Limited multi-objective optimization method applied to antenna design
Through the combination of multi-port network and non-dominant sorting genetic algorithm, the problem of not meeting performance constraints in multi-objective optimization in antenna design is solved, and a fast and effective antenna design is achieved, meeting the return loss requirements and improving signal transmission efficiency is improved. and reliability.
Patent Information
- Application Number
- CN202411031496.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2024-05-29
- Filing Date
- 2024-07-30
- Publication Date
- 2024-11-12
AI Technical Summary
The existing multi-objective optimization method for antenna design has solutions that do not meet the needs when considering actual performance constraints, and the calculation cost is high, so it is impossible to quickly obtain a design solution that meets the return loss requirements.
Using a multi-port network method to establish a physical model of antenna reflection coefficient, combined with the non-dominant sorting genetic algorithm II (NSGA-II), two objective function optimization algorithms are used to ensure that the Pareto frontier solution meets the return loss requirements of the antenna and reduces electromagnetics Simulation calculation to improve optimization efficiency.
It realizes finding multiple different trade-off solutions while meeting the antenna design requirements, reducing the cost of optimization time, and all solutions meet the antenna performance requirements and improving signal transmission efficiency and reliability.
Smart Images

Figure CN118940633A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of antenna technology, and in particular to a constrained multi-objective optimization method applied to antenna design. Background Art
[0002] Antenna design is a very challenging task because it is essentially a typical multi-objective optimization problem. During the design process, engineers need to carefully weigh multiple interrelated and potentially conflicting objectives. First, the choice of antenna size is crucial to achieving stability, efficient energy transmission, and wide frequency coverage. A suitable size not only ensures that the antenna can work stably in various environments, but also achieves efficient energy conversion and signal transmission. On the other hand, return loss, as one of the important indicators of antenna performance, must be given sufficient attention. Return loss refers to the energy loss caused by the signal radiated by the antenna being reflected back into the antenna system. The size of this indicator directly affects the overall performance and stability of the antenna system. Therefore, during the antenna design process, the return loss needs to be reduced as much as possible to improve the efficiency and reliability of signal transmission. In addition, frequency response, beam pointing accuracy, etc. are also factors that cannot be ignored in antenna design. These parameters are crucial to ensure the performance of the antenna in a specific frequency band, and play a decisive role in achieving accurate beam pointing and signal coverage.
[0003] However, in actual design, these design goals often restrict and influence each other. Adjusting one goal may have an adverse effect on the realization of other goals. Therefore, antenna design uses multi-objective optimization methods, such as multi-objective genetic algorithms or multi-objective particle swarm optimization algorithms, to find the best design solution. These methods can find an optimal balance point while meeting multiple objectives, thereby meeting various performance indicators under different requirements. In this way, engineers can design antenna systems with excellent performance, stability and reliability, providing strong support for wireless communications, radar detection and other fields.
[0004] In recent years, multi-objective optimization methods for antenna design have been studied. In 2022, X.Li and Y.-X.Guo, et al. published "The Grey Wolf Optimizer for Antenna Optimization Designs: Continuous, binary, single-objective, and multi-objective implementations" in the journal "IEEE Antennas and Propagation Magazine" No. 64, which introduced the use of the Grey Wolf optimizer to achieve multi-objective optimization of antennas. However, each individual in this method needs to be simulated using electromagnetic simulation software, which has a high computational cost. In 2023, W.Zheng and H.Li, et al. published "Designing Antennas With Quasi-Isotropic Radiation Patterns Using Pixel Structures" in the journal "IEEE Transactions on Antennas and Propagation" No. 71, No. 10. This method uses a multi-port network to establish an alternative model for the electromagnetic simulation model, greatly reducing the optimization time. However, the actual design requirements of the antenna design are not taken into account, so that some of the solutions in the Pareto frontier obtained by multi-objective optimization do not meet the actual performance requirements of the antenna, and these solutions cannot be used for antenna processing. Therefore, there is an urgent need for a fast multi-objective optimization algorithm that takes into account the actual performance constraints of the antenna. Summary of the invention
[0005] The purpose of the present invention is to provide a constrained multi-objective optimization method for antenna design, which solves the problems in the above-mentioned background technology.
