Design method and system of double-frequency yagi antenna

Abstract

The invention relates to the technical field of antennas, in particular to a design method and system of a double-frequency yagi antenna, and the method comprises the steps: constructing an initial antenna model based on the design requirements of the double-frequency yagi antenna; selecting a plurality of groups of antenna design parameters in the antenna design space as input samples and inputting the input samples to the initial antenna model for simulation to obtain an antenna model response corresponding to each input sample; calculating a coupling compensation value based on each input sample and the corresponding antenna model response, and training a preset model by taking the input sample as an input and the corresponding antenna model response and the coupling compensation value as an output to obtain an antenna performance prediction model; designing a target function, searching antenna design parameters and inputting the antenna design parameters into the model, calculating a target function value according to output, and determining whether a preset optimization standard is reached or not based on the target function value to determine optimal antenna design parameters. The method can be specially designed for the dual-frequency yagi antenna so as to determine the optimal antenna parameter meeting the dual-frequency cooperation requirement.

Description

Technical Field

[0001] The present application relates to the field of antenna technology, and in particular to a design method and system for a dual-frequency Yagi antenna. Background Art

[0002] In modern wireless communications, the Internet of Things, 5G and other application scenarios, supporting the transmission of two predetermined working frequency bands at the same time can significantly improve the flexibility and performance of the system. For example, different frequency bands may have their own advantages: one frequency band can achieve longer-distance coverage, while the other frequency band can provide higher data transmission rates or better anti-interference capabilities; in addition, through the coordinated operation of dual frequencies, load balancing can be achieved, reducing the risk of congestion and interference in a single frequency band. The Yagi antenna has always been regarded as an ideal choice for long-distance communication and directional transmission because of its simple structure, easy installation, high forward gain and excellent directivity. In order to meet the requirements of dual-frequency operation, the dual-drive oscillator method is usually used on the basis of the traditional Yagi antenna structure. By setting two independent drive oscillators at the front end of the antenna to correspond to different target frequency bands, and configuring matching reflectors and directors, the antenna can achieve simultaneous resonance and efficient radiation in two frequency bands.

[0003] After adopting the dual-drive vibrator method, the basic hardware structure of the dual-band Yagi antenna is determined. The key issue turns to how to accurately determine the optimal size, spacing, and number of each component, such as the reflector, two independent drive vibrators, and the director, to ensure that the antenna can achieve optimal performance in both frequency bands. In the prior art, the following method is often used: first, an initial antenna model is constructed, and several sample antenna parameters are collected in the design parameter space, and then these parameters are input into the initial antenna model, and the corresponding antenna response data is obtained through simulation, and then a mapping relationship between the sample antenna parameters and the antenna response data is established, and finally, a multi-objective optimization algorithm is combined to perform parameter screening for each performance indicator to determine the optimal antenna parameters that meet the design requirements.

[0004] However, there are obvious shortcomings in directly applying the above traditional methods to the dual-frequency Yagi antenna constructed by the dual-drive oscillator method. Since the dual-frequency Yagi antenna is required to work simultaneously in two frequency bands, there may be cross-coupling and mutual interference between the two frequency bands. Its performance response is no longer a simple single-frequency band problem, but a complex mapping of design parameters to the response of the two frequency bands, and even a compensation effect of mutual influence. In this case, it is difficult for the general antenna parameter optimization method to accurately capture and deal with the coupling problem between frequency bands, resulting in the optimization results not being able to take into account the best performance of each frequency band at the same time. Therefore, it is urgent to develop a method specifically for the design of dual-frequency Yagi antennas to determine the optimal antenna parameters that meet the dual-frequency coordination requirements. Summary of the invention

[0005] The present application provides a design method and system for a dual-frequency Yagi antenna, which can be specifically designed for a dual-frequency Yagi antenna to determine the optimal antenna parameters that meet the dual-frequency coordination requirements. The present application provides the following technical solutions: In a first aspect, the present application provides a design method for a dual-frequency Yagi antenna, the method comprising: Build an initial antenna model based on the predefined operating frequency band and design requirements of the dual-band Yagi antenna; A hierarchical clustering sampling method is used to select several groups of antenna design parameters in the antenna design space as input samples and input them into the initial antenna model for simulation to obtain an antenna model response corresponding to each input sample; Calculating a coupling compensation value based on each input sample and its corresponding antenna model response, wherein the coupling compensation value is used to measure the degree of mutual coupling between two operating frequency bands of the dual-band Yagi antenna; Taking the input sample as input and the corresponding antenna model response and coupling compensation value as output, a preset neural network model is trained to obtain an antenna performance prediction model; The objective function is designed based on the antenna model response and the coupling compensation value, the antenna design parameters are searched and input into the antenna performance prediction model, the objective function value is calculated according to the model output, and it is judged based on the objective function value whether the preset optimization standard is met to determine the optimal antenna design parameters.

[0006] In a specific implementation scheme, constructing an initial antenna model based on a predefined working frequency band and design requirements of a dual-band Yagi antenna includes: The two target operating frequency bands of the dual-frequency Yagi antenna are defined as and ; The initial antenna model is used to describe the basic structure of the dual-band Yagi antenna, which includes two independent driven oscillators, a reflector, and one or more directors. The antenna design parameters include the length of the two driving vibrators and their relative spacing, the length of the reflector and its actual distance from the driving vibrator, the number of directors, their length and the spacing between them and the driving vibrator.

