Frequency control array radar target positioning method based on deep learning

Through deep learning methods, feature extraction and nonlinear mapping of FDA radar echo signals is solved, and the problem of decoupling of azimuth angle and distance information in FDA radar is achieved, and high-precision targets with low signal-to-noise ratio and few snapshots are achieved position.

CN119001648APending Publication Date: 2024-11-22GUILIN UNIV OF ELECTRONIC TECH +1
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411095378.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-09
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

The prior art is difficult to effectively decouple azimuth and distance information from FDA radar echo data, resulting in blurred target positioning, especially in the case of low signal-to-noise ratio and few snapshot counts.

Method used

Using the frequency controlled array radar target positioning method based on deep learning, the radar echo signal covariance matrix is ​​characterized by using convolutional layer, maximum pooling layer and full connection layer to perform feature extraction and nonlinear mapping of radar echo signal covariance matrix. Achieve decoupled positioning of azimuth angle and distance.

Benefits of technology

Under the conditions of low signal-to-noise ratio and few snaps, the target azimuth angle and distance information are efficiently extracted to achieve accurate positioning of far-field targets, significantly improving positioning accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119001648A_ABST
    Figure CN119001648A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of array signal processing, in particular to a frequency diverse array (FDA) radar target positioning method based on deep learning. According to the method, a convolutional neural network (CNN) structure is adopted, the characteristics of beam characteristics of FDA and CNN weight sharing, local connection and pooling dimension reduction are utilized, a network model of which the input data size is matched with the network structure complexity is established, a target positioning problem is converted into a nonlinear mapping problem from an echo signal covariance matrix to a target azimuth angle and distance information, and the target positioning accuracy is improved. The invention provides an FDA-CNN positioning algorithm. According to the method, the azimuth angle distance decoupling problem is solved in the FDA-CNN network model, the overfitting risk is effectively reduced, azimuth angle distance decoupling is efficiently completed, azimuth angle and distance information is directly obtained, and meanwhile, the target positioning precision is also improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of array signal processing, and in particular to a frequency-controlled array radar target positioning method based on deep learning. Background Art

[0002] As an electronic detection device, radar technology has developed rapidly and is widely used in many fields, including military, aviation, maritime, meteorological observation, traffic monitoring, earth science research, and disaster prevention. For example, in the military field, radar is used for reconnaissance, surveillance, and target positioning and tracking; in the civilian field, it can help aircraft and ships avoid collisions, and is also used to detect storms and other weather phenomena in weather forecasts. Especially today, when the cause of strengthening the military is booming, the research on radar target positioning technology is in full swing, and the radar echo signal is processed in different ways and means to obtain the specific location of the target and achieve accurate positioning of the target.

[0003] In the early stages of radar technology, the radar system's interference suppression capability was insufficient. When electromagnetic interference existed in the working environment, the radar's performance would drop significantly. With the advancement of radar technology, especially the development of phased array radar technology, the performance has been significantly improved. Phased array radar achieves rapid scanning and shape change of antenna beams by adjusting the signal phase and amplitude of each unit of the array antenna. With its flexible beam adjustment, powerful interference blocking and high resolution, phased array radar provides technical advantages for expanding radar coverage, improving measurement accuracy and detecting low-altitude visible targets including stealth targets. It has been widely used in target positioning for a long time. However, the application of phased array in radar target positioning has certain limitations: its beam pattern corresponds to different distances at the same azimuth angle, and it is difficult to directly obtain the distance information of the target from the change of beam direction, resulting in distance ambiguity, and it is not possible to achieve the distance measurement and angle measurement of the target at the same time. Specifically, radar target positioning is to estimate the azimuth and distance information of the target's spatial position by processing the radar echo signal.

