A radar aerial target detection method and system based on particle swarm optimization

Abstract

The present invention provides a radar aerial target detection method and system based on particle swarm optimization, belonging to the field of radar target detection technology. The method comprises the following steps: pre-processing radar raw echo data to extract data slices of moving targets; then transforming the target slice data into a two-dimensional time domain and a range frequency domain; initializing a particle swarm and setting an image entropy error; compensating the target slice data using particle parameters; eliminating Doppler spread in the two-dimensional time domain, performing an azimuth Fourier transform to obtain a range Doppler map of the slice data; calculating image entropy; determining whether the image entropy value meets a condition; and finally performing secondary detection and screening on the focused suspicious moving targets to eliminate false alarm point detections and estimate the target's motion parameters based on the particle values. The method effectively improves the signal-to-noise ratio of the moving target and the detection rate of the system, while more accurately estimating the parameters of the moving target.

Description

Technical Field

[0001] The present invention belongs to the technical field of radar target detection, and in particular relates to a radar aerial target detection method and system based on particle swarm optimization. Background Art

[0002] Radar can detect moving targets by emitting electromagnetic waves and receiving echo signals, offering advantages such as all-day, all-weather, long-distance, and wide-area detection. However, in practical systems, the received signal contains not only the target's scattered echo but also thermal noise. The target's echo energy must compete with the thermal noise signal to be detected by the radar system.

[0003] The classic radar signal processing method for moving targets is coherent integration. However, this method ignores the range drift caused by target motion and the cross-Doppler spread caused by high-order velocity components. This is particularly true for high-speed, maneuvering targets, such as aerial targets, where direct coherent integration results in poor radar system detection performance. Existing techniques such as the Keystone transform and Radon Fourier transform can compensate to some extent, but they all suffer from the search grid problem. The mismatch between the actual target parameters and the discretized grid points leads to a loss of target signal-to-noise ratio. While fine sampling can overcome this limitation, it significantly increases the computational effort. Summary of the Invention

[0004] To solve the above technical problems, the present invention provides a radar aerial target detection method and system based on particle swarm optimization. By utilizing the entropy characteristics of the target image and taking the local minimum entropy as the optimization function, a range migration and Doppler spread compensation signal model for high-speed moving targets is constructed. The particle swarm algorithm is used to accurately estimate the target motion parameters while achieving refined target refocusing, effectively improving the detection performance and computational efficiency of aerial targets.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A radar aerial target detection method based on particle swarm optimization includes the following steps:

[0007] Step 1: Preprocess the radar raw echo data to obtain preliminary information of the moving target;

[0008] Step 2: Using low-threshold constant false alarm rate detection in the range Doppler domain, extract data slices of the moving target based on preliminary information of the moving target;

[0009] Step 3: transform the data slices of the moving target into the two-dimensional time domain through the azimuth inverse Fourier transform, and transform them into the range frequency domain through the range Fourier transform;

[0010] Step 4: Initialize the particle swarm and set the image entropy error;

[0011] Step 5: Use particle swarm parameters to compensate the data slices of the moving target in the range frequency domain and azimuth time domain, correct the target range migration, and convert the data back to the two-dimensional time domain;

[0012] Step 6: Eliminate Doppler spread in the two-dimensional time domain and perform azimuth Fourier transform to obtain a range Doppler map of the data slice;

[0013] Step 7: Calculate the data image entropy of the data slice of the moving target in the range Doppler domain;

[0014] Step 8: Determine whether the entropy value of two adjacent data images is less than the set image entropy value error or whether the number of iterations is greater than the set maximum number of iterations. If so, end the loop; if not, update the particle swarm parameters, and cyclically calculate the entropy value of the data image entropy of the data slice of the moving target in the Doppler domain until the conditions are met.

[0015] Step 9: When the value of the particle swarm parameter is consistent with the true motion parameter of the moving target, the range Doppler domain of the moving target is well focused and the image entropy value of the moving target is minimized. At this time, a secondary detection and screening is performed on the focused suspicious moving target in the range Doppler image to eliminate false alarm point detection. If it is a real moving target, the motion parameter of the real moving target is estimated based on the particle swarm parameter value.

