Integrated Method and System for Multi-Domain Detection and Localization of Moving Targets in Strong Clutter Background

Abstract

This invention provides an integrated method and system for multi-domain detection and localization of moving targets in strong clutter backgrounds. The method includes acquiring multi-channel echo data from a space-based early warning radar and performing data rearrangement and range-dependent error compensation; then constructing a first set of sub-channel data and a second set of sub-channel data; performing a first-level adaptive filtering process on both sets of sub-channel data to obtain the first-level processing result; performing a second-level moving target detection on the first-level processing result, outputting the range cell and Doppler cell when a moving target is detected; calculating the angle of the moving target using the single-pulse sum-difference beam angle measurement principle; and finally calculating the three-dimensional spatial position of the moving target in the geodetic coordinate system through the transformation relationship between the radar coordinate system and the geodetic coordinate system. This invention, through a two-level adaptive processing structure, effectively achieves integrated processing of moving target detection, angle estimation, and localization in strong clutter backgrounds.

Description

Technical Field

[0001] This invention relates to the field of moving target detection and localization in space-based early warning radar, specifically to an integrated method and system for multi-domain detection and localization of moving targets under strong clutter backgrounds. Background Technology

[0002] Space-based early warning radar plays a crucial role in long-range, wide-area surveillance, with moving target detection and localization being its core tasks. However, space-based early warning radar typically operates in a high-altitude, down-looking mode, where the detection environment is characterized by high-intensity and wide-ranging ground and sea clutter, resulting in strong clutter and weak moving target echoes. Furthermore, due to the Earth's rotation and the high-speed movement of the radar platform itself, ground clutter exhibits significant range dependence and Doppler spectrum broadening, further submerging moving target echo signals in the strong clutter background and posing significant challenges to detection.

[0003] To address strong clutter interference, dimension-reduction space-time adaptive processing techniques are widely used for clutter suppression. Traditional methods typically perform clutter suppression first, followed by constant false alarm rate (CFAR) detection to achieve multi-domain detection of moving targets. This sequential processing approach of suppression followed by detection results in a loss of target detection performance. Furthermore, after target detection, the angle estimation required for localization often necessitates additional spatial processing or search procedures with high computational complexity.

[0004] A patent search revealed an invention patent with publication number CN103197297A, which discloses a radar moving target detection method based on a cognitive framework. The implementation process is as follows: separating ground clutter components from the echo; obtaining a clutter map of the current frame based on the ground clutter components; dividing the current frame into noise and clutter zones based on the clutter map; calculating correlation coefficients based on the ground clutter components, using them to correct antenna pointing errors, and storing the corrected echo data in an environmental dynamic database (EDDB); estimating the interference covariance matrix using the echo data in the EDDB based on the current frame division; obtaining adaptive filter coefficients based on the estimated interference covariance matrix and performing adaptive filtering on the input data; and performing constant false alarm rate (CFAR) detection on the filter output. This patent lacks a multi-channel collaborative processing mechanism, does not consider range-dependent error compensation, and lacks target localization functionality. It can only perform detection, limiting its applicability to complex multi-channel radar scenarios and exhibiting insufficient functional completeness.

[0005] In summary, given the problems of the existing technologies, researching an integrated method and system for multi-domain detection and localization of moving targets under strong clutter backgrounds has become a critical task that urgently needs to be addressed. Summary of the Invention

[0006] To address the shortcomings of existing technologies, the purpose of this invention is to provide an integrated method and system for multi-domain detection and localization of moving targets against a background of strong clutter.

[0007] The present invention provides an integrated method for multi-domain detection and localization of moving targets under strong clutter background, comprising the following steps:

[0008] Step S1: Acquire multi-channel echo data from the space-based early warning radar and rearrange the multi-channel echo data to obtain the rearranged echo data matrix.

[0009] Step S2: Using the compensation matrix, distance dependence error compensation is performed on the rearranged echo data matrix to obtain the compensated echo data matrix.

[0010] Step S3: Combine the channels of the compensated echo data matrix to construct the first set of sub-channel data and the second set of sub-channel data;

[0011] Step S4: Based on the improved extended factorization algorithm (EFA) with additional auxiliary channels, perform first-level adaptive filtering on the first group of sub-channel data and the second group of sub-channel data respectively to obtain the first-level processing result;

[0012] Step S5: Based on the results of the first-level processing, the second-level moving target detection is performed using an adaptive matched filter detector combined with an improved extended factorization algorithm (EFA). When a moving target is detected, the range cell and Doppler cell where the moving target is located are output.

[0013] Step S6: When a moving target is detected, the corresponding first set of sub-channel data values ​​and second set of sub-channel data values ​​are extracted from the first-level processing results according to the range cell and Doppler cell where the moving target is located, and the angle of the moving target is calculated using the single-pulse sum-difference beam angle measurement principle.

[0014] Step S7: Based on the range cell where the moving target is located and the angle of the moving target, calculate the three-dimensional spatial position of the moving target in the geodetic coordinate system through the transformation relationship between the radar coordinate system and the geodetic coordinate system.

[0015] Preferably, in step S1, the multi-channel echo data is rearranged to obtain a rearranged echo data matrix. :

[0016]

[0017] In the formula, , This represents the total number of distance units. The number of pulses. For the number of channels, the first... Echo data vector of each distance cell , Indicates the first The distance unit, the first The pulse, the first Echo data from each channel, This indicates the matrix transpose.

