A target joint angle estimation method and device based on projection noise modeling

Abstract

The application provides a target joint angle estimation method and device based on projection noise modeling, radar observation data is acquired, and a corresponding signal model is constructed; a strong target set is extracted based on the observation data, and first data is obtained through projection; a non-target noise region is extracted based on the observation data, and second data is calculated; a sparse covariance model and a target optimization function are constructed according to the first data and the second data, iteration update is performed to obtain angle power parameters; the angle power parameters are screened according to a set threshold to obtain a weak target set; and a final target joint angle estimation result is obtained according to the strong target set and the weak target set. Through the combination of strong target suppression and sparse covariance fitting, and the modeling of the non-white noise generated after the strong target is stripped, the robust angle estimation in a complex scene is realized, and the overall estimation precision and robustness are improved.

Description

Technical Field

[0001] This invention relates to the field of radar signal processing technology, and in particular to a method and apparatus for joint target angle estimation based on projection noise modeling. Background Technology

[0002] With the development of millimeter-wave radar and array signal processing technology, angle estimation methods based on array observation data have been widely applied in fields such as autonomous driving, intelligent perception, and target detection. In practical applications, radar echoes are usually composed of multiple scattering sources of different intensities, meaning they simultaneously contain both strong and weak targets. The amplitude differences between different targets can reach tens of decibels or more, thus forming a distinct mixed scene of strong and weak targets.

[0003] In the above scenarios, strong targets often have a significant energy advantage, and their corresponding array responses produce strong main lobe and side lobe structures in the spatial spectrum, leading to the following problems:

[0004] 1. Sidelobe flooding effect

[0005] The side lobes generated by strong targets may cover the main lobe of weak targets, making the weak targets difficult to distinguish in the spatial spectrum;

[0006] 2. Dynamic range limitation problem

[0007] Traditional spectral estimation methods struggle to simultaneously ensure the detection performance of both strong and weak targets within a limited dynamic range.

[0008] 3. Covariance matrix distortion problem

[0009] Strong targets dominate the sample covariance structure, making it difficult to effectively extract subspace information corresponding to weak targets;

[0010] 4. Noise modeling mismatch problem

[0011] In the process of removing or suppressing strong targets, additional statistical distortion is often introduced, causing the noise to no longer satisfy the independent and identically distributed assumption.

[0012] Therefore, achieving robust angle estimation remains one of the key issues in current array signal processing when strong and weak targets coexist.

[0013] Existing methods for angle estimation mainly include spectral estimation methods, subspace methods, and sparse reconstruction methods, but they all have certain limitations in strong and weak target scenarios.

[0014] 1. Traditional Spectral Estimation Methods

[0015] Typical methods include Bartlett beamforming and Fast Fourier Transform (FFT), which estimate angles by calculating the array output response at different angles. The advantages of this method are low computational complexity, ease of engineering implementation, and robustness to model errors; the disadvantages are that resolution is limited by the array aperture, strong targets have significant sidelobes, weak targets are easily obscured, and it is difficult to handle high dynamic range scenes.

[0016] 2. Subspace class methods

[0017] Typical methods include MUSIC and ESPRIT, which achieve high-resolution angle estimation based on the orthogonality of the signal and noise subspaces. The advantages of this method are its super-resolution capability and superior performance under high signal-to-noise ratio conditions. The disadvantages include sensitivity to the number of samples and the accuracy of covariance estimation; when strong and weak targets coexist, the strong target dominates the covariance matrix, making it difficult to separate the weak target subspace; and sensitivity to array errors and model mismatch.

[0018] 3. Sparse Reconstruction Class Methods

[0019] In recent years, angle estimation methods based on sparse representations (such as SPICE, LASSO, and atomic norm methods) have received widespread attention. These methods estimate target parameters by constructing sparse optimization problems. The advantages of these methods are that they can achieve high-resolution estimation with limited samples, require no prior assumptions about the number of targets, and have a certain robustness to the statistical properties of noise (such as the SPICE method). The disadvantages are that when strong and weak targets coexist, the strong target may still dominate the optimization process; they are highly dependent on the noise model, which is usually assumed to be white noise; and when strong target interference exists, weak targets are easily suppressed or misjudged by sparse regularization terms.