[0004] In the traditional phased array architecture, it is quite difficult to solve the above problems. FDA radar came into being. The small frequency offset between array elements can produce a beam pattern with azimuth distance dependence that is different from the phased array. Such a coupled beam brings opportunities for target positioning, so FDA radar is widely used in target positioning. FDA radar effectively avoids the large bandwidth and nonlinear problems associated with broadband signals by making each array element emit a single-frequency continuous wave of different frequencies. In addition, only a single linear array is needed to simultaneously achieve the azimuth and distance positioning of the target, effectively simplifying the complexity of the array structure. However, it is precisely because the azimuth distance of the FDA beam pattern is coupled that the azimuth and distance information of the target cannot be directly extracted from the radar echo data by traditional methods. Therefore, exploring how to decouple the azimuth distance from the noisy echo data, extract the target azimuth and distance information, and thus complete the positioning of the target has become a key issue that needs to be solved in the field of array signal processing.

[0005] At present, with the rapid progress of artificial intelligence, especially the widespread application of deep learning methods, outstanding results have been achieved in many fields. The integration and intersection of disciplines have now become the main development direction of scientific research in the future. Deep learning can learn the important information of input data to form a function mapping. Compared with traditional methods, it can learn the effective information of radar echo signals and then directly map them to target-related parameters. It does not require the consistency of input and output data, has excellent feature extraction and generalization capabilities, and can be generalized to various unknown scenarios. Deep learning methods provide an innovative solution to overcome the core problems in radar signal processing.

[0006] In summary, how to combine deep learning methods with FDA radar to achieve high-precision target positioning has become an important topic with great application value. Summary of the invention

[0007] In order to solve the above problems, the present invention provides a frequency-controlled array radar target positioning method based on deep learning. The specific technical solution is as follows:

[0008] A frequency-controlled array radar target positioning method based on deep learning includes the following steps:

[0009] Step S1: Construct a data set. Build a single-base FDA radar transceiver array signal model, set various parameters of the signal model, obtain the FDA radar echo signal matrix, calculate the FDA radar echo matrix covariance, obtain the covariance matrix, separate the real from the virtual and normalize the covariance matrix, and obtain the network input data;

[0010] Step S2: Build the FDA-CNN network model. The model structure consists of two convolutional layers, two maximum pooling layers, two fully connected layers and one Softmax layer. The first layer is the convolutional layer (Conv.1); followed by a maximum pooling layer (Maxpool); then, the second convolutional layer (Conv.2); followed by the same maximum pooling layer; and finally, two fully connected layers FC1 and FC2;

[0011] Step S3: training model and outputting results. The built FDA-CNN network model is trained using the training set of input data. After the model training is completed, the model is tested using the test set of input data. Finally, the azimuth and distance of the far-field target are output to complete the estimation of the target position.

[0012] Preferably, the step S1 specifically includes the following steps:

[0013] In step S11, the selected FDA is a uniform linear array containing 11 array elements, each array element is arranged along a straight line, and the distance between every two adjacent array elements is λ / 2.

[0014] Step S12, consider that there are 24 far-field narrowband incoherent target signal sources with an initial carrier frequency of 10GHZ and a frequency offset of 5kHz incident on the array. Assume that the signal-to-noise ratio is a constant 10dB, the number of snapshots is a constant 100, and the noise is independent and obeys a zero-mean Gaussian distribution of white noise. The target azimuth angle range is [5°, 40°], the change interval is 5°, and there are 8 angles in total; the target distance range is [4km, 12km], the change interval is 4km, and there are 3 distances in total. The combination of 8 angles and 3 distances results in 24 target positions.

[0015] R=E[Y(t)Y H (t)] (1-1)

[0016] R′=[R1′,R2′,R3′,...,R′ N ] (1-2)

[0017] Among them, Y(t)=[y1(t),y2(t),···,y N (t)] T is the received signal vector, H represents the Hermitian operator, E[·] represents the mathematical expectation, R is the echo signal covariance matrix, and R′ represents the data obtained after the covariance matrix R is preprocessed by performing operations such as virtual-real separation and normalization.