[0006] To achieve the above object, the present invention provides the following technical solution: a constrained multi-objective optimization method applied to antenna design, comprising the following steps:
[0007] S1: Use multi-port network method to establish the physical model of antenna reflection coefficient;
[0008] S2: Establish the first objective function using the reflection coefficient of the antenna;
[0009] S3: Establish the second objective function using the antenna working size;
[0010] S4: The constraints of antenna performance are integrated into the non-dominated sorting genetic algorithm II (NSGA-II) to form a constrained multi-objective optimization algorithm to ensure that the Pareto frontiers of the antenna design meet the antenna return loss requirements.
[0011] Preferably, the specific content of step S1 is:
[0012] S11: Establish the antenna mother structure to be optimized, including one external port (feeding port) and N internal ports, and use CST Microwave Studio full-wave electromagnetic simulation to obtain the impedance matrix Z of the N+1 ports of the antenna mother structure;
[0013] S12: Based on the impedance matrix Z of the multi-port and the load impedance Z of the internal port L =[Z L1 ,Z L2 ,…,Z LN ] Calculate the antenna input impedance matrix Z in ;
[0014] Preferably, the specific method of step S2 is:
[0015] S21: According to the input impedance Z in Get the antenna reflection coefficient S 11 ,
[0016] S22: Determine the reflection coefficient target function f1(x) according to the operating frequency of the antenna.
[0017] Preferably, the specific method of step S3 is:
[0018] S31: Design x=Z L =[Z L1 ,Z L2 ,…,Z LN ] A simple transformation gives the matrix M describing the pixel patch operation. a , and then the matrix M a The maximum number of antenna working rows and the maximum number of antenna working columns are obtained by conversion;
[0019] S32: Establish an objective function f2(x) of the antenna working size according to the maximum number of working rows and the maximum number of working columns of the antenna.
[0020] Preferably, the specific method of step S4 is:
[0021] S31: Set the population size and randomly generate the initial population. Set the total number of algorithm runs, the number of runs to eliminate solutions that do not meet the constraints, and related parameters such as elite selection, tournament selection, crossover operation, and mutation operation;
[0022] S32: Calculate the objective function value of the initial population, and perform non-dominated sorting and crowding calculation on the initial population;
[0023] S33: using tournament selection to obtain parents, applying the crossover operator to the parents to generate sub-populations, and applying the mutation operator to the sub-populations;
[0024] S34: Calculate the objective function value of the sub-population, and perform non-dominated sorting and crowding calculation on the combined parent population and sub-population;
[0025] S35: When the number of cycles reaches the set number, the solutions with reflection coefficients greater than -10dB are downgraded to the lowest level to eliminate the solutions that do not meet the constraint conditions;
[0026] S36: Select the elites based on the rankings and select the outstanding groups;
[0027] S37: Repeat steps S31-S36 until the total number of algorithm runs reaches the limit;
[0028] S38: Return to the Pareto front.
[0029] The present invention provides a constrained multi-objective optimization method for antenna design. The constrained multi-objective optimization method for antenna design has the following beneficial effects:
[0030] (1) Compared with the traditional method, the present invention adds the practical requirements of antenna performance to the search conditions of the Pareto frontier and automatically screens out the Pareto solutions that do not meet the requirements. This enables the algorithm to find multiple different trade-off solutions to the problem, rather than just a single optimal solution, and all the solutions found meet the design requirements of the antenna.
