Dual-band same-aperture sparse layout optimization method and device, storage medium and terminal
Through the alternating optimization methods of convex optimization and genetic algorithm, the problem of poor optimization performance of sparse layout of dual-band antenna arrays is solved, and global and local joint optimization is achieved, and spectrum utilization and system performance are improved.
Patent Information
- Application Number
- CN202411175002.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-26
- Publication Date
- 2024-12-03
AI Technical Summary
In the prior art, the spatial sparse layout optimization performance of dual-band same-diameter antenna arrays is poor, the computational complexity is high, it is difficult to achieve real-time optimization, and it is easy to fall into local optimal solutions, resulting in the performance failure to reach the optimal level.
The alternating optimization method of convex optimization and genetic algorithm is adopted to optimize the excitation weight variables of high-frequency and low-frequency arrays, and optimize the conflicting position in combination with the genetic algorithm to achieve global and local joint optimization to obtain the optimized array.
It improves antenna layout and beam effect, improves spectrum utilization and system performance, and achieves better dual-band sparse layout optimization in the same diameter.
Smart Images

Figure CN119066822A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of antenna arrays, and relates to a dual-band same-caliber sparse layout optimization method, and in particular to a dual-band same-caliber sparse layout optimization method and device, storage medium and terminal. Background Art
[0002] In modern communication and radar systems, dual-band same-aperture antenna array design is widely used to significantly improve spectrum utilization and system performance. Dual-band same-aperture antenna array is an antenna system that can simultaneously support the transmission and reception of signals in two different frequency bands in the same physical antenna structure. This design helps to achieve multi-band communication and radar detection functions, and has many technical advantages and application values.
[0003] However, in practical applications, achieving spatial sparse layout of dual-bands with the same aperture faces many challenges, especially the complexity of the optimization process and the limitations of existing methods. Traditional optimization methods often have the following defects when dealing with such complex problems: in the sparse layout optimization of high-frequency and low-frequency arrays, the existing methods have high computational complexity, low efficiency when dealing with large arrays, and cannot meet real-time requirements; at the same time, traditional optimization methods are mostly local optimizations, which are easy to fall into local optimal solutions and difficult to find the optimal solution in the global scope. This limits the optimization effect of antenna array design and may result in performance failing to reach the optimal level. Summary of the invention
[0004] The purpose of the present invention is to provide a dual-band same-aperture sparse layout optimization method and device, storage medium and terminal, which are used to solve the technical problem of poor performance of dual-band same-aperture array spatial sparse layout optimization in the prior art.
[0005] In a first aspect, the present invention provides a dual-band same-caliber sparse layout optimization method, comprising:
[0006] Acquire a low-frequency array and a high-frequency array in a preset aperture space in an initial state, and combine the low-frequency array and the high-frequency array as an array to be optimized;
[0007] Using convex optimization and genetic algorithm to alternately optimize the array to be optimized to obtain an optimized array;
[0008] Determine whether the optimized array meets the preset conditions. If so, output the array layout of the optimized array as the optimization result. Otherwise, use the optimized array as the array to be optimized and perform alternate optimization again.
[0009] Wherein, alternately optimizing the array to be optimized by using convex optimization and genetic algorithm includes:
[0010] Convex optimization is performed on the array to be optimized to obtain a convex optimized array, and conflicting positions in the convex optimized array are optimally allocated using a genetic algorithm to obtain an optimized array.
[0011] In one embodiment of the present invention, performing convex optimization on the array to be optimized to obtain a convex optimized array includes:
[0012] Acquire a high-frequency array excitation weight variable based on a high-frequency array in the array to be optimized, and acquire a low-frequency array excitation weight variable based on a low-frequency array in the array to be optimized;
[0013] The first step convex optimization is performed on the high-frequency array excitation weight variable and the low-frequency array excitation weight variable to obtain the first step convex optimization result; the objective function of the first step convex optimization is to minimize the weight sum, and the constraint conditions of the first step convex optimization include that the gain of the array pattern in the target direction is a preset gain and the amplitude of the sidelobe area pattern is less than a preset amplitude;
[0014] A second step convex optimization is performed on the sparsity optimization variables in the first step convex optimization result to obtain the second step convex optimization result as the convex optimized array; the objective function of the second step convex optimization is to minimize the sparsity optimization variables, and the constraint conditions of the second step convex optimization include that the number of retained elements in the array is within a preset range.
[0015] In one embodiment of the present invention, using a genetic algorithm to optimally allocate conflicting positions in the convex optimized array to obtain an optimized array includes:
[0016] Obtaining preset algorithm parameters of a genetic algorithm, and obtaining the conflict position in the convexly optimized array as original conflict position data, wherein the preset algorithm parameters include population size, number of iterations, and mutation rate;
[0017] Based on the preset algorithm parameters and the original conflict position data, a genetic algorithm is used to iterate the conflict position allocation scheme for multiple times to obtain a conflict position allocation result;
[0018] Based on the conflict position allocation result, the directional diagram of the high-frequency array and the directional diagram of the low-frequency array in the convex optimized array are updated respectively to obtain an updated array as the optimized array.