[0018] Step S13, substitute the above 24 targets into formula (1-1) and formula (1-2) in turn for calculation to obtain 24 sets of data. It should be noted that the real part and the imaginary part of the signal are separated in the process of constructing the data, so the number of experimental samples is twice the size of the beam space. Repeat the same method 100 times, and obtain 2424 sets of data as the input sample data of the network. Select the training set and the validation set at a ratio of 100:1, and select the results of the first 100 calculations from the 24 targets, that is, 2400 groups, a total of 580,800 sample data, as the input data for network model training. Then select the results of the last calculation from the 24 targets, that is, 24 groups, a total of 5808 sample data, as the input data for verifying the performance of the network model. The data used in network training are all equipped with the corresponding target true azimuth and distance values ​​as training labels. The verification data is predicted by the network model to obtain the target azimuth and distance estimates. The actual azimuth and distance of the 24 targets are the test labels of the network;

[0019] Preferably, the FDA-CNN structure and algorithm flow of step S2 are as follows:

[0020] The FDA-CNN structure consists of two convolutional layers, two maximum pooling layers, two fully connected layers and one Softmax layer.

[0021] In step S21, the input data enters the first convolutional layer (Conv.1), which uses a 3×3 convolution kernel with a stride of 1 and contains 80 convolution kernels in total. The convolution kernel is used to slide the input data to capture and extract useful features. Subsequently, batch normalization is performed to standardize the output feature map activation value. Finally, nonlinearity is introduced into the network by applying an activation function, so that the network can capture more complex features.

[0022] Step S22, the data then enters a maximum pooling layer (Maxpool), which uses a pooling kernel of size 2×2 (pool size), and the stride is set to 2. Combined with the characteristics of radar echo data, the maximum pooling operation mode is adopted to perform specific processing on the input feature map, thereby achieving data dimensionality reduction, further reducing the number of parameters that the network needs to learn during the training process, and effectively preventing the network from overfitting;

[0023] Step S23, after that, the data enters the second convolution layer (Conv.2), which also uses a 3×3 convolution kernel, the stride is also set to 1, and also contains 80 convolution kernels; continue to capture and extract the most useful features;

[0024] Step S24, the data then passes through the same maximum pooling layer again;

[0025] Step S25, finally, after passing through two fully connected layers FC1 and FC2, it is passed to the Softmax layer output. First, the output of the pooling layer is integerized, and the output multidimensional data is flattened to convert it into a one-dimensional array. Then, it is input into the first fully connected layer. This layer consists of 256 neurons, and the feature vector is mapped to a high-dimensional space through linear transformation to facilitate the learning of complex relationships. In order to maintain the nonlinearity of the model and the ability to learn complex functions, batch normalization and ReLU activation are performed immediately. The ReLU-activated feature vector is passed to the second fully connected layer, which maps the output of the hidden layer to an 8-dimensional space, with each dimension corresponding to a specific output category. After the fully connected layer, the Softmax function is used to ensure that the sum of all output values ​​is maintained at 1, and each element value is between 0 and 1. At the output of the model, the output of the Softmax layer is multiplied by the node and summed along a specific dimension to obtain the final azimuth and distance prediction values.

[0026] Preferably, the training of the network model in step S3 specifically includes the following steps:

[0027] Step S31, input the sample data into the built network model to extract features and train learning, the batch size batch_size is set to 64, the number of network iterations is 200, the learning rate is 0.0001, MSE is selected as the loss function, and the optimizer is Adam. In the forward propagation stage, the model will calculate the output value according to the input data according to the FDA-CNN algorithm process;

[0028] Step S32, in the error back propagation stage, by comparing the difference between the output value and the true value, the error is calculated and back propagated layer by layer to update the network parameters, thereby optimizing the model. Through repeated iterations, the model gradually learns and improves the accuracy of target positioning. Once the training error approaches the established threshold or reaches the preset number of iterations, the training is considered complete.

