Dual-parameter adjustable detection method and system for subspace signal in complex gaussian clutter

Abstract

The present application relates to radar signal detection technical field, for realizing the flexible control of realizing the robust detection or sensitive inhibition of mismatch signal in the background of inverse gamma texture composite Gaussian clutter, and guaranteeing that the detector under no signal mismatch detects the target with higher detection rate, the present application, composite Gaussian clutter subspace signal double parameter adjustable detection method and system, including radar transmitter, transceiver switch, antenna, receiver, signal processor and display; first, the radar transmitter generates energy pulse, which is converted and transmitted to the antenna through the transceiver switch to direct the energy to a certain direction in space, the energy of the target reflected electromagnetic wave is converted and sent to the receiver through the antenna and the transceiver switch, and then sent to the computer, the computer adopts the subspace signal double parameter adjustable detection method for processing and displays the result on the display. The present application is mainly applied to the design and manufacture of radar.

Description

Technical Field

[0001] This invention relates to the field of radar signal detection technology, and in particular to a dual-parameter adjustable detection method and system for composite Gaussian clutter neutron space signals. Background Technology

[0002] Target detection is one of the main functions of radar. Besides potential target echoes, radar echo data often contains clutter echoes. With the continuous improvement of radar resolution, the statistical characteristics of clutter no longer conform to a Gaussian distribution, exhibiting severe tailing. Experiments have shown that modeling clutter using a composite Gaussian distribution can effectively solve this problem. The composite Gaussian distribution models clutter as the product of speckle and texture components. The speckle component follows a zero-mean complex Gaussian distribution, while the texture component is a non-negative random variable. This invention assumes that the texture component follows an inverse gamma distribution.

[0003] Furthermore, in radar target detection, wavefront distortion, multipath effects, sidelobe targets, and other factors can lead to signal mismatch. The sensitivity to mismatch signals categorizes detectors into robust detectors and mismatch-sensitive detectors. Robust detectors are needed when the radar operates in search mode, while mismatch-sensitive detectors are required when the radar operates in tracking mode. Traditional detectors lack flexibility in detecting mismatch signals, while adjustable detectors allow for flexible control over robustness and mismatch sensitivity by adjusting adjustable parameters.

[0004] However, no research has been conducted on tunable detectors in the context of complex Gaussian clutter. Therefore, a subspace signal dual-parameter tunable detector was designed in the context of complex Gaussian clutter with inverse gamma texture. Summary of the Invention

[0005] To overcome the shortcomings of existing technologies and achieve robust detection or sensitive suppression of mismatched signals against a background of complex Gaussian clutter with inverse gamma texture, and to ensure that the detector can detect the target with high accuracy even in the absence of signal mismatch, the present invention provides a dual-parameter adjustable detection method for neutron space signals in complex Gaussian clutter, comprising the following steps:

[0006] Step 1: Construct the data to be detected, the signal matrix, and the training sample matrix;

[0007] Step 2: Construct the shape and scale parameters of the inverse gamma distribution based on the data to be detected and the training sample matrix;

[0008] Step 3: Construct an intermediate variable matrix based on the data to be detected, the signal matrix, and the training sample matrix;

[0009] Step 4: Determine the adjustable parameters that represent the robustness and sensitivity characteristics of the mismatch signal, respectively;

[0010] Step 5: Construct the detection statistic based on the adjustable parameters determined in Step 4;

[0011] Step 6: Determine the detection threshold using the aforementioned detection statistics and false alarm probability;

[0012] Step 7: Compare the magnitude of the detection statistic with the detection threshold, and determine whether the target exists.

[0013] The detailed steps are as follows:

[0014] Step 1: Construct the detection data, signal matrix, and training sample matrix, represented as z, H, and Y, respectively, where the dimension of z is N×1, and z = [z1, z2, ..., z...]. N ] T ,symbol[·] T This represents the transpose, where H has a dimension of N×p, and the training sample matrix Y has a dimension of N×L, where N is the number of system channels and p represents the dimension of the signal subspace; Y can be represented as Y=[y1,L,y l ,L,y L ], y l Let y be the l-th training sample with dimension N×1. l =[y l,1 ,y l,2 ,L,y l,N ] T L is the number of training samples.