[0016] The present invention also provides a radar air target detection system based on particle swarm optimization, comprising the following modules:

[0017] The preliminary information acquisition module pre-processes the radar raw echo data to obtain preliminary information of the moving target;

[0018] The data slice extraction module uses low-threshold constant false alarm rate detection in the range Doppler domain to extract data slices of the moving target based on the preliminary information of the moving target;

[0019] The transformation module transforms the data slices of the moving target into the two-dimensional time domain through the azimuth dimension inverse Fourier transform, and transforms it into the range frequency domain through the range dimension Fourier transform;

[0020] Initialization module, initializes the particle swarm, and sets the image entropy error;

[0021] The correction module uses particle swarm parameters to compensate the data slices of the moving target in the range frequency domain and azimuth time domain, corrects the target range migration, and converts the data back to the two-dimensional time domain;

[0022] Elimination module, which eliminates Doppler spread in the two-dimensional time domain and performs azimuth Fourier transform to obtain the range Doppler map of the data slice;

[0023] A calculation module, calculating the data image entropy of the data slice of the moving target in the range Doppler domain;

[0024] The judgment module judges whether the entropy value of two adjacent data images is less than the set image entropy value error or whether the number of iterations is greater than the set maximum number of iterations. If so, the loop ends; if not, the particle swarm parameters are updated and the entropy value of the data image entropy of the data slice of the moving target in the Doppler domain is calculated cyclically until the conditions are met;

[0025] In the motion parameter acquisition module, when the value of the particle swarm parameter is consistent with the real motion parameter of the moving target, the range Doppler domain of the moving target is well focused and the image entropy value of the moving target is minimized. At this time, the focused suspicious moving target is screened for secondary detection in the range Doppler image to eliminate false alarm point detection. If it is a real moving target, the motion parameter of the real moving target is estimated based on the particle swarm parameter value.

[0026] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the above-mentioned radar aerial target detection method based on particle swarm optimization are implemented.

[0027] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-mentioned radar aerial target detection method based on particle swarm optimization.

[0028] Beneficial effects:

[0029] The present invention utilizes the particle swarm optimization algorithm in swarm intelligence to achieve global optimization of moving target imaging by iteratively updating the position of particles. Unlike traditional grid-based technologies, the particle swarm optimization algorithm performs optimization in a continuous space, avoiding the problem of converging to a suboptimal local minimum, and its gridless characteristics can achieve accurate target parameter estimation in a shorter computing time. In addition, the present invention also combines the target local entropy information with the particle swarm optimization algorithm to construct a gridless target motion parameter compensation function to maximize the target signal energy and restore its resolution, thereby effectively improving the signal-to-noise ratio of the moving target and the detection rate of the system, while more accurately estimating the parameters of the moving target. The present invention has significant advantages such as strong environmental adaptability, high detection rate of moving targets, and high accuracy of target parameter estimation, and can utilize existing parallel processing system resources, making it suitable for real-time processing in engineering. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a flow chart of a radar air target detection method based on particle swarm optimization of the present invention;

[0031] Figure 2 This is a schematic diagram of the geometric observation of aerial targets by bistatic radar;

[0032] Figure 3 Iteration curve graph for particle swarm optimization algorithm;

[0033] Figure 4a , Figure 4b This is the Doppler result diagram of the range before and after the target is focused; Figure 4a is the result of traditional target detection processing. Figure 4b The processing result of the present invention;

[0034] Figure 5 This is a schematic diagram of a radar air target detection system based on particle swarm optimization of the present invention. DETAILED DESCRIPTION

[0035] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only intended to illustrate the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.

[0036] like Figure 1 As shown, a radar air target detection method based on particle swarm optimization of the present invention includes the following steps:

[0037] Step 1: Preprocess the radar raw echo data, including range compression and azimuth Fourier transform;

[0038] Step 2: Using low threshold constant false alarm rate (CFAR) detection in the range Doppler domain, extract the data slice of the moving target (i.e. Figure 1 Low threshold constant false alarm rate detection, extracting suspicious moving target data slices);

[0039] Step 3: transform the target slice into the two-dimensional time domain through the azimuth inverse Fourier transform, and transform it into the range frequency domain through the range Fourier transform;

[0040] Step 4: Initialize the particle swarm, including the number of particles, initial global optimal position, initial local optimal position, inertia weight, cognitive coefficient, etc., and set the image entropy error;

