Robust Intelligent Target Detection Method under Structured Interference and Clutter

Abstract

The present invention discloses a robust intelligent target detection method under structured interference and clutter, belonging to the field of wideband radar signal processing. In the case of the existence of external interference and a small amount of uniform auxiliary data, aiming at the problem that it is difficult to balance the detection performance and mismatch robustness in the detection of range-extended targets by multi-channel wideband radars, the skew-symmetric structure information of the clutter covariance matrix is fully exploited, reducing the requirement for the amount of auxiliary data and improving the estimation accuracy of the unknown clutter covariance matrix, providing strong support for realizing the adaptive detection of targets under the condition of small samples with interference; a closed-form robust intelligent target detection method under structured interference and clutter is constructed, while ensuring the constant false alarm rate characteristic, taking into account the computational complexity, intelligent anti-interference, mismatch robustness, detection performance and other multi-faceted requirements of the algorithm, improving the detection performance of multi-channel wideband radars for weak targets and mismatch targets in complex interference environments, and having the value of popularization and application.

Description

Technical Field

[0001] The present invention belongs to the field of radar signal processing, and particularly relates to a robust intelligent target detection method under structured interference and clutter. Background Art

[0002] With the increase of radar bandwidth, its range resolution is further improved. Wideband radars are widely used in modern military and civilian fields such as anti-jamming, anti-reconnaissance, precise detection and imaging, high-precision tracking, and target recognition. The adaptive detection of range-extended targets for wideband radars has become one of the hot issues in the radar community. Different from the fact that the target echo signal of a narrowband radar usually occupies only one range resolution cell, the energy of the target scattering points of a wideband radar may spread to adjacent range cells, presenting as a "one-dimensional range profile" and forming a range-extended target. If the point target detection method is still used to detect the echo signal for each single range cell and the statistical characteristics of the background clutter are estimated by sampling the adjacent range cells; on the one hand, the strong scattering point energy of the range-extended target is likely to leak into the adjacent range cells, resulting in signal contamination, and further causing a masking effect on the target signal of each single range cell to be detected, making the point target detection method ineffective; on the other hand, in practical applications, radar detection faces a complex electromagnetic environment, and there may be natural or man-made interference sources such as electronic countermeasure signals or various civilian electromagnetic signals. In addition, the environment where the target is located is complex and changeable, making the non-uniformity of the background clutter increase, and the number of pure clutter auxiliary data that satisfies independent and identically distributed is relatively limited. Compared with narrowband radars, this problem is particularly prominent in the target detection scenario of wideband radars, resulting in the existing range-extended target detection methods being difficult to achieve ideal detection effects.

[0003] In addition, multi-channel adaptive target detection under Gaussian clutter with an unknown covariance matrix has always been a hot research topic. It is usually assumed that the clutter components in the observation data (also known as primary data) from multiple range cells to be detected have the same clutter covariance matrix as the reference range cell data containing only pure clutter (also known as auxiliary data), and it is assumed that there is a set of auxiliary data without target signals available to estimate the unknown clutter covariance matrix. In practical application scenarios, due to reasons such as wavefront distortion and array calibration errors, the target signal steering vector may be mismatched. For application scenarios such as radar search mode, the detector needs to have strong robustness to mismatched signals; for the commonly used rank-1 signal model, the target steering vector is fixed and completely known, making it difficult to handle the aforementioned mismatched problems. If a detector is constructed based on the primary data of multiple range cells to be detected for range-extended targets with a known clutter covariance matrix, and then the unknown clutter covariance matrix is estimated using the auxiliary data, a two-step Rao detector (abbreviated as 2S-Rao) for range-extended targets can be obtained. This detector can achieve a certain detection robustness when the amount of auxiliary data is sufficient, but in practical application scenarios, the amount of auxiliary data is often limited, and this detector is difficult to play an effective role. Considering that it is difficult to obtain sufficient pure clutter auxiliary data in the actual environment, for a radar receiver using a centrosymmetric linear array or centrosymmetric staggered pulse trains, its clutter covariance matrix has a special skew-symmetric structure. If based on the generalized likelihood ratio test criterion and considering the skew-symmetric structure information of the clutter covariance matrix, a skew-symmetric one-step generalized likelihood ratio test (P1S-GLRT) detector for range-extended targets can be obtained; compared with the 2S-Rao detector, the detection performance of this detector is improved, and the requirement for the amount of auxiliary data can be reduced, but its robustness to mismatched signals is poor, and the computational complexity of its detection statistic is relatively high, which is not convenient for engineering implementation.