[0019] In one embodiment of the present invention, based on the preset algorithm parameters and the original conflict position data, using a genetic algorithm to iterate the conflict position allocation scheme multiple times to obtain a conflict position allocation result includes:
[0020] Randomly generate an initial population of a conflict allocation scheme, and obtain the fitness value of each individual in the initial population based on a preset fitness function as an initial fitness set;
[0021] According to the preset algorithm parameters, the genetic algorithm is used to perform iterative optimization of multiple generations of the genetic algorithm on the initial population, and the optimization result of the last generation of the genetic algorithm is output as the conflict position allocation result;
[0022] Wherein, the genetic algorithm optimization of any generation includes:
[0023] Based on the preset fitness function, the fitness value of each individual in the previous generation population is obtained as a screening basis;
[0024] Screening the previous generation population according to the screening criteria to obtain a parent generation individual set corresponding to the current generation iteration;
[0025] Based on the parent generation individual set corresponding to the current generation iteration, the previous generation population is sequentially subjected to crossover operation, mutation operation and population update operation to obtain the current generation population as the conflict optimization result of the current generation.
[0026] In one embodiment of the present invention, obtaining the fitness value of a preset individual in a preset generation population based on a preset fitness function includes:
[0027] Acquire a basic matrix based on the original conflict position data;
[0028] Parsing the preset individuals into a distribution matrix, and obtaining a high-frequency array square difference and a low-frequency array square difference based on the distribution matrix;
[0029] Obtaining the reciprocal of the sum of the high frequency array square difference and the low frequency array square difference as a fitness value;
[0030] The preset generation population is any generation population, and the preset individual is any individual in the first generation population;
[0031] Getting the square difference of a preset array includes:
[0032] Obtaining the initial matrix, initial weight matrix and theoretical direction diagram corresponding to the preset array;
[0033] Based on the allocation matrix, the initial matrix, the initial weight matrix and the basic matrix, obtaining a corresponding allocated weight matrix;
[0034] According to the initial directional map and the weight matrix after allocation, obtaining the directional map after allocation corresponding to the preset matrix;
[0035] Obtaining a square difference between the assigned directional pattern and the theoretical directional pattern as the square difference of the preset array;
[0036] The preset array is a high-frequency array or a low-frequency array in the convex optimized array.
[0037] In one embodiment of the present invention, the screening method used in screening the previous generation population is tournament selection.
[0038] In one embodiment of the present invention, the preset condition includes that the number of times of the alternating optimization reaches a number threshold.
[0039] In a second aspect, the present invention further provides a dual-band same-caliber sparse layout optimization device, characterized in that it includes:
[0040] An initialization module, used to obtain a low-frequency array and a high-frequency array in a preset aperture space in an initial state, and to collect the low-frequency array and the high-frequency array as an array to be optimized;
[0041] An alternating optimization module, used for performing alternating optimization on the array to be optimized by using convex optimization and genetic algorithm to obtain an optimized array;
[0042] A condition judgment module is used to judge whether the optimized array meets the preset conditions. If so, the array layout of the optimized array is output as the optimization result. Otherwise, the optimized array is used as the array to be optimized and then alternately optimized again.
[0043] Wherein, alternately optimizing the array to be optimized by using convex optimization and genetic algorithm includes:
[0044] Convex optimization is performed on the array to be optimized to obtain a convex optimized array, and conflicting positions in the convex optimized array are optimally allocated using a genetic algorithm to obtain an optimized array.
[0045] In a third aspect, the present invention further provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the dual-band same-caliber sparse layout optimization method as described above.
[0046] In a fourth aspect, the present invention further provides a terminal, comprising a processor and a memory, wherein the memory is communicatively connected to the processor;
[0047] The memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory, so that the terminal executes the dual-band same-caliber sparse layout optimization method as described above.
[0048] As described above, the dual-band same-caliber sparse layout optimization method and device, storage medium and terminal described in the present invention have the following beneficial effects:
[0049] By using convex optimization and genetic algorithm for alternating optimization, global and local joint optimization is achieved, thereby achieving the best antenna layout and beam effect, which can improve spectrum utilization and system performance. BRIEF DESCRIPTION OF THE DRAWINGS
[0050] Figure 1 A schematic flow chart of the dual-band same-caliber sparse layout optimization method described in an embodiment of the present invention is shown.
[0051] Figure 2 A schematic diagram of the process of two-step convex optimization in the dual-band same-caliber sparse layout optimization method described in an embodiment of the present invention is shown.
[0052] Figure 3 A schematic diagram of a process of using a genetic algorithm to perform conflict allocation in a dual-band same-caliber sparse layout optimization method according to an embodiment of the present invention is shown.
[0053] Figure 4 A schematic diagram of a flow chart of any one genetic algorithm in the dual-band same-caliber sparse layout optimization method described in an embodiment of the present invention is shown.
[0054] Figure 5 The sparse array pattern in the low frequency band of the 144-element dual-band same-aperture array described in an embodiment of the present invention is shown.
[0055] Figure 6 The sparse array pattern in the high frequency band of the 144-element dual-band same-aperture array described in an embodiment of the present invention is shown.
[0056] Figure 7 A schematic diagram of the structure of a dual-band same-caliber sparse layout optimization device according to an embodiment of the present invention is shown.
[0057] Figure 8 A schematic diagram of the structure of a terminal according to an embodiment of the present invention is shown. DETAILED DESCRIPTION
[0058] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.
