Horizontal array intelligent target direction estimation method and system based on subarray division

Abstract

The application discloses a horizontal array intelligent target direction estimation method and system based on subarray division, relates to the technical field of target direction estimation, divides a full linear array into a plurality of subarrays, inputs CBF and MVDR beam domain features of the subarrays into a subarray DOA estimation model based on multi-stage fusion of multi-beam domain features, realizes fusion of a plurality of beam forming methods, and guarantees DOA estimation precision of a single subarray; then, the DOA estimation result of the single subarray, subarray information and array element parameters are taken as inputs of a conditional convolution DOA estimation model based on multi-source feature fusion, a convolution kernel in the model is parameterized, convolution kernel parameters are dynamically generated according to the inputs, all convolution kernel parameters in a backbone network are adaptively adjusted, and the information is fused to obtain a final DOA estimation result. According to the application, the full linear array is divided into a plurality of subarrays, multipath effects and noise interference can be effectively reduced, and the estimation precision of the target direction is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of target orientation estimation, and more particularly to a method and system for horizontal array intelligent target orientation estimation based on subarray division. Background Art

[0002] At present, using underwater acoustic information to accurately estimate the direction of arrival (DOA) of underwater acoustic targets is an important research direction in the field of array signal processing. The basic problem of DOA estimation is to determine the position of the signal of interest in space at the same time. Generally, key information of the target can be obtained through the DOA estimation method, such as the azimuth angle of the target relative to the reference element of the receiving array. The DOA process is generally achieved by processing the array signal through beamforming technology. Array signal refers to the spatial sampling signal received by a sensor array (such as a horizontal array) composed of multiple sensors placed in different spatial positions. Beamforming is a spatial processing technology for data received by multiple sensor arrays, which aims to enhance the signal reception capability in a specific direction while suppressing interference from other directions.

[0003] In real-world ocean environments, for long horizontal arrays, when the noise source level of a target is low, array elements farther from the target may not receive the target noise. Furthermore, arrays placed on the seafloor for extended periods may experience issues such as element failure or reduced sensitivity. Direction estimation using all array elements directly would compromise the final accuracy. Therefore, large arrays often employ subarray partitioning. For example, hundreds or thousands of frame elements are divided into a dozen or so subarrays, beamforming is performed within the subarrays, and the beamforming results across subarrays are then combined to produce the final result. By designing appropriate subarray weights, DOA estimation results can often be superior to those achieved with a single subarray.

[0004] Existing beamforming algorithms include conventional beamforming (CBF) and minimum variance distortionless response (MVDR). However, these algorithms have their own advantages and disadvantages. For example, MVDR is a high-resolution beamforming technique, but its robustness is low, and even a small mismatch can lead to large positioning errors. CBF does not rely on an accurate estimate of the noise covariance matrix and is suitable for rapidly changing environments, but its accuracy is inferior to MVDR.

[0005] Therefore, how to provide a new technology to further optimize the existing DOA algorithm to improve the accuracy of target direction information estimation is a problem that those skilled in the art urgently need to solve. SUMMARY

[0006] Therefore, the application provides a horizontal array intelligent target azimuth estimation method and system based on subarray division to solve the problems in the background art.

[0007] To achieve the above-mentioned purpose, the application adopts the following technical solutions:

[0008] A horizontal array intelligent target azimuth estimation method based on subarray division, comprising:

[0009] Step 1, dividing the signal direction of arrival into A sectors, setting the label as 1 if there is a target source in the sector, and 0 otherwise, and finally representing the label as a one-dimensional matrix with a length of A dimensions and only one value of 1 and other values of 0;

[0010] Step 2, randomly reading a full linear array data, dividing the full linear array P into several subarrays [P1, P2, P3,.., P N ];

[0011] Step 3, subarray data preprocessing, using two beam forming algorithms CBF and MVDR to process the several subarrays into two beam domain features, single-frequency CBF output power P CBF (f, θ) and single-frequency diagonal loading MVDR output power P Mv (f, θ), normalizing the two beam domain features and taking the modulus value to obtain and as the input of the subarray DOA estimation model based on multi-stage fusion of multiple beam domain features;

[0012] Step 4, inputting and to the pre-trained subarray DOA estimation model based on multi-stage fusion of multiple beam domain features to obtain the DOA estimation result

[0013] Step 5, executing steps 3 to 4 on all subarrays in step 2 to obtain all subarray DOA estimation results [y1, y2,.., y N ];

[0014] Step 6, inputting all subarray DOA estimation results [y1, y2,.., y N ], several subarray information [P1, P2, P3,.., P N ], and array element parameters to the pre-trained conditional convolution DOA estimation model based on multi-source feature fusion to obtain the DOA estimation result of the full linear array.

[0015] Optionally, the step 3 specifically comprises:

[0016] By scanning the existence range angles of several sub-array DOAs, the single-frequency CBF output power is obtained as follows:

[0017]

[0018] in is the weighted vector of CBF, obtained according to different weighting methods The values ​​are also different, for example, a uniform weighted approach can be used. R x is the cross-spectral density matrix of the received signal, R n With R x The unit noise matrix has the same size, and SNR is the signal-to-noise ratio.

