Radar anti-interference algorithm optimization platform

Abstract

The invention, which relates to the technical field of radar anti-interference algorithm optimization, discloses a radar anti-interference algorithm optimization platform comprising an environment sensing module, an interference classification module, a dynamic filtering module, a waveform optimization module, an array processing module and an efficiency feedback module. The environment sensing module is used for carrying out multi-dimensional sampling on radar receiving signals by adopting a broadband digital channelization technology to obtain original electromagnetic environment data; the interference classification module is used for analyzing the original electromagnetic environment data by adopting a deep convolutional neural network to obtain a feature parameter set; the dynamic filtering module is used for carrying out adaptive processing on the characteristic parameter set by adopting an improved RLS-NLMS hybrid algorithm to obtain a baseband signal; the waveform optimization module is used for carrying out waveform parameter inversion on the baseband signal by adopting a genetic algorithm to obtain a waveform code; and the array processing module is used for performing spatial domain synthesis on the waveform codes by adopting a distributed MVDR algorithm to obtain an anti-interference beam forming weight.

Description

Technical Field

[0001] The present invention relates to the technical field of radar anti-interference algorithm optimization, in particular to a radar anti-interference algorithm optimization platform. Background Art

[0002] Radar anti-interference algorithm optimization technology integrates multiple advanced algorithms and techniques to improve radar systems' target detection capabilities and anti-interference performance in complex electromagnetic environments. Therefore, leveraging advanced technologies to enhance the intelligence and safety of radar anti-interference algorithm optimization has become a pressing issue.

[0003] In the field of radar anti-interference algorithm optimization, traditional methods cannot provide sufficiently accurate classification results for interference signals of different types and intensities, resulting in inefficient or poor results in subsequent processing steps. In addition, traditional waveform design methods are mostly based on fixed parameters and lack the ability to flexibly adapt to different working environments, which limits the overall performance of the radar system. At the same time, existing adaptive filtering algorithms often face problems such as slow convergence and poor steady-state performance, which are particularly evident in rapidly changing interference environments. Summary of the Invention

[0004] In view of the above-mentioned existing problems, the present invention provides a radar anti-interference algorithm optimization platform to solve the problem that traditional methods cannot provide sufficiently accurate classification results, resulting in inefficient or poor results in subsequent processing steps, and traditional waveform design methods are mostly based on fixed parameters and lack flexible adaptability to different working environments, which limits the overall performance of the radar system.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] In a first aspect, the present invention provides a radar anti-interference algorithm optimization platform, which includes:

[0007] Environmental perception module, interference classification module, dynamic filtering module, waveform optimization module, array processing module, and performance feedback module;

[0008] The environment perception module is used to perform multi-dimensional sampling of radar received signals using broadband digital channelization technology to obtain original electromagnetic environment data;

[0009] The interference classification module is used to analyze the original electromagnetic environment data using a deep convolutional neural network to obtain a feature parameter set;

[0010] The dynamic filtering module is used to adaptively process the characteristic parameter set using an improved RLS-NLMS hybrid algorithm to obtain a baseband signal;

[0011] The waveform optimization module is used to use a genetic algorithm to perform waveform parameter inversion on the baseband signal to obtain waveform coding;

[0012] The array processing module is used to perform spatial domain synthesis on the waveform coding using a distributed MVDR algorithm to obtain anti-interference beamforming weights;

[0013] The performance feedback module is used to evaluate the performance of anti-interference beamforming weights using digital twin technology, obtain optimized parameters and provide closed-loop feedback to the environmental perception module.

[0014] As a preferred solution of the radar anti-interference algorithm optimization platform of the present invention, wherein: the broadband digital channelization technology is used to perform multi-dimensional sampling of the radar received signal to obtain the original electromagnetic environment data, and the specific steps are:

[0015] The analog signal received by the radar is digitized using a high-speed analog-to-digital converter (ADC) to obtain a time-domain radar received signal;

[0016] By applying Fast Fourier Transform (FFT), the time domain radar receiving signal is converted to the frequency domain, and the information of different frequency bands is separated to obtain the frequency domain information.