[0041] Step 5: Use the particle parameters to construct a range migration compensation function in the range frequency domain, perform compensation processing on the range frequency domain-azimuth time domain target slice data, correct the target range migration, and convert the data back to the two-dimensional time domain;

[0042] Step 6: Construct a Doppler quadratic phase compensation factor in the two-dimensional time domain using the particle parameters, perform compensation processing on the data, eliminate Doppler spread, and transform the data into the range Doppler domain;

[0043] Step 7: Calculate the image entropy of the moving target slice data in the range Doppler domain;

[0044] Step 8: If the entropy value of two adjacent images is less than the set image entropy value error or the number of iterations is greater than the set maximum number of iterations, the loop ends. If it is greater than the image entropy value error, the particle swarm parameters are updated and the image entropy value of the moving target slice data in the range Doppler domain is calculated repeatedly until the entropy value of the adjacent images is less than the set image entropy value error or the number of iterations is greater than the set maximum number of iterations (i.e. Figure 1 If the entropy value of the adjacent images is less than the entropy error, the training ends; otherwise, the particle swarm parameters are updated and the target entropy value is calculated cyclically until the entropy value of the adjacent images is less than the error);

[0045] Step 9: When the particle swarm parameter value is consistent with the target's true motion parameter, the target is well focused in the range Doppler domain and the target image entropy is also the smallest. At this time, the focused suspicious moving target is screened for secondary detection in the range Doppler image to eliminate false alarm detection points. If it is a real moving target, the target's motion parameter is estimated based on the particle value (i.e. Figure 1 The focused suspicious moving targets are screened for secondary detection to eliminate false alarms. If they are real targets, the motion parameters of the target are estimated based on the particle value).

[0046] Specifically, the step 1 includes:

[0047] Taking bistatic radar aerial target detection as an example, its observation geometry is as follows Figure 2 As shown, the radiation source is responsible for transmitting the signal. The path of the signal directly reaching the receiver is called the direct wave. The distance between the receiver and the radiation source is L. At the same time, the signal will also reach the target and then be reflected back. This part of the reflected signal is called the target echo. The distance between the receiver and the target is The receiver is used to receive the target echo. Figure 2 The target is the object that the radar system needs to detect, and the distance between the target and the radiation source is .

[0048] By analyzing the direct wave and the target echo, the radar system can determine the position and distance of the target. Assuming that the radar transmits a linear frequency modulation signal, the echo signal received at time t is After demodulation, it can be expressed as:

[0049] ;

[0050] in, is the scattering cross-section of the target, is the pulse slowing time, For distance to fast time, is the radar carrier frequency, To adjust the frequency, is the speed of light, is the distance pulse pressure envelope, is the pulse envelope, is the bistatic distance between the radar system and the target, that is ; exp() represents the exponential function, j represents the imaginary unit, and π is the ratio of circumference to circumference.

[0051] Assuming that the target moves toward the radar and only considering the radial velocity component, the distance to the target is a polynomial function of time, which is expanded by Taylor series and retained to the quadratic term to obtain:

[0052] ;

[0053] in, is the initial bistatic distance, is the target bistatic radial velocity, is the target bistatic radial acceleration.

[0054] Echo signal after pulse compression It can be expressed as:

[0055] ; in, is the distance pulse pressure envelope, is the radar wavelength.

[0056] After pulse compression, the echo signal is subjected to azimuth fast Fourier transform to obtain the signal in the range Doppler domain. , which can be expressed as:

[0057] ;

[0058] in, represents the azimuth fast Fourier transform, Indicates the azimuth frequency.

[0059] Specifically, the step 2 includes:

[0060] according to The value of , low threshold constant false alarm rate detection is performed to extract the moving target echo slice:

[0061] ;

[0062] in, is the detection threshold, which can be calculated by a fixed constant false alarm rate. When it is greater than the threshold, situation, that is, there may be a moving target in the unit. When it is less than the threshold, situation, that is, there is no moving target in the unit, is the target slice extraction function, It is the extracted slice data of suspected moving target signal.