[0059] The following will explain in detail the principles and implementation methods of the dual-band same-caliber sparse layout optimization method and device, storage medium and terminal of this embodiment, so that technical personnel in this field can understand the dual-band same-caliber sparse layout optimization method, device, storage medium and terminal of this embodiment without creative work.
[0060] In order to solve the above technical problems existing in the prior art, an embodiment of the present invention provides a dual-band same-caliber sparse layout optimization method.
[0061] Figure 1 FIG. 1 shows a flow chart of a dual-band same-caliber sparse layout optimization method according to an embodiment of the present invention, referring to FIG. Figure 1 As shown, the dual-band same-caliber sparse layout optimization method of the embodiment of the present invention mainly includes the following steps:
[0062] Step S100: obtaining a low-frequency array and a high-frequency array in a preset aperture space in an initial state, and combining the low-frequency array and the high-frequency array as an array to be optimized.
[0063] Specifically, the basic parameters of the low-frequency array and the high-frequency array in the preset aperture space are initialized, the basic parameters including frequency, wavelength, array size and scanning angle, the beam and pattern matrix of the high-frequency array are calculated based on the basic parameters, and then the wave number and pattern matrix of the low-frequency array are calculated based on the basic parameters. At this point, the initialization of the low-frequency array and the high-frequency array can be completed, that is, the low-frequency array and the high-frequency array in the preset aperture space in the initial state are obtained, so that the low-frequency array and the high-frequency array together constitute the array to be optimized in the preset aperture space.
[0064] Step S200: alternately optimizing the array to be optimized using convex optimization and genetic algorithm to obtain an optimized array.
[0065] Specifically, alternately optimizing the array to be optimized using convex optimization and genetic algorithm includes: performing convex optimization on the array to be optimized to obtain a convex optimized array, and using genetic algorithm to optimally allocate conflicting positions in the convex optimized array to obtain an optimized array.
[0066] Optionally, Figure 2 FIG. 1 is a flow chart showing a two-step convex optimization process in the dual-band same-caliber sparse layout optimization method according to an embodiment of the present invention. Figure 2 As shown, performing convex optimization on the array to be optimized to obtain the convex optimized array includes the following steps:
[0067] Step S201: obtaining a high-frequency array excitation weight variable based on a high-frequency array in the array to be optimized, and obtaining a low-frequency array excitation weight variable based on a low-frequency array in the array to be optimized.
[0068] Specifically, each unit in the antenna array will emit a fixed beam, that is, excitation. The excitation weight variable is the weight assigned to the excitation. Multiplying the excitation weight variable and the corresponding excitation can obtain the output beam of the corresponding unit. The high-frequency array includes multiple units, so the high-frequency array excitation weight variable also includes the excitation weight variables of multiple units. The high-frequency array excitation weight variable is a matrix, in which each value represents the excitation weight variable of a unit. Similarly, the low-frequency array excitation weight variable also includes the excitation weight variables of multiple units, and the low-frequency array excitation weight variable is also a matrix.
[0069] Step S202: Perform the first step convex optimization on the high-frequency array excitation weight variables and the low-frequency array excitation weight variables to obtain the first step convex optimization results; the objective function of the first step convex optimization is to minimize the weight sum, and the constraints of the first step convex optimization include that the gain of the array pattern in the target direction is a preset gain and the amplitude of the sidelobe area pattern is less than a preset amplitude.
[0070] Specifically, the variables of the first step of convex optimization are defined (i.e., determining the decision variables) as high-frequency array excitation, high-frequency array excitation weight variables, low-frequency array excitation, and low-frequency array excitation weight variables; the objective function is to minimize the weight sum, and the constraint condition is that the array pattern of the array to be optimized has a preset gain in the target direction, and the amplitude of the side lobe area of the array to be optimized is less than the preset amplitude, thereby preventing the generation of high side lobes while ensuring the strength of the main beam, reducing interference and noise. Furthermore, the preset gain is 1. Among them, the specific calculation method of the weight sum is to multiply each excitation by the corresponding excitation weight variable, and then sum all the products to obtain the weight sum, that is, the following formula:
[0071] Weight sum = high frequency array excitation * high frequency array excitation weight variable + low frequency array excitation * low frequency array excitation weight variable
[0072] Step S203: performing a second step convex optimization on the sparsity optimization variables in the first step convex optimization result to obtain the second step convex optimization result as the convex optimized array; the objective function of the second step convex optimization is to minimize the sparsity optimization variables, and the constraint condition of the second step convex optimization includes that the number of retained elements in the array is within a preset range.
[0073] The second step of convex optimization uses the result of the first step of convex optimization as the data basis for further processing. Sparse optimization variables are a key concept in the optimization of sparse layout of antenna arrays. They are mainly used to indicate which antenna units are selected or activated to achieve the purpose of sparse layout. The variable definition (that is, the determination of decision variables) of the second step of convex optimization is high-frequency array sparse optimization variables and low-frequency array sparse optimization variables; the objective function is to minimize the absolute value weighted sum of the sparse optimization variables. Specifically, the absolute value weighted sum of the sparse optimization variables is:
[0074] Min_xs=g_nt*abs(wg)+k_nt*abs(wk)
[0075] Among them, Min_xs is the absolute value weighted sum of the sparsity optimization variables, wg is the high-frequency array excitation weight variable, g_nt is the high-frequency array sparsity optimization variable, wk is the low-frequency array excitation weight variable, and k_nt is the low-frequency array sparsity optimization variable. The constraint condition is that the value of the sparsity optimization variable is greater than 0 and less than 1. And the number of retained elements in the array is within the preset range, so as to ensure that the number of activated antennas in the array is within a fixed range. The array here refers to the overall array targeted by the second step convex optimization, not the high-frequency array or the low-frequency array. .