[0017] Calculate the energy distribution of each frequency point in the frequency domain signal and mark the significant energy peak position as the key characteristic parameter, wherein the key characteristic parameter includes the spectrum peak and energy distribution;

[0018] All the information obtained after the above steps is integrated into the original electromagnetic environment data.

[0019] As a preferred solution of the radar anti-interference algorithm optimization platform of the present invention, wherein: the deep convolutional neural network is used to analyze the original electromagnetic environment data to obtain the characteristic parameter set, and the specific steps are as follows:

[0020] Perform format conversion and normalization on the original electromagnetic environment data to obtain normalized data;

[0021] The normalized data is subjected to feature extraction by multi-convolution kernel scanning method to obtain the preliminary feature map z j ;

[0022] The maximum pooling method is used to transform the initial feature map z j Perform dimensionality reduction processing to obtain a compressed feature map;

[0023] The compressed feature map is flattened into a one-dimensional vector v through the fully connected layer operation, and the flattened one-dimensional vector v is subjected to global feature extraction and classification to obtain the classification result y;

[0024] Based on the classification results, the category index with the highest probability is calculated and marked as the type of interference signal;

[0025] Based on the interference signal type, combined with the input feature vector c of the fully connected layer and the corresponding weight matrix W f , extract the characteristic parameter set F related to the interference signal c .

[0026] As a preferred solution of the radar anti-interference algorithm optimization platform of the present invention, wherein: the improved RLS-NLMS hybrid algorithm is used to adaptively process the characteristic parameter set to obtain the baseband signal, and the specific steps are:

[0027] The parameters of the RLS-NLMS hybrid algorithm are configured through the initialization method to obtain the initial weight vector and covariance matrix;

[0028] The recursive least squares method is used to quickly converge the initial weight vector and covariance matrix to obtain the preliminary weight update result;

[0029] The initial weight update results are optimized in a steady state using the normalized least mean square algorithm to obtain the final weight vector;

[0030] The hybrid strategy method fuses the results of the RLS and NLMS algorithms to obtain the optimized weight vector w final (n), and the optimized weight vector w is output by filtering method final (n) is applied to obtain the baseband signal.

[0031] As a preferred solution of the radar anti-interference algorithm optimization platform of the present invention, wherein: the waveform parameter inversion of the baseband signal using a genetic algorithm to obtain waveform coding is performed, and the specific steps are:

[0032] The waveform parameter range of the baseband signal is randomly sampled by the initialization population method to generate an initial waveform parameter set;

[0033] Combining the randomly sampled waveform parameter vectors into an initial population;

[0034] The signal matching degree of each individual waveform parameter is calculated through the fitness function to obtain the fitness value;

[0035] Select high-quality individuals from the current population through the roulette wheel selection method to form the parent population;

[0036] Through single-point crossover, two individuals in the parent population are genetically recombined to obtain offspring individuals, and the individuals in the offspring population are randomly disturbed to obtain mutated offspring individuals;

[0037] Set the termination condition, evaluate the fitness of the current population, and obtain the final waveform code;

[0038] When the termination condition is met, the individual with the highest fitness in the current population is output as the final waveform code, otherwise, the iteration continues.

[0039] As a preferred solution of the radar anti-interference algorithm optimization platform of the present invention, wherein: the distributed MVDR algorithm is used to perform spatial domain synthesis on the waveform coding to obtain the anti-interference beamforming weights, and the specific steps are:

[0040] Extracting waveform parameters p from genetic algorithm output i , generate the corresponding waveform matrix S;

[0041] Each waveform parameter set p i Contains pulse width and frequency offset information used to construct a representation of the baseband signal;

[0042] By performing statistical analysis on the input signal x(n), the covariance matrix R is calculated xx ;

[0043] Generate a steering vector based on the desired direction and calculate the optimal weight using the MVDR criterion;

[0044] The calculated optimal weights are applied to the waveform matrix to obtain the interference-resistant beamforming result.