[0004] In the presence of external interference and a small amount of uniform auxiliary data, aiming at the problem that it is difficult to balance detection performance and mismatch robustness for multi-channel wideband radar range-extended target detection, how to make full use of the skew-symmetric structure information, reduce the actual requirement for the amount of auxiliary data, improve the estimation accuracy of the unknown clutter covariance matrix, and then construct a robust intelligent detection method for range-extended targets with a closed form, effectively suppress interference signals while maintaining the constant false alarm rate (CFAR) characteristic, and balance the mismatch robustness, algorithm computational complexity and detection performance is the key to improving the detection ability of wideband radar in complex interference environments, and it is also one of the difficult problems faced by multi-channel wideband radar range-extended target adaptive detection. Summary of the Invention

[0005] 1. Technical Problems to be Solved

[0006] Aiming at the problem that the existing adaptive detectors for wideband radar range extended targets are difficult to balance the algorithm computational complexity, CFAR characteristics, detection performance, and mismatch robustness, and considering the problem that it is difficult to obtain pure clutter auxiliary data due to actual clutter non-uniformity, how to fully exploit the structural information of the clutter covariance matrix, further reduce the demand for auxiliary data volume, improve the estimation accuracy of the unknown clutter covariance matrix, and then construct a robust intelligent detection method for targets under structured interference and clutter with a closed form, while ensuring CFAR characteristics, taking into account the computational complexity, intelligent anti-interference, mismatch robustness, and detection performance of the range extended target adaptive detection algorithm, etc., to improve the adaptive detection performance of multi-channel wideband radar for weak targets and mismatch targets in complex interference environments.

[0007] 2. Technical Solutions

[0008] The robust intelligent detection method for targets under structured interference and clutter of the present invention includes the following technical measures:

[0009] Step 1: Obtain the main data from K to-be-detected range cells For perform a skew-symmetric unitary transformation to obtain the skew-symmetric transformed main data Z; under the condition of the known clutter skew-symmetric covariance matrix, use the complex Gaussian probability density function of Z under the target hypothesis to take the partial derivative of the target parameter vector, and then solve the sub-matrix corresponding to the target parameter vector in the inverse matrix of the Fisher information matrix, and combine the maximum likelihood estimation of the unknown coordinates of the interference subspace under the no-target hypothesis to construct the range extended target Rao detection statistic under the condition of the known clutter skew-symmetric covariance matrix; the specific steps include:

[0010] For a coherent radar system with N space-time joint channels, consider the binary hypothesis testing problem of H 0 and H 1 , where under the hypothesis of H 0 the target does not exist and only pure clutter exists; under the hypothesis of H 1 both the target, clutter, and interference exist. Assume that the target may occupy K consecutive to-be-detected range cells. Under the hypothesis of H 1 , the N×1-dimensional received complex signal in the t-th to-be-detected range cell is expressed as where the N×1-dimensional target complex signal vector and the N×1-dimensional interference complex vector are both assumed to be deterministic and can be expressed as and and are the known full column rank N×p-dimensional target signal subspace complex matrix and N×q-dimensional interference signal subspace complex matrix respectively, the p×1-dimensional complex vector p t and the q×1-dimensional complex vector qt The unknown complex coordinate vectors representing the target signal and the interference signal respectively, and the master data can be represented as an N×K dimensional complex matrix Note the subspace and are linearly independent, and a column full-rank augmented matrix is constructed and p + q ≤ N is satisfied. The N×1 dimensional clutter vector in the t-th distance cell to be detected is a zero-mean complex circular Gaussian vector, denoted as and the clutter vectors between different distance cells are independent and identically distributed, where the N×N dimensional clutter covariance matrix is an unknown Hermitian positive definite complex matrix.