[0076] Optionally, Figure 3 FIG. 1 shows a schematic diagram of a process for conflict allocation using a genetic algorithm in a dual-band same-caliber sparse layout optimization method according to an embodiment of the present invention, with reference to FIG. Figure 3 As shown, using a genetic algorithm to optimally allocate conflicting positions in the convex optimized array to obtain the optimized array includes the following steps:
[0077] Step S211: obtaining preset algorithm parameters of the genetic algorithm, obtaining the conflict position in the array after convex optimization as the original conflict position data, the preset algorithm parameters including population size, number of iterations, and mutation rate;
[0078] The preset algorithm parameters of the genetic algorithm include population size, number of iterations, and mutation rate. The conflict positions in the array after convex optimization are obtained as original conflict position data, and the subsequent genetic algorithm mainly iteratively optimizes the allocation of conflict positions in the array after convex optimization based on the original conflict position data.
[0079] In one implementation of step S211, conflict detection is used to obtain conflict positions in the convexly optimized array. Specifically, firstly, a corresponding retention matrix is generated based on the high-frequency array sparsity optimization variables and the low-frequency array sparsity optimization variables in the convexly optimized array; then, a conflict matrix is generated after screening based on the retention matrix, as shown in the following code:
[0080] gg=reshape(g,xline_num,yline_num)
[0081] kk=reshape(k,xline_num,yline_num)
[0082] GG=(gg<=eplison); / / 0 is discarded, 1 is retained
[0083] KK=(kk<=eplison);
[0084] Cd=(GG&KK); / / 0 means no position conflict, 1 means position conflict
[0085] Among them, g represents the high-frequency array sparsity optimization variable in the array after convex optimization, k represents the low-frequency array sparsity optimization variable in the array after convex optimization, xline_num and yline_num represent the array size of the array after convex optimization, reshape represents the matrix reshaping operation, gg represents the high-frequency retention matrix, and kk represents the low-frequency retention matrix. eplison represents the judgment threshold, GG represents the high-frequency screening result after screening the high-frequency retention matrix gg based on the judgment threshold eplison, KK represents the low-frequency screening result after screening the low-frequency retention matrix kk based on the judgment threshold eplison, Cd represents the conflicting position in the array after convex optimization, & represents the operation of logical AND, if both are 1 or 0 in the same position, it means that the position is conflicting, otherwise there is no conflict. In the conflicting position Cd, if the number corresponding to a unit is 0, it means that the corresponding position of the unit is not conflicting, and if the number is 1, it means that there is a conflict. Optionally, the judgment threshold Eplison is 1e-5.
[0086] Step S212: Based on preset algorithm parameters and original conflict position data, a genetic algorithm is used to iterate the conflict position allocation scheme multiple times to obtain a conflict position allocation result;
[0087] By simulating the process of natural selection and genetic variation, the genetic algorithm can search for the optimal antenna conflict position allocation scheme on a global scale. Multiple iterations help to gradually improve the allocation scheme and eventually find the optimal allocation scheme. The conflict optimization result finally output by the genetic algorithm after multiple iterations is used as the conflict position allocation result.
[0088] In one implementation of step S212, based on preset algorithm parameters and original conflict position data, a genetic algorithm is used to iterate the conflict position allocation scheme for multiple times to obtain the conflict position allocation result, including: randomly generating a primary population of a conflict allocation scheme, and obtaining the fitness value of each individual in the primary population based on a preset fitness function as a primary fitness set, specifically, the primary population is a binary matrix. According to the preset algorithm parameters, the genetic algorithm is used to iterate and optimize the primary population for multiple generations of genetic algorithms, and the conflict optimization result of the last generation of genetic algorithms is output as the conflict position allocation result, where the number of iterations of the genetic algorithm is the number of iterations in the preset algorithm parameters.
[0089] Figure 4 A schematic diagram of the process flow of any generation of genetic algorithm in the dual-band same-caliber sparse layout optimization method according to an embodiment of the present invention is shown, referring to Figure 4 As shown, any generation of genetic algorithm optimization includes the following steps:
[0090] S2121: Based on a preset fitness function, the fitness value of each individual in the previous generation population is obtained as a screening basis.
[0091] In a genetic algorithm, it is necessary to calculate the fitness of each individual, and the fitness function is defined according to the goal of the problem. The fitness function is used to evaluate the quality of the individual. In this embodiment, a preset fitness function is used to optimize the genetic algorithm. It should be noted that the previous generation population corresponding to the optimization of the first generation genetic algorithm is the first generation population, the optimization result of the first generation genetic algorithm optimization is the first generation population, the previous generation population corresponding to the optimization of the second generation genetic algorithm is the first generation population, and so on. For example, the previous generation population corresponding to the optimization of the tenth generation genetic algorithm is the ninth generation population, and the optimization result of the optimization of the tenth generation genetic algorithm is the tenth generation population.