[0015] Step 2: Construct the shape and scale parameters of the inverse gamma distribution based on the data to be detected and the training sample matrix; specifically, obtain the shape parameter λ and scale parameter β of the inverse gamma distribution using the fractional moment estimation method and the data to be detected and the training sample matrix.

[0016]

[0017] Here, λ and β are considered as unknown parameters, the symbol |·| represents the absolute value, and Γ(·) is the gamma function, expressed as:

[0018]

[0019] The shape parameter λ and scale parameter β of the clutter distribution are obtained by solving the equation for fractional moment estimation.

[0020] Step 3: The intermediate variable matrix constructed based on the data to be detected, the signal matrix, and the training sample matrix includes Φ, Σ, and Ψ. Specifically, the estimated value Φ of the covariance matrix is ​​obtained through an iterative method of approximate maximum likelihood estimation (AML). The iterative process is as follows:

[0021]

[0022] Where tr(·) represents the matrix trace operation, i = 0, 1, ..., Nit represents the number of iterations, and Φ AML (0) represents the initial value of the covariance matrix Φ, Φ AML (i+1) represents the iterative value of the covariance matrix Φ in the (i+1)th iteration, when N it The iteration terminates when Σ = 5, yielding the estimated value Φ of the covariance matrix; further, the other two intermediate variables are obtained as: Σ = z H ΦH(H H ΦH) -1 H H Φz and Ψ=z H Φz, superscript (·) H This indicates the conjugate transpose, (·) -1 Represents the inverse of a matrix;

[0023] Step 4: Based on system requirements, two adjustable coefficients are determined, denoted as a and b, where 0 ≤ a ≤ 2 and 0 ≤ b ≤ 2. If the detector is to be robust to mismatch signals, a should be set relatively small and b should be set relatively large. If the detector is to be mismatch sensitive to mismatch signals, a should be set relatively large and b should be set relatively small. If the detector is to have a high detection probability for signals without mismatch, a should be set relatively small and b should be set relatively large.

[0024] Step 5, the constructed detection statistic is shown in the following formula:

[0025]

[0026] Step 6: Determine the detection threshold using the aforementioned detection statistics and false alarm probability, obtained by the following formula.

[0027] η=t(n * )

[0028] In the formula, M represents the number of Monte Carlo simulations, and μ represents the false alarm probability of the system. For the floor operation, t(i) is the sequence. Arrange the i-th maximum value in descending order, and Φ(k) represents the k-th iteration process for achieving the complete approximate maximum likelihood estimation (AML).

[0029] Ψ(k)=z H (k)Φ(k)z(k), Σ(k)=z H (k)Φ(k)H(H H Φ(k)H) -1 H H Φ(k)z(k),

[0030] z(k) is the kth realization of the data to be detected containing only clutter components, k = 1, 2, L, M;

[0031] In step 7, the detection statistic is compared with the detection threshold, and the existence of the target is determined based on the following two cases:

[0032] If the detection statistic Detector is greater than or equal to the detection threshold η, then the target is determined to exist;

[0033] If the detection statistic Detector is less than the detection threshold η, then the target is determined to not exist.

[0034] A dual-parameter adjustable detection system for composite Gaussian clutter neutron space signals includes a radar transmitter, a transceiver switch, an antenna, a receiver, a signal processor, and a display. First, the radar transmitter generates a short-duration, high-power radio frequency energy pulse, which is transmitted to the antenna via the transceiver switch. The antenna then directs the energy in a specific direction in space. Targets in that direction reflect the electromagnetic waves they encounter. The reflected energy is transmitted through the antenna and transceiver switch to the receiver, forming a radar echo signal. The receiver then filters, amplifies, and performs analog-to-digital conversion on the echo signal before sending it to a computer. The computer contains the following modules:

[0035] The data construction module is used to construct the data to be detected, the signal matrix, and the training sample matrix.