[0018] Preferably, in step S2, for the first The compensation matrix for each distance cell is:

[0019]

[0020] In the formula, For the first The clutter Doppler center frequency of each distance cell, for 3D identity matrix This indicates the creation of a diagonal matrix. Indicates the Kronecker product; j Represents the imaginary unit. d It stands for doppler.

[0021] Using the first The compensation matrix of the distance unit is for the first distance unit. The data from each distance cell is compensated to obtain the compensated data vector. :

[0022]

[0023] In the formula, This indicates the conjugate transpose.

[0024] Preferably, in step S2, the compensated data vectors of all range cells are combined to obtain the compensated echo data matrix. :

[0025]

[0026] In the formula, .

[0027] Preferably, in step S3, the first group of sub-channel data Second group of sub-channel data They are respectively:

[0028]

[0029]

[0030] In the formula, the first Echo data vectors of distance units in the first group of sub-channels With the Echo data vectors of distance units in the second group of sub-channels They are respectively:

[0031]

[0032]

[0033] In the formula, Indicates the first The distance unit, the first The pulse, the first The compensated echo data of each channel, and .

[0034] Preferably, step S4 includes the following sub-steps:

[0035] Step S4.1: Construct the first-level dimensionality reduction matrix :

[0036]

[0037] In the formula, Represents the normalized Doppler frequency. express An identity matrix of order 1. Indicates first-level processing;

[0038] And based on the preset target normalized Doppler frequency and target normalized spatial frequency Construct the first-level target time-domain steering vector respectively. and target airspace steering vector :

[0039]

[0040]

[0041] In the formula, t Representing the time domain, s Indicates airspace.

[0042] Then, the target orientation vector after the first level of dimensionality reduction is calculated. :

[0043]

[0044] In the formula, target Indicates the goal.

[0045] Step S4.2: Use the training samples from the first group of sub-channels and the second group of sub-channels respectively. Through the first-level dimensionality reduction matrix A dimensionality reduction transformation is performed, and the maximum likelihood estimation method is used to calculate the clutter plus noise covariance matrix after the first-level dimensionality reduction for each group of data. :

[0046]

[0047] In the formula, Indicates the number of training samples;

[0048] Step S4.3: Based on the linear constraint minimum variance criterion, utilize the target guidance vector after the first-level dimensionality reduction. And the clutter plus noise covariance matrix after first-level dimensionality reduction Calculate the clutter suppression weight vectors for the first and second sub-channel data respectively. :

[0049]

[0050] In the formula, This represents finding the inverse of a matrix.

[0051] Step S4.4, using clutter suppression weight vector Sub-channel data respectively Filtering is performed to obtain the first-stage clutter suppression result. :

[0052]

[0053] In the formula, ; This indicates that after the first level of processing, the... Group sub-channel, first Data vector of one Doppler unit; This indicates that after the first level of processing, the... Group sub-channel, first The first Doppler unit, the first Data for each distance unit.

[0054] The results of the first-level clutter suppression are as follows: and Together constitute;

[0055] The results of the first-level clutter suppression were analyzed separately. Perform an inverse Fourier transform to obtain the time-domain data. :

[0056]

[0057] Wherein, IFFT() represents the inverse Fourier transform operation. , This indicates that after the first level of processing, the... Group sub-channel, first A data vector of pulses; This indicates that after the first level of processing, the... Group sub-channel, first The pulse, the first Data for each distance unit;

[0058] Finally, the data from all channels and pulses are reassembled sequentially into the first-level processing result. :

[0059]

[0060] in, , , This indicates the number of pulses after the first stage of processing.

[0061] Preferably, step S5 includes the following sub-steps:

[0062] Step S5.1: Construct the second-level dimensionality reduction matrix :

[0063]

[0064] In the formula, Represents the normalized Doppler frequency. Represents a second-order identity matrix. This indicates the second level of processing;

[0065] And based on the target normalized Doppler frequency and target normalized spatial frequency Construct the second-level target time-domain steering vector respectively and target airspace steering vector :

[0066]

[0067]

[0068] Then, the target orientation vector after the second-level dimensionality reduction is calculated. :

[0069]

[0070] Step S5.2, utilize the results of the first-level processing. training samples Through the second-level dimensionality reduction matrix A dimensionality reduction transformation is performed, and the maximum likelihood estimation method is used to calculate the clutter plus noise covariance matrix after the second-level dimensionality reduction. :

[0071]

[0072] Step S5.3: Process the results of the first stage. The target orientation vector after second-level dimensionality reduction The clutter-noise covariance matrix after second-order dimensionality reduction Substitute the values ​​into the adaptive matched filter detector and calculate the test statistic. :

[0073]

[0074] In the formula, AMF This indicates adaptive matched filtering;

[0075] And With preset reliable detection threshold If a comparison is made, Greater than or equal to If a moving target is detected, the distance unit where the moving target is located will be output. and Doppler unit The moving target detection results were obtained. .

[0076] Preferably, in step S6, based on the moving target detection result... Extract the corresponding first set of sub-channel data values ​​from the first-level processing results. Second group of sub-channel data values Based on the principle of single-pulse sum-difference beam angle measurement, the angle of a moving target is calculated. :

[0077]

[0078] In the formula, For the signal wavelength, This represents the spacing between sub-channels.

[0079] Preferably, in step S7, the radial slant range of the moving target is calculated based on the range cell where the moving target is located. Combined with the angle of the moving target, the three-dimensional coordinates of the moving target in the geodetic coordinate system are calculated through the transformation relationship between the radar coordinate system and the geodetic coordinate system.