[0092] Optionally, obtaining the fitness value of a preset individual in a preset generation population based on a preset fitness function includes:
[0093] Based on the original conflict position data, a basic matrix is obtained, which represents the elements involved in the allocation during the optimization process, that is, the conflict position to be optimized; the preset individual is parsed into an allocation matrix, and the high-frequency array square difference and the low-frequency array square difference are obtained based on the allocation matrix; the inverse of the sum of the high-frequency array square difference and the low-frequency array square difference is obtained as the fitness value. The specific implementation method of parsing the preset individual into an allocation matrix is to reshape the preset individual into a matrix of the form [N,N] as an allocation reshaping matrix, and then round off each value in the allocation reshaping matrix to obtain the allocation matrix, where N is the size of the directional pattern matrix (the directional pattern matrix is a square of N*N, and the directional pattern matrix size of the high-frequency array and the low-frequency array is the same).
[0094] The preset generation population is any generation population, the preset individual is any individual in the first generation population, and obtaining the square difference of the preset array includes:
[0095] Digital acquisition: Get the initial matrix, initial weight matrix and theoretical direction diagram corresponding to the preset array. Refer to the following formula for details:
[0096] matrix C = Cd
[0097] mapA=aa*InitWA
[0098] Among them, matrixC represents the initial matrix corresponding to the preset array, Cd represents the conflict position in the array after convex optimization (that is, the conflict matrix of the array after convex optimization), InitWA is the matrix formed by the excitation weight variables of the preset array, mapA represents the theoretical directional map, aa represents the array factor, and the array factor is the response of each element in the array to different directions.
[0099] Weight allocation: Based on the allocation matrix, initial matrix, initial weight matrix and basic matrix, obtain the corresponding weight matrix after allocation. For details, refer to the following formula:
[0100] finalWA=(InitialmatrixA-matrixC.*(1-allocationMatrix)).*InitWA
[0101] Among them, finalWA represents the weight matrix after allocation, matrixC represents the basic matrix, InitialmatrixA represents the initial matrix, InitWA represents the initial weight matrix, allocationmatrix represents the allocation matrix, and .* represents the dot product operation of the matrix.
[0102] Directional pattern acquisition: According to the initial directional pattern and the weight matrix after allocation, the allocated directional pattern corresponding to the preset matrix is obtained. For details, refer to the following formula:
[0103] actualMapA=initialMapA*finalWA(:)
[0104] Among them, actualMapA represents the directional map after allocation, initialMapA represents the initial directional map, and finalWA represents the weight matrix after allocation.
[0105] Square difference calculation: obtain the square difference between the assigned directional pattern and the theoretical directional pattern as the square difference of the preset array;
[0106] The preset array is the high frequency array or the low frequency array in the array after convex optimization. According to the above steps, the high frequency array square difference and the low frequency array square difference of the preset array can be obtained respectively.
[0107] S2122: Filter the previous generation population according to the screening criteria to obtain the parent generation individual set corresponding to the current generation iteration.
[0108] Specifically, the parent array is initialized, and the length of the parent array is equal to the length of the matrix formed by the previous generation population (that is, the number of individuals in the population). The individuals in the previous generation population are screened multiple times, and each screening selects an individual with the highest fitness value and adds the individual to the parent array. The number of screenings is equal to the number of individuals in the previous generation population. The parent array after screening is used as the parent individual set. This ensures that excellent individuals are selected for reproduction and excellent genes are retained.
[0109] Optionally, the screening method used in the previous generation population is the tournament selection method. The tournament selection method randomly selects multiple individuals for comparison. The tournament selection method can avoid premature convergence of individuals, help maintain the diversity of the population, and thus prevent falling into a local optimal solution. Furthermore, the selection parameter of the tournament selection method is 5, that is, 5 random individuals are randomly obtained each time the screening is performed and the individual with the highest fitness value is selected to be added to the parent generation array.
[0110] S2123: Based on the parent generation individual set corresponding to the current generation iteration, the previous generation population is sequentially subjected to crossover operation, mutation operation and population update operation to obtain the current generation population as the conflict optimization result of the current generation.
[0111] Crossover operation: Initialize a zero matrix with the same size as the parent individual set, randomly select a crossover point, use the number of individuals in the parent individual set as the total number of loops, and use 2 as the step size of the loop to perform crossover value selection. In each loop, the genes of the preset number of individuals in the zero matrix are composed of the genes before the crossover point of the preset number of individuals in the parent individual set and the genes after the crossover point of the preset number of individuals plus 1 in the parent individual set, and the genes of the preset number of individuals plus 1 in the zero matrix are composed of the genes before the crossover point of the preset number of individuals plus 1 in the parent individual set and the genes after the crossover point of the preset number of individuals in the parent individual set. For example, when in the i-th loop, the specific code is as shown in the following:
[0112] parent1=parents(i,:)
[0113] parent2=parents(i+1,:)
[0114] offspring(i) = the first crossoverPoint gene of parent1 + the subsequent gene of parent2
[0115] offspring(i+1) = the first crossoverPoint gene of parent2 + the subsequent gene of parent1
[0116] Among them, parents represents the parent individual set, parent1 represents the i-th individual in the parent individual set parents, parent2 represents the i+1-th individual in the parent individual set parents, offspring(i) represents the i-th individual in the zero matrix, offspring(i+1) represents the i+1-th individual in the zero matrix, and crossoverPoint represents the crossover point.