[0007] In a specific implementation scheme, the step of selecting a plurality of groups of antenna design parameters in the antenna design space as input samples by using a hierarchical clustering sampling method and inputting the samples into the initial antenna model for simulation includes: Perform sensitivity analysis on antenna design parameters and divide the antenna design parameters into layers based on sensitivity weights; Based on the stratification results, multidimensional parameter intervals are constructed, and representative samples are selected in each multidimensional interval using clustering method; Integrate the selected representative samples into a candidate sample set , perform weighted random sampling based on the sampling probability to obtain input samples.

[0008] In a specific possible implementation scheme, performing sensitivity analysis on antenna design parameters and dividing antenna design parameters into layers based on sensitivity weights includes: Based on the initial antenna model, a small perturbation is made to each antenna design parameter and the change in return loss is recorded. , defining its sensitivity coefficient as follows: ; in, is the change in return loss after disturbance, Design parameter variations for antennas; Normalize the sensitivity coefficient to the sensitivity weight as follows: ; in, The number of types of antenna design parameters; The antenna design parameters are divided into layers according to the sensitivity weight. Antenna design parameters with sensitivity weight greater than the preset threshold are identified as key parameters and their parameter range is divided into 10 layers. Antenna design parameters with sensitivity weight less than or equal to the preset threshold are identified as secondary parameters and their parameter range is divided into 2 layers.

[0009] In a specific implementation scheme, the selected representative samples are integrated into a candidate sample set. , weighted random sampling is performed based on the sampling probability, and the input samples are obtained including; The representative samples selected from all multidimensional intervals are integrated into the final candidate sample set, and the representative score of each candidate sample is calculated using the following formula: : ; in, For the candidate sample The value of the dimension, and They are the first The mean and standard deviation of the parameters, For the The sensitivity weights of the parameters, For sample The cluster center The distance A scaling factor to adjust the effect of cluster distance; Calculate the sampling probability of each sample as follows: ; Based on the calculated sampling probability All candidate samples are randomly sampled and the representative sample set is used as the input sample.

[0010] In a specific implementation scheme, calculating the coupling compensation value based on each input sample and its corresponding antenna model response includes: Antenna model response including frequency band Response and frequency band Response, frequency band Response includes Antenna return loss, gain and standing wave ratio under frequency band Response includes Antenna return loss, gain and standing wave ratio under Coupling compensation value The calculation formula is as follows: ; in, and Respectively represent the working frequency band and The return loss under and Respectively represent the working frequency band and The antenna gain under and Respectively represent the working frequency band and The standing wave ratio under is a positive constant, is the hyperbolic tangent function.

[0011] In a specific implementation scheme, the design of the objective function based on the antenna model response and the coupling compensation value includes: The objective function is as follows: ; in, is the coupling compensation value output by the antenna performance prediction model, is the adjustment coefficient, and The working frequency bands are and The performance error under , is calculated as follows: ; ; in, and They are the outputs of the antenna performance prediction model in the working frequency band and The return loss under and They are the outputs of the antenna performance prediction model in the working frequency band and The antenna gain under and They are the outputs of the antenna performance prediction model in the working frequency band and The standing wave ratio under and In the working frequency band and The preset return loss target value is and In the working frequency band and The preset antenna gain target value is and In the working frequency band and The preset VSWR target value is set.

[0012] In a second aspect, the present application provides a design system for a dual-frequency Yagi antenna, which adopts the following technical solution: A design system for a dual-frequency Yagi antenna, comprising: An initial antenna model building module is used to build an initial antenna model based on the predefined working frequency band and the design requirements of the dual-band Yagi antenna; An antenna design parameter simulation module is used to select a plurality of groups of antenna design parameters as input samples in the antenna design space by using a hierarchical clustering sampling method and input them into the initial antenna model for simulation to obtain an antenna model response corresponding to each input sample; A coupling compensation value calculation module, used to calculate a coupling compensation value based on each input sample and its corresponding antenna model response, wherein the coupling compensation value is used to measure the degree of mutual coupling between two working frequency bands of the dual-frequency Yagi antenna; A prediction model training module, used to train a preset neural network model using the input sample as input and the corresponding antenna model response and coupling compensation value as output to obtain an antenna performance prediction model; The antenna design parameter determination module is used to design an objective function based on the antenna model response and the coupling compensation value, search for antenna design parameters and input them into the antenna performance prediction model, calculate the objective function value according to the model output, judge whether the preset optimization standard is met based on the objective function value, and determine the optimal antenna design parameters.

[0013] In a third aspect, the present application provides an electronic device, comprising a processor and a memory; the memory stores a program, and the program is loaded and executed by the processor to implement a design method for a dual-band Yagi antenna as described in the first aspect.

[0014] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the storage medium stores a program, and when the program is executed by a processor, it is used to implement a design method for a dual-frequency Yagi antenna as described in the first aspect.

[0015] In summary, the beneficial effects of this application include at least: (1) An optimization method based on adaptive weighted hierarchical clustering sampling is proposed. Compared with the traditional Latin hypercube sampling method, it can optimize the design parameters more accurately. This method first performs a sensitivity analysis on the antenna design parameters and assigns different weights to each parameter according to the degree of influence on the antenna performance. Parameters with higher sensitivity are divided into more levels, and the key parameter space with a greater impact on antenna performance is explored with a finer resolution, thereby effectively improving the pertinence of the sampling strategy. Through this sensitivity weighting and hierarchical clustering sampling method, the exploration accuracy of the key performance impact area can be significantly improved, avoiding the problem of insufficient attention to key parameters in traditional sampling methods. This method can more effectively identify the important factors affecting antenna performance and reduce the exploration of invalid design space, thereby achieving a more accurate design optimization process and ensuring that the obtained antenna design parameters achieve the optimal performance.