[0080] This invention also provides an integrated system for multi-domain detection and localization of moving targets under strong clutter backgrounds, employing the aforementioned integrated method for multi-domain detection and localization of moving targets under strong clutter backgrounds, comprising:

[0081] Module M1 acquires multi-channel echo data from the space-based early warning radar and rearranges the multi-channel echo data to obtain the rearranged echo data matrix.

[0082] Module M2 uses a compensation matrix to perform distance-dependent error compensation on the rearranged echo data matrix, resulting in a compensated echo data matrix.

[0083] Module M3 performs channel combination on the compensated echo data matrix to construct the first set of sub-channel data and the second set of sub-channel data.

[0084] Module M4, based on the improved extended factorization algorithm (EFA) with additional auxiliary channels, performs first-level adaptive filtering on the first group of sub-channel data and the second group of sub-channel data respectively to obtain the first-level processing result;

[0085] Module M5, based on the results of the first-level processing, uses an adaptive matched filter detector combined with an improved extended factorization algorithm (EFA) to perform the second-level moving target detection. When a moving target is detected, it outputs the range cell and Doppler cell where the moving target is located.

[0086] Module M6, when a moving target is detected, extracts the corresponding first set of sub-channel data values ​​and the second set of sub-channel data values ​​from the first-level processing results according to the range unit and Doppler unit where the moving target is located, and calculates the angle of the moving target using the single-pulse sum-difference beam angle measurement principle;

[0087] Module M7 calculates the three-dimensional spatial position of the moving target in the geodetic coordinate system based on the range cell where the moving target is located and the angle of the moving target, through the transformation relationship between the radar coordinate system and the geodetic coordinate system.

[0088] Compared with the prior art, the present invention has the following beneficial effects:

[0089] 1. This invention employs a two-stage adaptive processing structure. By constructing a first-stage dimensionality reduction matrix and a second-stage dimensionality reduction matrix, the improved Extended Factorization Algorithm (EFA) is used in the first-stage processing to perform dimensionality reduction and clutter suppression on the two sets of sub-channel data, respectively. In the second-stage processing, adaptive matched filtering detection is performed based on the output of the first stage. This structure avoids the target detection performance loss caused by performing clutter suppression before using constant false alarm rate (CFAR) detection in traditional methods, thereby improving the target detection probability in strong clutter and complex Doppler environments.

[0090] 2. After obtaining the precise range-Doppler unit of the moving target, this invention directly extracts the complex values ​​of the corresponding channels from the first-level processing results and efficiently calculates the target angle using the single-pulse sum-difference beam angle measurement principle. This method requires no additional spatial domain search or large-scale matrix operations, resulting in low computational complexity. By combining the target range and angle information, three-dimensional spatial positioning of the target can be achieved through a mature coordinate transformation model, forming a complete and efficient processing chain from detection to positioning. Attached Figure Description

[0091] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0092] Figure 1 This is a flowchart of an integrated method for multi-domain detection and localization of moving targets under strong clutter background, as described in an embodiment of the present invention.

[0093] Figure 2 This is the clutter range-Doppler spectrum before range-dependent error compensation in this embodiment of the invention;

[0094] Figure 3 This is the clutter spatiotemporal spectrum before distance-dependent error compensation in this embodiment of the invention;

[0095] Figure 4 This is the clutter range-Doppler spectrum after range-dependent error compensation in an embodiment of the present invention;

[0096] Figure 5 This is the clutter spatiotemporal spectrum after distance-dependent error compensation in an embodiment of the present invention;

[0097] Figure 6 This is the range Doppler spectrum obtained after the first-stage clutter suppression in the integrated processing method for multi-domain detection and localization of moving targets under strong clutter background in the embodiments of the present invention;

[0098] Figure 7 This is a comparison of the first-stage clutter suppression of the integrated processing method for multi-domain detection and localization of moving targets under strong clutter background in this embodiment of the invention and the output signal-to-clutter-to-noise ratio curve obtained by using the traditional EFA for the first-stage clutter suppression.

[0099] Figure 8 This is a diagram of the second-level moving target detection result obtained by the integrated processing method for multi-domain detection and localization of moving targets under strong clutter background in this embodiment of the invention.

[0100] Figure 9 The image shows the moving target detection results obtained by using the improved EFA for clutter suppression based on the first-level processing results and then using CA-CFAR for the second-level moving target multi-domain detection.

[0101] Figure 10 This is a comparison of the moving target detection probability curves obtained by the integrated processing method for multi-domain detection and localization of moving targets under strong clutter background in this embodiment of the invention and the traditional method for clutter suppression using an improved EFA and then CA-CFAR for second-level multi-domain detection of moving targets based on the first-level processing results;

[0102] Figure 11This is an angle estimation result diagram of the integrated processing method for multi-domain detection and localization of moving targets under strong clutter background in this embodiment of the invention;

[0103] Figure 12 The image shows the target positioning accuracy results of the integrated processing method for multi-domain detection and positioning of moving targets under strong clutter background provided in this embodiment of the invention. Detailed Implementation

[0104] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.