[0011] In addition, when the radar receiving system uses a centrosymmetric linear array or a centrosymmetric pulse train, and both have skew-symmetric structures. Based on the skew-symmetric structure information in the received data, the complex matrices and can be respectively transformed into real matrices H, J and M through the skew-symmetric unitary transformation T. At the same time, the original master data can be transformed through the skew-symmetric unitary transformation, that is

[0012]

[0013] where and respectively represent the set of m×n dimensional real matrices and the set of complex matrices; the skew-symmetric unitary transformation matrix T can be represented as

[0014]

[0015] In the above formula, I m represents the m×m dimensional identity matrix, and the imaginary unit The m×m dimensional skew-symmetric matrix D m is represented as

[0016]

[0017] Under the assumptions of H 0 and H 1 The complex Gaussian probability density function (PDF) of the skew-symmetric transformed master data Z can be respectively represented as

[0018]

[0019]

[0020] where the unknown coordinate matrix of the target subspace Unknown coordinate matrix of the interference subspace Superscript (·) T And (·) H Denote transpose and conjugate transpose respectively, det(·) represents the determinant of a square matrix, and the tr function represents the trace of a square matrix.

[0021] Under the condition that the clutter skew-symmetric covariance matrix M is known, a robust intelligent detector for targets under structured interference and clutter is constructed based on the Rao detection criterion. The Rao detection statistic of the range-extended target can be expressed as

[0022]

[0023] Where Target parameter vector Is unknown, and the interference parameter vector Is unknown; Represents the maximum likelihood estimate of Θ under the hypothesis H 0 The vec function realizes the vectorization of the matrix. In addition, I -1 (Θ) is the inverse matrix of I(Θ), Represents the inverse matrix I -1 (Θ) where Θ r The corresponding pK×pK-dimensional submatrix of the component, and the Fisher information matrix I(Θ) can be expressed as

[0024]

[0025] Where, E{·} represents taking the mathematical expectation; the superscript (·) * Represents conjugation. I(Θ) can be block-represented according to the submatrices corresponding to the Θ r Component and the Θ s Component as

[0026]

[0027] Note that Is a zero matrix. According to the matrix inversion lemma, the submatrix of the inverse matrix I -1 (Θ) Can be expressed as

[0028]

[0029] Next, using the PDF of equation (5) to take partial derivatives of Θ r And Respectively, we can obtain

[0030]

[0031]

[0032] Substituting Equation (10) and Equation (11) into Equation (7), and according to Equation (9), the sub - matrix of the inverse matrix of the Fisher information matrix can be expressed as

[0033]

[0034] where represents the Kronecker product.

[0035] Substituting Equation (10), Equation (11) and Equation (12) into Equation (6), the Rao detection statistic of the range - extended target under the given P and Q can be obtained as

[0036]

[0037] Note that H 0 Assume that the target does not exist under the hypothesis, so the maximum - likelihood estimates of P and Θ r are both zero matrices, then the maximum - likelihood estimate of Q under the H 0 hypothesis can be expressed as

[0038]

[0039] where M -1 / 2 represents the square - root matrix of the inverse matrix M -1 .

[0040] Substituting Equation (14) into D in Equation (13), and through multiple algebraic operations, when the clutter skew - symmetric covariance matrix M is known, the Rao detection statistic of the robust intelligent detector for the target under structured interference and clutter is

[0041]

[0042] where

[0043] Step 2: Obtain auxiliary data from R reference range cells adjacent to the range cell to be detected Perform a skew - symmetric unitary transformation on to obtain the skew - symmetric transformation auxiliary data Y. Use the complex Gaussian probability density function of Y to take the derivative of the clutter skew - symmetric covariance matrix and set it to zero to obtain the maximum - likelihood estimate of the clutter skew - symmetric covariance matrix based on the auxiliary data. Substitute the maximum - likelihood estimate of the clutter skew - symmetric covariance matrix into the Rao detection statistic of the range - extended target obtained in Step 1, replace the unknown clutter skew - symmetric covariance matrix therein, and construct the detection statistic λ of the robust intelligent detection method for the target under structured interference and clutter. The specific steps include:

[0044] To estimate Obtain R observation data from R reference range cells adjacent to the range cell to be detected Assume Only contains pure clutter components, then the auxiliary data can be expressed as an N×R dimensional complex matrix where the N×1 dimensional complex vector of the t-th reference range cell satisfies and it is also independently and identically distributed among different range cells.