[0019] By scanning the existence range angles of several sub-array DOAs, the single-frequency diagonally loaded MVDR output power is obtained as follows:

[0020]

[0021] Where W MV (f, θ) is the optimal weight vector of MVDR, which can be obtained by using the Lagrange operator method. R x is the cross-spectral density matrix of the received signal, R n With R x The unit noise matrix of the same size, SNR is the signal-to-noise ratio, when no noise is added, R = R x +ρI, ρ is the diagonal loading, I is the size and R x Same identity matrix;

[0022] The beamformed output P CBF (f, θ) and P Mv (f, θ) is normalized and modulo value is obtained and Input into the subarray DOA estimation model based on multi-stage fusion multi-beam domain features; the input dimension is N f ×N θ , as the input of the model, where N f is the number of frequency points, N θ is the number of scanning angles during beamforming, and the matrix form of the two inputs is

[0023]

[0024]

[0025] For each input matrix Corresponding to a label vector

[0026] Optionally, before processing in step 3, the following steps may be further included:

[0027] Modeling of a single full-line array receiving signal, K signals of center frequency f S(t) = [s1(t), s2(t), ..., s K (t)] T Since the wave direction θ=[θ1,θ2,…,θ K ] T The signal is incident on a single uniform linear array. The number of sensors on the linear array is N, N>K, the array element spacing is d, and the signal satisfies the far-field narrowband assumption. The signal envelope does not change during the passage through the array. X(t) = [x1(t), x2(t), ..., x K (t)] T For each array element receiving signal, the array receiving signal model is expressed as:

[0028] X(t)=B(f,θ)S(t)+N(t);

[0029] Among them, B(f, θ)=[a(f, θ1), a(f, θ2), .., a(f, θ K )] is the array manifold vector matrix, and the incoming wave direction is θ k The corresponding array manifold vector is a(f,θ k )=[exp(-jk T p1), exp(-jk T p2), ..., exp(-jk T p N )] T , the array element position is [p1, p2, ..., p N ], uniform linear array [p1, p2,,.., p N ]=[0,d,...,(N-1)d],k is the wave number vector, k=2πf / c·v(θ),b(θ) is the unit vector of the acoustic signal propagation direction, c is the signal propagation speed, N(t)=[n1(t),n2(t),...,n N (t)] T is the noise vector matrix received by each element of the linear array.

[0030] Optionally, step 4 specifically includes:

[0031] Step 4-1, get The input is fed into a sub-array DOA estimation model that fuses multi-beam domain features in multiple stages, and an encoder is used to encode the input features.

[0032] Step 4-2: Use two feature extraction paths to extract CBF beam domain features G respectively r and MVDR beam domain characteristics L r , the two feature extraction paths are composed of R cascaded downsampling modules, which are composed of Conv Block and Identity Block:

[0033] Step 4-3: For the CBF beam domain feature G obtained at each stage r and MVDR beam domain characteristics L r , the hierarchical feature fusion module is used to fuse the CBF beam domain features and MVDR beam domain features of different scales obtained at different stages, and the fused feature F after the R stage is obtained. R ;

[0034] Step 4-4: The feature F after the R stage fusion R Input into the classifier to obtain the sub-array DOA estimation result

[0035] Optionally, the encoder is specifically:

[0036]

[0037] Among them, e CBF , e MV Represents the low-level features obtained by encoding, ε represents the encoder, which consists of two-dimensional convolution Conv2d, one-dimensional convolution Conv1d, normalization layer LayerNorm, ReLU activation function, and maximum pooling MaxPool.

[0038] Optionally, the downsampling module is composed of a convolution block and an identity block, Conv Block is a convolution block, and IdentitvBlock is an identity block. The specific structure is:

[0039] Conv Block(x)=ReLU(BatchNorm(Conv2d

[0040] (ReLU(BatchNorm(Conv2d(ReLU(BatchNorm(Conb2d(x))))))))+

[0041] BatchNorm(Conv2d(x)));

[0042] Identity Block=BatchNorm(Conv2d(ReLU(BatchNorm

[0043] (Conv2d(ReLU(BatchNorm(Conv2d(x))))))))+x;

[0044] in, represents the rth downsampling module for processing CBF features, r∈(1,R), represents the parameters of the rth downsampling module,

[0045] The specific extraction of CBF beam domain features and MVDR beam domain features is as follows:

[0046]

[0047]

[0048] Optionally, the specific calculation process of the hierarchical feature fusion module is as follows:

[0049] Enter G r , L r and F r-1 , calculated using the following formula:

[0050]

[0051]

[0052]

[0053]

[0054]

[0055]

[0056] in, represents element-wise multiplication, Generated by the channel attention mechanism, it represents the CBF pathway features that have been enhanced in the channel dimension. Generated by the spatial attention mechanism, by paying attention to L r Different areas in the feature map are used to enhance the model's ability to extract MVDR pathway features. F obtained from the previous stage of HFFM r-1 The intermediate features generated by downsampling, F r-1 Indicates the fusion result of the hierarchical feature fusion module obtained in the previous stage; Conv1d represents the one-dimensional convolution layer, Avgpo0l is the average pooling layer, Conv2d represents the two-dimensional convolution layer, Concat represents the splicing operation, GELU represents the activation function, and FC represents the linear layer; is the intermediate fusion result, which is obtained by G r ,Lr , Calculated;

[0057] Will and Stitch together and pass Generate the fusion result F of the current stage r , It consists of the structure of an inverse residual multilayer perceptron.