[0105] This invention provides an integrated method and system for multi-domain detection and localization of moving targets in a cluttered environment. The method includes acquiring multi-channel echo data from a space-based early warning radar and performing data rearrangement and range-dependent error compensation; then, combining the channels of the compensated echo data matrix to construct a first set of sub-channel data and a second set of sub-channel data; employing an improved Extended Factorization Algorithm (EFA) based on additional auxiliary channels to perform a first-level adaptive filtering process on the first and second sets of sub-channel data respectively, obtaining the first-level processing result; using an adaptive matched filter detector combined with the improved EFA algorithm to perform a second-level moving target detection on the first-level processing result, outputting the range cell and Doppler cell where the moving target is located when a moving target is detected; then extracting the corresponding data from the first-level processing result and calculating the angle of the moving target using the single-pulse sum-difference beam angle measurement principle; finally, calculating the three-dimensional spatial position of the moving target in the geodetic coordinate system through the transformation relationship between the radar coordinate system and the geodetic coordinate system. This invention, through a two-level adaptive processing structure, effectively achieves integrated processing of moving target detection, angle estimation, and localization in a cluttered environment.

[0106] Example 1:

[0107] Figure 1 This is a flowchart of an integrated method for multi-domain detection and localization of moving targets under strong clutter background, as described in an embodiment of the present invention.

[0108] like Figure 1 As shown, this embodiment provides an integrated method for multi-domain detection and localization of moving targets against a strong clutter background. All steps in this embodiment are performed on the MATLAB 2020b simulation platform, and specifically include the following steps:

[0109] Step S1: Acquire multi-channel echo data from the space-based early warning radar and rearrange the multi-channel echo data to obtain the rearranged echo data matrix.

[0110] In this embodiment, the multi-channel echo data is rearranged to obtain a rearranged echo data matrix. :

[0111]

[0112] In the formula, , This represents the total number of distance units. The number of pulses. For the number of channels, the first... Echo data vector of each distance cell , Indicates the first The distance unit, the first The pulse, the first Echo data from each channel This indicates the matrix transpose.

[0113] Step S2: Using the compensation matrix, distance dependence error compensation is performed on the rearranged echo data matrix to obtain the compensated echo data matrix.

[0114] In this embodiment, for the first The compensation matrix for each distance cell is:

[0115]

[0116] In the formula, For the first The clutter Doppler center frequency of each distance cell, for 3D identity matrix This indicates the creation of a diagonal matrix. Indicates the Kronecker product; j Represents the imaginary unit. d It stands for doppler.

[0117] Using the first The compensation matrix of the distance unit is for the first distance unit. The data from each distance cell is compensated to obtain the compensated data vector. :

[0118]

[0119] In the formula, This indicates the conjugate transpose.

[0120] Furthermore, in step S2, the compensated data vectors of all range cells are combined to obtain the compensated echo data matrix. :

[0121]

[0122] In the formula, .

[0123] Step S3: Combine the channels of the compensated echo data matrix to construct the first set of sub-channel data and the second set of sub-channel data.

[0124] In this embodiment, in step S3, the compensated first set of sub-channel data and the compensated second set of sub-channel data They are respectively:

[0125]

[0126]

[0127] In the formula, the first Echo data vectors of distance units in the first group of sub-channels With the Echo data vectors of distance units in the second group of sub-channels They are respectively:

[0128]

[0129]

[0130] In the formula, Indicates the first The distance unit, the first The pulse, the first The compensated echo data of each channel, and .

[0131] Step S4: Based on the improved extended factorization algorithm (EFA) with additional auxiliary channels, perform first-level adaptive filtering on the first group of sub-channel data and the second group of sub-channel data respectively to obtain the first-level processing result.

[0132] Specifically, step S4 includes the following sub-steps:

[0133] Step S4.1: Construct the first-level dimensionality reduction matrix :

[0134]

[0135] In the formula, Represents the normalized Doppler frequency. express An identity matrix of order 1. Indicates first-level processing;

[0136] And based on the preset target normalized Doppler frequency and target normalized spatial frequency Construct the first-level target time-domain steering vector respectively. and target airspace steering vector :

[0137]

[0138]

[0139] In the formula, t Representing the time domain, s Indicates airspace.

[0140] Then, the target orientation vector after the first level of dimensionality reduction is calculated. :

[0141]

[0142] In the formula, target Indicates the goal.

[0143] Step S4.2: Use the training samples from the first group of sub-channels and the second group of sub-channels respectively. Through the first-level dimensionality reduction matrix A dimensionality reduction transformation is performed, and the maximum likelihood estimation method is used to calculate the clutter plus noise covariance matrix after the first-level dimensionality reduction for each group of data. :

[0144]

[0145] In the formula, Indicates the number of training samples;

[0146] Step S4.3: Based on the linear constraint minimum variance criterion, utilize the target guidance vector after the first-level dimensionality reduction. And the clutter plus noise covariance matrix after first-level dimensionality reduction Calculate the clutter suppression weight vectors for the first and second sub-channel data respectively. :

[0147]

[0148] In the formula, This represents finding the inverse of a matrix.

[0149] Step S4.4, using clutter suppression weight vector Sub-channel data respectively Filtering is performed to obtain the first-stage clutter suppression result. :

[0150]

[0151] In the formula, ; This indicates that after the first level of processing, the... Group sub-channel, first Data vector of one Doppler unit; This indicates that after the first level of processing, the... Group sub-channel, first The first Doppler unit, the first Data for each distance unit.

[0152] The results of the first-level clutter suppression are as follows: and Together constitute;

[0153] The results of the first-level clutter suppression were analyzed separately. Perform an inverse Fourier transform to obtain the time-domain data. :

[0154]

[0155] Wherein, IFFT() represents the inverse Fourier transform operation. ; This indicates that after the first level of processing, the... Group sub-channel, first A data vector of pulses; This indicates that after the first level of processing, the... Group sub-channel, first The pulse, the first Data for each distance unit;

[0156] Finally, the data from all channels and pulses are reassembled sequentially into the first-level processing result. :

[0157]

[0158] in, , , This indicates the number of pulses after the first stage of processing.