[0063] Specifically, step 3 includes:

[0064] The target slice is transformed into the two-dimensional time domain through the azimuth inverse Fourier transform, and then transformed into the range frequency domain through the distance Fourier transform. Therefore, the azimuth time domain and the range frequency domain signal It can be expressed as:

[0065] ;

[0066] in, represents the distance frequency, represents the azimuth inverse fast Fourier transform, represents the fast Fourier transform of distance.

[0067] Specifically, step 4 includes:

[0068] The number of PSO (particle swarm optimization) iterations is recorded as , Denotes the maximum number of iterations of the particle swarm. Starting from L randomly generated particle solutions, it is expressed as ( ), is the target parameter with random inertia , set the initial position of each particle to the optimal position , initialize the inertia weight , cognitive coefficient , social coefficient , set the image entropy error to . Among them, r represents the target radial component, Indicates the Iteration No. The target radial velocity value for this example is, Indicates the Iteration No. Target radial acceleration value for each example.

[0069] Specifically, step 5 includes:

[0070] Using target parameters Construct the distance migration compensation function in the distance frequency domain, for the Iteration, particles, whose target parameters are , construct the distance migration compensation function in the distance frequency domain , which can be expressed as:

[0071] ;

[0072] Compensate the range-frequency-azimuth-time domain target slice data, correct the target range migration, and convert the data back to the two-dimensional time domain to obtain:

[0073] ;

[0074] in, Indicates the target The signal of the position time domain and distance time domain under the parameters, represents the inverse fast Fourier transform of distance.

[0075] Specifically, step 6 includes:

[0076] On target Constructing the azimuth time domain range time domain Doppler secondary phase compensation signal under the parameters , which can be expressed as:

[0077] ;

[0078] By processing the target data, eliminating Doppler spread, and performing azimuth Fourier transform, the range Doppler map of the slice data can be obtained:

[0079] ;

[0080] in, Indicates the target The target signal in the frequency domain and time domain is located under the parameters.

[0081] Specifically, step 7 includes:

[0082] Calculate the entropy of the moving target slice image, its entropy value It can be expressed as:

[0083] ;

[0084] in, express The entropy value of the image entropy of the moving target under the parameter, is the range Doppler signal Center the window on the target The local value of is the normalization coefficient, which is the energy of the entire image . Indicates the number of orientation units, Indicates the number of distance units, Indicates from 1 to The integer value of Indicates from 1 to The integer value of .

[0085] The local minimum entropy problem of the range-Doppler graph is expressed as follows:

[0086] ;

[0087] Among them, F is the objective function that needs to be optimized by particle swarm optimization, is the target motion parameter support set.

[0088] Specifically, step 8 includes:

[0089] If the number of iterations is greater than the maximum number of iterations Or the two image entropy values ​​are less than the set image entropy error, then the loop ends. If it is greater than the image entropy error and less than the maximum number of iterations , update the particle swarm parameters, each particle will follow the following three rules from the Iterations to +1 iteration to update its position in the search domain.

[0090] a) pass coefficient (inertia weight) and its previous inertia Linked.

[0091] b) by a constant coefficient (cognitive coefficient) attracts it from the 0th iteration to the The best position of the individual visited so far in iterations is expressed as , is the individual minimum entropy;

[0092] ;

[0093] c) By the entire group from iteration 0 to iteration The best position visited by the iteration is attracted, which is expressed as , whose attraction factor is a constant coefficient (social coefficient), is the minimum entropy of the group;

[0094] ;

[0095] Where L represents the number of particles.

[0096] Then update the particle's inertia :

[0097] ;

[0098] in, and is a random variable in the range of [0, 1], which represents the randomness of the particles. The first part ensures the global convergence of the algorithm, and the second and third parts ensure the local convergence ability. Larger, the global convergence ability is strong and the local convergence ability is weakened; Small, the global convergence ability is weakened, the local convergence ability is strong, and it can also be used for searching. Make dynamic adjustments.

[0099] ;

[0100] in, Indicates the maximum inertia weight; Indicates the minimum inertia weight.

[0101] Therefore, the updated potential solution It is expressed as follows:

[0102] ;

[0103] Repeat steps 5 to 8 above, iteratively calculating the target local entropy value until the two image entropy values ​​are less than the set image entropy value error or the maximum number of iterations .