[0045] Perform a transformation on the original auxiliary data through the skew-symmetric unitary transformation T to obtain the skew-symmetric transformation auxiliary data Y

[0046]

[0047] The complex Gaussian PDF of Y can be expressed as

[0048]

[0049] Based on the skew-symmetric structure information of the clutter covariance matrix, take the derivative of M with respect to the formula (17) and set it to zero, that is, set the corresponding derivative to zero, and the maximum likelihood estimate of M based on the auxiliary data can be obtained as

[0050]

[0051] where denotes taking the real part. To ensure the non-singularity of the covariance matrix estimate, set denotes the smallest integer greater than or equal to the given parameter. Note that the existing range extended target detectors such as 2S-Rao that do not consider the skew-symmetric structure information require R≥N. The detection method of the present invention reduces the requirement for the amount of auxiliary data, providing strong support for realizing the adaptive detection of range extended targets under the small sample condition of the presence of interference.

[0052] Substitute the maximum likelihood estimate of formula (18) into formula (15), replace the unknown matrix M in formula (15), and the robust intelligent detector for the target under structured interference and clutter can be obtained, and its detection statistic:

[0053]

[0054] where

[0055]

[0056] The method of the present invention constructs a two-step Rao skew-symmetric intelligent robust detector for range-extended targets. It can be seen from Equation (19) that the proposed robust intelligent target detection method under structured interference and clutter has a closed-form expression for the detection statistic, without iterative operations. Additionally, it is worth noting that compared with the P1S-GLRT detector for range-extended targets, the robust intelligent target detection method under structured interference and clutter has a lower algorithm computational complexity and stronger detection robustness for signals with steering vector mismatch. Generally speaking, the robust intelligent target detection method under structured interference and clutter of the present invention can effectively balance the algorithm computational complexity, mismatch robustness, and detection performance while maintaining the CFAR characteristic.

[0057] Step 3 is to set the detection threshold T according to the preset false alarm probability to maintain the CFAR characteristic of the detection method; compare the detection statistic λ with the detection threshold T. If λ≥T, it is determined that there is a target in the currently to-be-detected range cell, and the main data is not used as the auxiliary data for other subsequent to-be-detected range cells; conversely, if λ<T, it is determined that there is no target in the currently to-be-detected range cell, and the main data is used as the auxiliary data for other subsequent to-be-detected range cells.

[0058] 3. Beneficial effects

[0059] Compared with the background art, the beneficial effects of the present invention are:

[0060] 1) Make full use of the skew-symmetric information of the clutter covariance matrix, improve the estimation accuracy of the unknown clutter covariance matrix, reduce the requirement for the amount of auxiliary data, and provide strong support for realizing the adaptive detection of range-extended targets under the condition of small samples with interference;

[0061] 2) Construct a two-step Rao skew-symmetric intelligent robust detection method for range-extended targets. Its detector has a closed-form expression and has a lower computational complexity, which is convenient for engineering implementation;

[0062] 3) For the interference environment with subspace structure, the two-step Rao skew-symmetric intelligent robust detection method for range-extended targets of the present invention can effectively suppress interference signals with different intensities and has good intelligent anti-interference performance;

[0063] 4) For the case of steering vector mismatch of the target signal, the two-step Rao skew-symmetric intelligent robust detection method for range-extended targets of the present invention can effectively detect the mismatched signal and has strong detection robustness for the mismatched signal;

[0064] 5) The detection method of the present invention balances the algorithm computational complexity, detection performance, and mismatch robustness while maintaining the CFAR characteristic, and improves the adaptive detection performance of multi-channel broadband radar for weak targets and mismatched targets in complex environments;

[0065] 6) The method of the present invention is applicable to some non-wideband radar detection scenarios. For example, it is used to detect large targets using low / medium-resolution radars or to detect groups of spatially adjacent point targets moving at the same speed (such as ship formations, aircraft formations, vehicle formations, etc.), and has good application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 is a functional module diagram of the robust intelligent target detection method under structured interference and clutter of the present invention;