[0045] As a preferred solution of the radar anti-interference algorithm optimization platform of the present invention, wherein: the digital twin technology is used to evaluate the anti-interference beamforming weight performance, obtain the optimized parameters and feedback them to the environmental perception module in a closed loop. The specific steps are as follows:

[0046] Build a digital twin model to generate a high-fidelity virtual environment based on the radar's hardware configuration and operating environment characteristics;

[0047] The radar operating environment characteristics include terrain and meteorological conditions;

[0048] Apply the anti-interference beamforming weights derived from the MVDR algorithm in the generated virtual environment to generate a simulation data set;

[0049] Calculate its performance indicators based on simulation data;

[0050] By using comparative analysis methods, the performance indicators obtained by simulation are compared with the pre-set target values to identify the parameter set that needs to be optimized;

[0051] Applying genetic algorithm to the identified parameter set to obtain the optimized parameter set;

[0052] Through a closed-loop feedback mechanism, the optimized parameter set is sent back to the environmental perception module to update the radar configuration.

[0053] As a preferred solution of the radar anti-interference algorithm optimization platform of the present invention, wherein: the anti-interference beamforming weights obtained from the MVDR algorithm are applied in the generated virtual environment to generate a simulation data set, specifically as follows:

[0054] The radar includes an antenna array structure, a signal processing module and a noise model;

[0055] Simulate interference sources, target echo signals, and noise distribution in actual operating environments;

[0056] In the virtual environment, the input signal is composed of a superposition of a target signal, an interference signal and noise;

[0057] Apply the anti-interference beamforming weights calculated by the MVDR algorithm to the received signal in the virtual environment to obtain the output signal after beamforming;

[0058] After running the simulation in the virtual environment, record the output signal y after beamforming beam (n) and related performance indicators to generate simulation data sets;

[0059] The simulation data set includes time domain signals, frequency domain characteristics and performance indicators.

[0060] In a second aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the radar anti-interference algorithm optimization platform as described in the first aspect of the present invention is implemented.

[0061] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the radar anti-interference algorithm optimization platform as described in the first aspect of the present invention is implemented.

[0062] The beneficial effects of the present invention are as follows: by adopting broadband digital channelization technology to perform multi-dimensional sampling of radar received signals, the acquisition of original electromagnetic environment data is realized, comprehensive coverage of radar received signals is ensured, various frequency components and their energy distribution can be accurately captured, and a high-quality data basis is provided for subsequent interference classification; by improving the RLS-NLMS hybrid algorithm, the characteristic parameter set is adaptively processed to achieve the optimization of the baseband signal, which can maintain good tracking capability and stability in a dynamically changing environment, effectively suppress the influence of interference signals, and improve the quality of the baseband signal; by using the genetic algorithm to invert the waveform parameters of the baseband signal, the generation of the optimal waveform code is achieved; the global search capability of the genetic algorithm enables it to find the optimal or near-optimal waveform code in a larger solution space, thereby improving the anti-interference capability and target detection accuracy of the radar system; by using the distributed MVDR algorithm to perform spatial domain synthesis of the waveform code, the calculation of the anti-interference beamforming weight is achieved, so that the radar can still maintain high sensitivity to the target signal in a complex environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0064] Figure 1 This is a system diagram of the radar anti-interference algorithm optimization platform in Example 1. DETAILED DESCRIPTION

[0065] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0066] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0067] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0068] Example 1, with reference to Figure 1, which is the first embodiment of the present invention, provides a radar anti-interference algorithm optimization platform, including:

[0069] Environmental perception module, interference classification module, dynamic filtering module, waveform optimization module, array processing module, and performance feedback module;

[0070] The environmental perception module is used to perform multi-dimensional sampling of radar received signals using broadband digital channelization technology to obtain raw electromagnetic environment data;

[0071] Furthermore, a high-speed analog-to-digital converter (ADC) is used to digitize the analog signal received by the radar to obtain a time-domain radar received signal;

[0072] By applying Fast Fourier Transform (FFT), the time domain radar receiving signal is converted to the frequency domain, and the information of different frequency bands is separated to obtain the frequency domain information.