[0159] Step S5: Based on the results of the first-level processing, the second-level moving target detection is performed using an adaptive matched filter detector that combines the improved extended factorization algorithm (EFA). When a moving target is detected, the range cell and Doppler cell where the moving target is located are output.

[0160] Specifically, step S5 includes the following sub-steps:

[0161] Step S5.1: Construct the second-level dimensionality reduction matrix :

[0162]

[0163] In the formula, Represents the normalized Doppler frequency. Represents a second-order identity matrix. This indicates the second level of processing;

[0164] And based on the target normalized Doppler frequency and target normalized spatial frequency Construct the second-level target time-domain steering vector respectively and target airspace steering vector :

[0165]

[0166]

[0167] Then, the target orientation vector after the second-level dimensionality reduction is calculated. :

[0168]

[0169] Step S5.2, utilize the results of the first-level processing. training samples Through the second-level dimensionality reduction matrix A dimensionality reduction transformation is performed, and the maximum likelihood estimation method is used to calculate the clutter plus noise covariance matrix after the second-level dimensionality reduction. :

[0170]

[0171] Step S5.3: Process the results of the first stage. The target orientation vector after second-level dimensionality reduction The clutter-noise covariance matrix after second-order dimensionality reduction Substitute the values ​​into the adaptive matched filter detector and calculate the test statistic. :

[0172]

[0173] In the formula, AMF This indicates adaptive matched filtering;

[0174] And With preset reliable detection threshold If a comparison is made, Greater than or equal to If a moving target is detected, the distance unit where the moving target is located will be output. and Doppler unit The moving target detection results were obtained. .

[0175] Specifically, It is the first The test statistics for each Doppler unit and all distance units are given, with a data dimension of [missing information]. Traversing when detecting moving targets One Doppler unit is obtained. The test statistic matrix, such as Figure 8 As shown, the final distance cell where the moving target is located is obtained. and Doppler unit .

[0176] Step S6: When a moving target is detected, the corresponding first set of sub-channel data values ​​and the second set of sub-channel data values ​​are extracted from the first-level processing results according to the range cell and Doppler cell where the moving target is located, and the angle of the moving target is calculated using the single-pulse sum-difference beam angle measurement principle.

[0177] Specifically, in step S6, based on the moving target detection results... Extract the corresponding first set of sub-channel data values ​​from the first-level processing results. Second group of sub-channel data values Based on the principle of single-pulse sum-difference beam angle measurement, the angle of a moving target is calculated. :

[0178]

[0179] In the formula, For the signal wavelength, This represents the spacing between sub-channels.

[0180] Step S7: Based on the range cell where the moving target is located and the angle of the moving target, calculate the three-dimensional spatial position of the moving target in the geodetic coordinate system through the transformation relationship between the radar coordinate system and the geodetic coordinate system.

[0181] Specifically, in step S7, the radial slant range of the moving target is calculated based on the range cell where the moving target is located. Combined with the angle of the moving target, the three-dimensional coordinates of the moving target in the geodetic coordinate system are calculated through the transformation relationship between the radar coordinate system and the geodetic coordinate system.

[0182] All steps in this embodiment were performed on the MATLAB 2020b simulation platform, and the experimental results are as follows:

[0183] Figure 2 and Figure 3These are the clutter range-Doppler spectrum and space-time spectrum before compensation, respectively. Figure 4 and Figure 5 The figures show the corresponding results after compensation. As can be seen from the comparison, the distance dependence of the clutter spectrum is effectively corrected after processing with the compensation matrix of this invention. Figure 6 The distance-Doppler spectrum after the first stage of processing according to the present invention is shown, and... Figure 4 A comparison of the clutter range-Doppler spectra before and after the first-stage processing shows that the range-Doppler spectra after the first-stage processing can clearly distinguish moving targets, and the strong clutter that masks moving targets is significantly suppressed. Figure 7 The output signal-to-clutter ratio (SNR) curves of the improved EFA of this invention and the traditional EFA after the first-stage processing were compared. The results show that the improved EFA algorithm based on an additional auxiliary channel used in this invention has better clutter suppression performance. Figure 8 The results of the second-level moving target detection of the present invention are shown, in which the target is clearly detected. Figure 9 The results of the conventional method (first using improved EFA clutter suppression, then using CA-CFAR detection) are presented as a comparison. Figure 10 The diagram shows a comparison of the moving target detection probability curves obtained by the present invention and the traditional method that uses an improved EFA for clutter suppression based on the first-level processing results and then uses CA-CFAR for the second-level moving target multi-domain detection. It shows that the integrated two-level processing method of the present invention achieves better moving target detection performance. Figure 11 An angle diagram of the moving target obtained by the present invention is shown. Figure 12 The final target positioning accuracy result is shown in the figure. The results show that the method of the present invention can stably and accurately realize the angle estimation and three-dimensional spatial positioning of moving targets within the main lobe beamwidth of the antenna.

[0184] Example 2:

[0185] The present invention also provides an integrated system for multi-domain detection and localization of moving targets under strong clutter background. The integrated system for multi-domain detection and localization of moving targets under strong clutter background can be implemented by executing the process steps of the integrated method for multi-domain detection and localization of moving targets under strong clutter background. That is, those skilled in the art can understand the integrated method for multi-domain detection and localization of moving targets under strong clutter background as a preferred embodiment of the integrated system for multi-domain detection and localization of moving targets under strong clutter background.