[0031] (2) The multi-objective optimization algorithm proposed in the present invention is based on a multi-port network physical model. Only the extraction of the multi-port network impedance parameters requires a full-wave electromagnetic simulation. In the entire optimization process, the electromagnetic simulation model is not needed, and the time cost is low. BRIEF DESCRIPTION OF THE DRAWINGS
[0032] Figure 1 is a flow chart of the research algorithm of the present invention;
[0033] Figure 2 A schematic diagram of the antenna mother structure tested by the present invention;
[0034] Figure 3 It is a schematic diagram of the optimization process of the multi-objective optimization algorithm used in the present invention, and three optimization results are taken in the Pareto front for verification;
[0035] Figure 4 This is a schematic diagram of the antenna structure corresponding to the first solution of the present invention;
[0036] Figure 5 The reflection coefficient of the antenna corresponding to the first solution of the present invention;
[0037] Figure 6 A schematic diagram of an antenna structure corresponding to the second solution of the present invention;
[0038] Figure 7 The reflection coefficient of the antenna corresponding to the second solution of the present invention;
[0039] Figure 8 This is a schematic diagram of the antenna structure corresponding to the third solution of the present invention;
[0040] Fig. 9 is the reflection coefficient of the antenna corresponding to the third solution of the present invention; DETAILED DESCRIPTION
[0041] In order to have a clearer understanding of the technical features, purposes and effects of the present invention, specific embodiments of the present invention are now described with reference to the accompanying drawings.
[0042] The implementation of the constrained multi-objective optimization method for antenna design provided by the present invention is as follows: Figures 1 to 9 As shown: The constrained multi-objective optimization method applied to antenna design includes the following steps:
[0043] S1: Use the multi-port network method to establish the physical model of the antenna reflection coefficient. The specific content is:
[0044] S11: Establish the antenna mother structure to be optimized, including an external port (feeding port) and N internal ports. Use CST Microwave Studio full-wave electromagnetic simulation to obtain the impedance matrix Z of the N+1 ports of the antenna mother structure, as shown in formula (1).
[0045]
[0046] In the formula, Z 00 is the self-impedance of the antenna feed port, is the vector of mutual impedance between the feeding port and each interior port, It's Z 0N The transpose of is the impedance between all internal ports and ω is the antenna operating frequency.
[0047] S12: According to the multi-port impedance matrix Z and the load impedance x of the internal port = Z L =[Z L1 ,Z L2 ,...,Z LN ] Calculate the antenna input impedance matrix Z in , as shown in formula (2) and formula (3).
[0048] Z L =diag(Z L ) (2)
[0049] Z in (ω)=Z 00 (ω)-Z 1N (ω)·(Z L (ω)+Z NN (ω) -1 ·Z N1 (ω) (3)
[0050] S2: Use the reflection coefficient of the antenna to establish the first objective function, the specific content is:
[0051] S21: According to the input impedance Z in Get the antenna reflection coefficient S 11 , the calculation formula is shown in formula (4):
[0052]
[0053] S22: Determine the reflection coefficient objective function f1(x) according to the working frequency of the antenna. The first optimization objective function is to minimize the reflection coefficient of the working frequency band, which is defined as follows:
[0054] f1(x)=|S 11 (x,ω) dB (7)
[0055] Where |S 11 (·) dB The reflection coefficient of the antenna, in dB.
[0056] S3: Use the antenna working size to establish the second objective function, the specific content of which is:
[0057] S31: Set the design variable x=Z L =[Z L1 ,Z L2 ,...,Z LN ] is converted into a matrix M that describes the pixel patch operation a , according to M a Get the maximum number of antenna working rows C(x) and the maximum number of antenna working columns R(x);
[0058] S32: According to the maximum number of working rows and columns of the antenna, establish the objective function f2(x) of the antenna working size, which is defined as follows:
[0059] f2(x)=C(x)·R(x) (8)
[0060] S4: The constraints of antenna performance are integrated into the non-dominated sorting genetic algorithm II (NSGA-II) to form a constrained multi-objective optimization algorithm to ensure that the Pareto frontiers of the antenna design meet the antenna return loss requirements. The specific contents are as follows:
[0061] S31: Set the population size and randomly generate the initial population. Set the total number of algorithm runs, the number of runs to eliminate solutions that do not meet the constraints, and related parameters such as elite selection, tournament selection, crossover operation, and mutation operation;
[0062] S32: Calculate the objective function value of the initial population, and perform non-dominated sorting and crowding calculation on the initial population;
[0063] S33: using tournament selection to obtain parents, applying the crossover operator to the parents to generate sub-populations, and applying the mutation operator to the sub-populations;
[0064] S34: Calculate the objective function value of the sub-population, and perform non-dominated sorting and crowding calculation on the combined parent population and sub-population;
[0065] S35: When the number of cycles reaches the set number, the solutions with reflection coefficients greater than -10dB are downgraded to the lowest level to eliminate the solutions that do not meet the constraint conditions;
[0066] S36: Select the elites based on the rankings and select the outstanding groups;
[0067] S37: Repeat steps S31-S36 until the total number of algorithm runs reaches the limit;
[0068] S38: Return to the Pareto front.