[0073] Calculate the energy distribution of each frequency point in the frequency domain signal and mark the significant energy peak position as the key characteristic parameter, which includes the spectrum peak and energy distribution;

[0074] Integrate all the information obtained after the above steps into raw electromagnetic environment data;

[0075] It should be noted that broadband digital channelization technology can not only efficiently process multi-dimensional sampling of radar receive signals, but also adapt to the dynamic changes of signals in different frequency bands. Through the combination of high-speed ADC and FFT, the module can capture subtle changes in the electromagnetic environment in real time, providing accurate data support for subsequent interference classification. In addition, marking the positions of significant energy peaks helps to quickly identify potential interference sources, improving the system's response speed and accuracy.

[0076] Interference classification module, which uses deep convolutional neural network to analyze the original electromagnetic environment data and obtain a set of characteristic parameters;

[0077] Furthermore, the original electromagnetic environment data is format converted and normalized to obtain normalized data;

[0078] The normalized data is subjected to feature extraction by multi-convolution kernel scanning method to obtain the preliminary feature map z j , the expression is:

[0079] z j =f(W j *x+b j )

[0080] Among them, z j is the feature map output by the jth convolution kernel, representing the preliminary feature map, W jis the weight matrix of the jth convolution kernel, x is the input data, b j is the bias term, * represents the convolution operation, and f is the ReLU activation function;

[0081] The maximum pooling method is used to transform the initial feature map z j Perform dimensionality reduction processing to obtain a compressed feature map;

[0082] The compressed feature map is flattened into a one-dimensional vector v through the fully connected layer operation, and the flattened one-dimensional vector v is subjected to global feature extraction and classification to obtain the classification result y, which is expressed as:

[0083] y=softmax(W f ·v+b f )

[0084] Among them, y is the classification result, W f is the weight matrix of the fully connected layer, v is the flattened eigenvector, b f is the bias term;

[0085] Based on the classification results, the category index with the highest probability is calculated and marked as the type of interference signal;

[0086] Based on the interference signal type, combined with the input feature vector c of the fully connected layer and the corresponding weight matrix W f , extract the characteristic parameter set F related to the interference signal c , the expression is:

[0087] F c ={f i |=W f,ci ·v+b f,c}

[0088] Among them, F c is the feature parameter set extracted for interference signal type c, f i is the i-th characteristic parameter, W f,ci is the weight subset related to category c in the fully connected layer, v is the flattened feature vector, b f,c is the bias term related to category c;

[0089] It should be noted that the deep convolutional neural network (CNN) demonstrates powerful feature extraction capabilities when processing complex raw electromagnetic environment data. By converting and normalizing the data format, it ensures that the data input to the CNN has good consistency and stability. The multi-convolution kernel scanning method can effectively capture features of different scales, and the maximum pooling method further enhances the model's ability to focus on key features. The final fully connected layer operation not only realizes global feature extraction, but also can accurately extract the feature parameter set related to the interference signal based on the classification results, providing a reliable foundation for the dynamic filtering module.