[0058] Optionally, step 6 specifically includes:

[0059] Step 6-1: calculate all the sub-array DOA estimation results [y1, y2, ..., y N ] and several sub-array information [P1, P2, P3, .., P N ] Input the backbone network and use two consecutive backbone modules to encode it to obtain the low-level encoding feature y in , where the backbone module consists of residual and convolution operations;

[0060] At the same time, the array element parameters, including the number of array elements S F and spacing S D Input to the first fully connected layer of the model to obtain the input feature f FC , where the fully connected layer contains n FC neurons, using ReLU activation function;

[0061] Step 6-2, f FC Input fully connected layer 1:

[0062] Fully connected layer 1 to f FC Processing to obtain output feature f FC1 , where the fully connected layer 1 contains n FC1 neurons, using Sigmoid activation function;

[0063] f FC1 As the first convolution parameter to generate the weight coefficient of the block, for n FC1 The convolution kernels of the groups are respectively in n EXP1 Perform weighted summation on the output channels to obtain a set of n output channels EXP1 The convolution kernel of

[0064] In the conditional convolution layer 1, the convolution kernel is used to in Do convolution, the output is P CC1 ;

[0065] Step 6-3: Further, f FC Input fully connected layer 2:

[0066] Fully connected layer 2 pairs fFC Processing to obtain output feature f FC2 , where the fully connected layer 2 contains n FC2 neurons, using Sigmoid activation function;

[0067] f FC2 As the weight coefficient of the second convolution parameter generation block, for n FC2 The convolution kernels of the groups are respectively in n EXP2 Perform weighted summation on the output channels to obtain a set of n output channels EXP2 The convolution kernel of

[0068] In the conditional convolution layer 2, the convolution kernel is used to obtain P CC1 Do convolution, the output is P CC2 ;

[0069] Step 6-4, f FC Input fully connected layer 3:

[0070] Fully connected layer 3 pairs of f FC Processing to obtain output feature f FC3 , where the fully connected layer 3 contains n FC3 neurons, using Sigmoid activation function;

[0071] f FC3 As the third convolution parameter to generate the weight coefficient of the block, FC3 The convolution kernels of the groups are respectively in n EXP3 Perform weighted summation on the output channels to obtain a set of n output channels EXP3 The convolution kernel of

[0072] In the conditional convolution layer 3, the convolution kernel is used to obtain P CC2 Do convolution, the output is P CC3 ;

[0073] Step 6-5, f FC Input fully connected layer 4:

[0074] Fully connected layer 4 pairs of f FC Processing to obtain output feature f FC4 , where the fully connected layer 4 contains n FC4 neurons, using Sigmoid activation function;

[0075] f FC4 As the weight coefficient of the fourth convolution parameter generation block, for n FC4 The convolution kernels of the groups are respectively in n EXP4 Perform weighted summation on the output channels to obtain a set of n output channels EXP4 The convolution kernel of

[0076] In the conditional convolution layer 4, the convolution kernel is used to CC3 Do convolution, the output is P CC4 ;

[0077] Step 6-6, until f FC Input fully connected layer M:

[0078] The fully connected layer M is used to FCM Processing to obtain output feature f FCM , where FCM contains n FCM neurons, using Sigmoid activation function;

[0079] f FCM As the weight coefficient of the Mth convolution parameter generation block, for n FCM The convolution kernels of the groups are respectively in n EXPM Perform weighted summation on the output channels to obtain a set of n output channels EXPM The convolution kernel of

[0080] In the conditional convolution layer M, the convolution kernel P is obtained. CC(M-1) Do convolution, the output is P CCM ;

[0081] Step 6-7, P CCM Input into the classifier to obtain the DOA estimation result of the entire linear array The classifier consists of a fully connected layer and a Softmax function layer, with P CCM As input, output

[0082] A horizontal array intelligent target azimuth estimation system based on subarray division, comprising:

[0083] The signal direction division module divides the signal arrival direction into A sectors. The label is set to 1 if the target source exists in the sector, and 0 otherwise. The final label is represented as a one-dimensional matrix with a length of A dimensions and only one value of 1 and the other values ​​of 0.

[0084] The subarray acquisition module randomly reads a full line array data and divides the full line array P into several subarrays [P1, P2, P3, .., P N ];

[0085] The subarray data preprocessing module uses two beamforming algorithms, CBF and MVDR, to process several subarrays into two beam domain characteristics, single-frequency CBF output power P CBF (f, θ) and the single-frequency diagonally loaded MVDR output power PMv(f, θ), the two beam domain characteristics are normalized and modulo the values ​​to obtain and As the input of the sub-array DOA estimation model based on multi-stage fusion of multi-beam domain features;

[0086] Subarray DOA estimation module, and Input it into the pre-trained sub-array DOA estimation model based on multi-stage fusion multi-beam domain features to obtain the DOA estimation results of each sub-array of the full linear array

[0087] The sub-array estimation result output module performs DOA estimation on all sub-arrays and obtains the DOA estimation results of all sub-arrays [y1, y2, ..., y N ];

[0088] The DOA estimation module of the full array takes all the DOA estimation results of the sub-arrays [y1, y2, ..., y N ], several sub-array information [P1, P2, P3, .., P N ] and the array element parameters are input into the pre-trained conditional convolution DOA estimation model based on multi-source feature fusion to obtain the DOA estimation result of the entire linear array.