[0186] Specifically, this integrated system for multi-domain detection and localization of moving targets under strong clutter backgrounds includes:

[0187] Module M1 acquires multi-channel echo data from the space-based early warning radar and rearranges the multi-channel echo data to obtain the rearranged echo data matrix.

[0188] Module M2 uses a compensation matrix to perform distance-dependent error compensation on the rearranged echo data matrix, resulting in a compensated echo data matrix.

[0189] Module M3 performs channel combination on the compensated echo data matrix to construct the first set of sub-channel data and the second set of sub-channel data.

[0190] Module M4, based on the improved extended factorization algorithm (EFA) with additional auxiliary channels, performs first-level adaptive filtering on the first group of sub-channel data and the second group of sub-channel data respectively to obtain the first-level processing result;

[0191] Module M5, based on the results of the first-level processing, uses an adaptive matched filter detector combined with an improved extended factorization algorithm (EFA) to perform the second-level moving target detection. When a moving target is detected, it outputs the range cell and Doppler cell where the moving target is located.

[0192] Module M6, when a moving target is detected, extracts the corresponding first set of sub-channel data values ​​and the second set of sub-channel data values ​​from the first-level processing results according to the range unit and Doppler unit where the moving target is located, and calculates the angle of the moving target using the single-pulse sum-difference beam angle measurement principle;

[0193] Module M7 calculates the three-dimensional spatial position of the moving target in the geodetic coordinate system based on the range cell where the moving target is located and the angle of the moving target, through the transformation relationship between the radar coordinate system and the geodetic coordinate system.

[0194] Specifically, in module M1, the multi-channel echo data is rearranged to obtain the rearranged echo data matrix. :

[0195]

[0196] In the formula, , This represents the total number of distance units. The number of pulses. For the number of channels, the first... Echo data vector of each distance cell , Indicates the first The distance unit, the first The pulse, the first Echo data from each channel This indicates the matrix transpose.

[0197] Specifically, in module M2, for the first The compensation matrix for each distance cell is:

[0198]

[0199] In the formula, For the first The clutter Doppler center frequency of each distance cell, for 3D identity matrix This indicates the creation of a diagonal matrix. Indicates the Kronecker product; j Represents the imaginary unit. d It stands for doppler.

[0200] Using the first The compensation matrix of the distance unit is for the first distance unit. The data from each distance cell is compensated to obtain the compensated data vector. :

[0201]

[0202] In the formula, This indicates the conjugate transpose.

[0203] Furthermore, in module M2, the compensated data vectors of all range cells are combined to obtain the compensated echo data matrix. :

[0204]

[0205] In the formula, .

[0206] Specifically, in module M3, the first group of sub-channel data Second group of sub-channel data They are respectively:

[0207]

[0208]

[0209] In the formula, the first Echo data vectors of distance units in the first group of sub-channels With the Echo data vectors of distance units in the second group of sub-channels They are respectively:

[0210]

[0211]

[0212] In the formula, Indicates the first The distance unit, the first The pulse, the first The compensated echo data of each channel, and .

[0213] Specifically, module M4 includes the following sub-modules:

[0214] Module M4.1 constructs the first-level dimensionality reduction matrix. :

[0215]

[0216] In the formula, Represents the normalized Doppler frequency. express An identity matrix of order 1. Indicates first-level processing;

[0217] And based on the preset target normalized Doppler frequency and target normalized spatial frequency Construct the first-level target time-domain steering vector respectively. and target airspace steering vector :

[0218]

[0219]

[0220] In the formula, t Representing the time domain, s Indicates airspace.

[0221] Then, the target orientation vector after the first level of dimensionality reduction is calculated. :

[0222]

[0223] In the formula, target Indicates the goal.

[0224] Module M4.2 utilizes the training samples from the first group of sub-channels and the second group of sub-channels, respectively. Through the first-level dimensionality reduction matrix A dimensionality reduction transformation is performed, and the maximum likelihood estimation method is used to calculate the clutter plus noise covariance matrix after the first-level dimensionality reduction for each group of data. :

[0225]

[0226] In the formula, Indicates the number of training samples;

[0227] Module M4.3, based on the linear constraint minimum variance criterion, utilizes the target orientation vector after the first-level dimensionality reduction. And the clutter plus noise covariance matrix after first-level dimensionality reduction Calculate the clutter suppression weight vectors for the first and second sub-channel data respectively. :

[0228]

[0229] In the formula, This represents finding the inverse of a matrix.

[0230] Module M4.4 utilizes clutter suppression weight vectors Sub-channel data respectively Filtering is performed to obtain the first-stage clutter suppression result. :

[0231]

[0232] In the formula, ; This indicates that after the first level of processing, the... Group sub-channel, first Data vector of one Doppler unit; This indicates that after the first level of processing, the... Group sub-channel, first The first Doppler unit, the first Data for each distance unit.

[0233] The results of the first-level clutter suppression are as follows: and Together constitute;

[0234] The results of the first-level clutter suppression were analyzed separately. Perform an inverse Fourier transform to obtain the time-domain data. :

[0235]

[0236] Wherein, IFFT() represents the inverse Fourier transform operation. This indicates that after the first level of processing, the... Group sub-channel, first A data vector of pulses; This indicates that after the first level of processing, the... Group sub-channel, first The pulse, the first Data for each distance unit;

[0237] Finally, the data from all channels and pulses are reassembled sequentially into the first-level processing result. :

[0238]

[0239] in, , , This indicates the number of pulses after the first stage of processing.