[0090] Dynamic filtering module, used to adaptively process the characteristic parameter set using an improved RLS-NLMS hybrid algorithm to obtain a baseband signal;

[0091] Furthermore, the parameters of the RLS-NLMS hybrid algorithm are configured through the initialization method to obtain the initial weight vector and covariance matrix;

[0092] The recursive least squares method is used to quickly converge the initial weight vector and covariance matrix to obtain the preliminary weight update result, which is expressed as:

[0093] w(n+1)=w(n)+K(n)e(n)

[0094]

[0095] Where w(n+1) is the updated weight vector, K(n) is the gain vector, e(n) is the error signal, P(n+1) is the updated covariance matrix, λ is the forgetting factor, and x(n) is the input feature vector at the current moment;

[0096] The initial weight update result is optimized in a steady state by the normalized least mean square algorithm to obtain the final weight vector, which is expressed as:

[0097]

[0098] Among them, w(n+1) is the final updated weight vector, μ is the step size parameter, e(n) is the error signal, ∥x(n)∥ 2 represents the square norm of the input feature vector;

[0099] The hybrid strategy method fuses the results of the RLS and NLMS algorithms to obtain the optimized weight vector w final (n), and the optimized weight vector w is output by filtering method final (n) is applied to obtain the baseband signal, which is expressed as:

[0100]

[0101] Among them, y(n) is the output baseband signal, is the transpose of the final optimized weight vector, and x(n) is the input feature vector at the current moment;

[0102] It should be noted that the improved RLS-NLMS hybrid algorithm combines the advantages of two classic adaptive filtering algorithms, achieving both rapid convergence and excellent steady-state performance. Initialization methods configure the initial weight vector and covariance matrix, ensuring the algorithm's stability and effectiveness during startup. Recursive least squares (RLS) is used to rapidly adjust the weight vector, while the normalized least mean squares (NLMS) algorithm further optimizes the weight vector's steady-state performance. This hybrid strategy not only improves baseband signal quality but also effectively copes with complex and changing interference environments, enhancing the system's anti-interference capabilities.

[0103] The waveform optimization module is used to invert the waveform parameters of the baseband signal using a genetic algorithm to obtain waveform coding;

[0104] Furthermore, the waveform parameter range of the baseband signal is randomly sampled by an initialization population method to generate an initial waveform parameter set;

[0105] Combining the waveform parameter vectors obtained by random sampling into an initial population;

[0106] The fitness function is used to calculate the signal matching degree of each individual waveform parameter to obtain the fitness value, which is expressed as:

[0107]

[0108] Among them, f(p i ) represents individual p i The fitness value of y(n) represents the original baseband signal, y′(n; p i ) represents the waveform parameter p i The inverse baseband signal is a small positive number;

[0109] Select high-quality individuals from the current population through the roulette wheel selection method to form the parent population;

[0110] Through single-point crossover, two individuals in the parent population are genetically recombined to obtain offspring individuals, and the individuals in the offspring population are randomly perturbed to obtain mutated offspring individuals;

[0111] Set the termination condition, evaluate the fitness of the current population, and obtain the final waveform code, which is expressed as:

[0112]

[0113] When the termination condition is met, the individual with the highest fitness in the current population is output as the final waveform code, otherwise, the iteration continues;

[0114] It should be noted that as a global optimization method, the genetic algorithm can search for the optimal solution in a larger solution space and avoid falling into the local optimum. By initializing the population and applying the fitness function to evaluate the signal matching degree of each individual, this module can gradually optimize the waveform parameters and ultimately obtain high-quality waveform coding. Genetic operations such as roulette selection, single-point crossover and mutation ensure the diversity of the population and the correctness of the evolutionary direction. Setting termination conditions and outputting the best individual as the final waveform coding when the conditions are met ensures the effectiveness and efficiency of the waveform optimization process.