[0089] Through the above technical solution, it can be seen that compared with the existing technology, the present invention discloses a method and system for intelligent target direction estimation based on subarray division for a horizontal array. First, a multi-stage fusion method is used to achieve the fusion of multiple beamforming methods, thereby obtaining rich beam domain features and ensuring the accuracy of the DOA estimation of a single subarray. Then, the DOA estimation result of a single subarray, subarray information, and array element parameters are used as inputs to a conditional convolution DOA estimation model based on multi-source feature fusion. The conditional convolution in the model can parameterize its convolution kernel and dynamically generate convolution kernel parameters based on the input conditions using a linear combination of several expert modules. All convolution kernel parameters in the backbone network are adaptively adjusted, and the information is integrated to obtain the final DOA estimation result. The DOA estimation result obtained by the present invention not only integrates multiple beamforming methods, but also forms an adaptive subarray fusion based on subarray information and array element parameters, which can solve the problem of insufficient beam domain features generated by current single beamforming methods. In addition, by dividing the full linear array into multiple subarrays, the present method can effectively reduce multipath effects and noise interference, improve the estimation accuracy of target direction, and is suitable for target detection and positioning in complex environments. At the same time, the independence of the sub-arrays is utilized to achieve effective signal processing, thereby enhancing the robustness of the system in strong interference environments and ensuring the accuracy of target direction information. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0091] Figure 1 A schematic diagram of the target direction estimation method provided by the present invention;

[0092] Figure 2 Schematic diagram of data tag and target incident angle calculation provided by the present invention;

[0093] Figure 3 This is a structural diagram of the sub-array DOA estimation model based on multi-stage fusion of multi-beam domain features provided by the present invention;

[0094] Figure 4 Detailed diagrams of the encoder, convolution block, and identity block modules in the subarray DOA estimation model provided by the present invention, wherein (1) is the encoder detail diagram, (2) is the convolution block detail diagram, and (3) is the identity block detail diagram;

[0095] Figure 5 Detailed diagram of the HFFM module in the sub-array DOA estimation model provided by the present invention;

[0096] Figure 6 This is a structural diagram of the conditional convolution DOA estimation model based on multi-source feature fusion provided by the present invention;

[0097] Figure 7 Detailed diagram of the backbone module in the DOA estimation model. DETAILED DESCRIPTION

[0098] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0099] The embodiment of the present invention discloses a method for estimating the direction of a horizontal array intelligent target based on sub-array division, the schematic diagram of which is shown in FIG. Figure 1 As shown, the method includes:

[0100] Step 1: The present invention defines the DOA estimation problem of the sub-array and the full linear array as a classification problem. Figure 2As shown in the figure, the signal arrival direction is divided into A sectors. The label is 1 when the target source exists in the sector, otherwise it is 0. The final label is represented in the form of one-hot encoding, that is, a one-dimensional matrix with a length of A dimensions and only one value of 1 and other values ​​of 0.

[0101] Step 2: Randomly read a full line array data and divide the full line array P into several sub-arrays [P1, P2, P3, .., P N ], each sub-array can be overlapping or non-overlapping, which is set and adjusted according to the array division criteria and the actual situation of the array. For example, the division scheme can be: P1: [1, 2, 3, 4], P2: [5, 6, 7, 8], P3: [9, 10, 11, 12], P4: [13, 14, 15, 16];

[0102] Step 3: Sub-array data preprocessing: using two beamforming algorithms, CBF and MVDR, to process the sub-array data into two beam domain features P CBF (f,θ) and P MV (f,θ), the two beam domain features are normalized and modulo valued to obtain and As the input of the sub-array DOA estimation model based on multi-stage fusion of multi-beam domain features;

[0103] Step 4: and Input into the pre-trained sub-array DOA estimation model based on multi-stage fusion multi-beam domain features. The model structure is as follows: Figure 3 As shown, the DOA estimation results of each sub-array of the full linear array are obtained.

[0104] Step 5: Perform steps 3 to 4 on all sub-arrays in step 2 to obtain the DOA estimation results of all sub-arrays [y1, y2, ..., y N ];

[0105] In step 6) replace [y1, y2, ..., y N ]、Sub-array information [P1,P2,P3,..,P N ]、Array element parameters (including the number of array elements S F , spacing S D ) is input into the pre-trained conditional convolution DOA estimation model based on multi-source feature fusion. The model structure is as follows Figure 6 As shown, the DOA estimation result of the full linear array is obtained;

[0106] Furthermore, step 3 specifically includes:

[0107] The present invention models the signal received by a single full-line array, assuming that K signals of center frequency f are S(t)=[s1(t), s2(t), ..., S K (t)] T Since the wave direction θ=[θ1,θ2,…,θ K ] T The signal is incident on a single uniform linear array with N sensors (N>K) and an element spacing of d. The signal satisfies the far-field narrowband assumption and the signal envelope does not change while passing through the array.