[0016] (2) By introducing the concept of coupling compensation value, the interference problem caused by mutual coupling between frequency bands in dual-frequency working mode is further solved. The defined coupling compensation value comprehensively considers the key performance parameters of the antenna in different frequency bands, such as return loss, gain and standing wave ratio, and quantifies the degree of coupling between the dual-frequency antenna in each working frequency band through a specific calculation formula. In the process of antenna optimization design, by incorporating the coupling compensation value into the performance evaluation system, the actual performance of the antenna in the dual-frequency working state can be more accurately reflected. By giving priority to design parameters with low coupling degree and stable performance, the signal coupling and interference between different working frequency bands can be effectively reduced, thereby improving the working stability and overall performance of the antenna in the dual-frequency mode. This has significant technical advantages for signal quality assurance and signal loss suppression in dual-frequency communication applications.

[0017] First, an initial antenna model that only determines the basic structure and geometric layout is constructed. Then, representative samples are selected in the design parameter space through the adaptive weight hierarchical clustering sampling method. The antenna performance response data in two working frequency bands are obtained by electromagnetic simulation. The coupling compensation value is calculated based on the antenna performance response data, and the key performance parameters of the two frequency bands are comprehensively mapped into a quantitative indicator to reflect the performance balance and coordination between the frequency bands. Subsequently, the input samples and their corresponding dual-frequency responses and coupling compensation values ​​are used as outputs to train the preset neural network model, and an antenna performance prediction model that can simultaneously predict the performance and coupling compensation values ​​in two working frequency bands is constructed. Finally, a multi-objective optimization algorithm is used to search for candidate solutions in the entire design parameter space, and an objective function is designed. The objective function not only takes into account the performance errors of the two frequency bands, but also strengthens the constraints on insufficient coupling through an exponential penalty mechanism, thereby screening out one or more groups of optimal design parameters that meet the best performance requirements of each frequency band and achieve good collaborative work. This design not only makes up for the deficiency of traditional methods in capturing complex nonlinear interactions between frequency bands, but also realizes the overall optimization of the collaborative performance of dual-frequency antennas through the introduction of coupling compensation values, effectively solving the problem of performance imbalance caused by cross-coupling and mutual interference between frequency bands during dual-frequency operation in the background technology.

[0018] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application and implement it according to the contents of the specification, the following is a detailed description of the preferred embodiments of the present application in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a flow chart of the design method of the dual-frequency Yagi antenna in the embodiment of the present application.

[0020] Figure 2 It is a schematic diagram of the overall process of the design method of the dual-frequency Yagi antenna in the embodiment of the present application.

[0021] Figure 3 It is a data diagram for verifying the effect of the design method of the dual-band Yagi antenna in the embodiment of the present application.

[0022] Figure 4 It is a structural block diagram of the design system of the dual-frequency Yagi antenna in the embodiment of the present application.

[0023] Figure 5 It is a block diagram of an electronic device designed with a dual-band Yagi antenna in an embodiment of the present application. DETAILED DESCRIPTION

[0024] The specific implementation methods of the present application are further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present application but are not intended to limit the scope of the present application.

[0025] Optionally, the present application uses the design method of the dual-band Yagi antenna provided in each embodiment as an example for use in an electronic device, where the electronic device is a terminal or a server. The terminal may be a mobile phone, a computer, a tablet computer, etc. This embodiment does not limit the type of electronic device.

[0026] Reference Figure 1 , is a flow chart of a method for designing a dual-frequency Yagi antenna provided by an embodiment of the present application, the method comprising at least the following steps: Step S101: construct an initial antenna model based on a predefined working frequency band and design requirements of a dual-band Yagi antenna.

[0027] In step S101, two target operating frequency bands of the dual-frequency Yagi antenna are first predefined, which are respectively denoted as and Then, the initial antenna model of the dual-frequency Yagi antenna is constructed based on the design concept of the dual-driven oscillator method. The initial antenna model is used to describe the basic structure of the dual-frequency Yagi antenna, which includes two independent driven oscillators, a reflector, and one or more directors.

[0028] Among them, each driving oscillator corresponds to the working frequency band and In its preliminary design, the length of the driving vibrator is set according to the half-wavelength of each working frequency band, the reflector is located behind the driving vibrator, and is usually slightly longer than the driving vibrator in design, and the director is set in front of the driving vibrator, and its number and size are reserved as adjustable parameters. After the initial antenna model is built, the initial antenna model only determines the basic framework and geometric layout of the dual-band Yagi antenna, and the antenna design parameters have not yet been determined. The antenna design parameters include the length of the two driving vibrators and their relative spacing, the length of the reflector and its actual distance from the driving vibrator, and the number, length, and spacing of the directors from the driving vibrator.