[0117] In each loop, the zero matrix is written in an overwriting manner, and the zero matrix after the loop is completed is used as the offspring individual set. Through the crossover operation, some genes (elements in the matrix) of the selected parent individuals are exchanged to generate new offspring individuals. The crossover operation introduces genetic diversity and helps to explore different solution spaces.
[0118] Mutation operation: Initialize the mutated population to obtain the initial data of the mutated population; for each individual in the initial data of the mutated population, perform a gene flip operation (i.e., change 0 to 1, and 1 to 0), and use the initial data of the mutated population after the flip operation as the mutated individual set. The mutation operation can further increase the diversity of the population and avoid falling into a local optimal solution.
[0119] Update operation: merge the mutated individual set with the parent individual set to form a new population.
[0120] The above steps S2051-S2053 are iterated multiple times. Each generation of genetic algorithm optimization will improve the allocation scheme. The result of step S205 is the conflict allocation result. The conflict allocation scheme represented by the result is close to the global optimal solution.
[0121] Specifically, the final conflict allocation result is the conflict optimization result of the genetic algorithm after the last iteration, that is, the last generation of population.
[0122] Step S213: based on the conflict position allocation result, the directional diagram of the high frequency array and the directional diagram of the low frequency array in the convex optimized array are updated respectively to obtain an updated array as the optimized array.
[0123] Based on the conflict position allocation results, the directional diagrams of the high-frequency array and the low-frequency array in the convex optimization array are updated respectively. The conflict position allocation results need to be parsed into a matrix form to obtain the conflict position allocation results in matrix form. Then, the conflict position allocation results in matrix form are used to obtain the intermediate allocation matrix, which includes the high-frequency intermediate allocation matrix and the low-frequency intermediate allocation matrix. The method of obtaining the intermediate allocation matrix refers to the following formula and code:
[0124] bestAllocation=population(idx,:)
[0125] D1=thebestAllocation.*matrixC
[0126] D2=(1-thebestAllocation).*matrixC
[0127] Among them, bestAllocation represents the optimal allocation result, population represents the conflict allocation result, idx represents the index corresponding to the individual with the smallest fitness value in the conflict allocation result population. matrixC represents the basic matrix, thebestAllocation represents the conflict position allocation result in matrix form (that is, reshaping bestAllocation into a matrix of [N,N] form), D1 represents the high-frequency intermediate allocation matrix, and D2 represents the low-frequency intermediate allocation matrix.
[0128] Based on the intermediate allocation matrix and the theoretical directional pattern, the actual directional pattern is obtained. For details, refer to the following formula:
[0129] xfinalWA=(xInitMapA-D2).*reshape(xinitWA,[N,N])
[0130] actualMapA=xinitMapA*xfinalWA(:)
[0131] xfinalWB=(xInitMapB-D1).*reshape(xinitWB,[N,N])
[0132] actualMapB=xinitMapB*xfinalWB(:)
[0133] Among them, xfinalWA represents the weight matrix after allocation corresponding to the high-frequency array in the conflict allocation result, xinitWA represents the initial weight matrix corresponding to the high-frequency array in the conflict allocation result, xinitMapA represents the initial directional pattern corresponding to the high-frequency array in the conflict allocation result; xfinalWB represents the weight matrix after allocation corresponding to the low-frequency array in the conflict allocation result, xinitWB represents the initial weight matrix corresponding to the low-frequency array in the conflict allocation result, N represents the size of the directional pattern matrix, xinitMapB represents the initial directional pattern corresponding to the low-frequency array in the conflict allocation result. actualMapA represents the directional pattern of the high-frequency array in the optimized array, actualMapB represents the directional pattern of the low-frequency array in the optimized array, and actualMapA and actualMapB together constitute the array directional pattern of the optimized array, as the optimized array.
[0134] Step S300: determining whether the optimized array meets the preset conditions. If so, outputting the array layout of the optimized array as the optimization result. Otherwise, taking the optimized array as the array to be optimized, performing alternate optimization again.
[0135] Optionally, the preset condition includes that the number of alternating optimizations reaches a threshold number.
[0136] Furthermore, the present application verifies and illustrates the technical effect through the following simulation examples.
[0137] The low-frequency band planar array of this embodiment is placed in the XOY plane, the size of the low-frequency array and the high-frequency array are both 12*12, the array element spacing dx=dy=6 / 1000, the frequency of the low-frequency array is 18GHz, the frequency of the high-frequency array is 36GHz, and the wavelength of the low-frequency array and the high-frequency array is 0.3. Specifically, Figure 5 The sparse array pattern of the low frequency band in the 144-element dual-band same-aperture array according to the embodiment of the present invention is shown. Figure 6 The sparse array pattern of the high frequency band of the 144-element dual-band same-aperture array described in the embodiment of the present invention is shown. Figure 5 and Figure 6 It can be seen that under the set parameter conditions, this embodiment can optimize the radiation pattern to meet the specific sidelobe level requirements, and can handle the conflicts between array elements, optimize the weight distribution at different frequencies, and flexibly and efficiently realize the global optimization of the dual-band same-aperture sparse array.