[0067] Figure 2 is a comparison diagram of the detection performance of the method of the present invention and the existing detection method for the matched signal;

[0068] Figure 3 is a comparison diagram of the detection performance of the method of the present invention and the existing detection method for the mismatched signal;

[0069] Figure 2 where N = 12, K = 15, R = 13, p = 4, q = 4, false alarm probability P fa = 10 -3 , interference clutter power ratio ICR = 15 dB;

[0070] Figure 3 where N = 12, K = 15, R = 13, p = 4, q = 4, P fa = 10 -3 , ICR = 15 dB, square value of the mismatch angle cos 2 φ = 0.5. DETAILED DESCRIPTION OF THE INVENTION

[0071] The present invention will be further described below with reference to the accompanying drawings. The embodiments of the present invention are used to explain the present invention, rather than to limit the present invention. Any modifications and changes made to the present invention within the spirit and scope of the claims of the present invention fall within the protection scope of the present invention.

[0072] To verify the effectiveness of the method of the present invention, two embodiments are given in this detailed description. The first embodiment is for the sea detection environment, and the second embodiment is for the land detection environment.

[0073] Embodiment 1:

[0074] Referring to the attached Figure 1 of the specification, the specific implementation of Embodiment 1 is divided into the following steps:

[0075] Step A1: Use a sea detection radar to irradiate the sea area to be detected, and obtain the main data of K distance units to be detected The main data Sent to the unitary transformation module; in the unitary transformation module, the skew-symmetric transformation main data Z is obtained according to Equation (1). The skew-symmetric transformation main data Z is sent to H 1 The derivative module of the assumed probability density function, the sub-block inverse module of the Fisher information matrix, and H 0 In the maximum likelihood estimation solving module under the assumption; in H 1 In the derivative module of the assumed probability density function, H is obtained according to Equations (10) and (11) 1 The derivative result of the complex Gaussian probability density function of Z under the assumption with respect to Θ r and In the sub-block inverse module of the Fisher information matrix, the sub-matrix of the inverse matrix of the Fisher information matrix is obtained according to Equation (12) In H 0 In the maximum likelihood estimation solving module under the assumption, the maximum likelihood estimation of Q under the assumption is obtained according to Equation (14) 0 The maximum likelihood estimation of Q under the assumption Send the results obtained above in the derivative module of the probability density function under the assumption, the sub-block inverse module of the Fisher information matrix, and the maximum likelihood estimation solving module under the assumption to the Rao detection statistic construction module under the condition of the known covariance matrix. According to Equation (15), the Rao detection statistic of the range-extended target under the condition of the known clutter skew-symmetric covariance matrix is constructed and sent to the target robust intelligent detector construction module under structured interference and clutter 1 The derivative module of the probability density function under the assumption, the sub-block inverse module of the Fisher information matrix, and 0 In the maximum likelihood estimation solving module under the assumption, the results obtained are sent to the Rao detection statistic construction module under the condition of the known covariance matrix. According to Equation (15), the Rao detection statistic of the range-extended target under the condition of the known clutter skew-symmetric covariance matrix is constructed and sent to the target robust intelligent detector construction module under structured interference and clutter

[0076] It should be noted that in step A1, the complex Gaussian distribution is used to model the sea clutter component. At the same time, considering that there may be external interference in the actual ocean environment, which has an adverse impact on the adaptive detection of the range-extended target, the external interference is also considered in the detector design process. The subspace signal is used to model the interference to reduce the possible mismatch impact of the interference signal. For the interference environment with subspace structure, the two-step Rao skew-symmetric intelligent robust detection method for range-extended targets of the present invention can effectively suppress interference signals of different intensities and has good intelligent anti-interference performance. For the case of target signal steering vector mismatch, the two-step Rao skew-symmetric intelligent robust detection method for range-extended targets of the present invention can effectively detect the mismatched signal and has strong detection robustness for the mismatched signal