[0115] An array processing module is used to perform spatial domain synthesis of waveform codes using a distributed MVDR algorithm to obtain anti-interference beamforming weights;

[0116] Furthermore, the waveform parameters p are extracted from the genetic algorithm output. i , generate the corresponding waveform matrix S;

[0117] Each waveform parameter set p i Contains pulse width and frequency offset information used to construct a representation of the baseband signal;

[0118] By performing statistical analysis on the input signal x(n), the covariance matrix R is calculated xx , the expression is:

[0119] R xx =E[x(n)x H (n)]

[0120] Where E[.] represents the mathematical expectation operation, x(n) is the received signal vector, and x H (n) is its conjugate transpose;

[0121] Generate a steering vector based on the desired direction and calculate the optimal weight using the MVDR criterion;

[0122] Applying the calculated optimal weights to the waveform matrix to obtain the interference-resistant beamforming result;

[0123] It should be noted that the distributed MVDR algorithm realizes the precise calculation of the anti-interference beamforming weights by calculating the covariance matrix and generating the steering vector. The method can not only provide maximum gain in the desired direction, but also effectively suppress interference signals in other directions. After the optimal weights calculated based on the MVDR criterion are applied to the waveform matrix, the generated anti-interference beamforming results can significantly improve the target detection capability and anti-interference performance of the radar system. Especially in complex environments, this spatial domain synthesis method can significantly improve the overall performance of the system.

[0124] The performance feedback module is used to evaluate the performance of the anti-interference beamforming weights using digital twin technology, obtain optimized parameters, and provide closed-loop feedback to the environmental perception module;

[0125] Furthermore, a digital twin model is constructed to generate a high-fidelity virtual environment based on the radar’s hardware configuration and operating environment characteristics;

[0126] Radar operating environment characteristics include terrain and meteorological conditions;

[0127] Apply the anti-interference beamforming weights derived from the MVDR algorithm in the generated virtual environment to generate a simulation data set;

[0128] The radar includes antenna array structure, signal processing module and noise model;

[0129] Simulate interference sources, target echo signals, and noise distribution in actual operating environments;

[0130] In a virtual environment, the input signal is composed of the superposition of target signal, interference signal and noise;

[0131] The anti-interference beamforming weights calculated by the MVDR algorithm are applied to the received signal in the virtual environment to obtain the output signal after beamforming, which is expressed as:

[0132]

[0133] in, is the conjugate transpose of the optimal weight, x(n) is the received signal simulated in the virtual environment;

[0134] After running the simulation in the virtual environment, record the output signal y after beamforming beam (n) and related performance indicators to generate simulation data sets;

[0135] The simulation data set includes time domain signals, frequency domain characteristics, and performance indicators;

[0136] Calculate its performance indicators based on simulation data;

[0137] By using comparative analysis methods, the performance indicators obtained by simulation are compared with the pre-set target values to identify the parameter set that needs to be optimized;

[0138] Applying genetic algorithm to the identified parameter set to obtain the optimized parameter set;

[0139] Through a closed-loop feedback mechanism, the optimized parameter set is sent back to the environmental perception module to update the radar configuration;

[0140] It should be noted that digital twin technology constructs a high-fidelity virtual environment to simulate various factors in the actual operating environment, such as terrain, meteorological conditions, and interference sources. It applies the anti-interference beamforming weights derived from the MVDR algorithm in the virtual environment to generate a simulation data set, and calculates performance indicators based on these data. It identifies the parameter set that needs to be optimized through comparative analysis methods, and applies genetic algorithms for optimization. Finally, the optimized parameter set is sent back to the environmental perception module through a closed-loop feedback mechanism. This method not only improves the flexibility and adaptability of the system, but also can quickly iterate and optimize system parameters without affecting actual operation, significantly improving the robustness and reliability of the system.

[0141] This embodiment also provides a computer device, which is applicable to the case of a radar anti-interference algorithm optimization platform, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the radar anti-interference algorithm optimization platform proposed in the above embodiment.

[0142] The computer device may be a terminal, comprising a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner may be achieved through Wi-Fi, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a button, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse.