[0020] 4. CLEAN class methods

[0021] The CLEAN method reduces the interference of strong targets on other targets by iteratively detecting the strongest targets and gradually removing their contributions from the observation data. Its advantages include effectively suppressing strong target sidelobes, improving the detectability of weak targets, and its simple structure, making it easy to implement. Its disadvantages include reliance on successive estimation, the potential for error accumulation, high requirements for the accuracy of target parameter estimation, and difficulty in guaranteeing global optimum in multi-target scenarios.

[0022] In summary, existing methods generally suffer from the following problems in scenarios where strong and weak targets coexist:

[0023] 1. Lack of a unified modeling framework

[0024] CLEAN-type methods focus on strong target suppression, while sparse methods focus on overall estimation; the two have not been effectively integrated.

[0025] 2. The noise model assumptions are inaccurate.

[0026] After stripping away strong targets, the noise in the residual data usually exhibits non-independent and identically distributed characteristics, while existing methods still use the white noise assumption, leading to estimation bias.

[0027] 3. Insufficient weak target detection capability

[0028] In high dynamic range scenarios, weak targets are easily disturbed by strong targets or suppressed by sparse constraints.

[0029] 4. Insufficient robustness

[0030] It has limited adaptability to model errors, array non-idealities, and complex environments.

[0031] Therefore, there is an urgent need for a target joint angle estimation method and apparatus based on projection noise modeling to improve the above problems. Summary of the Invention

[0032] The purpose of this invention is to provide a target joint angle estimation method and apparatus based on projection noise modeling, which can achieve robust angle estimation in complex scenarios and improve the overall estimation accuracy and robustness.

[0033] In a first aspect, the present invention provides a joint target angle estimation method based on projection noise modeling, comprising the steps of: acquiring radar observation data and constructing a corresponding signal model; extracting a set of strong targets based on the observation data and obtaining first data through projection; extracting target-free noise regions based on the observation data and calculating second data; constructing a sparse covariance model and a target optimization function based on the first and second data, and iteratively updating them to obtain power parameters for each angle; filtering the power parameters for each angle based on a set threshold to obtain a set of weak targets; and obtaining the final joint target angle estimation result based on the set of strong targets and the set of weak targets.

[0034] Optionally, the observation data includes strong targets, weak targets, and Gaussian white noise; and / or the signal model is:

[0035]

[0036] in, This refers to observation data within the radar's medium-range Doppler cell, and , The number of array elements in the radar; To increase the number of strong targets; Number of weak targets; For the target complex range; From the perspective of the target; For angle Array steering vector; It is complex Gaussian white noise.

[0037] Optionally, extracting a set of strong targets based on the observation data and obtaining first data through projection includes: performing spatial spectrum estimation based on the observation data, extracting all strong targets corresponding to the spatial spectrum peaks to obtain a set of strong targets, and constructing a corresponding strong target steering matrix based on the set of strong targets; constructing an orthogonal complement projection matrix of the strong target subspace based on the strong target steering matrix, and mapping the observation data to the orthogonal complement space of the strong target subspace based on the orthogonal complement projection matrix to obtain the first data; and / or the first data is a residual vector, wherein the residual vector is:

[0038]

[0039] in, For spatial spectral functions; This refers to observation data within the radar's medium-range Doppler unit.

[0040] Optionally, the target-free noise region is a distant, echo-free region or a Doppler interval without significant targets; and / or the second data is the raw noise power, wherein the raw noise power is:

[0041]

[0042] Among them, median is used to suppress the influence of outliers; This is a region with no target noise. These are the observations for the corresponding range-Doppler cells.

[0043] Optionally, the sparse covariance model is:

[0044]

[0045] in, , for angle Corresponding power parameters; For a strong target set; For angle Array steering vector; For angle The conjugate transpose of the array guiding vector; , is the noise covariance matrix after the second data is projected; The spatial spectrum function is; and / or the objective optimization function is:

[0046]

[0047] The iterative update method is as follows:

[0048]

[0049] in, Find the minimum value under the constraint that all angular power parameters are greater than or equal to 0; The residual vector; This is the conjugate transpose of the residual vector; For sparse covariance matrix The inverse matrix; For sparse covariance matrix traces; For the first The first angle The power update value for the next iteration; This represents the current iteration step. This is the number of the next iteration after the update; For angle Array steering vector; For angle The conjugate transpose of the guiding vector.