[0104] like Figure 3 As shown in the figure, as the particle swarm iterates, the entropy of the target local image becomes smaller. When the image entropy value is consistent with the target true parameters, the loop is exited.

[0105] Specifically, step 9 includes:

[0106] When the particle swarm parameter values ​​are consistent with the target's true motion parameters, the target is well focused in the Doppler domain, e.g. Figure 4a , Figure 4b As shown ( Figure 4a is the result of traditional target detection processing. Figure 4b (This is the processing result of the present invention). At this time, the target image entropy value is minimized. At this time, a secondary detection and screening is performed on the focused suspicious moving target in the range Doppler map to eliminate false alarm detection points. If it is a real target, the target's motion parameters are estimated based on the particle value. Since the imaging result is related to the target's motion parameters, this can be used to estimate the target's two-dimensional motion parameters:

[0107] ;

[0108] in, Indicates Search the set for possible parameters of the target and ,exist When it is the smallest, take out the and , as an estimate of the target and . is the target motion parameter support set. When different target motion parameters are used for imaging, the image entropy When it is the smallest, the parameter is considered to be the parameter of the moving target. and are the estimated radial velocity and radial acceleration of the target.

[0109] like Figure 5 As shown, the present invention also provides a radar air target detection system based on particle swarm optimization, comprising the following modules:

[0110] The preliminary information acquisition module pre-processes the radar raw echo data to obtain preliminary information of the moving target;

[0111] The data slice extraction module uses low-threshold constant false alarm rate detection in the range Doppler domain to extract data slices of the moving target based on the preliminary information of the moving target;

[0112] The transformation module transforms the data slices of the moving target into the two-dimensional time domain through the azimuth dimension inverse Fourier transform, and transforms it into the range frequency domain through the range dimension Fourier transform;

[0113] Initialization module, initializes the particle swarm, and sets the image entropy error;

[0114] The correction module uses particle swarm parameters to compensate the data slices of the moving target in the range frequency domain and azimuth time domain, corrects the target range migration, and converts the data back to the two-dimensional time domain;

[0115] Elimination module, which eliminates Doppler spread in the two-dimensional time domain and performs azimuth Fourier transform to obtain the range Doppler map of the data slice;

[0116] A calculation module, calculating the data image entropy of the data slice of the moving target in the range Doppler domain;

[0117] The judgment module judges whether the entropy value of two adjacent data images is less than the set image entropy value error or whether the number of iterations is greater than the set maximum number of iterations. If so, the loop ends; if not, the particle swarm parameters are updated and the entropy value of the data image entropy of the data slice of the moving target in the Doppler domain is calculated cyclically until the conditions are met;

[0118] In the motion parameter acquisition module, when the value of the particle swarm parameter is consistent with the real motion parameter of the moving target, the range Doppler domain of the moving target is well focused and the image entropy value of the moving target is minimized. At this time, the focused suspicious moving target is screened for secondary detection in the range Doppler image to eliminate false alarm point detection. If it is a real moving target, the motion parameter of the real moving target is estimated based on the particle swarm parameter value.

[0119] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the above-mentioned radar aerial target detection method based on particle swarm optimization are implemented.

[0120] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-mentioned radar aerial target detection method based on particle swarm optimization.

[0121] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk drives, CD-ROMs, optical storage devices, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention may be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0122] The present invention is described by flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0123] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0124] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0125] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0126] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A radar air target detection method based on particle swarm optimization, characterized in that: The steps include: Step 1: Preprocess the radar raw echo data to obtain preliminary information of the moving target; Step 2: Using low-threshold constant false alarm rate detection in the range Doppler domain, extract data slices of the moving target based on preliminary information of the moving target; Step 3: transform the data slices of the moving target into the two-dimensional time domain through the azimuth inverse Fourier transform, and transform them into the range frequency domain through the range Fourier transform; Step 4: Initialize the particle swarm and set the image entropy error; Step 5: Use particle swarm parameters to compensate the data slices of the moving target in the range frequency domain and azimuth time domain, correct the target range migration, and convert the data back to the two-dimensional time domain; Step 6: Eliminate Doppler spread in the two-dimensional time domain and perform azimuth Fourier transform to obtain a range Doppler map of the data slice; Step 7: Calculate the image entropy of the data slice of the moving target in the range Doppler domain; Step 8: Determine whether the entropy value of two adjacent data images is less than the set image entropy value error or whether the number of iterations is greater than the set maximum number of iterations. If so, end the loop; if not, update the particle swarm parameters, and cyclically calculate the entropy value of the data image entropy of the data slice of the moving target in the Doppler domain until the conditions are met. Step 9: When the value of the particle swarm parameter is consistent with the true motion parameter of the moving target, the range Doppler domain of the moving target is well focused and the image entropy value of the moving target is minimized. At this time, a secondary detection and screening is performed on the focused suspicious moving target in the range Doppler image to eliminate false alarm point detection. If it is a real moving target, the motion parameter of the real moving target is estimated based on the particle swarm parameter value.