[0029] It should be noted that although the predefined working frequency band and This provides a theoretical basis for the preliminary design of the driving vibrator, that is, the ideal length of each driving vibrator is half a wavelength, but in actual design, the precise size of the driving vibrator still needs to be further optimized. This is because the actual antenna performance is not only affected by the operating frequency band, but also by multiple factors such as the cross-coupling between the dual driving vibrators, the antenna environment, manufacturing errors, and the interaction with the reflector and director configuration. Therefore, although the length of the driving vibrator can be determined by half a wavelength in theory, its specific size still needs to be fine-tuned through subsequent simulation and optimization to achieve the best practical effect.

[0030] In addition, in the present application, the dual driving vibrators adopt a symmetrical layout, so the distance between the reflector and the two driving vibrators is defined as the distance from the reflector to the symmetry center of the driving vibrators. Similarly, the distance between the guide and the two driving vibrators is also the distance from the guide to the symmetry center of the driving vibrators.

[0031] Step S102: A hierarchical clustering sampling method is used to select several groups of antenna design parameters in the antenna design space as input samples and input them into the initial antenna model for simulation to obtain an antenna model response corresponding to each input sample.

[0032] In step S102, firstly, several groups of antenna design parameters are selected as input samples in the antenna design space, and the input samples are input into the initial antenna model. The initial antenna model of the input samples is simulated and solved using an electromagnetic simulation tool to obtain the antenna model response corresponding to the input sample. The antenna model response includes but is not limited to the antenna return loss, gain and standing wave ratio.

[0033] In the implementation, the input samples are input into the initial antenna model, and the electromagnetic simulation tool is used to perform dual-frequency joint simulation on the initial antenna model of the input samples, that is, the simulation solution of each input sample needs to be simultaneously in the working frequency band. and The antenna model response of the two frequency bands is obtained. For each input sample, the antenna model response includes the frequency band Response and frequency band Response, specifically, Antenna return loss, gain and standing wave ratio under Antenna return loss, gain and standing wave ratio under.

[0034] In addition, as a preferred embodiment, when selecting several groups of antenna design parameters as input samples in the antenna design space, the prior art usually adopts the Latin hypercube sampling method for sample selection. This method can evenly cover the parameter space, but in the high-dimensional dual-band Yagi antenna design problem with significant nonlinearity and interaction between parameters, it is often difficult to fully capture the key performance sensitive areas, resulting in the possible omission of important candidate solutions in subsequent optimization. Therefore, in order to make up for the shortcomings of Latin hypercube sampling in exploring key areas, this application proposes a new adaptive weighted hierarchical clustering sampling method, which performs sensitivity analysis on each design parameter, assigns different importance weights to each parameter, focuses on the key variables that affect dual-band performance, partitions the design parameter space using hierarchical and clustering techniques, and then selects representative samples from each area, thereby ensuring that the samples not only evenly cover the entire design space, but also focus on highly sensitive areas to avoid repeated sampling.

[0035] Among them, the specific method of the adaptive weight stratified clustering sampling method is as follows: Step S1021: perform sensitivity analysis on antenna design parameters and divide the antenna design parameters into layers based on sensitivity weights.

[0036] Specifically, based on the initial antenna model, a small perturbation is performed on each antenna design parameter. , and record the change in return loss for each antenna design parameter , defining its sensitivity coefficient as follows: ; in, is the change in return loss after disturbance, The change in antenna design parameters.

[0037] It should be noted that the return loss directly reflects the radiation efficiency and matching degree of the antenna, and is the core performance indicator that best reflects the quality of the design. In a dual-band Yagi antenna, ensuring that the return loss in both operating frequency bands is less than -10dB is a core requirement.

[0038] The sensitivity coefficient is then normalized to the sensitivity weight as follows: ; in, is the number of types of antenna design parameters. The sensitivity weight reflects the relative influence of each antenna design parameter on the antenna performance. The higher the sensitivity weight, the greater the influence of the parameter on the antenna performance.

[0039] Finally, the antenna design parameters are hierarchically divided according to the sensitivity weights. For antenna design parameters with sensitivity weights greater than a preset threshold, they are identified as key parameters and their parameter intervals are divided into more layers, 10 layers in this application. For antenna design parameters with sensitivity weights less than or equal to the preset threshold, they are identified as secondary parameters and their parameter intervals are divided into fewer layers, 2 layers in this application. The above hierarchical strategy can ensure a more detailed exploration of key antenna design parameters to a certain extent, while reducing the sampling complexity of antenna design parameters with less impact.

[0040] Step S1022: construct a multidimensional parameter interval based on the stratification result, and select representative samples in each multidimensional interval using a clustering method.

[0041] Specifically, after sensitivity analysis of each antenna design parameter and stratification according to sensitivity weight, each parameter has been divided into several levels. By combining the layer divisions of antenna design parameters, the entire design parameter space can be divided into A small multidimensional interval, where For the For example, if there are 3 key parameters (10 layers) and 2 secondary parameters (2 layers), the total number of layers can be divided into small intervals. The multidimensional interval is the discretization result of the antenna design parameter space. Each interval covers a set of similar parameter values, ensuring finer subdivision in highly sensitive areas, so as to more fully capture the impact of key parameter changes on performance indicators such as return loss. For each generated multidimensional interval, all possible candidate parameter points in the interval are extracted through coarse sampling. If there are many candidate points in the interval, a clustering algorithm is used to divide these points into several subclusters, and the point closest to the cluster center is selected from each subcluster as a representative sample. If there are fewer candidate points in the interval, all points are directly selected or the regional center value is used as the representative sample. In this way, the number of layers set in the stratification directly determines the number of divisions of the multidimensional interval, and clustering in each interval can avoid repeated sampling in areas with similar parameter values, ensuring that the selected representative samples can not only evenly cover the entire design space, but also focus on capturing the change information of highly sensitive areas.

