighted generalized inverse Gaussian clutter by running the non-volatile software programs and instructions stored in the memory 22.

[0141] Memory 22 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, memory 22 may optionally include memory remotely located relative to processor 21, which can be connected to processor 21 via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0142] The program instructions / modules are stored in the memory 22. When executed by one or more processors 21, they perform the robust detection method for extended radar targets under weighted generalized inverse Gaussian clutter as described in the above embodiments. For example, they perform each step of the robust detection method for extended radar targets under weighted generalized inverse Gaussian clutter as described in the above embodiments of the present invention.

[0143] This invention also provides a non-volatile computer storage medium storing computer-executable instructions that are executed by one or more processors, for example... Figure 8 A processor 21 may enable one or more of the processors to execute the robust detection method for extended radar targets under weighted generalized inverse Gaussian clutter as described in the specific embodiments of the present invention. For example, it may execute the various steps of the robust detection method for extended radar targets under weighted generalized inverse Gaussian clutter as described in the embodiments of the present invention above; it may also implement... Figure 8 The various modules and units described above; or the robust detection method for extended radar targets under weighted generalized inverse Gaussian clutter as described in the specific embodiments of the present invention, for example, executing the various steps of the robust detection method for extended radar targets under weighted generalized inverse Gaussian clutter as described above in the embodiments of the present invention; can also be implemented Figure 8 The various modules and units mentioned above.

[0144] It is worth noting that the information interaction and execution process between the modules and units in the above-mentioned device and system are based on the same concept as the processing method embodiment of the present invention. For details, please refer to the description in the method embodiment of the present invention, and will not be repeated here.

[0145] Those skilled in the art will understand that all or part of the steps in the various methods of the embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, which may include: read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk, etc.

[0146] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A weighted generalized inverse radar clutter below Gaussian extended target robust detection method, characterized in that, include: The first set of equations is constructed using the clutter prior information vector and the moment estimation order vector; Solve the first set of equations to obtain the parameter moment estimate; determine the parameter moment estimate as the initial value of the loss function; minimize the loss value of the loss function, and when the loss value satisfies a preset condition, obtain the estimated clutter parameter vector; The first set of equations is as follows: ； in, , This represents the weighted generalized inverse Gaussian distribution. dimensional weight vector, , Let represent the right-scale parameter vector corresponding to the weighted generalized inverse Gaussian distribution. , Let represent the left-scale parameter vector corresponding to the weighted generalized inverse Gaussian distribution. , This represents the shape parameter vector corresponding to the weighted generalized inverse Gaussian distribution; Representing the prior information vector of clutter First-order sample moments; , represents the moment estimation order vector, where Indicates the weighted number; , represents the clutter prior information vector, where This represents the length of the clutter prior information vector. Indicates transpose; Indicates the order is The second type of modified Bessel function; Based on the estimated values ​​of the training sample matrix, the data matrix to be detected, and the clutter parameter vector, estimate the inverse expectation of the texture components; A test statistic is constructed based on the training sample matrix, the data matrix to be detected, the signal steering vector, and the estimated value of the inverse expectation of the texture components; The target state is determined based on the test statistic.

2. The method of claim 1, wherein, The expression for the loss function is: ； wherein, , denotes the length of the vector, , , denotes the number of intervals representing the clutter amplitude range, denotes the amplitude value corresponding to each interval.

3. The method of claim 1, wherein, The step of estimating the inverse expectation of the texture components based on the estimated values ​​of the training sample matrix, the data matrix to be detected, and the clutter parameter vector includes: The covariance matrix is ​​estimated iteratively using the training sample matrix; Using the estimated value of the covariance matrix, the data matrix to be detected, and the estimated value of the clutter parameter vector, the inverse expectation of the texture components is estimated.

4. The robust detection method of weighted generalized inverse radar with Gaussian clutter and extended target according to claim 3, characterized in that, The construction of the test statistic based on the estimated values ​​of the training sample matrix, the data matrix to be detected, the signal steering vector, and the inverse expectation of the texture components includes: A test statistic is constructed using the estimated value of the covariance matrix, the data matrix to be detected, the signal steering vector, and the estimated value of the inverse expectation of the texture components.

5. The method of claim 4, wherein, Determining the target state based on the test statistic includes: The detection threshold is determined based on the test statistic and the preset false alarm probability. If the test statistic is greater than the detection threshold, the target state is that the target exists; otherwise, the target state is that the target does not exist.

6. The method of claim 5, wherein, The expression for the test statistic is: ； in, , Represents absolute value. Indicates the first One data point to be detected. This indicates the conjugate transpose. Indicates the first The estimated value of the reciprocal expectation of the texture components in the data to be detected. Represents the signal steering vector. This represents the iterative estimate of the covariance matrix; The expression for the detection threshold is: ； in, , For the number of Monte Carlo simulations, To preset the false alarm probability, For rounding operations, For sequence Arrange from largest to smallest The maximum value, , This indicates the first data point of the test data containing only clutter components. The first experiment The estimated value of each data point to be detected. This indicates the first data point of the test data containing only clutter components. The estimated value of the inverse expectation of the texture components in this experiment. This indicates the first data point of the test data containing only clutter components. The estimated value of the covariance matrix for this experiment.

7. A device for robust detection of radar extended targets in clutter with weighted generalized inverse, characterized in that, The weighted generalized inverse Gaussian clutter radar extended target robust detection device includes at least one processor and a memory, which are connected via a data bus. The memory stores instructions that can be executed by the at least one processor. After being executed by the processor, the instructions are used to implement the weighted generalized inverse Gaussian clutter radar extended target robust detection method according to any one of claims 1-6.

8. A non-transitory computer storage medium, comprising, The computer storage medium stores computer-executable instructions, which are executed by one or more processors to perform the robust radar target detection method under weighted generalized inverse Gaussian clutter as described in any one of claims 1-6.

Images