[0050] Optionally, the process of filtering the power parameters of each angle to obtain a set of weak targets based on a set threshold includes: filtering angles whose power parameters exceed the set threshold as a set of weak targets; and / or obtaining a final joint angle estimation result for the targets based on the set of strong targets and the set of weak targets includes: merging the set of strong targets and the set of weak targets to obtain a final joint angle estimation result for the targets.

[0051] Secondly, the present invention provides a joint target angle estimation apparatus based on projection noise modeling, the apparatus comprising modules / units for performing any of the possible design methods described in the first aspect above. These modules / units can be implemented in hardware or by hardware executing corresponding software.

[0052] Thirdly, the present invention provides an electronic device including a memory and a processor, wherein the memory stores a program executable on the processor, and when the program is executed by the processor, the electronic device implements a method for performing any of the possible designs described above.

[0053] Fourthly, the present invention provides a readable storage medium storing a program, which, when executed, implements a method of any possible design of any of the above aspects.

[0054] Fifthly, the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0055] The beneficial effects of the method of this invention are as follows: acquiring radar observation data and constructing a corresponding signal model; extracting a set of strong targets based on the observation data and obtaining first data through projection; extracting target-free noise regions based on the observation data and calculating second data; constructing a sparse covariance model and a target optimization function based on the first and second data, and iteratively updating to obtain power parameters for each angle; filtering the power parameters for each angle based on a set threshold to obtain a set of weak targets; and obtaining the final joint angle estimation result of the targets based on the set of strong targets and the set of weak targets. By combining strong target suppression with sparse covariance fitting and modeling the non-white noise generated after strong target stripping, robust angle estimation under complex scenarios is achieved, improving the overall estimation accuracy and robustness. Attached Figure Description

[0056] Figure 1 A flowchart illustrating a joint target angle estimation method based on projection noise modeling provided in an embodiment of the present invention;

[0057] Figure 2 This is a schematic diagram of a target joint angle estimation device based on projection noise modeling provided in an embodiment of the present invention;

[0058] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0059] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions in the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without inventive effort are within the scope of protection of this invention. Unless otherwise defined, the technical or scientific terms used herein should have the ordinary meaning understood by those skilled in the art. The terms "comprising" and similar expressions used herein mean that the element or object preceding the word covers the element or object listed following the word and its equivalents, but do not exclude other elements or objects.

[0060] The technical solutions of the embodiments of the present invention will be described below with reference to the accompanying drawings. In the description of the embodiments of the present invention, the terminology used in the following embodiments is for the purpose of describing specific embodiments only and is not intended to limit the present invention. The singular expressions “a,” “the,” “the,” and “this” are intended to also include expressions such as “one or more,” unless the context clearly indicates otherwise. It should also be understood that in the following embodiments of the present invention, “at least one” and “one or more” refer to one or more (including two). The term “and / or” is used to describe the relationship between related objects, indicating that three relationships can exist; for example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. The character “ / ” generally indicates that the preceding and following related objects are in an “or” relationship.

[0061] References to "one embodiment" or "some embodiments" in this specification mean that a particular feature, structure, or characteristic described in connection with that embodiment is included in one or more embodiments of the invention. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," and "in still other embodiments" appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including, but not limited to," unless otherwise specifically emphasized. The term "connection" includes both direct and indirect connections, unless otherwise stated. "First" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated.

[0062] In embodiments of the present invention, "exemplarily" or "for example" are used to indicate that they are examples, illustrations, or descriptions. Any embodiment or design described as "exemplarily" or "for example" in embodiments of the present invention should not be construed as being more preferred or advantageous than other embodiments or design solutions. Rather, the use of "exemplarily" or "for example" is intended to present the relevant concepts in a specific manner.

[0063] like Figure 1 As shown, this invention provides a joint target angle estimation method based on projection noise modeling, including the following steps:

[0064] S101 acquires radar observation data and constructs the corresponding signal model.

[0065] In some embodiments, the observation data includes strong targets, weak targets, and Gaussian white noise.

[0066] In other embodiments, the signal model is:

[0067]

[0068] in, This refers to observation data within the radar's medium-range Doppler cell, and , The number of array elements in the radar; To increase the number of strong targets; Number of weak targets; For the target complex range; From the perspective of the target; For angle Array steering vector; It is complex Gaussian white noise.

[0069] S102, extract a set of strong targets based on the observation data, and obtain the first data through projection.