[0069] The above description is only an illustrative specific implementation mode of the present invention and is not intended to limit the scope of the present invention. Any equivalent changes and modifications made by any technician in the field without departing from the concept and principle of the present invention should fall within the scope of protection of the present invention. It should also be noted that the various components of the present invention are not limited to the above-mentioned overall application. The various technical features described in the specification of the present invention can be selected one by one or multiple ones can be selected and used in combination according to actual needs. Therefore, the present invention should naturally cover other combinations and specific applications related to the present invention.
Claims
1. A constrained multi-objective optimization method for antenna design, characterized by: The following steps are involved: S1: Use multi-port network method to establish the physical model of antenna reflection coefficient; S2: Establish the first objective function using the reflection coefficient of the antenna; S3: Establish the second objective function using the antenna working size; S4: The constraints of antenna performance are integrated into the non-dominated sorting genetic algorithm II (NSGA-II) to form a constrained multi-objective optimization algorithm to ensure that the Pareto frontiers of the antenna design meet the antenna return loss requirements.
2. The constrained multi-objective optimization method for antenna design according to claim 1, characterized in that: The specific contents of step S1 are: S11: Establish the antenna mother structure to be optimized, including one external port (feeding port) and N internal ports, and use CST Microwave Studio full-wave electromagnetic simulation to obtain the impedance matrix Z of the N+1 ports of the antenna mother structure; S12: Based on the impedance matrix Z of the multi-port and the load impedance Z of the internal port L =[Z L1 ,Z L2 ,…,Z LN ] Calculate the antenna input impedance matrix Z in .
3. The constrained multi-objective optimization method for antenna design according to claim 1, characterized in that: The specific contents of step S2 are: S21: According to the input impedance Z in Get the antenna reflection coefficient S 11 , S22: Determine the reflection coefficient target function f1(x) according to the operating frequency of the antenna.
4. The constrained multi-objective optimization method for antenna design according to claim 1, characterized in that: The specific method of step S3 is: S31: Design x=Z L =[Z L1 ,Z L2 ,…,Z LN ] A simple transformation gives the matrix M describing the pixel patch operation. a , and then the matrix M a The maximum number of antenna working rows and the maximum number of antenna working columns are obtained by conversion; S32: Establish an objective function f2(x) of the antenna working size according to the maximum number of working rows and the maximum number of working columns of the antenna.
5. The constrained multi-objective optimization method for antenna design according to claim 1, characterized in that: The specific method of step S4 is: S31: Set the population size, randomly generate the initial population, set the total number of algorithm runs, the number of runs to eliminate solutions that do not meet the constraints, and related parameters such as elite selection, tournament selection, crossover operation, and mutation operation; S32: Calculate the objective function value of the initial population, and perform non-dominated sorting and crowding calculation on the initial population; S33: using tournament selection to obtain parents, applying the crossover operator to the parents to generate sub-populations, and applying the mutation operator to the sub-populations; S34: Calculate the objective function value of the sub-population, and perform non-dominated sorting and crowding calculation on the combined parent population and sub-population; S35: When the number of cycles reaches the set number, the solutions with reflection coefficients greater than -10dB are downgraded to the lowest level to eliminate the solutions that do not meet the constraint conditions; S36: Select the elites based on the rankings and select the outstanding groups; S37: Repeat steps S31-S36 until the total number of algorithm runs reaches the limit; S38: Return to the Pareto front.
Citation Information
Cited By
Pixel antenna design method, system and device based on adaptive updating and medium
CN120145471A
Adaptive update-based pixel antenna design method, system, device, and medium
CN120145471B