[0029] The beneficial effects of the present invention are as follows: the frequency-controlled array (FDA) radar beam is favored in target positioning because of its azimuth-distance dependence. When the FDA radar accurately locates far-field targets, two aspects of the problem are mainly considered: one is how to efficiently complete the coupling of azimuth distance; the other is how to accurately extract the azimuth and distance information of the target under low signal-to-noise ratio and few snapshots. The existing methods are difficult to extract the azimuth and distance information of the target from the FDA radar echo data under low signal-to-noise ratio and few snapshots. In view of the above two problems, the present invention mainly utilizes the beam characteristics of the FDA radar, combines the deep learning method, studies and explores the target positioning method, transforms the target positioning problem into a nonlinear mapping problem from the echo signal covariance matrix to the target azimuth and distance information, and proposes a FDA radar target positioning method based on deep learning. The experimental results show that the method proposed by the present invention can extract features from the echo data under low signal-to-noise and few snapshots, efficiently complete the decoupling of azimuth distance, directly obtain azimuth and distance information, and realize the accurate positioning of far-field targets. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the specific embodiments or the description of the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual scale.

[0031] Figure 1 It is a uniform linear FDA structure;

[0032] Figure 2 It is the FDA signal receiving and sending model;

[0033] Figure 3 It is the FDA-CNN network model;

[0034] Figure 4 Comparison of target azimuth and distance estimation results;

[0035] Figure 5 The influence of different signal-to-noise ratios on the accuracy of the algorithm in estimating the target azimuth angle;

[0036] Figure 6 The influence of different signal-to-noise ratios on the accuracy of the algorithm in estimating target distance;

[0037] Figure 7 The influence of different snapshot numbers on the accuracy of the algorithm in estimating the target azimuth;

[0038] Figure 8 The impact of different snapshot numbers on the accuracy of the algorithm in estimating target distance; DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are 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 creative work are within the scope of protection of the present invention.

[0040] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0041] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0042] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0043] like Figure 1 , Figure 2 As shown, a specific embodiment of the present invention provides a frequency-controlled array radar target positioning method based on deep learning, comprising the following steps:

[0044] Step S1: Build a single-base FDA radar transceiver array signal model, set various parameters of the signal model, obtain the FDA radar echo signal matrix, calculate the FDA radar echo matrix covariance, obtain the covariance matrix, separate the real from the virtual and normalize the covariance matrix, and obtain the network input data;

[0045] The specific steps include:

[0046] In step S11, the selected FDA is a uniform linear array containing 11 array elements, each array element is arranged along a straight line, and the distance between every two adjacent array elements is λ / 2.

[0047] Step S12, consider that there are 24 far-field narrowband incoherent target signal sources with an initial carrier frequency of 10 GHZ and a frequency offset of 5 kHz incident on Figure 2In the array shown. Assume that the signal-to-noise ratio is constant at 10dB, the number of snapshots is constant at 100, and the noise is independent and white noise with zero-mean Gaussian distribution. The target azimuth range is [5°, 40°], the change interval is 5°, and there are 8 angles in total; the target distance range is [4km, 12km], the change interval is 4km, and there are 3 distances in total. The combination of 8 angles and 3 distances results in 24 target positions.

[0048] R=E[Y(t)Y H (t)] (1-1)

[0049] R′=[R1′,R2′,R3′,...,R′ N ] (1-2)

[0050] Among them, Y(t)=[y1(t),y2(t),···,y N (t)] T is the received signal vector, H represents the Hermitian operator, E[·] represents the mathematical expectation, R is the echo signal covariance matrix, and R′ represents the data obtained after the covariance matrix R is preprocessed by performing operations such as virtual-real separation and normalization.