[0042] Optionally, in the present application, k-means or a hierarchical clustering algorithm is used as a specific clustering algorithm to divide the points into several subclusters. Other types of clustering algorithms may also be used. The present application does not limit the specific type of clustering algorithm.

[0043] Step S1023: Integrate the selected representative samples into a candidate sample set , perform weighted random sampling based on the sampling probability to obtain input samples.

[0044] Specifically, the representative samples selected from all multidimensional intervals are integrated into the final candidate sample set, and then the representative score of each candidate sample is calculated using the following formula: : ; in, For the candidate sample The value of the dimension, and They are the first The mean and standard deviation of the parameters, For the The sensitivity weights of the parameters, For sample The cluster center The distance A scaling factor to adjust the effect of cluster distance. In the above formula, the standardized deviation of the sample from the mean of all candidate samples in each dimension is first calculated, and then weighted by the sensitivity weight, so that the parameters that have a greater impact on antenna performance can receive higher attention. This helps to ensure that key parameter areas can be more fully explored. Secondly, the exponential term is used to penalize samples that are close to the center of the cluster to which they belong, thereby encouraging sampling dispersion and avoiding too many redundant samples, so that the selected samples represent the key areas and cover the diversity of the parameter space. The formula combines the degree of deviation of each parameter with the uniqueness of the sample in the cluster by multiplying the geometric mean with the exponential penalty term, thereby generating a comprehensive representative score, so that the sampling process can not only focus on performance-sensitive areas, but also avoid the bias caused by local extreme values, which helps to obtain a more balanced and representative set of candidate samples.

[0045] Then calculate the sampling probability of each sample as follows: ; Finally, based on the calculated sampling probability All candidate samples are randomly sampled and finally a representative sample set is obtained as input sample for subsequent dual-frequency joint simulation. After that, the sampling probability is calculated by normalization , ensuring that the sum of the sampling probabilities of all candidate samples is 1, so that the final input sample can be directly selected using the weighted random sampling method. This method ensures that high-scoring samples are selected first and that the overall sample is reasonably covered.

[0046] Step S103: Calculate a coupling compensation value based on each input sample and its corresponding antenna model response, where the coupling compensation value is used to measure the degree of mutual coupling between the two working frequency bands of the dual-band Yagi antenna.

[0047] In step S103, a coupling compensation value is calculated based on each input sample in step S102 and its corresponding antenna model response. The coupling compensation value is a value used to measure the coupling compensation of the dual-band Yagi antenna in two working frequency bands. and A quantitative indicator of the degree of mutual coupling between the two frequency bands, reflecting the balance and synergy between the performance of the two frequency bands. The calculation formula is as follows: ; in, and Respectively represent the working frequency band and The return loss under and Respectively represent the working frequency band and The antenna gain under and Respectively represent the working frequency band and The standing wave ratio under is a positive constant used to prevent the denominator from being zero. is the hyperbolic tangent function, and its range is , in this application, the positive value range is taken to ensure the coupling compensation value Falling Inside.

[0048] In the above formula, the numerator represents the difference in return loss in the two frequency bands. Return loss is an important indicator for measuring antenna input matching and radiation efficiency. The greater the difference, the more serious the performance imbalance and cross-coupling problem between the two frequency bands. The first part of the denominator represents the geometric mean of the gains of the two frequency bands, reflecting the overall radiation efficiency of the antenna. The latter part of the multiplication represents the average standing wave ratio of the two frequency bands, which is used to reflect the matching situation. When the antenna design performance is better, the denominator is larger, so that the ratio brought by the same return loss difference is lower; conversely, when the performance is poor, the denominator is smaller, and the compensation value drops more significantly. Finally, the hyperbolic tangent function is used to perform nonlinear compression on the above ratio to ensure that the output value is stable between 0 and 1 and has a smooth transition characteristic. Based on the traditional weighted summation or difference comparison, this formula introduces the geometric mean and average standing wave ratio as comprehensive performance indicators, and uses nonlinear functions for mapping, which not only retains the information of the interaction between key performance indicators, but also limits the output to 1 through normalization. In summary, the formula structure embodies the idea of ​​"overall performance adjustment-local difference amplification", that is, when the overall performance is good, the penalty for the difference in a single frequency band is small; when the overall performance is poor, even if the difference is not large, it will cause a large change in the coupling compensation value, reflecting a more sensitive coupling effect.

[0049] Specifically, the coupling compensation value Indicates the degree of coordination between the two frequency bands in terms of return loss, gain and standing wave ratio under the current antenna design parameters. Coupling compensation value The closer the value is to 1, the more balanced the performance of the two frequency bands are, the less interference there is between them, and the closer the antenna design is to the optimal state of dual-frequency coordination. The closer the value is to 0, the stronger the performance difference or coupling interference is between the two frequency bands, which needs to be compensated and optimized in the design. The formula takes into account the difference in return loss, gain level, and standing wave ratio, and can fully reflect the matching and radiation performance of the antenna in dual-frequency working state. The use of the hyperbolic tangent function ensures that the difference changes in a small range will not cause excessive fluctuations, while the imbalance between frequency bands can be quickly reflected in a large range, which is in line with the characteristics of nonlinear effects in actual antenna design. By normalizing each item and calculating the ratio, and then mapping it with an exponential function or a hyperbolic tangent function, the formula output is ensured to be stable and easy to interpret, which can be comprehensively considered as an important constraint indicator or objective function part in the subsequent optimization process.