[0240] Specifically, module M5 includes the following sub-modules:

[0241] Module M5.1, constructing the second-level dimensionality reduction matrix :

[0242]

[0243] In the formula, Represents the normalized Doppler frequency. Represents a second-order identity matrix. This indicates the second level of processing;

[0244] And based on the target normalized Doppler frequency and target normalized spatial frequency Construct the second-level target time-domain steering vector respectively and target airspace steering vector :

[0245]

[0246]

[0247] Then, the target orientation vector after the second-level dimensionality reduction is calculated. :

[0248]

[0249] Module M5.2 utilizes the results of the first-level processing. training samples Through the second-level dimensionality reduction matrix A dimensionality reduction transformation is performed, and the maximum likelihood estimation method is used to calculate the clutter plus noise covariance matrix after the second-level dimensionality reduction. :

[0250]

[0251] Module M5.3 will process the results of the first level of processing. The target orientation vector after second-level dimensionality reduction The clutter-noise covariance matrix after second-order dimensionality reduction Substitute the values ​​into the adaptive matched filter detector and calculate the test statistic. :

[0252]

[0253] In the formula, AMF This indicates adaptive matched filtering;

[0254] And With preset reliable detection threshold If a comparison is made, Greater than or equal to If a moving target is detected, the distance unit where the moving target is located will be output. and Doppler unit The moving target detection results were obtained. .

[0255] Specifically, in module M6, based on the moving target detection results... Extract the corresponding first set of sub-channel data values ​​from the first-level processing results. Second group of sub-channel data values Based on the principle of single-pulse sum-difference beam angle measurement, the angle of a moving target is calculated. :

[0256]

[0257] In the formula, For the signal wavelength, This represents the spacing between sub-channels.

[0258] Those skilled in the art will understand that, besides implementing the system and its various devices, modules, and units provided by this invention in the form of purely computer-readable program code, the same functions can be achieved entirely through logical programming of the method steps, making the system and its various devices, modules, and units of this invention function in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, the system and its various devices, modules, and units provided by this invention can be considered as a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered as structures within the hardware component; alternatively, the devices, modules, and units for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0259] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A method for integrated multi-domain detection and localization of moving targets under strong clutter background, characterized in that, Includes the following steps: Step S1: Acquire multi-channel echo data from the space-based early warning radar, and rearrange the multi-channel echo data to obtain the rearranged echo data matrix. Step S2: Using the compensation matrix, distance dependence error compensation is performed on the rearranged echo data matrix to obtain the compensated echo data matrix. Step S3: Combine the channels of the compensated echo data matrix to construct the first set of sub-channel data and the second set of sub-channel data; Step S4: Based on the improved extended factorization algorithm with additional auxiliary channels, perform first-level adaptive filtering on the first group of sub-channel data and the second group of sub-channel data respectively to obtain the first-level processing result; Step S4 includes the following sub-steps: Step S4.1: Construct the first-level dimensionality reduction matrix : In the formula, Represents the normalized Doppler frequency. express An identity matrix of order 1. Indicates first-level processing; And based on the preset target normalized Doppler frequency and target normalized spatial frequency Construct the first-level target time-domain steering vector respectively. and target airspace steering vector : In the formula, t Representing the time domain, s Indicates airspace; Then, the target orientation vector after the first level of dimensionality reduction is calculated. : In the formula, target Indicate the goal; Step S4.2: Use the training samples from the first group of sub-channels and the second group of sub-channels respectively. Through the first-level dimensionality reduction matrix A dimensionality reduction transformation is performed, and the maximum likelihood estimation method is used to calculate the clutter plus noise covariance matrix after the first-level dimensionality reduction for each group of data. : In the formula, Indicates the number of training samples; Step S4.3: Based on the linear constraint minimum variance criterion, utilize the target guidance vector after the first-level dimensionality reduction. And the clutter plus noise covariance matrix after the first-level dimensionality reduction Calculate the clutter suppression weight vectors for the first group of sub-channel data and the second group of sub-channel data, respectively. : In the formula, This represents finding the inverse of a matrix. Step S4.4, using the clutter suppression weight vector Sub-channel data respectively Filtering is performed to obtain the first-stage clutter suppression result. : In the formula, ; This indicates that after the first level of processing, the... Group sub-channel, first Data vector of one Doppler unit; This indicates that after the first level of processing, the... Group sub-channel, first The first Doppler unit, the first Data for each distance unit; The first-stage clutter suppression result is from and Together constitute; And respectively the clutter suppression results of the first stage Perform an inverse Fourier transform to obtain the time-domain data. : Wherein, IFFT() represents the inverse Fourier transform operation. This indicates that after the first level of processing, the... Group sub-channel, first A data vector of pulses; This indicates that after the first level of processing, the... Group sub-channel, first The pulse, the first Data for each distance unit; Finally, the data from all channels and pulses are reassembled sequentially into the first-level processing result. : in, , , This indicates the number of pulses after the first stage of processing; Step S5: Based on the first-level processing result, the second-level moving target detection is performed using an adaptive matched filter detector combined with an improved extended factorization algorithm. When a moving target is detected, the distance cell and Doppler cell where the moving target is located are output. Step S6: When a moving target is detected, according to the distance unit and Doppler unit where the moving target is located, the corresponding first set of sub-channel data values ​​and the second set of sub-channel data values ​​are extracted from the first-level processing results, and the angle of the moving target is calculated using the single-pulse sum-difference beam angle measurement principle. Step S7: Based on the distance cell where the moving target is located and the angle of the moving target, calculate the three-dimensional spatial position of the moving target in the geodetic coordinate system through the transformation relationship between the radar coordinate system and the geodetic coordinate system.