[0077] Step A2 irradiates the target-free area around the sea area to be detected by radar to obtain R reference range cells of auxiliary data containing only pure sea clutter Send the auxiliary data To the unitary transformation module; in the unitary transformation module, according to Equation (1) for Perform a skew-symmetric unitary transformation to obtain skew-symmetric transformation auxiliary data Y; send the skew-symmetric transformation auxiliary data Y to the maximum likelihood estimation module of the skew-symmetric clutter covariance matrix, take the derivative of the clutter skew-symmetric covariance matrix using the complex Gaussian probability density function of Y and set it to zero, and obtain the maximum likelihood estimation of the clutter skew-symmetric covariance matrix based on the auxiliary data according to Equation (18). Send to the construction module of the robust intelligent detector for the target under structured interference and clutter, and substitute into the range-extended target Rao detection statistic obtained in step A1, replace the unknown clutter skew-symmetric covariance matrix therein, construct the detection statistic λ of the robust intelligent detection method for the target under structured interference and clutter according to Equation (19), and send λ to the detection decision module.

[0078] It should be noted that in step A2, the maximum likelihood estimation of the clutter skew-symmetric covariance matrix utilizes the skew-symmetric structure information, improving the estimation accuracy of the unknown sea clutter covariance matrix; compared with the existing clutter covariance matrix estimation method that does not consider the skew-symmetric structure information, the covariance matrix maximum likelihood estimation constructed by the method of the present invention considering the skew-symmetric structure information reduces the requirement for the amount of pure sea clutter auxiliary data, providing strong support for realizing the adaptive detection of range-extended targets under the condition of small sample sea clutter with interference. Compared with the P1S-GLRT detector for range-extended targets, the algorithm of the method of the present invention has lower computational complexity and stronger detection robustness for steering vector mismatch signals. In addition, the constructed robust intelligent detection method for the target under structured interference and clutter has a closed-form expression. Compared with the existing adaptive detection methods for range-extended targets, while maintaining the CFAR characteristics, it takes into account the performance balance of algorithm computational complexity, detection performance, and mismatch robustness, improving the adaptive detection ability of multi-channel broadband radar for weak sea targets and mismatch targets in complex electromagnetic environments.

[0079] Step A3 sets the detection threshold T according to a preset false alarm probability: specifically, set the false alarm probability to P fa and, according to the Monte Carlo method, calculate the detection threshold T based on the 100 / P fa measured sea clutter data accumulated in the early stage. Considering the difficulty in obtaining sea clutter, if the amount R of actual measured pure sea clutter data obtained is less than 100 / P fa , then the missing 100 / P fa-R clutter data can be obtained by simulating with a sea clutter simulation model, and the model parameters are reasonably estimated and set according to the measured data of pure sea clutter that has been obtained. Further, the detection statistic λ is compared with the detection threshold T. If λ≥T, it is determined that there is a target in the current K range cells to be detected, and the main data is not used as auxiliary data for other subsequent range cells to be detected; conversely, if λ<T, it is determined that there is no target in the current K range cells to be detected, and the main data is used as auxiliary data for other subsequent range cells to be detected.

[0080] The comparison results of the detector performance under the target-oriented vector matching environment are shown in the appendix Figure 2 . The results show that, compared with the existing detectors for extended targets such as P1S-GLRT and 2S-Rao, the detector of the method of the present invention has better detection performance in the matching environment.

[0081] Example 2:

[0082] Refer to the appendix of the specification Figure 1 , and the specific implementation of Example 2 is divided into the following steps:

[0083] Step B1: Use a ground detection radar to irradiate the area to be detected with radar to obtain the main data of K range cells to be detected Send the main data to the unitary transformation module; in the unitary transformation module, obtain the skew-symmetric transformation main data Z according to Equation (1). Send the skew-symmetric transformation main data Z to the derivative module of the probability density function under the H 1 hypothesis, the sub-block inverse module of the Fisher information matrix, and the maximum likelihood estimation solving module under the H 0 hypothesis; in the derivative module of the probability density function under the H 1 hypothesis, obtain the derivative results of the complex Gaussian probability density function of Z with respect to Θ 1 and r according to Equations (10) and (11); in the sub-block inverse module of the Fisher information matrix, obtain the sub-matrix of the inverse matrix of the Fisher information matrix according to Equation (12) ; in the maximum likelihood estimation solving module under the H hypothesis, obtain the maximum likelihood estimation of Q under the H 0 hypothesis according to Equation (14) 0 ; Send the above derivative module of the probability density function under the H 1 hypothesis, the sub-block inverse module of the Fisher information matrix, and the H 0The result obtained by the maximum likelihood estimation solving module under the hypothesis is sent to the Rao detection statistic construction module under the known covariance matrix. According to Equation (15), the Rao detection statistic of the range extended target under the known clutter skew-symmetric covariance matrix is constructed and sent to the target robust intelligent detector construction module under structured interference and clutter.