[0036] The clutter parameter construction module constructs the shape and scale parameters of the inverse gamma distribution based on the data to be detected and the training sample matrix.

[0037] The intermediate variable matrix construction module constructs an intermediate variable matrix based on the data matrix to be detected, the signal matrix, and the training sample matrix, which is used to construct the intermediate variable matrix required for the detection statistics.

[0038] The adjustable coefficient determination module is used to determine the adjustable coefficient based on system requirements;

[0039] The detection statistic construction module is used to construct detection statistics based on the intermediate variable matrix and adjustable parameters.

[0040] The detection threshold determination module is used to determine the detection threshold using the detection statistics and the false alarm probability.

[0041] The target decision module is used to compare the detection statistic with the detection threshold and determine whether the target exists. Specifically, if the detection statistic is greater than the detection threshold, the target is determined to exist; if the detection statistic is less than or equal to the detection threshold, the target is determined to not exist. Finally, the detection result is displayed on the screen.

[0042] The features and beneficial effects of this invention are:

[0043] This invention is based on the concept of adaptive detection. The detector designed can simultaneously achieve clutter suppression and target detection without the need for a separate filtering process.

[0044] This invention utilizes training sample data and the approximate maximum likelihood estimation (AML) method to design a detector that can achieve clutter suppression and has constant false alarm rate (CFAR) characteristics, thus eliminating the need for additional CFAR processing steps.

[0045] This invention introduces two adjustable parameters, a and b, to give the adjustable detector greater flexibility, enabling better adjustment of the detector's effective suppression of mismatch signals and achieving detection of no-mismatch signals with a high probability. Specifically, if robustness to mismatch signals is desired, a should be set relatively small and b should be set relatively large; if mismatch sensitivity is desired, a should be set relatively large and b should be set relatively small; if a high probability of detecting no-mismatch signals is desired, a should be set relatively small and b should be set relatively large. Attached image description:

[0046] Figure 1 This is a flowchart illustrating the dual-parameter adjustable detection method and system for composite Gaussian clutter neutron space signals described in this invention.

[0047] Figure 2 This is a structural framework diagram of a dual-parameter adjustable detection method and system for composite Gaussian clutter neutron space signals according to the present invention.

[0048] Figure 3 This section compares the detection probabilities of the proposed method and existing methods under different signal mismatches, where cos 2 φ represents the signal mismatch, when cos 2 There is no mismatch when φ=1, and when cos 2 φ = 0, complete mismatch, SGLRT-CG represents a typical adaptive detector in a compound Gaussian clutter background. As can be seen from the figure, with appropriate adjustable parameter pairs, the provided detector can achieve both sensitive suppression of mismatch signals (the detector's detection probability decreases rapidly with increasing mismatch, thus failing to detect the mismatch signal) and robust detection of mismatch signals (the detector's detection probability does not decrease rapidly with increasing mismatch, thus detecting the mismatch signal with a relatively high probability).

[0049] Figure 4This figure compares the detection probabilities of the proposed method with existing methods under different signal-to-clutter ratios in the absence of signal mismatch, where SGLRT-CG represents a typical adaptive detector in a compound Gaussian clutter background. As can be seen from the figure, with appropriate parameter pairs, the proposed detector has a higher detection probability than the existing detector SGLRT-CG. Detailed Implementation

[0050] To achieve robust detection or sensitive suppression of mismatched signals in the context of complex Gaussian clutter with inverse gamma texture and to ensure that the detector can detect the target with high accuracy even without signal mismatch, this invention provides a dual-parameter adjustable detection method for neutron space signals in complex Gaussian clutter, comprising the following steps:

[0051] Step 1: Construct the data to be detected, the signal matrix, and the training sample matrix;

[0052] Step 2: Construct the shape and scale parameters of the inverse gamma distribution based on the data to be detected and the training sample matrix;

[0053] Step 3: Construct an intermediate variable matrix based on the data to be detected, the signal matrix, and the training sample matrix;

[0054] Step 4: Determine the adjustable parameters based on system requirements;