[0070] In some embodiments, extracting a set of strong targets based on the observation data and obtaining first data through projection includes: performing spatial spectrum estimation based on the observation data, extracting all strong targets corresponding to the spatial spectrum peaks to obtain a set of strong targets, and constructing a corresponding strong target guidance matrix based on the set of strong targets; constructing an orthogonal complement projection matrix of the strong target subspace based on the strong target guidance matrix, and mapping the observation data to the orthogonal complement space of the strong target subspace based on the orthogonal complement projection matrix to obtain the first data.

[0071] In other embodiments, the first data is a residual vector, wherein the residual vector is:

[0072]

[0073] in, For spatial spectral functions; This refers to observation data within the radar's medium-range Doppler unit.

[0074] S103, Based on the observation data, extract the target-free noise region and calculate the second data.

[0075] In some embodiments, the target-free noise region is a distant echo-free region or a Doppler interval without a significant target.

[0076] In other embodiments, the second data is the raw noise power, which is:

[0077]

[0078] Among them, median is used to suppress the influence of outliers; This is a region with no target noise. For the observations corresponding to the range-Doppler cell, i.e. the th The distance unit and the first Observational data corresponding to each Doppler unit.

[0079] S104, construct a sparse covariance model and objective optimization function based on the first data and the second data, and perform iterative updates to obtain the power parameters at each angle.

[0080] In some embodiments, the sparse covariance model is:

[0081]

[0082] in, , for angle Corresponding power parameters; For a strong target set; For angle Array steering vector; For angle The conjugate transpose of the array guiding vector; , is the noise covariance matrix after the second data is projected; is the spatial spectral function.

[0083] In some embodiments, the objective optimization function is:

[0084]

[0085] The iterative update method is as follows:

[0086]

[0087] in, Find the minimum value under the constraint that all angular power parameters are greater than or equal to 0; The residual vector; This is the conjugate transpose of the residual vector; For sparse covariance matrix The inverse matrix; For sparse covariance matrix traces; For the first The first angle The power update value for the next iteration; This represents the current iteration step. This is the number of the next iteration after the update; For angle Array steering vector; For angle The conjugate transpose of the guiding vector.

[0088] S105, a set of weak targets is obtained by filtering the power parameters of each angle according to the set threshold.

[0089] In some embodiments, filtering the power parameters of each angle according to a set threshold to obtain a set of weak targets includes: filtering angles whose power parameters exceed the set threshold as a set of weak targets.

[0090] S106, Based on the set of strong targets and the set of weak targets, the final joint angle estimation result of the targets is obtained.

[0091] In some embodiments, obtaining the final joint angle estimation result of the target based on the set of strong targets and the set of weak targets includes: merging the set of strong targets and the set of weak targets to obtain the final joint angle estimation result of the target.

[0092] The advantages of this invention are that by combining strong target suppression with sparse covariance fitting and modeling the non-white noise generated after strong target stripping, stable estimation can be achieved under the condition of coexistence of strong and weak targets, effectively suppressing the interference of strong targets on weak targets, reasonably modeling the statistical distortion introduced after strong target stripping, and improving the weak target detection capability and overall robustness.

[0093] To facilitate understanding, this embodiment further elaborates on the specific implementation process of the above method in conjunction with a specific application scenario. Taking vehicle-mounted radar as an example, the vehicle-mounted radar includes an antenna array, and specifically includes the following steps:

[0094] (I) Signal Model and Variable Definition

[0095] Assume the antenna array includes The observation vector (i.e., observation data) of each array element within a certain range-Doppler cell is:

[0096]

[0097] Its signal model is:

[0098]

[0099] in, To increase the number of strong targets; Number of weak targets; For the target complex range; From the perspective of the target; For angle Array steering vector; It is complex Gaussian white noise.

[0100] (II) Strong Target Detection and Subspace Construction

[0101] First, the initial angular spectrum is obtained using spatial spectrum estimation methods (such as FFT or beamforming):

[0102]

[0103] Select the set of strong targets corresponding to the spectral peaks:

[0104]

[0105] Construct a strong target orientation matrix:

[0106]

[0107] (III) Subspace Projection and Residual Construction

[0108] Construct the orthogonal complement projection matrix of the strong target subspace:

[0109]

[0110] Projecting the original observation vector onto the orthogonal complement space of the strong target subspace yields the residual vector (i.e., the first data):

[0111]

[0112] This operation eliminates strong target components while preserving weak target and noise information.