[0050] Step S104: using the input sample as input and the corresponding antenna model response and coupling compensation value as output to train the preset neural network model to obtain an antenna performance prediction model.

[0051] In step S104, a number of input samples are used as input, that is, a number of groups of numerical antenna design parameters, including the lengths of the two driving vibrators and their relative spacing, the length of the reflector and its actual distance from the driving vibrator, and the number, length, and spacing of the directors from the driving vibrator. The antenna model responses corresponding to the input samples and the calculated coupling compensation values ​​are used as outputs to train the preset neural network model to obtain an antenna performance prediction model. The trained antenna performance prediction model can output the corresponding frequency band according to the input antenna design parameters. Response, frequency band response and coupling compensation values.

[0052] Optionally, the fully connected feedforward neural network, i.e., FNN, is selected as the type of preset neural network model in this application. Since the input is a set of numerical antenna design parameters, the parameters are numerical features of fixed dimensions, and there is no local correlation in space or time, FNN is the most direct and efficient neural network architecture for processing nonlinear numerical mapping, and is particularly suitable for processing complex nonlinear relationships between input features and output responses. Other types of neural network models can also be selected, and this application does not limit the type of preset neural network model.

[0053] Step S105, design an objective function based on the antenna model response and the coupling compensation value, search for antenna design parameters and input them into the antenna performance prediction model, calculate the objective function value according to the model output, determine whether the preset optimization standard is met based on the objective function value, and determine the optimal antenna design parameters.

[0054] In step S105, it is first necessary to construct an objective function, which simultaneously considers the performance error in the two working frequency bands and the coupling synergy effect between the frequency bands. The objective function is as follows: ; in, is the coupling compensation value output by the antenna performance prediction model, is a positive adjustment coefficient used to enhance the penalty effect on insufficient coupling. and The working frequency bands are and The performance error under , is calculated as follows: ; ; in, and They are the outputs of the antenna performance prediction model in the working frequency band and The return loss under and They are the outputs of the antenna performance prediction model in the working frequency band and The antenna gain under and They are the outputs of the antenna performance prediction model in the working frequency band and The standing wave ratio under and In the working frequency band and The preset return loss target value is and In the working frequency band and The preset antenna gain target value is and In the working frequency band and The preset VSWR target value is set.

[0055] In the design of the above objective function, the first term represents the mean of the performance errors of the two frequency bands, which can fully reflect the deviations of the antenna design parameters in each frequency band, while the exponential term plays the role of a nonlinear amplifier. When the coupling effect between the two frequency bands is good, that is, the coupling compensation value is close to 1, the exponential term tends to 1, and has little effect on the objective function. When the coupling compensation value is low, the exponential term increases rapidly, thereby significantly improving the overall objective function value. This design not only pursues the optimal performance of a single frequency band during the optimization process, but also requires that the responses between the two frequency bands have high consistency and synergy.

[0056] After the objective function is constructed, an existing multi-objective optimization algorithm is used to iteratively search in the antenna design parameter space, which includes all adjustable parameters, such as the size and relative spacing of the two driving vibrators, the size of the reflector and the distance between it and the driving vibrator, and the number, size and distance between the directors and the driving vibrator. In each iteration, the candidate design parameters are input into the antenna performance prediction model trained in step S104 to obtain the corresponding frequency band. Response, frequency band The response and coupling compensation value are used to calculate the value of the objective function. When the objective function value is lower than the preset optimization standard, that is, less than or equal to the preset threshold, the performance errors in all frequency bands are within an acceptable range and the coupling compensation value is close enough to 1, which is considered to meet the dual-band collaborative work requirements, thereby determining that the candidate design parameters are the optimal antenna design parameters.

[0057] Optionally, NSGA-II or a multi-objective particle swarm algorithm is selected as the multi-objective optimization algorithm in the present application, and the present application does not limit the specific type of the multi-objective optimization algorithm.

[0058] It should be noted that during the optimization process, usually not only one set of optimal design parameters is obtained, but several sets of candidate parameters are screened out, all of which meet or are close to the preset optimization criteria, thereby forming an optimal solution set. Each set of design parameters in these solution sets can achieve lower performance error and higher frequency band coordination on the two working frequency bands.

[0059] In practice, the benefits of the objective function design are: first, it clearly integrates the performance deviation of the two frequency bands and the collaborative working effect between the frequency bands, measures the performance through the weighted root mean square error, and strengthens the penalty for insufficient coupling through the exponential penalty mechanism to ensure that the frequency band coordination is not sacrificed during the optimization process; second, this design can adaptively adjust the impact of each part on the overall objective function, with high flexibility and engineering applicability. Finally, the overall objective function structure is simple and creative, which not only meets the multi-objective requirements of dual-frequency performance optimization, but also ensures that the design parameters can achieve balanced coordination in actual antenna work, providing strong theoretical support for subsequent practical verification and application.