[0138] The protection scope of the dual-band same-caliber sparse layout optimization method of the embodiment of the present invention is not limited to the execution order of the steps listed in this embodiment. All solutions implemented by adding, reducing, and replacing steps in the prior art based on the principles of the present invention are included in the protection scope of the present invention.
[0139] The dual-band same-aperture sparse layout optimization method of the embodiment of the present invention uses a genetic algorithm to globally optimize the dual-band same-aperture array, and uses convex optimization to locally optimize the dual-band same-aperture array. By using convex optimization and genetic algorithm for alternating optimization, global and local joint optimization is achieved, thereby achieving the best antenna layout and beam effect, which can improve spectrum utilization and system performance.
[0140] In order to solve the above technical problems existing in the prior art, an embodiment of the present invention further provides a dual-band same-caliber sparse layout optimization device.
[0141] Figure 7 The schematic diagram of the structure of the dual-band same-caliber sparse layout optimization device according to the embodiment of the present invention is shown. Figure 7 As shown, the dual-band same-caliber sparse layout optimization device according to the embodiment of the present invention includes an initialization module, an alternating optimization module and a condition judgment module.
[0142] An initialization module is used to obtain a low-frequency array and a high-frequency array in a preset aperture space in an initial state, and to collect the low-frequency array and the high-frequency array as an array to be optimized;
[0143] An alternating optimization module is used to perform alternating optimization on the array to be optimized using convex optimization and genetic algorithm to obtain an optimized array;
[0144] A condition judgment module is used to judge whether the optimized array meets the preset conditions. If so, the array layout of the optimized array is output as the optimization result. Otherwise, the optimized array is used as the array to be optimized and then alternately optimized again.
[0145] Among them, using convex optimization and genetic algorithm to alternately optimize the array to be optimized includes:
[0146] The array to be optimized is convexly optimized to obtain a convexly optimized array, and the conflicting positions in the convexly optimized array are optimally allocated using a genetic algorithm to obtain an optimized array.
[0147] The dual-band same-aperture sparse layout optimization device of the embodiment of the present invention uses a genetic algorithm to globally optimize the dual-band same-aperture array, and uses convex optimization to locally optimize the dual-band same-aperture array. By using convex optimization and genetic algorithm for alternating optimization, global and local joint optimization is achieved, thereby achieving the best antenna layout and beam effect, which can improve spectrum utilization and system performance.
[0148] In order to solve the above-mentioned technical problems existing in the prior art, an embodiment of the present invention further provides a storage medium on which a computer program is stored, characterized in that when the program is executed by a processor, all steps of the dual-band same-caliber sparse layout optimization method of the embodiment are implemented.
[0149] The specific steps of the dual-band same-caliber sparse layout optimization method and the beneficial effects obtained by applying the readable storage medium provided by the embodiment of the present invention are the same as those in the above embodiment and will not be described in detail here.
[0150] A person of ordinary skill in the art can understand that all or part of the steps in the method for implementing the above-mentioned embodiment can be completed by instructing a processor through a program, and the program can be stored in a computer-readable storage medium, and the storage medium is a non-transitory medium, such as a random access memory, a read-only memory, a flash memory, a hard disk, a solid-state hard disk, a magnetic tape, a floppy disk, an optical disc, and any combination thereof. The above-mentioned storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that includes one or more available media. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a digital video disc (DVD)), or a semiconductor medium (for example, a solid-state disk (SSD)), etc.
[0151] In order to solve the above technical problems existing in the prior art, an embodiment of the present invention further provides a terminal. Figure 8 A schematic diagram of the structure of the terminal according to an embodiment of the present invention is shown, referring to Figure 8 As shown, the terminal of an embodiment of the present invention includes a processor and a memory, and the memory is communicatively connected to the processor; the memory is used to store computer programs, and the processor is used to execute the computer programs stored in the memory, so that the terminal executes all steps of the dual-band same-caliber sparse layout optimization method of the above embodiment.
[0152] The specific steps of the dual-band same-caliber sparse layout optimization method and the beneficial effects obtained by applying the terminal provided by the embodiment of the present invention are the same as those in the above embodiment and will not be described in detail here.
[0153] It should be noted that the memory may include random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage. Similarly, the processor may also be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components.
[0154] Although the embodiments disclosed in the present invention are as above, the contents described are only embodiments adopted for facilitating the understanding of the present invention and are not intended to limit the present invention. Any technician in the technical field to which the present invention belongs can make any modifications and changes in the form and details of the implementation without departing from the spirit and scope disclosed in the present invention, but the protection scope of the present invention shall still be subject to the scope defined in the attached claims.
Claims
1. A dual-band same-caliber sparse layout optimization method, comprising: Acquire a low-frequency array and a high-frequency array in a preset aperture space in an initial state, and combine the low-frequency array and the high-frequency array as an array to be optimized; Using convex optimization and genetic algorithm to alternately optimize the array to be optimized to obtain an optimized array; Determine whether the optimized array meets the preset conditions. If so, output the array layout of the optimized array as the optimization result. Otherwise, use the optimized array as the array to be optimized and perform alternate optimization again. Wherein, alternately optimizing the array to be optimized by using convex optimization and genetic algorithm includes: Convex optimization is performed on the array to be optimized to obtain a convex optimized array, and conflicting positions in the convex optimized array are optimally allocated using a genetic algorithm to obtain an optimized array.