[0051] Step S13, substitute the above 24 targets into formula (1-1) and formula (1-2) in turn for calculation to obtain 24 sets of data. It should be noted that the real part and the imaginary part of the signal are separated in the process of constructing the data, so the number of experimental samples is twice the size of the beam space. Repeat the same method 100 times, and obtain 2424 sets of data as the input sample data of the network. Select the training set and the validation set at a ratio of 100:1, and select the results of the first 100 calculations from the 24 targets, that is, 2400 groups, a total of 580,800 sample data, as the input data for network model training. Then select the results of the last calculation from the 24 targets, that is, 24 groups, a total of 5808 sample data, as the input data for verifying the performance of the network model. The data used in network training are all equipped with the corresponding target true azimuth and distance values ​​as training labels. The verification data is predicted by the network model to obtain the target azimuth and distance estimates. The actual azimuth and distance of the 24 targets are the test labels of the network;

[0052] Step S2: Figure 3 The figure shows the completed FDA-CNN network model. The model structure consists of two convolutional layers, two maximum pooling layers, two fully connected layers and one Softmax layer. The first layer is the convolutional layer (Conv.1); followed by a maximum pooling layer (Maxpool); then, the second convolutional layer (Conv.2); followed by the same maximum pooling layer; and finally, two fully connected layers FC1 and FC2.

[0053] The specific steps include:

[0054] The FDA-CNN structure consists of two convolutional layers, two maximum pooling layers, two fully connected layers and one Softmax layer.

[0055] In step S21, the input data enters the first convolutional layer (Conv.1), which uses a 3×3 convolution kernel with a stride of 1 and contains 80 convolution kernels in total. The convolution kernel is used to slide the input data to capture and extract useful features. Subsequently, batch normalization is performed to standardize the output feature map activation value. Finally, nonlinearity is introduced into the network by applying an activation function, so that the network can capture more complex features.

[0056] Step S22, the data then enters a maximum pooling layer (Maxpool), which uses a pooling kernel of size 2×2 (pool size), and the stride is set to 2. Combined with the characteristics of radar echo data, the maximum pooling operation mode is adopted to perform specific processing on the input feature map, thereby achieving data dimensionality reduction, further reducing the number of parameters that the network needs to learn during the training process, and effectively preventing the network from overfitting;

[0057] Step S23, after that, the data enters the second convolution layer (Conv.2), which also uses a 3×3 convolution kernel, the stride is also set to 1, and also contains 80 convolution kernels; continue to capture and extract the most useful features;

[0058] Step S24, the data then passes through the same maximum pooling layer again;

[0059] Step S25, finally, after passing through two fully connected layers FC1 and FC2, it is passed to the Softmax layer output. First, the output of the pooling layer is integerized, and the output multidimensional data is flattened to convert it into a one-dimensional array. Then, it is input into the first fully connected layer. This layer consists of 256 neurons, and the feature vector is mapped to a high-dimensional space through linear transformation to facilitate the learning of complex relationships. In order to maintain the nonlinearity of the model and the ability to learn complex functions, batch normalization and ReLU activation are performed immediately. The ReLU-activated feature vector is passed to the second fully connected layer, which maps the output of the hidden layer to an 8-dimensional space, with each dimension corresponding to a specific output category. After the fully connected layer, the Softmax function is used to ensure that the sum of all output values ​​is maintained at 1, and each element value is between 0 and 1. At the output of the model, the output of the Softmax layer is multiplied by the node and summed along a specific dimension to obtain the final azimuth and distance prediction values.

[0060] Step S3: train the model and output the results. The training learning parameters are set, the batch size batch_size is set to 64, the number of network iterations is 200, the learning rate is 0.0001, MSE is selected as the loss function, and the optimizer is Adam. The built FDA-CNN network model is trained with the training set of the input data. After the model training is completed, the model is tested with the test set of the input data. Finally, the far-field target azimuth and distance are output to complete the target position estimation. The target azimuth and distance estimation results are obtained, as well as the impact of different signal-to-noise ratios and snapshot numbers on the accuracy of the algorithm in estimating the target azimuth, such as Figure 4-Figure 8 shown.