[0084] It should be noted that in step B1, the ground clutter component is modeled using the complex Gaussian distribution. At the same time, considering that there may be external interference in the actual ground environment, which has an adverse effect on the adaptive detection of the range extended target, the external interference is also considered in the detector design process. The subspace signal is used to model the interference to reduce the possible mismatch effect of the interference signal. For the interference environment with subspace structure, the two-step Rao skew-symmetric intelligent robust detection method for range extended targets of the present invention can effectively suppress interference signals of different intensities and has good intelligent anti-interference performance. For the case of mismatch of the target signal steering vector, the two-step Rao skew-symmetric intelligent robust detection method for range extended targets of the present invention can effectively detect the mismatch signal and has strong detection robustness for the mismatch signal.

[0085] In step B2, the radar irradiates the target-free area around the area to be detected, and obtains R reference range cells of auxiliary data containing only pure ground clutter. The auxiliary data is sent to the unitary transformation module; in the unitary transformation module, according to Equation (1), is subjected to skew-symmetric unitary transformation to obtain the skew-symmetric transformation auxiliary data Y; the skew-symmetric transformation auxiliary data Y is sent to the maximum likelihood estimation module of the skew-symmetric clutter covariance matrix. The complex Gaussian probability density function of Y is used to derive and set to zero the clutter skew-symmetric covariance matrix, and the maximum likelihood estimation of the clutter skew-symmetric covariance matrix based on the auxiliary data is obtained according to Equation (18). Send to the target robust intelligent detector construction module under structured interference and clutter, and substitute into the Rao detection statistic of the range extended target obtained in step A1, replace the unknown clutter skew-symmetric covariance matrix therein, and construct the detection statistic λ of the target robust intelligent detection method under structured interference and clutter according to Equation (19), and send λ to the detection decision module.

[0086] It should be noted that in step B2, the maximum likelihood estimation of the clutter skew-symmetric covariance matrix utilizes the skew-symmetric structure information, improving the estimation accuracy of the unknown ground clutter covariance matrix. Compared with the existing clutter covariance matrix estimation methods that do not consider the skew-symmetric structure information, the covariance matrix maximum likelihood estimation constructed by the method of the present invention considering the skew-symmetric structure information reduces the requirement for the amount of pure ground clutter auxiliary data, providing strong support for the adaptive detection of range-extended targets under the condition of small sample ground clutter with interference. Compared with the P1S-GLRT detector for range-extended targets, the algorithm of the method of the present invention has lower computational complexity and stronger detection robustness for signals with steering vector mismatch. In addition, the constructed robust intelligent detection method for targets under structured interference and clutter has a closed-form expression. Compared with the existing adaptive detection methods for range-extended targets, while maintaining the CFAR characteristics, it takes into account the performance balance of algorithm computational complexity, detection performance, and mismatch robustness, improving the adaptive detection ability of multi-channel broadband radar for weak ground targets and mismatched targets in complex electromagnetic environments.

[0087] In step B3, the detection threshold T is set according to a preset false alarm probability: specifically, the false alarm probability is set to P fa , and according to the Monte Carlo method, based on the 100 / P fa measured ground clutter data accumulated previously, the detection threshold T is calculated. Further, the detection statistic λ is compared with the detection threshold T. If λ≥T, it is determined that there is a target in the current K range cells to be detected, and the main data is not used as the auxiliary data for other subsequent range cells to be detected; otherwise, if λ<T, it is determined that there is no target in the current K range cells to be detected, and the main data is used as the auxiliary data for other subsequent range cells to be detected.