[0055] Step 5: Construct the detection statistics;

[0056] Step 6: Determine the detection threshold using the aforementioned detection statistics and false alarm probability;

[0057] Step 7: Compare the detection statistic with the detection threshold, and determine whether the target exists;

[0058] In step 1, the constructed data to be detected, signal matrix, and training sample matrix are represented as z, H, and Y, respectively, where the dimension of z is N×1, and z = [z1, z2, ..., z...]. N ] T ,symbol[·] T This represents the transpose, where H has a dimension of N×p, and the training sample matrix Y has a dimension of N×L, where N is the number of system channels and p represents the dimension of the signal subspace; Y can be represented as Y=[y1,L,y l ,L,y L ], y l Let y be the l-th training sample with dimension N×1. l =[y l,1 ,y l,2 ,L,y l,N ] T L is the number of training samples.

[0059] Further, in step 2, the shape parameters and scale parameters of the inverse gamma distribution are constructed based on the data to be detected and the training sample matrix; specifically, the shape parameter λ and scale parameter β of the inverse gamma distribution are obtained using the data to be detected and the training sample matrix through the fractional moment estimation method.

[0060]

[0061] Here, λ and β are considered as unknown parameters, the symbol |·| represents the absolute value, and Γ(·) is the gamma function, expressed as:

[0062]

[0063] The shape parameter λ and scale parameter β of the clutter distribution are obtained by solving the equation for fractional moment estimation.

[0064] Further, in step 3, the intermediate variable matrix constructed based on the data to be detected, the signal matrix, and the training sample matrix includes: Φ, Σ, and Ψ; specifically, the estimated value Φ of the covariance matrix is ​​obtained through an iterative method of approximate maximum likelihood estimation (AML), and the iterative process is as follows:

[0065]

[0066] Where tr(·) represents the matrix trace operation, i = 0, 1, ..., Nit represents the number of iterations, and Φ AML (0) represents the initial value of the covariance matrix Φ, Φ AML (i+1) represents the iterative value of the covariance matrix Φ in the (i+1)th iteration, when N it The iteration terminates when Σ = 5, yielding the estimated value Φ of the covariance matrix; further, the other two intermediate variables are obtained as: Σ = z H ΦH(H H ΦH) -1 H H Φz and Ψ=z H Φz, superscript (·) H This indicates the conjugate transpose, (·) -1 This represents the inverse of a matrix.

[0067] Furthermore, in step 4, the adjustable coefficients determined according to system requirements include two, denoted as a and b, where 0 ≤ a ≤ 2 and 0 ≤ b ≤ 2. If it is desired that the detector has robust characteristics to mismatch signals, then a should be set relatively small and b should be set relatively large. If it is desired that the detector has mismatch sensitive characteristics to mismatch signals, then a should be set relatively large and b should be set relatively small. If it is desired that the detector has a high detection probability for no mismatch signals, then a should be set relatively small and b should be set relatively large.

[0068] Furthermore, in step 5, the constructed detection statistic is shown in the following formula:

[0069]

[0070] Furthermore, in step 6, the detection threshold is determined using the detection statistic and false alarm probability by the following formula.

[0071] η=t(n * )

[0072] In the formula, M represents the number of Monte Carlo simulations, and μ represents the false alarm probability of the system. For the floor operation, t(i) is the sequence. Arrange the i-th maximum value in descending order, and Φ(k) represents the k-th iteration process for achieving the complete approximate maximum likelihood estimation (AML).

[0073] Ψ(k)=z H (k)Φ(k)z(k), Σ(k)=z H (k)Φ(k)H(H H Φ(k)H) -1 H H Φ(k)z(k),

[0074] z(k) is the kth realization of the data to be detected containing only clutter components, where k = 1, 2, L, M.

[0075] Further, in step 7, the detection statistic is compared with the detection threshold, and the existence of the target is determined in two ways:

[0076] If the detection statistic Detector is greater than or equal to the detection threshold η, then the target is determined to exist;

[0077] If the detection statistic Detector is less than the detection threshold η, then the target is determined to not exist.