[0061] Figure 4 The experimental results of the method of the present invention for estimating the target azimuth and distance are compared. The experiment verifies whether the proposed algorithm is accurate in estimating the azimuth and distance of the set 24 far-field targets through simulation, and compares the estimation results with those of the traditional FDA-MUSIC algorithm and FDA-BP algorithm. The number of snapshots is fixed to 100, the SNR is fixed to 10dB, the number of training set samples is 580800, the number of test samples is 5808, and the size of a single estimation result is 48. In order to provide a more direct perception in "visual" and make the presentation of the target azimuth and distance estimation results clearer and more intuitive, the polar coordinate method is considered for data presentation, and the 0° direction is set to be consistent with the normal direction of the transmitting array, the radial range is set to [-36000m, 16000m], and the polar angle range is set to [0°, 45°]. The radial represents the distance of the target source, and the polar angle represents the azimuth of the target source. The estimation results of the FDA-CNN algorithm are represented by black circles, the real targets are depicted by brown stars, and the estimation results of the FDA-BP algorithm and the FDA-MUSIC algorithm are represented by black and red stars, respectively. The degree of overlap between the estimation results and the real positions is a direct indicator of the accuracy of the algorithm. As can be seen from the figure, the coordinates corresponding to the black circles and brown stars almost completely overlap, which shows that the FDA-CNN algorithm has achieved high accuracy in the estimation of the target azimuth and distance. Relatively speaking, the coordinates corresponding to the black stars and the black circles have a relatively high matching degree in the close-range and low-angle areas, which means that the FDA-BP algorithm has a relatively high estimation accuracy in the case of low-angle close-range, but it performs far worse than FDA-CNN in other cases. As for the red stars, there is a significant deviation from the coordinates corresponding to the black circles, which shows that the accuracy of the FDA-MUSIC algorithm in estimating the target azimuth and distance is far inferior to that of the FDA-CNN algorithm. In summary, the FDA-CNN algorithm shows the best accuracy, followed by the FDA-BP algorithm, while the accuracy of the FDA-MUSIC algorithm is relatively low.

[0062] Figure 5 The influence of different signal-to-noise ratios on the accuracy of the algorithm in estimating the target azimuth angle is shown in Figure 2. Figure 6 The influence of different signal-to-noise ratios on the accuracy of the algorithm in estimating the target distance. Figure 7 The influence of different snapshot numbers on the accuracy of the algorithm in estimating the target azimuth angle is shown in Figure 2. Figure 8 The effect of different snapshot numbers on the accuracy of the algorithm in estimating target distance.

[0063] Depend on Figure 5-Figure 6 The performance of the algorithm in estimating the target distance angle under different signal-to-noise ratios can be analyzed. The horizontal axis is the signal-to-noise ratio, ranging from -10dB to 15dB, and the vertical axis is the RMSE value of the azimuth and distance, ranging from 0° to 1° and 0m to 270m respectively; It can be clearly seen from the figure that under the same signal-to-noise ratio, the azimuth and distance RMSE values ​​corresponding to the FDA-CNN algorithm are smaller than those corresponding to the FDA-BP algorithm; With the increase of the signal-to-noise ratio, the RMSE of both algorithms is decreasing, whether it is azimuth or distance estimation, because with the gradual increase of the signal-to-noise ratio, the effective information in the signal gradually stands out, making it easier for the algorithm to extract target features; The sensitivity of the FDA-CNN algorithm to changes in the signal-to-noise ratio is also much lower than that of the FDA-BP algorithm. When the signal-to-noise ratio is increased to 0dB, the signal-to-noise ratio has little effect on its azimuth and distance RMSE values, and the rate of decline of the RMSE value begins to slow down, and is no longer mainly affected by the signal-to-noise ratio. As the signal-to-noise ratio continues to increase, the azimuth RMSE even tends to a stable value. Experimental results show that within the range of signal-to-noise ratio variation, the estimation performance of the FDA-CNN algorithm is better and more stable than that of the FDA-BP algorithm, especially in the case of a low signal-to-noise ratio of -10dB, where the noise is suppressed more significantly and higher estimation accuracy is shown.