[0143] This embodiment also provides a storage medium having a computer program stored thereon. When the program is executed by the processor, the radar anti-interference algorithm optimization platform proposed in the above embodiment is implemented. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0144] In summary, the present invention adopts broadband digital channelization technology to perform multi-dimensional sampling of radar received signals, thereby achieving the acquisition of original electromagnetic environment data, ensuring comprehensive coverage of radar received signals, and being able to accurately capture various frequency components and their energy distribution, providing a high-quality data basis for subsequent interference classification. By adaptively processing the feature parameter set through the improved RLS-NLMS hybrid algorithm, the optimization of the baseband signal is achieved, which can maintain good tracking capability and stability in a dynamically changing environment, effectively suppress the influence of interference signals, and improve the quality of the baseband signal. The waveform parameters of the baseband signal are inverted through a genetic algorithm to achieve the generation of optimal waveform coding. The global search capability of the genetic algorithm enables it to find the optimal or near-optimal waveform coding in a larger solution space, thereby improving the anti-interference capability and target detection accuracy of the radar system. The waveform coding is spatially synthesized through a distributed MVDR algorithm to achieve the calculation of anti-interference beamforming weights, so that the radar can still maintain high sensitivity to target signals in complex environments.

[0145] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. Radar anti-interference algorithm optimization platform, characterized by: include: Environmental perception module, interference classification module, dynamic filtering module, waveform optimization module, array processing module, and performance feedback module; The environment perception module is used to perform multi-dimensional sampling of radar received signals using broadband digital channelization technology to obtain original electromagnetic environment data; The interference classification module is used to analyze the original electromagnetic environment data using a deep convolutional neural network to obtain a feature parameter set; The dynamic filtering module is used to adaptively process the characteristic parameter set using an improved RLS-NLMS hybrid algorithm to obtain a baseband signal; The waveform optimization module is used to use a genetic algorithm to perform waveform parameter inversion on the baseband signal to obtain waveform coding; The array processing module is used to perform spatial domain synthesis on the waveform coding using a distributed MVDR algorithm to obtain anti-interference beamforming weights; The performance feedback module is used to evaluate the performance of anti-interference beamforming weights using digital twin technology, obtain optimized parameters and provide closed-loop feedback to the environmental perception module.

2. The radar anti-interference algorithm optimization platform according to claim 1, characterized in that: The broadband digital channelization technology is used to perform multi-dimensional sampling on the radar received signal to obtain the original electromagnetic environment data. The specific steps are as follows: The analog signal received by the radar is digitized using a high-speed analog-to-digital converter (ADC) to obtain a time-domain radar received signal; By applying Fast Fourier Transform (FFT), the time domain radar receiving signal is converted to the frequency domain, and the information of different frequency bands is separated to obtain the frequency domain information. Calculate the energy distribution of each frequency point in the frequency domain signal and mark the significant energy peak position as the key characteristic parameter, wherein the key characteristic parameter includes the spectrum peak and energy distribution; All the information obtained after the above steps is integrated into the original electromagnetic environment data.

3. The radar anti-interference algorithm optimization platform according to claim 2, characterized in that: The deep convolutional neural network is used to analyze the original electromagnetic environment data to obtain a characteristic parameter set. The specific steps are as follows: Perform format conversion and normalization on the original electromagnetic environment data to obtain normalized data; The normalized data is subjected to feature extraction by multi-convolution kernel scanning method to obtain the preliminary feature map z j ; The maximum pooling method is used to transform the initial feature map z j Perform dimensionality reduction processing to obtain a compressed feature map; The compressed feature map is flattened into a one-dimensional vector v through the fully connected layer operation, and the flattened one-dimensional vector v is subjected to global feature extraction and classification to obtain the classification result y; Based on the classification results, the category index with the highest probability is calculated and marked as the type of interference signal; Based on the interference signal type, combined with the input feature vector c of the fully connected layer and the corresponding weight matrix W f , extract the characteristic parameter set F related to the interference signal c .