[0088] The comparison results of the detector performance in the environment of target steering vector mismatch are shown in the appendix Figure 3 . The results show that compared with the existing detectors such as P1S-GLRT and 2S-Rao for range-extended targets, the detector of the method of the present invention has better detection robustness in the mismatch environment.

Claims

1. Robust intelligent target detection method under structured interference and clutter, Characterized in that, It includes the following steps: Step 1: Obtain the main data from K distance cells to be detected For perform a skew-symmetric unitary transformation to obtain the skew-symmetric transformed main data Z; under the condition of knowing the clutter skew-symmetric covariance matrix M, take the partial derivative of the target parameter vector using the complex Gaussian probability density function of Z under the target hypothesis, and then solve the sub-matrix corresponding to the target parameter vector in the inverse matrix of the Fisher information matrix. Combine the maximum likelihood estimate of the unknown coordinates of the interference subspace under the no-target hypothesis to construct a range extended target Rao detection statistic under the condition of knowing the clutter skew-symmetric covariance matrix; Step 2: Obtain auxiliary data from R reference range cells adjacent to the range cell to be detected For perform a skew-symmetric unitary transformation to obtain skew-symmetric transformation auxiliary data Y. Use the complex Gaussian probability density function of Y to take the derivative of the clutter skew-symmetric covariance matrix and set it to zero to obtain the maximum likelihood estimate of the clutter skew-symmetric covariance matrix based on the auxiliary data. Substitute the maximum likelihood estimate of the clutter skew-symmetric covariance matrix into the range extended target Rao detection statistic obtained in Step 1, replace the unknown clutter skew-symmetric covariance matrix therein, and construct the detection statistic λ of the target robust intelligent detection method under structured interference and clutter; Step 3 is to maintain the constant false alarm rate characteristic of the detection method, set the detection threshold T according to the preset false alarm probability; compare the detection statistic λ with the detection threshold T. If λ≥T, it is determined that there is a target in the current distance unit to be detected, and the main data is not used as the auxiliary data for other subsequent distance units to be detected; On the contrary, if λ<T, it is determined that there is no target in the current distance unit to be detected, and the main data is used as the auxiliary data for other subsequent distance units to be detected.

2. The robust intelligent target detection method under structured interference and clutter according to claim 1, Characterized in that, In the said step 1: Targeted H 1 Under the assumption, the partial derivative of the complex Gaussian probability density function of Z with respect to the target parameter vector Θ r is as follows: where, f 1 (Z|P,Q,M) represents the complex Gaussian probability density function of H 1 under the hypothesis, the superscript (·) T and (·) H represent transpose and conjugate transpose respectively, and the vec function realizes the vectorization of the matrix; B = [H J], D = [P T Q T T ; P represents the unknown coordinate matrix of the target subspace, Q represents the unknown coordinate matrix of the interference subspace, and T represents the skew-symmetric unitary transformation matrix, is a known full column rank N×p-dimensional complex matrix of the target signal subspace, is a known full column rank N×q-dimensional complex matrix of the interference signal subspace.​ 3. The robust intelligent target detection method under structured interference and clutter according to claim 1, Characterized in that, In the said step 1: The target parameter vector Θ in the inverse matrix of the Fisher information matrix r The corresponding submatrix Is denoted as Among them, I m represents an m×m dimensional unit matrix, represents the Kronecker product.

4. The robust intelligent target detection method under structured interference and clutter according to claim 1, Characterized in that, In the said step 1: The maximum likelihood estimate of the unknown coordinate Q of the interference subspace under the no-target hypothesis is expressed as Among them, .

5. The robust intelligent target detection method under structured interference and clutter according to claim 1, Characterized in that, In the said step 2: The maximum likelihood estimate of the clutter skew-symmetric covariance matrix M based on the auxiliary data is expressed as Among them, represents taking the real part.

6. The robust intelligent target detection method under structured interference and clutter according to claim 1, Characterized in that, In the said step 2: The detection statistic λ of the robust intelligent target detection method under structured interference and clutter is expressed as Among them, The tr function represents taking the trace of a square matrix.

Images