[0064] Depend on Figure 7-Figure 8The performance of the algorithm in estimating the target distance angle under different snapshot numbers can be analyzed. The horizontal axis is the number of snapshots, ranging from 100 to 600, and the vertical axis is the RMSE value of azimuth and distance, ranging from 0° to 1° and 0m to 64m respectively. It can be clearly seen from the figure that under the same number of snapshots, the azimuth and distance RMSE values ​​corresponding to the FDA-CNN algorithm are smaller than those corresponding to the FDA-BP algorithm. With the increase of the number of snapshots, the RMSE values ​​of both algorithms are decreasing, whether it is azimuth or distance estimation. This is because the increase in the number of snapshots will increase the number of sampling points in space and obtain more useful signal information. More sampling points enable the network model to more accurately determine the azimuth and distance of the target. The sensitivity of the FDA-CNN algorithm to the change of the number of snapshots is also much lower than that of the FDA-BP algorithm. When the number of snapshots increases to 200, the number of snapshots has little effect on its azimuth and distance RMSE values, and the rate of decline of RMSE begins to slow down, and is no longer mainly affected by the number of snapshots. When the number of snapshots is [200, 400], the azimuth RMSE value even tends to a stable value. Experimental results show that within the range of snapshot number variations, especially under the condition of small snapshot number, the estimation performance of the FDA-CNN algorithm is better and more stable than that of the FDA-BP algorithm.

[0065] Those of ordinary skill in the art will appreciate that the units of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition of each example has been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0066] In the embodiments provided in the present application, it should be understood that the division of units is merely a logical function division, and there may be other division methods in actual implementation, for example, multiple units may be combined into one unit, one unit may be split into multiple units, or some features may be ignored.

[0067] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

[0068] In summary, the present invention has been disclosed above with preferred embodiments.

Claims

1. A frequency-controlled array radar target positioning method based on deep learning, characterized in that: The following steps are involved: Step S1: Construct a data set. Build a single-base FDA radar transceiver array signal model, set various parameters of the signal model, obtain the FDA radar echo signal matrix, calculate the FDA radar echo matrix covariance, obtain the covariance matrix, separate the real from the virtual and normalize the covariance matrix, and obtain the network input data; Step S2: Build the FDA-CNN network model. The model structure consists of two convolutional layers, two maximum pooling layers, two fully connected layers and one Softmax layer. The first layer is the convolutional layer (Conv.1); followed by a maximum pooling layer (Maxpool); then, the second convolutional layer (Conv.2); followed by the same maximum pooling layer; and finally, two fully connected layers FC1 and FC2; Step S3: training model and outputting results. The built FDA-CNN network model is trained using the training set of input data. After the model training is completed, the model is tested using the test set of input data. Finally, the azimuth and distance of the far-field target are output to complete the estimation of the target position.

2. The frequency-controlled array radar target positioning method based on deep learning according to claim 1 is characterized in that: The step S1 specifically includes the following steps: In step S11, the selected FDA is a uniform linear array containing 11 array elements, each array element is arranged along a straight line, and the distance between every two adjacent array elements is λ / 2. Step S12, consider that there are 24 far-field narrowband incoherent target signal sources with an initial carrier frequency of 10GHZ and a frequency offset of 5kHz incident on the array. Assume that the signal-to-noise ratio is a constant 10dB, the number of snapshots is a constant 100, and the noise is independent and obeys a zero-mean Gaussian distribution of white noise. The target azimuth angle range is [5°, 40°], the change interval is 5°, and there are 8 angles in total; the target distance range is [4km, 12km], the change interval is 4km, and there are 3 distances in total. The combination of 8 angles and 3 distances results in 24 target positions. R=E[Y(t)Y H (t)] (1-1) R′=[R′1,R′2,R′3,...,R′ N ] (1-2) Among them, Y(t)=[y1(t),y2(t),···,y N (t)] T is the received signal vector, H represents the Hermitian operator, E[·] represents the mathematical expectation, R is the echo signal covariance matrix, and R′ represents the data obtained after the covariance matrix R is preprocessed by performing operations such as virtual-real separation and normalization. Step S13, the above 24 targets are substituted into formula (1-1) and formula (1-2) in turn for calculation to obtain 24 sets of data. It should be noted that the real part and the imaginary part of the signal are separated in the process of constructing the data, so the number of experimental samples is twice the size of the beam space. Repeat the same method 100 times, and the 2424 sets of data obtained are used as the input sample data of the network. The training set and the validation set are selected at a ratio of 100:1, and the results of the first 100 calculations are selected from the 24 targets, that is, 2400 groups, a total of 580,800 sample data, as the input data for network model training. Then the results of the last calculation from the 24 targets, that is, 24 groups, a total of 5,808 sample data, are used as input data to verify the performance of the network model. The data used in network training are all equipped with corresponding target real azimuth and distance values ​​as training labels. The validation data is predicted by the network model to obtain the target azimuth and distance estimation values. The actual azimuth and distance of the 24 targets are the test labels of the network.