[0113] (iv) Noise power estimation (noise interval method)

[0114] To obtain the raw noise power Selecting target-free areas from the observation data ,For example:

[0115] Long-distance echo-free region;

[0116] Doppler intervals without significant targets.

[0117] Define the original noise power (i.e., the second data) as:

[0118]

[0119] Among them, median is used to suppress the influence of outliers; This is a region with no target noise. These are the observations for the corresponding range-Doppler cells.

[0120] (v) Projection noise covariance modeling

[0121] Since the projection operation alters the statistical characteristics of the noise, the residual noise is:

[0122]

[0123] Its covariance (i.e., the noise covariance matrix after the second data projection) is:

[0124]

[0125] This model accurately describes the non-independent, identically distributed noise structure generated after the stripping of strong targets. To reduce computational complexity, a diagonal approximation can be used:

[0126]

[0127] in:

[0128]

[0129] (vi) Spatial covariance modeling of residual space (SPICE)

[0130] Construct a sparse covariance model in the residual space:

[0131]

[0132] in, , for angle Corresponding power parameters; For a strong target set; For angle Array steering vector; For angle The conjugate transpose of the array guiding vector; , is the noise covariance matrix after the second data is projected; is the spatial spectral function.

[0133] (vii) Parameter estimation (SPICE optimization)

[0134] Estimate by minimizing the following objective function :

[0135]

[0136] Iterative update method:

[0137]

[0138] in, Find the minimum value under the constraint that all angular power parameters are greater than or equal to 0; The residual vector; This is the conjugate transpose of the residual vector; For sparse covariance matrix The inverse matrix; For sparse covariance matrix traces; For the first The first angle The power update value for the next iteration; This represents the current iteration step. This is the number of the next iteration after the update; For angle Array steering vector; For angle The conjugate transpose of the guiding vector.

[0139] (viii) Output of angle estimation results

[0140] Filtering power parameters exceeding the threshold Angles as a set of weak targets:

[0141]

[0142] The set of strong targets With weak target set The final joint angle estimation result of the target is obtained by merging:

[0143]

[0144] This invention addresses the problem of limited angle estimation accuracy in scenarios where strong and weak targets coexist. Improvements are made primarily in three aspects: strong target suppression, noise modeling, and weak target recovery. The key improvements are as follows:

[0145] 1. Strong target elimination mechanism based on subspace projection

[0146] This invention constructs a strong target orientation matrix:

[0147]

[0148] And further construct its orthogonal complement projection matrix:

[0149]

[0150] Project the observation data onto the orthogonal complement space of the strong target subspace:

[0151]

[0152] This achieves the overall elimination of strong target components (not successive stripping), avoiding the error accumulation problem caused by the stepwise subtraction in the traditional CLEAN method. It can process multiple strong targets in one projection, improving stability and computational efficiency.

[0153] 2. Projection noise covariance modeling mechanism

[0154] This invention discovers that during the strong target stripping process, the statistical characteristics of noise change, and the residual noise no longer satisfies the independent and identically distributed assumption, but instead satisfies:

[0155]

[0156] Its covariance is:

[0157]

[0158] Accordingly, this invention, by directly incorporating the subspace projection matrix into noise covariance modeling, not only accurately characterizes the spatial correlation of noise after strong target stripping, but also overcomes the limitations of traditional methods that uniformly employ... This addresses the model mismatch issue and provides the correct weighting structure for subsequent parameter estimation.

[0159] 3. Original Noise Power Estimation Method Based on Noise Range

[0160] To obtain the scale parameters in the projection noise model This invention employs a noise interval estimation method independent of the target signal:

[0161]

[0162] This method has the following characteristics: it directly estimates the original noise power using a targetless region; it avoids interference from strong and weak targets on noise estimation; it enhances robustness to outliers through median calculation; and it decouples from subspace projection operations to improve overall stability.

[0163] 4. Residual Space Sparse Covariance Modeling Mechanism

[0164] After completing strong target suppression and noise modeling, this invention constructs a covariance model in the residual space:

[0165]

[0166] Parameters are estimated using the sparse optimization method (SPICE).

[0167]

[0168] This process enables the recovery of weak targets in a low-interference space where strong targets have been removed, automatically determines the number of targets using sparse constraints, and improves the detection capability and resolution of weak targets.