[0078] On the other hand, this invention provides a dual-parameter adjustable detection system for composite Gaussian clutter neutron space signals. This system includes a radar transmitter, a transceiver switch, an antenna, a receiver, a signal processor, and a display. First, the radar transmitter generates a short-duration high-power radio frequency energy pulse, which is transmitted to the antenna via the transceiver switch. The antenna then directs the energy in a specific direction in space. A target in that direction reflects the electromagnetic wave it encounters. The reflected energy is transmitted to the receiver via the antenna and the transceiver switch, forming a radar echo signal. The receiver then filters, amplifies, and performs analog-to-digital conversion on the echo signal before sending it to a computer. The computer includes the following modules:

[0079] The data construction module is used to construct the data to be detected, the signal matrix, and the training sample matrix.

[0080] The clutter parameter construction module constructs the shape and scale parameters of the inverse gamma distribution based on the data to be detected and the training sample matrix.

[0081] The intermediate variable matrix construction module constructs an intermediate variable matrix based on the data matrix to be detected, the signal matrix, and the training sample matrix, which is used to construct the intermediate variable matrix required for the detection statistics.

[0082] The adjustable coefficient determination module is used to determine the adjustable coefficient based on system requirements;

[0083] The detection statistic construction module is used to construct detection statistics based on the intermediate variable matrix and adjustable parameters.

[0084] The detection threshold determination module is used to determine the detection threshold using the detection statistics and the false alarm probability.

[0085] The target decision module is used to compare the detection statistic with the detection threshold and determine whether the target exists. Specifically, if the detection statistic is greater than the detection threshold, the target is determined to exist; if the detection statistic is less than or equal to the detection threshold, the target is determined to not exist. Finally, the detection result is displayed on the screen.

[0086] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0087] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0088] For a multi-channel radar with N channels, the received radar data can be represented by an N×1 dimensional vector z = [z1, z2, ..., z...]. N ] T Assume the target signal s is located in the subspace. <h>In, it can be represented as:

[0089] s=Hθ (1)

[0090] In the formula, H is an N×p-dimensional full-rank column matrix, and θ is a p×1-dimensional unknown coordinate vector of the target in the subspace. Under the H1 hypothesis test, the data to be detected can be expressed as:

[0091] z=Hθ+n (2)

[0092] In the formula, n represents N×1 dimensional clutter. Under the composite Gaussian model, clutter can be represented as the product of a non-negative texture component and a speckle component, i.e.:

[0093]

[0094] In the formula, x represents the speckle component, which follows a pattern with a mean of zero and a covariance of E(xx). H Let E(·) be a complex Gaussian distribution of Φ, where E(·) represents the statistical expectation, and the superscript [·] is used. H τ represents the conjugate transpose; τ represents the non-negative texture component, assuming τ follows an inverse gamma distribution, i.e.:

[0095]

[0096] In the formula, λ represents the shape parameter and β represents the scale parameter.

[0097] Suppose there are L training samples that do not contain the target, and let the l-th training sample be:

[0098] y l =n l (5)

[0099] In the formula, l = 1, 2, L, L, y l Let y be the l-th training sample with dimension N×1. l =[y l,1 ,y l,2 ,L,y l,N ] T The detection problem can be expressed as a binary hypothesis test as shown in the following formula:

[0100]

[0101] Regarding the detection problem in equation (6), the literature (F. Gini and A. Farina, "Vector subspace detection in compound-Gaussian clutter. Part I: survey and new results," IEEE Transactions on Aerospace and Electronic Systems, vol. 38, no. 4, pp. 1295-1311, 2002.) did not consider the prior information of inverse gamma texture distribution. Under the compound Gaussian model, a detector was obtained that addresses the assumption in (6). Since it was not named in the literature, it is named the Compound Gaussian Subspace Generalized Likelihood Ratio Detector (SGLRT-CG) in this invention for ease of comparison. However, the above detector is difficult to simultaneously satisfy robust detection of mismatched signals, sensitive suppression of mismatched signals, and high detection probability for signals without mismatch.