[0060] In summary, combined with Figure 2, this application first constructs an initial antenna model that only determines the basic structure and geometric layout, and then selects representative samples in the design parameter space through an adaptive weighted hierarchical clustering sampling method, and uses electromagnetic simulation to obtain antenna performance response data in two working frequency bands, and calculates the coupling compensation value based on the antenna performance response data, and comprehensively maps the key performance parameters of the two frequency bands into a quantitative indicator to reflect the performance balance and degree of coordination between the frequency bands. Subsequently, the preset neural network model is trained with the input samples and their corresponding dual-frequency responses and coupling compensation values ​​as outputs to construct an antenna performance prediction model that can simultaneously predict the performance and coupling compensation values ​​in two working frequency bands. Finally, a multi-objective optimization algorithm is used to search for candidate solutions in the entire design parameter space, and an objective function is designed. The objective function not only takes into account the performance errors of the two frequency bands, but also strengthens the constraints on insufficient coupling through an exponential penalty mechanism, thereby screening out one or more groups of optimal design parameters that meet the optimal performance requirements of each frequency band and achieve good collaborative work. This design not only makes up for the deficiency of traditional methods in capturing complex nonlinear interactions between frequency bands, but also realizes the overall optimization of the collaborative performance of dual-frequency antennas through the introduction of coupling compensation values, effectively solving the problem of performance imbalance caused by cross-coupling and mutual interference between frequency bands during dual-frequency operation in the background technology.

[0061] In addition, preferably, the combination Figure 3 In order to verify the technical effect of the present application, a simulation verification test was conducted. The figure is a set of sample data, which describes the objective function values ​​obtained by three different methods in 20 optimization iterations. The lower the value of the objective function value, the closer the performance is to the preset target. The first one is the traditional conventional technology, that is, the prior art described in the background technology. The second one is a solution that only adopts the objective function designed by the present application, but does not introduce the coupling compensation value. The third one is the complete solution of the present application, which includes both the design of the objective function and the introduction of the coupling compensation value. It can be seen from the data that as the number of iterations increases, the objective function values ​​of all methods show a downward trend, but there are obvious differences in the rate of decline and the final value. The optimization effect of the traditional conventional technology is the worst, and its objective function value is always high; the solution that only adopts the design objective function has achieved good improvements, but it still fails to fully solve the problem of cross-interference between frequency bands; and after the coupling compensation value is introduced, the objective function value of the complete solution of the present application is significantly reduced, and finally reaches about 97.1, proving that it has obvious advantages in achieving performance balance and collaborative work between the two frequency bands. The comparative line graph data intuitively proves that the method of the present application can achieve better synergy between frequency bands while maintaining low error performance, thereby obtaining optimal antenna design parameters that better meet preset requirements.

[0062] Figure 4 This is a structural block diagram of a design system for a dual-frequency Yagi antenna provided by an embodiment of the present application. The system includes at least the following modules: An initial antenna model building module is used to build an initial antenna model based on the predefined working frequency band and the design requirements of the dual-band Yagi antenna; The antenna design parameter simulation module is used to select several groups of antenna design parameters as input samples in the antenna design space by using a hierarchical clustering sampling method and input them into the initial antenna model for simulation to obtain the antenna model response corresponding to each input sample; A coupling compensation value calculation module, used for calculating a coupling compensation value based on each input sample and its corresponding antenna model response, wherein the coupling compensation value is used to measure the mutual coupling degree between two working frequency bands of the dual-band Yagi antenna; A prediction model training module is used to train a preset neural network model using input samples as input and corresponding antenna model responses and coupling compensation values ​​as output to obtain an antenna performance prediction model; The antenna design parameter determination module is used to design the objective function based on the antenna model response and the coupling compensation value, search for the antenna design parameters and input them into the antenna performance prediction model, calculate the objective function value according to the model output, and determine whether the preset optimization criteria are met based on the objective function value to determine the optimal antenna design parameters.

[0063] For relevant details, refer to the above method embodiment.

[0064] Figure 5 4 is a block diagram of an electronic device provided by an embodiment of the present application. The device at least includes a processor 401 and a memory 402.

[0065] The processor 401 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 401 may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). The processor 401 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 401 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 401 may also include an AI (Artificial Intelligence) processor, which is used to process computing operations related to machine learning.

[0066] The memory 402 may include one or more computer-readable storage media, which may be non-transitory. The memory 402 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 402 is used to store at least one instruction, which is used to be executed by the processor 401 to implement the design method of the dual-band Yagi antenna provided in the method embodiment of the present application.

[0067] In some embodiments, the electronic device may further optionally include: a peripheral device interface and at least one peripheral device. The processor 401, the memory 402 and the peripheral device interface may be connected via a bus or a signal line. Each peripheral device may be connected to the peripheral device interface via a bus, a signal line or a circuit board. Schematically, the peripheral devices include but are not limited to: a radio frequency circuit, a touch display screen, an audio circuit, and a power supply.

[0068] Of course, the electronic device may also include fewer or more components, which is not limited in this embodiment.

[0069] Optionally, the present application also provides a computer-readable storage medium, in which a program is stored, and the program is loaded and executed by a processor to implement the design method of the dual-band Yagi antenna of the above method embodiment.

[0070] Optionally, the present application also provides a computer product, which includes a computer-readable storage medium, in which a program is stored, and the program is loaded and executed by a processor to implement the design method of the dual-band Yagi antenna of the above method embodiment.