[0108] Let X(t)=[x1(t), x2(t),...,x K (t)] T The signal received by each array element generally includes the sound source signal and noise signal after propagation through the channel. The array receiving signal model is expressed as

[0109] X(t)=B(f,θ)S(t)+N(t);

[0110] Among them, B(f, θ)=[a(f, θ1), a(f, θ2), .., a(f, θ K )] is the array manifold vector matrix, and the array manifold vector corresponding to the incoming wave direction θk is:

[0111] a(f,θ k )=[exp(-jk T p1), exp(-jk T p2), ..., exp(-jk T p N )] T , the array element position is [p1, p2, ..., p N ], uniform linear array [p1, p2, ..., p N ]=[0,d,...,(N-1)d],k is the wave number vector, k=2πffc·v(θ),v(θ) is the unit vector of the acoustic signal propagation direction, c is the signal propagation speed, N(t)=[n1(t),n2(t),...,n N (t)] T is the noise vector matrix received by each element of the linear array.

[0112] Furthermore, by scanning the possible existence range angle of DOA, the single-frequency CBF output power is obtained as

[0113]

[0114] in, is the weighted vector of CBF, obtained according to different weighting methods The values ​​are also different, for example, a uniform weighted approach can be used. R x is the cross-spectral density matrix of the received signal, R n With R x The unit matrix has the same size, and SNR is the signal-to-noise ratio.

[0115] By scanning the possible existence range angle of DOA, the output power of the single-frequency diagonally loaded MVDR is obtained as follows:

[0116]

[0117] Among them, w MV (f,θ) is the optimal weight vector of MVDR, which can be obtained using the Lagrange operator method. R x is the cross-spectral density matrix of the received signal, R n With R x The unit noise matrix of the same size, SNR is the signal-to-noise ratio, when no noise is added, R = R x +ρI, ρ is the diagonal loading, I is the size and R x Same identity matrix;

[0118] Furthermore, the beamforming output P CBF (f,θ) and P MV (f,θ) is normalized and modulo value is obtained and Input into the subarray DOA estimation model based on multi-stage fusion of multi-beam domain features. The input dimensions are all N f ×N θ , as the input of the model, where N f is the number of frequency points, N θ is the number of scanning angles during beamforming, and the matrix form of the two inputs is:

[0119]

[0120]

[0121] For each input matrix Corresponding to a label vector

[0122] Furthermore, step 4 specifically includes:

[0123] Step 4-1: the The input is fed into the subarray DOA estimation model based on multi-stage fusion of multi-beam domain features. The model first uses an encoder to encode the input features. The encoder structure is as follows:Figure 4 (1) As shown:

[0124]

[0125] Among them, e CBF , e MV Represents the low-level features obtained by encoding, ε represents the encoder, which consists of two-dimensional convolution, one-dimensional convolution, normalization layer, ReLU activation function, and maximum pooling;

[0126] Step 4-2: Further, two feature extraction pathways are used to extract CBF beam domain features and MVDR beam domain features respectively. The two feature extraction pathways are composed of R cascaded downsampling modules. The downsampling module is composed of a convolution block and an identity block. Its structure is as follows: Figure 4 (2) and Figure 4 (3) can be calculated by the following formula:

[0127]

[0128]

[0129] Conv Block(x)=ReLU(BatchNorm(Conv2d

[0130] (ReLU(BatchNorm(Conv2d(ReLU(BatchNorm(Conv2d(x))))))))+

[0131] BatchNorm(Conv2d(x)));

[0132] Identity Block=BatchNorm(Conv2d(ReLU(BatchNorm

[0133] (Conv2d(ReLU(BatchNorm(Conv2d(x))))))))+x;

[0134] in, represents the rth downsampling module for processing CBF features, r∈(1,R), Represents the parameters of the rth downsampling module;

[0135] Step 4-3: For each stage of G r and L r ,like Figure 5 As shown in Figure 2, the Hierarchical Feature Fusion Module (HFFM) is used to fuse the CBF features and MVDR features of different scales obtained at different stages.r-1 Indicates the fusion result of the HFFM module obtained in the previous stage, input G r , L r and F r-1 , can be calculated by the following formula:

[0136]

[0137]

[0138]

[0139]

[0140]

[0141]

[0142] in, represents element-wise multiplication, Generated by the channel attention mechanism, it represents the CBF pathway features that have been enhanced in the channel dimension. Generated by the spatial attention mechanism, by paying attention to L r Different areas in the feature map are used to enhance the model's ability to extract MVDR pathway features. F obtained from the previous stage of HFFM r-1 Intermediate features generated by downsampling;

[0143] Conv1d represents a one-dimensional convolution layer, Avgp00l is an average pooling layer, Conv2d represents a two-dimensional convolution layer, Concat represents a concatenation operation, GELU represents an activation function, and FC represents a linear layer;

[0144] is the intermediate fusion result, which is obtained by G r , L r , Calculated;

[0145] Will and Stitch together and pass Generate the fusion result F of the current stage r ,

[0146] It consists of the structure of inverse residual multilayer perceptron;

[0147] Step 4-4: Further, the feature F fused in the R stage R Input into the classifier to obtain the sub-array DOA estimation result

[0148] Further, step 6 specifically comprises:

[0149] Step 6-1, input the subarray DOA estimation result calculated in step 5, [y1, y2,..., y N ] and the subarray information [P1, P2, P3,.., P N ] into the backbone network, and encode the same by using two consecutive backbone modules to obtain low-level coding features y in , wherein the backbone module is composed of residual and convolution operations;

[0150] At the same time, input the array element parameters (including the number of array elements S F , the spacing S D ) into the first layer full connection layer of the model to obtain input features f FC , wherein the full connection layer contains n FC neurons and adopts ReLU activation function;

[0151] Step 6-2, further, input f FC into the full connection layer 1:

[0152] Step 6-2-1, the full connection layer 1 processes f FC to obtain output features f FC1 , wherein the full connection layer 1 contains n FC1 neurons and adopts Sigmoid activation function;

[0153] Step 6-2-2, take f FC1 as the weight coefficient of the first convolution parameter generation block, and perform weighted summation on n FC1 groups of convolution kernels in n EXP1 output channels to obtain a group of convolution kernels with nE XP1 output channels;

[0154] Step 6-2-3, in the conditional convolution layer 1, perform convolution on P in (input subarray information) by using the obtained convolution kernel, and output P CC1 ;

[0155] Step 6-3, further, input f FC into the full connection layer 2:

[0156] Step 6-3-1, the full connection layer 2 processes f FC to obtain output features f FC2 , wherein the full connection layer 2 contains n FC2 neurons and adopts Sigmoid activation function;

[0157] Step 6-3-2, take f FC2As the weight coefficient of the second convolution parameter generation block, for n FC2 The convolution kernels of the groups are respectively in n EXP2 Perform weighted summation on the output channels to obtain a set of n output channels EXP2 The convolution kernel of

[0158] Step 6-3-3, use the convolution kernel obtained in the conditional convolution layer 2 to check P CC1 Do convolution, the output is P CC2 ;

[0159] Step 6-4: Further, f FC Input fully connected layer 3:

[0160] Step 6-4-1, fully connected layer 3 pairs f FC Processing to obtain output feature f FC3 , where the fully connected layer 3 contains n FC3 neurons, using Sigmoid activation function;

[0161] Step 6-4-2, f FC3 As the third convolution parameter to generate the weight coefficient of the block, FC3 The convolution kernels of the groups are respectively in n EXP3 Perform weighted summation on the output channels to obtain a set of n output channels EXP3 The convolution kernel of

[0162] Step 6-4-3, use the convolution kernel obtained in the conditional convolution layer 3 to check P CC2 Do convolution, the output is P CC3 ;

[0163] Step 6-5: Further, f FC Input fully connected layer 4:

[0164] Step 6-5-1, fully connected layer 4 pairs f FC Processing to obtain output feature f FC4 , where the fully connected layer 4 contains n FC4 neurons, using Sigmoid activation function;

[0165] Step 6-5-2, f FC4 As the weight coefficient of the fourth convolution parameter generation block, for n FC4 The convolution kernels of the groups are respectively in n EXP4 Perform weighted summation on the output channels to obtain a set of n output channels EXP4 The convolution kernel of

[0166] Step 6-5-3, use the obtained convolution kernel to check P in the conditional convolution layer 4 CC3 Do convolution, the output is P CC4;

[0167] Step 6-6, until f FC Input fully connected layer M:

[0168] Step 6-6-1, fully connected layer M to f FCM Processing to obtain output feature f FCM , where FCM contains n FCM neurons, using Sigmoid activation function;

[0169] Step 6-6-2, f FCM As the weight coefficient of the Mth convolution parameter generation block, for n FCM The convolution kernels of the groups are respectively in n EXPM Perform weighted summation on the output channels to obtain a set of n output channels EXPM The convolution kernel of

[0170] Step 6-6-3, use the obtained convolution kernel to check P in the conditional convolution layer M CC(M-1) Do convolution, the output is P CCM ;

[0171] Step 6-7, P CCM Input into the classifier to obtain the DOA estimation result of the entire linear array The classifier consists of a fully connected layer and a Softmax function layer, with P CCM As input, output The details of the backbone module in the DOA estimation model are shown in the figure below. Figure 7 shown.

[0172] A horizontal array intelligent target azimuth estimation system based on subarray division, comprising:

[0173] The signal direction division module divides the signal arrival direction into A sectors. The label is set to 1 if the target source exists in the sector, and 0 otherwise. The final label is represented as a one-dimensional matrix with a length of A dimensions and only one value of 1 and the other values ​​of 0.

[0174] The subarray acquisition module randomly reads a full line array data and divides the full line array P into several subarrays [P1, P2, P3, .., P N ];

[0175] The subarray data preprocessing module uses two beamforming algorithms, CBF and MVDR, to process several subarrays into two beam domain characteristics, single-frequency CBF output power P CBF (f, θ) and the single-frequency diagonally loaded MVDR output power PMv(f, θ), the two beam domain characteristics are normalized and modulo the values ​​to obtain and As the input of the sub-array DOA estimation model based on multi-stage fusion of multi-beam domain features;

[0176] Subarray DOA estimation module, and Input it into the pre-trained sub-array DOA estimation model based on multi-stage fusion multi-beam domain features to obtain the DOA estimation results of each sub-array of the full linear array

[0177] The sub-array estimation result output module performs DOA estimation on all sub-arrays and obtains the DOA estimation results of all sub-arrays [y1,y2,...,y N ];

[0178] The DOA estimation module of the full array takes all the DOA estimation results of the sub-arrays [y1,y2,...,y N ], several sub-array information [P1, P2, P3, .., P N ] and the array element parameters are input into the pre-trained conditional convolution DOA estimation model based on multi-source feature fusion to obtain the DOA estimation result of the entire linear array.