3. The frequency-controlled array radar target positioning method based on deep learning according to claim 1 is characterized in that: The FDA-CNN structure and algorithm flow of step S2 are as follows: The FDA-CNN structure consists of two convolutional layers, two maximum pooling layers, two fully connected layers and one Softmax layer. In step S21, the input data enters the first convolutional layer (Conv.1), which uses a 3×3 convolution kernel with a stride of 1 and contains 80 convolution kernels in total. The convolution kernel is used to slide the input data to capture and extract useful features. Subsequently, batch normalization is performed to standardize the output feature map activation value. Finally, nonlinearity is introduced into the network by applying an activation function, so that the network can capture more complex features. Step S22, the data then enters a maximum pooling layer (Maxpool), which uses a pooling kernel of size 2×2 (pool size), and the stride is set to 2. Combined with the characteristics of radar echo data, the maximum pooling operation mode is adopted to perform specific processing on the input feature map, thereby achieving data dimensionality reduction, further reducing the number of parameters that the network needs to learn during the training process, and effectively preventing the network from overfitting; Step S23, after that, the data enters the second convolution layer (Conv.2), which also uses a 3×3 convolution kernel, the stride is also set to 1, and also contains 80 convolution kernels; continue to capture and extract the most useful features; Step S24, the data then passes through the same maximum pooling layer again; Step S25, finally, after passing through two fully connected layers FC1 and FC2, it is passed to the Softmax layer output. First, the output of the pooling layer is integerized, and the output multidimensional data is flattened to convert it into a one-dimensional array. Then, it is input into the first fully connected layer. This layer consists of 256 neurons, and the feature vector is mapped to a high-dimensional space through linear transformation to facilitate the learning of complex relationships. In order to maintain the nonlinearity of the model and the ability to learn complex functions, batch normalization and ReLU activation are performed immediately. The ReLU-activated feature vector is passed to the second fully connected layer, which maps the output of the hidden layer to an 8-dimensional space, with each dimension corresponding to a specific output category. After the fully connected layer, the Softmax function is used to ensure that the sum of all output values ​​is maintained at 1, and each element value is between 0 and 1. At the output of the model, the output of the Softmax layer is multiplied by the node and summed along a specific dimension to obtain the final azimuth and distance prediction values.

4. The frequency-controlled array radar target positioning method based on deep learning according to claim 1 is characterized in that: The training of the network model in step S3 specifically includes the following steps: Step S31. Input the sample data into the built network model to extract features and train the model. The batch size is set to 64, the number of network iterations is 200, the learning rate is 0.0001, MSE is selected as the loss function, and the optimizer is Adam. In the forward propagation stage, the model will calculate the output value according to the input data according to the FDA-CNN algorithm process; Step S32, in the error back propagation stage, by comparing the difference between the output value and the true value, the error is calculated and back propagated layer by layer to update the network parameters, thereby optimizing the model. Through repeated iterations, the model gradually learns and improves the accuracy of target positioning. Once the training error approaches the established threshold or reaches the preset number of iterations, the training is considered complete.

Citation Information

Cited By

  • FDA radar target positioning method based on Transform neural network

    CN119936831A