[0102] To resolve the above issues, please refer to Figure 1 As shown, this invention provides a dual-parameter adjustable detection method and system for composite Gaussian clutter neutron space signals, comprising:

[0103] Step 1: Construct the data to be detected, the signal matrix, and the training sample matrix;

[0104] Step 2: Construct the shape and scale parameters of the inverse gamma distribution based on the data to be detected and the training sample matrix;

[0105] Step 3: Construct an intermediate variable matrix based on the data to be detected, the signal matrix, and the training sample matrix;

[0106] Step 4: Determine the adjustable parameters based on system requirements;

[0107] Step 5: Construct the detection statistics;

[0108] Step 6: Determine the detection threshold using the aforementioned detection statistics and false alarm probability;

[0109] Step 7: Compare the detection statistic with the detection threshold, and determine whether the target exists;

[0110] In step 1, the constructed data to be detected, signal matrix, and training sample matrix are represented as z, H, and Y, respectively, where the dimension of z is N×1, and z = [z1, z2, ..., z...]. N ] T ,symbol[·] T This represents the transpose, where H has a dimension of N×p, and the training sample matrix Y has a dimension of N×L, where N is the number of system channels and p represents the dimension of the signal subspace; Y can be represented as Y=[y1,L,y l ,L,y L ], y l Let y be the l-th training sample with dimension N×1. l =[y l,1 ,y l,2 ,L,y l,N ] T L is the number of training samples.

[0111] Further, in step 2, the shape parameters and scale parameters of the inverse gamma distribution are constructed based on the data to be detected and the training sample matrix; specifically, the shape parameter λ and scale parameter β of the inverse gamma distribution are obtained using the data to be detected and the training sample matrix through the fractional moment estimation method.

[0112]

[0113] Here, λ and β are considered as unknown parameters, the symbol |·| represents the absolute value, and Γ(·) is the gamma function, expressed as:

[0114]

[0115] The shape parameter λ and scale parameter β of the clutter distribution are obtained by solving the equation for fractional moment estimation.

[0116] Further, in step 3, the intermediate variable matrix constructed based on the data to be detected, the signal matrix, and the training sample matrix includes: Φ, Σ, and Ψ; specifically, the estimated value Φ of the covariance matrix is ​​obtained through an iterative method of approximate maximum likelihood estimation (AML), and the iterative process is as follows:

[0117]

[0118] Where tr(·) represents the matrix trace operation, i = 0, 1, ..., Nit represents the number of iterations, and Φ AML (0) represents the initial value of the covariance matrix Φ, Φ AML (i+1) represents the iterative value of the covariance matrix Φ in the (i+1)th iteration, when N it The iteration terminates when Σ = 5, yielding the estimated value Φ of the covariance matrix; further, the other two intermediate variables are obtained as: Σ = z H ΦH(H H ΦH) -1 H H Φz and Ψ=z H Φz, superscript (·) H This indicates the conjugate transpose, (·) -1 This represents the inverse of a matrix.

[0119] Furthermore, in step 4, the adjustable coefficients determined according to system requirements include two, denoted as a and b, where 0 ≤ a ≤ 2 and 0 ≤ b ≤ 2. If it is desired that the detector has robust characteristics to mismatch signals, then a should be set relatively small and b should be set relatively large. If it is desired that the detector has mismatch sensitive characteristics to mismatch signals, then a should be set relatively large and b should be set relatively small. If it is desired that the detector has a high detection probability for no mismatch signals, then a should be set relatively small and b should be set relatively large.

[0120] Furthermore, in step 5, the constructed detection statistic is shown in the following formula:

[0121]

[0122] Furthermore, in step 6, the detection threshold is determined using the detection statistic and false alarm probability by the following formula.

[0123] η=t(n * )

[0124] In the formula, M represents the number of Monte Carlo simulations, and μ represents the false alarm probability of the system. For the floor operation, t(i) is the sequence. Arrange the i-th maximum value in descending order, and Φ(k) represents the k-th iteration process for achieving the complete approximate maximum likelihood estimation (AML).