[0179] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0180] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for estimating the direction of a horizontal array intelligent target based on sub-array division, characterized in that: include: Step 1: Divide the signal arrival direction into Sector angle, set the label as 1 when the target source exists in the sector angle, otherwise it is 0, and the final label is expressed as length A one-dimensional matrix with only one value of 1 and other values ​​of 0; Step 2: Randomly read a full line array data and convert the full line array Divide into several sub-arrays ; Step 3: Sub-array data preprocessing, using two beamforming algorithms CBF and MVDR to process several sub-arrays into two beam domain features: single frequency CBF output power and single-frequency diagonally loaded MVDR output power , normalize the two beam domain features and take the modulus value to obtain and As the input of the sub-array DOA estimation model based on multi-stage fusion of multi-beam domain features; Step 4: and Input it into the pre-trained sub-array DOA estimation model based on multi-stage fusion multi-beam domain features to obtain the DOA estimation results of each sub-array of the full linear array ; Step 5: Perform steps 3 to 4 on all sub-arrays in step 2 to obtain the DOA estimation results of all sub-arrays. ; Step 6: DOA estimation results of all sub-arrays , several sub-array information , the array element parameters are input into the pre-trained conditional convolution DOA estimation model based on multi-source feature fusion to obtain the DOA estimation result of the entire linear array; The step 3 specifically includes: obtaining the single-frequency CBF output power by scanning the existence range angles of several sub-array DOAs: ； in, f is the center frequency, is the incoming wave direction, N is the number of sensors on the linear array, is the array manifold vector, is the weight vector of CBF, , , is the cross-spectral density matrix of the received signal, and The unit noise matrix of the same size, is the signal-to-noise ratio; By scanning the existence range angles of several sub-array DOAs, the single-frequency diagonally loaded MVDR output power is obtained as follows: ； in, is the optimal weight vector of MVDR, , , is the cross-spectral density matrix of the received signal, and The unit noise matrix of the same size, is the signal-to-noise ratio. When no noise is added, , is the diagonal loading, Is the size and Same identity matrix; The beamformed output and Normalize and modulo the value to get and Input into the subarray DOA estimation model based on multi-stage fusion multi-beam domain features; the input dimensions are , which is used as the input of the model, is the number of frequency points, is the number of scanning angles during beamforming, and the matrix form of the two inputs is ； ； For each input matrix Corresponding to a label vector .

2. The method for horizontal array intelligent target direction estimation based on subarray division according to claim 1, characterized in that: Before step 3 is processed, it also includes: Modeling of a single full-line array receiving signal, Center frequency signal Incoming wave direction The incident light is incident on a single uniform linear array, and the number of sensors on the linear array is , the array element spacing is , the signal satisfies the far-field narrowband assumption, and the signal envelope does not change while passing through the array. For each array element receiving signal, the array receiving signal model is expressed as: ； in, is the array manifold vector matrix, the incoming wave direction The corresponding array manifold vector is , the array element position is , uniform linear array , is the wave number vector, , is the unit vector of the acoustic signal propagation direction, is the signal propagation speed, is the noise vector matrix received by each element of the linear array.

3. The method for horizontal array intelligent target direction estimation based on subarray division according to claim 1, characterized in that: The step 4 specifically includes: Step 4-1, get The input is fed into a sub-array DOA estimation model based on multi-stage fusion of multi-beam domain features, and an encoder is used to encode the input features; Step 4-2: Use two feature extraction paths to extract CBF beam domain features separately and MVDR beam domain characteristics , the two feature extraction pathways are composed of The downsampling module is composed of a cascaded downsampling module, which consists of a convolution block and an identity block: Step 4-3: CBF beam domain features obtained at each stage and MVDR beam domain characteristics , the hierarchical feature fusion module is used to fuse the CBF beam domain features and MVDR beam domain features of different scales obtained at different stages to obtain the first Features after stage fusion ; Step 4-4, Features after stage fusion Input into the classifier to obtain the sub-array DOA estimation result .

4. The method for horizontal array intelligent target direction estimation based on subarray division according to claim 3 is characterized in that: The encoder is specifically: ； in Represents the low-level features obtained by encoding, Represents the splicing operation, represents the encoder, which consists of two-dimensional convolution , one-dimensional convolution , Standardization layer , ReLU activation function, maximum pooling composition.