2. The integrated method for multi-domain detection and localization of moving targets under strong clutter background as described in claim 1, characterized in that, In step S1, the multi-channel echo data is rearranged to obtain a rearranged echo data matrix. : In the formula, , This represents the total number of distance units. The number of pulses. For the number of channels, the first... Echo data vector of each distance cell , Indicates the first The distance unit, the first The pulse, the first Echo data from each channel This indicates the matrix transpose.

3. The integrated method for multi-domain detection and localization of moving targets under strong clutter background as described in claim 2, characterized in that, In step S2, for the first The compensation matrix for each distance cell is: In the formula, For the first The clutter Doppler center frequency of each distance cell, for 3D identity matrix This indicates the creation of a diagonal matrix. Indicates the Kronecker product; j Represents the imaginary unit. d Indicates Doppler; Using the first The compensation matrix of the distance unit is for the first distance unit. The data from each distance cell is compensated to obtain the compensated data vector. : In the formula, This indicates the conjugate transpose.

4. The integrated method for multi-domain detection and localization of moving targets under strong clutter background as described in claim 3, characterized in that, In step S2, the compensated data vectors of all range cells are combined to obtain the compensated echo data matrix. : In the formula, .

5. The integrated method for multi-domain detection and localization of moving targets under strong clutter background as described in claim 4, characterized in that, In step S3, the first group of sub-channel data and the second group of sub-channel data They are respectively: In the formula, the first Echo data vectors of distance units in the first group of sub-channels With the Echo data vectors of distance units in the second group of sub-channels They are respectively: In the formula, Indicates the first The distance unit, the first The pulse, the first The compensated echo data of each channel, and .

6. The integrated method for multi-domain detection and localization of moving targets under strong clutter background as described in claim 5, characterized in that, Step S5 includes the following sub-steps: Step S5.1: Construct the second-level dimensionality reduction matrix : In the formula, Represents the normalized Doppler frequency. Represents a second-order identity matrix. This indicates the second level of processing; And based on the target normalized Doppler frequency and target normalized spatial frequency Construct the second-level target time-domain steering vector respectively. and target airspace steering vector : Then, the target orientation vector after the second-level dimensionality reduction is calculated. : Step S5.2, utilize the results of the first-level processing. training samples Through the second-level dimensionality reduction matrix A dimensionality reduction transformation is performed, and the maximum likelihood estimation method is used to calculate the clutter plus noise covariance matrix after the second-level dimensionality reduction. : Step S5.3: Process the results of the first stage. The target orientation vector after the second level of dimensionality reduction and the clutter plus noise covariance matrix after the second-level dimensionality reduction Substitute the values ​​into the adaptive matched filter detector and calculate the test statistic. : In the formula, AMF This indicates adaptive matched filtering; And With preset reliable detection threshold If a comparison is made, Greater than or equal to If a moving target is detected, the distance unit of the moving target will be output. and Doppler unit The moving target detection results are obtained. .

7. The integrated method for multi-domain detection and localization of moving targets under strong clutter background as described in claim 6, characterized in that, In step S6, based on the moving target detection result Extract the corresponding first set of sub-channel data values ​​from the first-level processing results. Second group of sub-channel data values Based on the principle of single-pulse sum-difference beam angle measurement, the angle of the moving target is calculated. : In the formula, For the signal wavelength, This represents the spacing between sub-channels.

8. The integrated method for multi-domain detection and localization of moving targets under strong clutter background as described in claim 7, characterized in that, In step S7, the radial slant range of the moving target is calculated based on the distance cell where the moving target is located. Combined with the angle of the moving target, the three-dimensional coordinates of the moving target in the geodetic coordinate system are calculated through the transformation relationship between the radar coordinate system and the geodetic coordinate system.

9. A multi-domain detection and localization integrated system for moving targets under strong clutter background, employing the multi-domain detection and localization integrated method for moving targets under strong clutter background as described in any one of claims 1-8, characterized in that, include: Module M1 acquires multi-channel echo data from the space-based early warning radar and rearranges the multi-channel echo data to obtain a rearranged echo data matrix. Module M2 uses a compensation matrix to perform distance-dependent error compensation on the rearranged echo data matrix to obtain a compensated echo data matrix. Module M3 performs channel combination on the compensated echo data matrix to construct a first set of sub-channel data and a second set of sub-channel data; Module M4, based on an improved extended factorization algorithm with additional auxiliary channels, performs first-level adaptive filtering on the first group of sub-channel data and the second group of sub-channel data respectively to obtain the first-level processing result; Module M5, based on the first-level processing results, uses an adaptive matched filter detector combined with an improved extended factorization algorithm to perform the second-level moving target detection. When a moving target is detected, it outputs the distance cell and Doppler cell where the moving target is located. Module M6, when a moving target is detected, extracts the corresponding first set of sub-channel data values ​​and the second set of sub-channel data values ​​from the first-level processing results according to the distance unit and Doppler unit where the moving target is located, and calculates the angle of the moving target using the single-pulse sum-difference beam angle measurement principle; Module M7 calculates the three-dimensional spatial position of the moving target in the geodetic coordinate system based on the distance cell where the moving target is located and the angle of the moving target, through the transformation relationship between the radar coordinate system and the geodetic coordinate system.

Images