[0125] Ψ(k)=z H (k)Φ(k)z(k), Σ(k)=z H (k)Φ(k)H(H H Φ(k)H) -1 H H Φ(k)z(k),

[0126] z(k) is the kth realization of the data to be detected containing only clutter components, where k = 1, 2, L, M.

[0127] Further, in step 7, the detection statistic is compared with the detection threshold, and the existence of the target is determined in two ways:

[0128] If the detection statistic Detector is greater than or equal to the detection threshold η, then the target is determined to exist;

[0129] If the detection statistic Detector is less than the detection threshold η, then the target is determined to not exist.

[0130] On the other hand, this invention provides a dual-parameter adjustable detection system for composite Gaussian clutter neutron space signals, mainly composed of a radar transmitter, a transceiver switch, an antenna, a receiver, a signal processor, and a display. First, the radar transmitter generates a short-duration high-power radio frequency energy pulse, which is converted and transmitted to the antenna via the transceiver switch. Then, the antenna directs the energy in a certain direction in space. Targets in this direction reflect the electromagnetic waves they encounter. The reflected energy is converted and sent to the receiver via the transceiver switch, forming a radar echo signal. The receiver then filters, amplifies, and performs analog-to-digital conversion on the echo signal, before sending it to the computer. The computer then executes any of the dual-parameter adjustable detection methods for composite Gaussian clutter neutron space signals described in steps 1-8 for signal processing and target detection. Finally, the detection results are displayed on the display.

[0131] The computer includes at least one processor; and a memory communicatively connected to the at least one processor, the program of which may be stored in a computer-readable storage medium, which may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc. The memory stores instructions executable by the at least one processor, the instructions being programmed to execute the dual-parameter adjustable detection method for composite Gaussian clutter neutron space signals described in any of steps 1-8.

[0132] The signal processing and target detection of the composite Gaussian clutter neutron space signal dual-parameter adjustable detection method described in any of steps 1-8 are achieved using the following modules in a computer:

[0133] The data construction module is used to construct the data to be detected, the signal matrix, and the training sample matrix.

[0134] The clutter parameter construction module constructs the shape and scale parameters of the inverse gamma distribution based on the data to be detected and the training sample matrix.

[0135] The intermediate variable matrix construction module constructs an intermediate variable matrix based on the data matrix to be detected, the signal matrix, and the training sample matrix, which is used to construct the intermediate variable matrix required for the detection statistics.

[0136] The adjustable coefficient determination module is used to determine the adjustable coefficient based on system requirements;

[0137] The detection statistic construction module is used to construct detection statistics based on the intermediate variable matrix and adjustable parameters.

[0138] The detection threshold determination module is used to determine the detection threshold using the detection statistics and the false alarm probability.

[0139] The target decision module is used to compare the detection statistic with the detection threshold and determine whether the target exists. Specifically, if the detection statistic is greater than the detection threshold, the target is determined to exist; if the detection statistic is less than or equal to the detection threshold, the target is determined to not exist.