[0169] 5. A phased joint estimation mechanism for strong and weak targets

[0170] This invention adopts the following phased strategy:

[0171] Phase 1: Strong target extraction and suppression through spectral analysis and subspace projection;

[0172] Second stage: Recover the weak target in the residual space using sparse methods;

[0173] Final fusion: Outputs joint estimation results for strong and weak targets.

[0174] This process decouples "strong target suppression" from "weak target recovery," avoiding the performance limitations of a single method in high dynamic range scenarios and improving the overall estimation accuracy and robustness.

[0175] 6. Search space optimization mechanism based on structural constraints

[0176] In the sparse estimation process, this invention will use a set of strong targets. Remove from the search space:

[0177]

[0178] This avoids strong targets being repeatedly estimated, reduces the optimization dimensionality and computational complexity, and improves the stability of weak target detection.

[0179] 7. Scalable noise modeling and computation implementation mechanism

[0180] The projection noise modeling method proposed in this invention has good engineering scalability and can be implemented in a complete matrix form. To improve accuracy, or to reduce computational complexity by using diagonal approximation, it can be applied to different array structures (uniform arrays, non-uniform arrays, and virtual arrays) and is easy to combine with parallel computing architectures to achieve efficient solutions.

[0181] The advantages of this invention are that, addressing the limitation in angle estimation performance under conditions of coexistence of strong and weak targets, it achieves significant results at both the theoretical and engineering levels by introducing subspace projection noise modeling and a staged joint estimation mechanism, as detailed below:

[0182] (i) Significantly improve the ability to distinguish between strong and weak targets

[0183] This invention achieves holistic elimination of strong target components by constructing a subspace projection matrix:

[0184]

[0185] This effectively suppresses interference from the main lobe and side lobes of strong targets on weak targets. Compared with traditional spectral estimation methods, this invention not only significantly reduces side lobe leakage of strong targets, but also improves the visibility of weak targets in the spatial spectrum, achieving stable resolution under high dynamic range conditions.

[0186] (II) Improve the accuracy and stability of weak target detection

[0187] Introducing sparse covariance modeling in the residual space:

[0188]

[0189] By combining the SPICE optimization method, this invention ensures that: weak targets are no longer masked by the covariance structure dominated by strong targets; sparse constraints effectively suppress spurious peaks; and it still maintains good detection capabilities under low signal-to-noise ratio conditions. Compared to traditional subspace methods, this invention exhibits higher robustness to weak targets.

[0190] (III) Establish a noise statistical model that conforms to reality

[0191] This invention proposes, by analyzing the statistical characteristics after strong target stripping:

[0192]

[0193] This model can accurately describe the spatial correlation of noise after projection, thus overcoming the model mismatch problem caused by the white noise assumption in traditional methods, improving the accuracy of parameter estimation, and avoiding angle deviation and false detection caused by improper noise modeling.

[0194] (iv) Reduce error accumulation and improve algorithm stability

[0195] Unlike the traditional successive CLEAN method, this invention achieves strong target elimination through one-time subspace projection, which not only avoids error propagation and accumulation during successive subtraction and reduces dependence on the estimation accuracy of individual targets, but also improves the overall stability in multi-target scenarios.

[0196] (v) Achieving unified modeling and joint estimation of strong and weak targets

[0197] This invention employs a phased strategy of "strong target suppression + residual estimation." Strong targets are eliminated through subspace projection, while weak targets are recovered through sparse covariance modeling, ultimately outputting a unified angle estimation result. This process effectively decouples strong and weak targets, unifies the modeling of targets with different intensities, and improves the overall estimation performance.

[0198] (vi) Enhance adaptability to complex environments

[0199] By introducing noise interval estimation and projected noise modeling, this invention can accurately estimate noise power and decouple it from the target signal, suppress the influence of environmental interference on parameter estimation, and maintain stable performance in complex backgrounds.

[0200] (vii) Reduce computational complexity and improve implementation efficiency

[0201] This invention has good computational characteristics in its structural design. Subspace projection can be efficiently achieved through matrix operations. The sparse covariance model supports iterative structure updates. Diagonal approximation can be used to reduce the complexity of matrix operations. It is suitable for parallel computing architectures and improves real-time performance.