4. The radar anti-interference algorithm optimization platform according to claim 3, characterized in that: The improved RLS-NLMS hybrid algorithm is used to adaptively process the characteristic parameter set to obtain the baseband signal. The specific steps are as follows: The parameters of the RLS-NLMS hybrid algorithm are configured through the initialization method to obtain the initial weight vector and covariance matrix; The recursive least squares method is used to quickly converge the initial weight vector and covariance matrix to obtain the preliminary weight update result; The initial weight update results are optimized in a steady state using the normalized least mean square algorithm to obtain the final weight vector; The hybrid strategy method fuses the results of the RLS and NLMS algorithms to obtain the optimized weight vector w final (n), and the optimized weight vector w is output by filtering method final (n) is applied to obtain the baseband signal.

5. The radar anti-interference algorithm optimization platform according to claim 4, characterized in that: The method uses a genetic algorithm to invert the waveform parameters of the baseband signal to obtain waveform coding, and the specific steps are: The waveform parameter range of the baseband signal is randomly sampled by the initialization population method to generate an initial waveform parameter set; Combining the waveform parameter vectors obtained by random sampling into an initial population; The signal matching degree of each individual waveform parameter is calculated through the fitness function to obtain the fitness value; Select high-quality individuals from the current population through the roulette wheel selection method to form the parent population; Through single-point crossover, two individuals in the parent population are genetically recombined to obtain offspring individuals, and the individuals in the offspring population are randomly disturbed to obtain mutated offspring individuals; Set the termination condition, evaluate the fitness of the current population, and obtain the final waveform code; When the termination condition is met, the individual with the highest fitness in the current population is output as the final waveform code, otherwise, the iteration continues.

6. The radar anti-interference algorithm optimization platform according to claim 5, characterized in that: The distributed MVDR algorithm is used to perform spatial domain synthesis on the waveform code to obtain the anti-interference beamforming weight. The specific steps are: Extracting waveform parameters p from genetic algorithm output i , generate the corresponding waveform matrix S; Each waveform parameter set p i Contains pulse width and frequency offset information used to construct a representation of the baseband signal; By performing statistical analysis on the input signal x(n), the covariance matrix R is calculated xx ; Generate a steering vector based on the desired direction and calculate the optimal weight using the MVDR criterion; The calculated optimal weights are applied to the waveform matrix to obtain the interference-resistant beamforming result.

7. The radar anti-interference algorithm optimization platform according to claim 6, characterized in that: The digital twin technology is used to evaluate the performance of the anti-interference beamforming weights, obtain the optimized parameters and feedback them to the environment perception module in a closed loop. The specific steps are as follows: Build a digital twin model to generate a high-fidelity virtual environment based on the radar's hardware configuration and operating environment characteristics; The radar operating environment characteristics include terrain and meteorological conditions; Apply the anti-interference beamforming weights derived from the MVDR algorithm in the generated virtual environment to generate a simulation data set; Calculate its performance indicators based on simulation data; By using comparative analysis methods, the performance indicators obtained by simulation are compared with the pre-set target values to identify the parameter set that needs to be optimized; Applying genetic algorithm to the identified parameter set to obtain the optimized parameter set; Through a closed-loop feedback mechanism, the optimized parameter set is sent back to the environmental perception module to update the radar configuration.

8. The radar anti-interference algorithm optimization platform according to claim 7, characterized in that: The anti-interference beamforming weights obtained from the MVDR algorithm are applied in the generated virtual environment to generate a simulation data set, as follows: The radar includes an antenna array structure, a signal processing module and a noise model; Simulate interference sources, target echo signals, and noise distribution in actual operating environments; In the virtual environment, the input signal is composed of a superposition of a target signal, an interference signal and noise; Apply the anti-interference beamforming weights calculated by the MVDR algorithm to the received signal in the virtual environment to obtain the output signal after beamforming; After running the simulation in the virtual environment, record the output signal y after beamforming beam (n) and related performance indicators to generate simulation data sets; The simulation data set includes time domain signals, frequency domain characteristics and performance indicators.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the radar anti-interference algorithm optimization platform according to any one of claims 1 to 8 are implemented.

10. A 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 anti-interference algorithm optimization platform according to any one of claims 1 to 8 are implemented.

Images