5. The method for horizontal array intelligent target direction estimation based on subarray division according to claim 4 is characterized in that: The downsampling module consists of a convolution block and an identity block. is the convolution block, It is an identity block with the following structure: ； ； in, Indicates the first A downsampling module, , Indicates the The parameters of the downsampling module, The specific extraction of CBF beam domain features and MVDR beam domain features is as follows: ； 。 6. The method for horizontal array intelligent target direction estimation based on subarray division according to claim 3 is characterized in that: The specific calculation process of the hierarchical feature fusion module is as follows: enter 、 and , calculated using the following formula: ； ； ； ； ； ； in, represents element-wise multiplication, Generated by the channel attention mechanism, it represents the CBF pathway features that have been enhanced in the channel dimension. Generated by the spatial attention mechanism, by paying attention to Different areas in the feature map are used to enhance the model's ability to extract MVDR pathway features. Obtained by the previous stage of the hierarchical feature fusion module The intermediate features generated by downsampling, Represents the fusion result of the hierarchical feature fusion module obtained in the previous stage; represents a one-dimensional convolutional layer, is the average pooling layer, represents a two-dimensional convolutional layer, Represents the splicing operation, represents the activation function, represents the linear layer; is the intermediate fusion result, Calculated; Will 、 and Stitch together and pass Generate the fusion result of the current stage , It consists of the structure of an inverse residual multilayer perceptron.

7. The method for horizontal array intelligent target direction estimation based on subarray division according to claim 1 is characterized in that: The step 6 specifically includes: Step 6-1: The DOA estimation results of all sub-arrays are calculated and several sub-array information Input the backbone network and use two consecutive backbone modules to encode it to obtain low-level encoding features , where the backbone module consists of residual and convolution operations; At the same time, the array element parameters, including the number of array elements and spacing Input to the first fully connected layer of the model to obtain input features , where the fully connected layer contains neurons, using ReLU activation function; Step 6-2, Input fully connected layer 1: Fully connected layer 1 pair Processing to obtain output features , where the fully connected layer 1 contains neurons, using Sigmoid activation function; Will As the first convolution parameter to generate the weight coefficient of the block, The convolution kernels of the groups are respectively Perform weighted summation on the output channels to obtain a set of output channels with the number of The convolution kernel of In the conditional convolution layer 1, the convolution kernel is obtained Do convolution and the output is ; Step 6-3: Further, Input fully connected layer 2: Fully connected layer 2 pairs Processing to obtain output features , where the fully connected layer 2 contains neurons, using Sigmoid activation function; Will As the weight coefficient of the second convolution parameter generation block, The convolution kernels of the groups are respectively Perform weighted summation on the output channels to obtain a set of output channels with the number of The convolution kernel of In the conditional convolution layer 2, the convolution kernel is used to Do convolution and the output is ; Step 6-4, Input fully connected layer 3: Fully connected layer 3 pairs Processing to obtain output features , where the fully connected layer 3 contains neurons, using Sigmoid activation function; Will As the weight coefficient of the third convolution parameter generation block, The convolution kernels of the groups are respectively Perform weighted summation on the output channels to obtain a set of output channels with the number of The convolution kernel of In the conditional convolution layer 3, the convolution kernel is used to Do convolution and the output is ; Step 6-5, Input fully connected layer 4: 4 pairs of fully connected layers Processing to obtain output features , where the fully connected layer 4 contains neurons, using Sigmoid activation function; Will As the weight coefficient of the fourth convolution parameter generation block, The convolution kernels of the groups are respectively Perform weighted summation on the output channels to obtain a set of output channels with the number of The convolution kernel of In the conditional convolution layer 4, the convolution kernel is used to Do convolution and the output is ; Step 6-6, until Input fully connected layer M: Fully connected layer M pairs Processing to obtain output features , among which FCM contains neurons, using Sigmoid activation function; Will As the weight coefficient of the Mth convolution parameter generation block, The convolution kernels of the groups are respectively Perform weighted summation on the output channels to obtain a set of output channels with the number of The convolution kernel of In the conditional convolution layer M, the convolution kernel is used to Do convolution and the output is ; Step 6-7, Input into the classifier to obtain the DOA estimation result of the entire linear array , the classifier consists of a fully connected layer and a Softmax function. As input, output .

8. A horizontal array intelligent target direction estimation system based on sub-array division, characterized in that: The method for estimating the horizontal array intelligent target direction based on subarray division according to any one of claims 1 to 7 is applied, comprising: Signal direction division module divides the signal arrival direction into Sector angle, set the label as 1 when the target source exists in the sector angle, otherwise it is 0, and the final label is expressed as length A one-dimensional matrix with only one value of 1 and other values ​​of 0; The sub-array acquisition module randomly reads a full-line array data and Divide into several sub-arrays ; The subarray data preprocessing module uses two beamforming algorithms, CBF and MVDR, to process several subarrays into two beam domain features: single-frequency CBF output power and single-frequency diagonally loaded MVDR output power , normalize the two beam domain features and take the modulus value to obtain and As the input of the sub-array DOA estimation model based on multi-stage fusion of multi-beam domain features; Subarray DOA estimation module, and Input it into the pre-trained sub-array DOA estimation model based on multi-stage fusion multi-beam domain features to obtain the DOA estimation results of each sub-array of the full linear array ; The sub-array estimation result output module performs DOA estimation on all sub-arrays and obtains the DOA estimation results of all sub-arrays. ; The DOA estimation module of the full array will estimate the DOA of all sub-arrays. , several sub-array information The array element parameters are input into the pre-trained conditional convolution DOA estimation model based on multi-source feature fusion to obtain the DOA estimation result of the entire linear array.

Images