[0202] (viii) Possesses good engineering scalability

[0203] The method of this invention has strong versatility and scalability, and can be applied to different array structures (uniform arrays and non-uniform arrays). It can be extended to multiple snapshots and multi-dimensional parameter estimation scenarios, and can also be combined with other high-resolution methods to further improve performance.

[0204] like Figure 2 As shown, based on the above method, the present invention provides a target joint angle estimation device based on projection noise modeling, comprising: an acquisition unit 201 for acquiring radar observation data and constructing a corresponding signal model; a projection unit 202 for extracting a set of strong targets based on the observation data and obtaining first data through projection; an extraction unit 203 for extracting target-free noise regions based on the observation data and calculating second data; an iteration unit 204 for constructing a sparse covariance model and a target optimization function based on the first and second data, and iteratively updating to obtain power parameters for each angle; a filtering unit 205 for filtering the power parameters for each angle based on a set threshold to obtain a set of weak targets; and an estimation unit 206 for obtaining the final target joint angle estimation result based on the set of strong targets and the set of weak targets.

[0205] It should be understood that all relevant content of each step involved in the above method embodiments can be referenced to the functional description of the corresponding functional module, and will not be repeated here. Furthermore, the use of suffixes such as "module," "component," or "unit" to represent elements is merely for the purpose of illustrative purposes and has no specific meaning in itself. Therefore, "module," "component," or "unit" can be used interchangeably. Terminals can be implemented in various forms. For example, the terminals described in this invention may include mobile terminals such as mobile phones, tablets, laptops, handheld computers, personal digital assistants (PDAs), portable media players (PMPs), navigation devices, wearable devices, smart bracelets, pedometers, etc., as well as fixed terminals such as digital TVs and desktop computers. The following description will use mobile terminals as examples; those skilled in the art will understand that, in addition to elements specifically designed for mobile purposes, the construction according to embodiments of the present invention can also be applied to fixed-type terminals.

[0206] In other embodiments of the present invention, an electronic device 300 is disclosed, such as... Figure 3 As shown, the device may include: one or more processors 301; memory 302; display 303; one or more application programs (not shown); and one or more computer programs 304. These devices can be connected via one or more communication buses 305. The one or more computer programs 304 are stored in the memory 302 and configured to be executed by the one or more processors 301. The one or more computer programs 304 include instructions that can be used to perform actions such as... Figure 1 Each step in the corresponding embodiment.

[0207] Processor 301 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0208] The memory 302 can be an internal storage unit of the electronic device 300, such as a hard disk or RAM of the electronic device 300. The memory 302 can also be an external storage device of the electronic device 300, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, or FlashCard equipped on the electronic device 300. Furthermore, the memory 302 can include both internal and external storage units of the electronic device 300. The memory 302 is used to store computer programs and other programs and data required by the electronic device. The memory 302 can also be used to temporarily store data that has been output or will be output.

[0209] The computer program 304 can be divided into one or more modules / units. The one or more modules / units can be a series of computer program instruction segments that can perform a specific function. The instruction segments are used to describe the execution process of the computer program 304 in the electronic device 300.

[0210] In addition to the above-described structure, those skilled in the art will understand that Figure 3This is merely an example of electronic device 300 and does not constitute a limitation on electronic device 300. Electronic device 300 may include more or fewer components than shown, or combine certain components, or different components. For example, electronic device may also include input / output devices, network access devices, buses, etc.

[0211] Those skilled in the art will understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the functions described above can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this invention. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0212] Based on the above embodiments, the present invention also discloses a computer-readable storage medium having at least one computer program stored thereon, wherein the computer program, when executed by a processor, implements the methods described in the foregoing embodiments.

[0213] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing a processor. The program can be stored in a computer-readable storage medium, which is a non-transitory medium, such as random access memory, read-only memory, flash memory, hard disk, solid-state drive, magnetic tape, floppy disk, optical disk, and any combination thereof. The storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. This available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., digital video disc (DVD)), or a semiconductor medium (e.g., solid-state drive (SSD)).

[0214] The descriptions of the processes or structures corresponding to the above figures each have their own emphasis. For parts of a process or structure that are not described in detail, please refer to the relevant descriptions of other processes or structures.