2. The method according to claim 1, characterized in that Performing convex optimization on the array to be optimized to obtain a convex optimized array includes: Acquire a high-frequency array excitation weight variable based on a high-frequency array in the array to be optimized, and acquire a low-frequency array excitation weight variable based on a low-frequency array in the array to be optimized; The first step convex optimization is performed on the high-frequency array excitation weight variable and the low-frequency array excitation weight variable to obtain the first step convex optimization result; the objective function of the first step convex optimization is to minimize the weight sum, and the constraint conditions of the first step convex optimization include that the gain of the array pattern in the target direction is a preset gain and the amplitude of the sidelobe area pattern is less than a preset amplitude; A second step convex optimization is performed on the sparsity optimization variables in the first step convex optimization result to obtain the second step convex optimization result as the convex optimized array; the objective function of the second step convex optimization is to minimize the sparsity optimization variables, and the constraint conditions of the second step convex optimization include that the number of retained elements in the array is within a preset range.
3. The method according to claim 1, characterized in that: Using a genetic algorithm to optimally allocate conflicting positions in the convex optimized array to obtain an optimized array includes: Obtaining preset algorithm parameters of a genetic algorithm, and obtaining the conflict position in the convexly optimized array as original conflict position data, wherein the preset algorithm parameters include population size, number of iterations, and mutation rate; Based on the preset algorithm parameters and the original conflict position data, a genetic algorithm is used to iterate the conflict position allocation scheme for multiple times to obtain a conflict position allocation result; Based on the conflict position allocation result, the directional diagram of the high-frequency array and the directional diagram of the low-frequency array in the convex optimized array are updated respectively to obtain an updated array as the optimized array.
4. The method according to claim 3, characterized in that Based on the preset algorithm parameters and the original conflict position data, the conflict position allocation scheme is iterated multiple times using a genetic algorithm to obtain a conflict position allocation result, including: Randomly generate an initial population of a conflict allocation scheme, and obtain the fitness value of each individual in the initial population based on a preset fitness function as an initial fitness set; According to the preset algorithm parameters, the genetic algorithm is used to perform iterative optimization of multiple generations of genetic algorithms on the initial population, and the conflict optimization result of the last generation of genetic algorithms is output as the conflict position allocation result; Wherein, the genetic algorithm optimization of any generation includes: Based on the preset fitness function, the fitness value of each individual in the previous generation population is obtained as a screening basis; Screening the previous generation population according to the screening criteria to obtain a parent generation individual set corresponding to the current generation iteration; Based on the parent generation individual set corresponding to the current generation iteration, the previous generation population is sequentially subjected to crossover operation, mutation operation and population update operation to obtain the current generation population as the conflict optimization result of the current generation.
5. The method according to claim 4, characterized in that Obtaining the fitness value of a preset individual in a preset generation population based on a preset fitness function includes: Acquire a basic matrix based on the original conflict position data; Parsing the preset individuals into a distribution matrix, and obtaining a high-frequency array square difference and a low-frequency array square difference based on the distribution matrix; Obtaining the reciprocal of the sum of the differences between the square of the high frequency array and the square of the low frequency array as a fitness value; The preset generation population is any generation population, and the preset individual is any individual in the first generation population; obtaining the square difference of the preset array includes: Obtaining the initial matrix, initial weight matrix and theoretical direction diagram corresponding to the preset array; Based on the allocation matrix, the initial matrix, the initial weight matrix and the basic matrix, obtaining a corresponding allocated weight matrix; According to the initial directional map and the weight matrix after allocation, obtaining the directional map after allocation corresponding to the preset matrix; Obtaining a square difference between the assigned directional pattern and the theoretical directional pattern as the square difference of the preset array; The preset array is a high-frequency array or a low-frequency array in the convex optimized array.
6. The method according to claim 4, characterized in that The screening method used in the previous generation population is tournament selection.
7. The method according to claim 1, characterized in that The preset condition includes that the number of times of the alternating optimization reaches a number threshold.
8. A dual-band same-caliber sparse layout optimization device, characterized in that: include: An initialization module, used to obtain a low-frequency array and a high-frequency array in a preset aperture space in an initial state, and to collect the low-frequency array and the high-frequency array as an array to be optimized; An alternating optimization module, used for performing alternating optimization on the array to be optimized by using convex optimization and genetic algorithm to obtain an optimized array; A condition judgment module is used to judge whether the optimized array meets the preset conditions. If so, the array layout of the optimized array is output as the optimization result. Otherwise, the optimized array is used as the array to be optimized and then alternately optimized again. Wherein, alternately optimizing the array to be optimized by using convex optimization and genetic algorithm includes: Convex optimization is performed on the array to be optimized to obtain a convex optimized array, and conflicting positions in the convex optimized array are optimally allocated using a genetic algorithm to obtain an optimized array.
9. A storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the dual-band same-caliber sparse layout optimization method according to any one of claims 1 to 7 is implemented.
10. A terminal, characterized in that: It includes a processor and a memory, and the memory is communicatively connected to the processor; the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory, so that the terminal executes the dual-band same-caliber sparse layout optimization method as described in any one of claims 1 to 7.
Citation Information
Cited By
Comprehensive method of dual-band mixed sparse array beam
CN119740485A
A comprehensive method for dual-band hybrid sparse array beamforming
CN119740485B