2. The radar air target detection method based on particle swarm optimization according to claim 1, characterized in that: In step 1, the radar transmits a linear frequency modulation signal, and determines the position and distance of the moving target by analyzing the direct wave and the target echo; the preprocessing includes range compression and azimuth Fourier transform.

3. The radar air target detection method based on particle swarm optimization according to claim 1, characterized in that: In step 2, low-threshold constant false alarm rate detection is performed according to the detection threshold to extract slice data of the moving target.

4. The radar air target detection method based on particle swarm optimization according to claim 1, characterized in that: In step 4, initializing the particle swarm includes initializing the number of particles, the initial global optimal position, the initial local optimal position, the inertia weight, and the cognitive coefficient; when initializing the particle swarm, the image entropy error is set to a preset value.

5. The radar air target detection method based on particle swarm optimization according to claim 1, characterized in that: In step 5, a range migration compensation function is constructed in the range frequency domain using the particle swarm parameters to perform compensation processing on the slice data of the moving target and correct the target range migration.

6. The radar air target detection method based on particle swarm optimization according to claim 1, characterized in that: In step 6, a Doppler quadratic phase compensation factor is constructed using particle swarm parameters in the two-dimensional time domain.

7. The radar air target detection method based on particle swarm optimization according to claim 1, characterized in that: In step 8, the particle swarm parameters are iteratively updated by the particle swarm optimization algorithm until the value of the entropy of two adjacent images is less than the set image entropy value error or the number of iterations is greater than the set maximum number of iterations, thereby achieving accurate estimation of the target parameters.

8. A radar air target detection system based on particle swarm optimization, characterized in that: Includes the following modules: The preliminary information acquisition module pre-processes the radar raw echo data to obtain preliminary information of the moving target; The data slice extraction module uses low-threshold constant false alarm rate detection in the range Doppler domain to extract data slices of the moving target based on the preliminary information of the moving target; The transformation module transforms the data slices of the moving target into the two-dimensional time domain through the azimuth dimension inverse Fourier transform, and transforms it into the range frequency domain through the range dimension Fourier transform; Initialization module, initializes the particle swarm, and sets the image entropy error; The correction module uses particle swarm parameters to compensate the data slices of the moving target in the range frequency domain and azimuth time domain, corrects the target range migration, and converts the data back to the two-dimensional time domain; Elimination module, which eliminates Doppler spread in the two-dimensional time domain and performs azimuth Fourier transform to obtain the range Doppler map of the data slice; A calculation module, calculating the data image entropy of the data slice of the moving target in the range Doppler domain; The judgment module judges whether the entropy value of two adjacent data images is less than the set image entropy value error or whether the number of iterations is greater than the set maximum number of iterations. If so, the loop ends; if not, the particle swarm parameters are updated and the entropy value of the data image entropy of the data slice of the moving target in the Doppler domain is calculated cyclically until the conditions are met; In the motion parameter acquisition module, when the value of the particle swarm parameter is consistent with the real motion parameter of the moving target, the range Doppler domain of the moving target is well focused and the image entropy value of the moving target is minimized. At this time, the focused suspicious moving target is screened for secondary detection in the range Doppler image to eliminate false alarm point detection. If it is a real moving target, the motion parameter of the real moving target is estimated based on the particle swarm parameter value.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the radar air target detection method based on particle swarm optimization according to any one of claims 1 to 7 are implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the radar air target detection method based on particle swarm optimization according to any one of claims 1 to 7 are implemented.

Images