[0215] Although the embodiments of the present invention have been described in detail above, it will be apparent to those skilled in the art that various modifications and variations can be made to these embodiments. The above descriptions are merely embodiments of the present invention and do not limit the patent scope of the present invention. However, it should be understood that such modifications and variations fall within the scope and spirit of the present invention. Moreover, the present invention described herein may have other embodiments and can be implemented or realized in various ways. All equivalent transformations made based on the description and drawings of the present invention, or direct or indirect applications in related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A joint angle estimation method for targets based on projection noise modeling, characterized in that, Including the following steps: Acquire radar observation data and construct corresponding signal models; Based on the observation data, a set of strong targets is extracted, and the first data is obtained by projection. Based on the observation data, a target-free noise region is extracted, and the second data is calculated. Based on the first and second data, a sparse covariance model and objective optimization function are constructed, and power parameters at each angle are obtained through iterative updates. A set of weak targets is obtained by filtering the power parameters of each angle according to the set threshold. The final joint angle estimation result of the targets is obtained based on the set of strong targets and the set of weak targets.

2. The method according to claim 1, characterized in that, The observation data includes strong targets, weak targets, and Gaussian white noise; And / or the signal model is: in, This refers to observation data within the radar's medium-range Doppler cell, and , The number of array elements in the radar; To increase the number of strong targets; Number of weak targets; For the target range; From the perspective of the target; For angle Array steering vector; It is complex Gaussian white noise.

3. The method according to claim 1, characterized in that, Based on the observation data, a set of strong targets is extracted, and the first data obtained through projection includes: Based on the observation data, spatial spectrum estimation is performed, and all strong targets corresponding to the spatial spectrum peaks are extracted to obtain a set of strong targets. A corresponding strong target steering matrix is ​​then constructed based on the set of strong targets. Construct an orthogonal complement projection matrix of the strong target subspace based on the strong target guidance matrix, and map the observation data to the orthogonal complement space of the strong target subspace based on the orthogonal complement projection matrix to obtain the first data; And / or the first data is a residual vector, wherein the residual vector is: in, For spatial spectral functions; This refers to observation data within the radar's medium-range Doppler unit.

4. The method according to claim 1, characterized in that, The target-free noise region is a distant region without echoes or a Doppler interval without significant targets. And / or the second data is the raw noise power, wherein the raw noise power is: Among them, median is used to suppress the influence of outliers; This is a region with no target noise. These are the observations for the corresponding range-Doppler cells.

5. The method according to claim 4, characterized in that, The sparse covariance model is as follows: in, , for angle Corresponding power parameters; For a strong target set; For angle Array steering vector; For angle The conjugate transpose of the array guiding vector; , is the noise covariance matrix after the second data is projected; For spatial spectral functions; and / or the objective optimization function is: The iterative update method is as follows: in, Find the minimum value under the constraint that all angular power parameters are greater than or equal to 0; It is the residual vector; This is the conjugate transpose of the residual vector; For sparse covariance matrix The inverse matrix; For sparse covariance matrix traces; For the first The first angle The power update value for the next iteration; This represents the current iteration step. This is the number of the next iteration after the update; For angle The conjugate transpose of the guiding vector.

6. The method according to any one of claims 1-5, characterized in that, The set of weak targets obtained by filtering the power parameters of each angle according to the set threshold includes: Angles whose power parameters exceed a set threshold are selected as weak targets. And / or the final joint angle estimation result of the targets obtained based on the set of strong targets and the set of weak targets includes: The strong target set and the weak target set are merged to obtain the final joint angle estimation result of the targets.

7. A joint target angle estimation device based on projection noise modeling, used in the method of any one of claims 1-6, characterized in that, include: The acquisition unit is used to acquire radar observation data and construct corresponding signal models; A projection unit is used to extract a set of strong targets based on the observation data and obtain first data through projection. An extraction unit is used to extract target-free noise regions based on the observation data and calculate second data. The iterative unit is used to construct a sparse covariance model and an objective optimization function based on the first data and the second data, and to perform iterative updates to obtain power parameters at each angle. A filtering unit is used to filter the power parameters of each angle according to a set threshold to obtain a set of weak targets; The estimation unit is used to obtain the final joint angle estimation result of the target based on the set of strong targets and the set of weak targets.

8. An electronic device, characterized in that, The device includes a memory and a processor, wherein the memory stores a program that can run on the processor, and when the program is executed by the processor, causes the electronic device to perform the method of any one of claims 1-6.

9. A readable storage medium storing a program, characterized in that, When the program is executed, it implements the method of any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1-6.

Images