[0140] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.< / h>

Claims

1. A dual-parameter adjustable detection method for composite Gaussian clutter neutron space signals, characterized in that the steps are as follows: as follows: Step 1: Construct the data to be detected, the signal matrix, and the training sample matrix; Step 2: Construct the shape and scale parameters of the inverse gamma distribution based on the data to be detected and the training sample matrix; Step 3: Construct an intermediate variable matrix based on the data to be detected, the signal matrix, and the training sample matrix; Step 4: Determine the adjustable parameters that represent the robustness and sensitivity characteristics of the mismatch signal, respectively; Step 5: Construct the detection statistic based on the adjustable parameters determined in Step 4; Step 6: Determine the detection threshold using the aforementioned detection statistics and false alarm probability; Step 7: Compare the detection statistic with the detection threshold, and determine whether the target exists; The detailed steps are as follows: Step 1, construct the detection data, signal matrix, and training sample matrix, respectively represented as follows: , and ,in The dimension is , ,symbol This indicates transpose. The dimension is Training sample matrix The dimension is ,in The number of system channels. The dimension of the signal subspace; It can be represented as , For the first There are training samples with dimension [ ]. , , This represents the number of training samples; Step 2: Construct the shape and scale parameters of the inverse gamma distribution based on the data to be detected and the training sample matrix; specifically, obtain the shape parameters of the inverse gamma distribution using the fractional moment estimation method with the data to be detected and the training sample matrix. With scale parameters : Among them, and Treat it as an unknown parameter, symbol It represents the absolute value. The gamma function is expressed as: The shape parameters of the clutter distribution are obtained by solving the equation for fractional moment estimation. and scale parameters ; Step 3, the intermediate variable matrix constructed based on the data to be detected, the signal matrix, and the training sample matrix includes: , and Specifically, the covariance matrix is ​​estimated using an iterative method of approximate maximum likelihood estimation (AML). The iterative process is as follows: in This represents the trace operation of a matrix. This indicates the number of iterations. This represents the covariance matrix. initial value, It indicates the first The covariance matrix at the next iteration The iteration value, when The iteration is terminated at time to obtain the estimated value of the covariance matrix. Furthermore, the other two intermediate variables are obtained as follows: and superscript This indicates the conjugate transpose. Represents the inverse of a matrix; Step 4, based on system requirements, determines two adjustable coefficients, denoted as . and ,and and If it is desired that the detector has robust characteristics to mismatch signals, then It should be set relatively small. It should be set relatively large; if it is desired that the detector has mismatch sensitivity to mismatch signals, then It should be set to a relatively large value. It should be set relatively small; if a high detection probability of the detector for signals without mismatch is desired, then It should be set relatively small. It should be set relatively large; Step 5, the constructed detection statistic is shown in the following formula: Step 6: Determine the detection threshold using the aforementioned detection statistics and false alarm probability, obtained by the following formula. In the formula, , For the number of Monte Carlo simulations, This represents the false alarm probability value of the system. For rounding operations, For sequence Arrange from largest to smallest The maximum value, For the first This completes the iterative process of approximate maximum likelihood estimation (AML). ， ， For the first detection data containing only clutter components This is the first implementation. ; In step 7, the detection statistic is compared with the detection threshold, and the existence of the target is determined based on the following two cases: If the detection statistic Greater than or equal to the detection threshold If so, then the target is determined to exist; If the detection statistic Less than the detection threshold If the target does not exist, then it is determined that the target does not exist.

2. A dual-parameter adjustable detection system for composite Gaussian clutter neutron space signals, characterized in that, The system includes a radar transmitter, transceiver switch, antenna, receiver, signal processor, and display. First, the radar transmitter generates a short, high-power radio frequency energy pulse, which is converted by the transceiver switch and transmitted to the antenna. The antenna then directs the energy in a specific direction. A target in that direction reflects the electromagnetic wave, and the reflected energy is transmitted through the antenna and transceiver switch to the receiver, forming a radar echo signal. The receiver then filters, amplifies, and performs analog-to-digital conversion on the echo signal before sending it to the computer. The computer contains the following modules: The data construction module is used to construct the data to be detected, the signal matrix, and the training sample matrix. The clutter parameter construction module constructs the shape and scale parameters of the inverse gamma distribution based on the data to be detected and the training sample matrix. The intermediate variable matrix construction module constructs an intermediate variable matrix based on the data matrix to be detected, the signal matrix, and the training sample matrix, which is used to construct the intermediate variable matrix required for the detection statistics. The adjustable coefficient determination module is used to determine the adjustable coefficient based on system requirements; The detection statistic construction module is used to construct detection statistics based on the intermediate variable matrix and adjustable parameters. The detection threshold determination module is used to determine the detection threshold using the detection statistics and the false alarm probability. The target decision module is used to compare the detection statistic with the detection threshold and decide whether the target exists. Specifically, if the detection statistic is greater than the detection threshold, the target is determined to exist; if the detection statistic is less than or equal to the detection threshold, the target is determined to not exist. Finally, the detection result is displayed on the screen. The computer is used to perform the steps of claim 1.

Images