[0071] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0072] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A design method for a dual-frequency Yagi antenna, characterized in that: The method comprises: Build an initial antenna model based on the predefined operating frequency band and design requirements of the dual-band Yagi antenna; A hierarchical clustering sampling method is used to select several groups of antenna design parameters in the antenna design space as input samples and input them into the initial antenna model for simulation to obtain an antenna model response corresponding to each input sample; Calculating a coupling compensation value based on each input sample and its corresponding antenna model response, wherein the coupling compensation value is used to measure the degree of mutual coupling between two operating frequency bands of the dual-band Yagi antenna; Taking the input sample as input and the corresponding antenna model response and coupling compensation value as output, a preset neural network model is trained to obtain an antenna performance prediction model; The objective function is designed based on the antenna model response and the coupling compensation value, the antenna design parameters are searched and input into the antenna performance prediction model, the objective function value is calculated according to the model output, and it is judged based on the objective function value whether the preset optimization standard is met to determine the optimal antenna design parameters.

2. The design method of the dual-frequency Yagi antenna according to claim 1, characterized in that: The initial antenna model is constructed based on the predefined working frequency band and the design requirements of the dual-frequency Yagi antenna, including: The two target operating frequency bands of the dual-frequency Yagi antenna are defined as and ; The initial antenna model is used to describe the basic structure of the dual-band Yagi antenna, which includes two independent driven oscillators, a reflector, and one or more directors. The antenna design parameters include the length of the two driving vibrators and their relative spacing, the length of the reflector and its actual distance from the driving vibrator, the number of directors, their length and the spacing between them and the driving vibrator.

3. The design method of the dual-frequency Yagi antenna according to claim 1, characterized in that: The method of selecting a plurality of groups of antenna design parameters in the antenna design space as input samples by adopting the hierarchical clustering sampling method and inputting the samples into the initial antenna model for simulation includes: Perform sensitivity analysis on antenna design parameters and divide the antenna design parameters into layers based on sensitivity weights; Based on the stratification results, multidimensional parameter intervals are constructed, and representative samples are selected in each multidimensional interval using clustering method; Integrate the selected representative samples into a candidate sample set , perform weighted random sampling based on the sampling probability to obtain input samples.

4. The design method of the dual-frequency Yagi antenna according to claim 3, characterized in that: The performing sensitivity analysis on the antenna design parameters and dividing the antenna design parameters into layers based on the sensitivity weights comprises: Based on the initial antenna model, a small perturbation is made to each antenna design parameter and the change in return loss is recorded. , defining its sensitivity coefficient as follows: ； in, is the change in return loss after disturbance, Design parameter variations for antennas; Normalize the sensitivity coefficient to the sensitivity weight as follows: ； in, The number of types of antenna design parameters; The antenna design parameters are divided into layers according to the sensitivity weight. Antenna design parameters with sensitivity weight greater than the preset threshold are identified as key parameters and their parameter range is divided into 10 layers. Antenna design parameters with sensitivity weight less than or equal to the preset threshold are identified as secondary parameters and their parameter range is divided into 2 layers.

5. The design method of the dual-frequency Yagi antenna according to claim 3, characterized in that: The selected representative samples are integrated into a candidate sample set , weighted random sampling is performed based on the sampling probability, and the input samples are obtained including; The representative samples selected from all multidimensional intervals are integrated into the final candidate sample set, and the representative score of each candidate sample is calculated using the following formula: : ； in, For the candidate sample The value of the dimension, and They are the first The mean and standard deviation of the parameters, For the The sensitivity weights of the parameters, For sample The cluster center The distance A scaling factor to adjust the effect of cluster distance; Calculate the sampling probability of each sample as follows: ； Based on the calculated sampling probability All candidate samples are randomly sampled and the representative sample set is used as the input sample.

6. The design method of the dual-frequency Yagi antenna according to claim 2, characterized in that: The calculating of the coupling compensation value based on each input sample and its corresponding antenna model response comprises: Antenna model response including frequency band Response and frequency band Response, frequency band Response includes Antenna return loss, gain and standing wave ratio under frequency band Response includes Antenna return loss, gain and standing wave ratio under Coupling compensation value The calculation formula is as follows: ； in, and Respectively represent the working frequency band and The return loss under and Respectively represent the working frequency band and The antenna gain under and Respectively represent the working frequency band and The standing wave ratio under is a positive constant, is the hyperbolic tangent function.

7. The design method of the dual-frequency Yagi antenna according to claim 6, characterized in that: The objective function designed based on the antenna model response and the coupling compensation value includes: The objective function is as follows: ； in, is the coupling compensation value output by the antenna performance prediction model, is the adjustment coefficient, and The working frequency bands are and The performance error under , is calculated as follows: ； ； in, and They are the outputs of the antenna performance prediction model in the working frequency band and The return loss under and They are the outputs of the antenna performance prediction model in the working frequency band and The antenna gain under and They are the outputs of the antenna performance prediction model in the working frequency band and The standing wave ratio under and In the working frequency band and The preset return loss target value is and In the working frequency band and The preset antenna gain target value is and In the working frequency band and The preset VSWR target value is set.

8. A design system for a dual-frequency Yagi antenna, characterized in that: include: An initial antenna model building module is used to build an initial antenna model based on the predefined working frequency band and the design requirements of the dual-band Yagi antenna; An antenna design parameter simulation module is used to select a plurality of groups of antenna design parameters as input samples in the antenna design space by using a hierarchical clustering sampling method and input them into the initial antenna model for simulation to obtain an antenna model response corresponding to each input sample; A coupling compensation value calculation module, used to calculate a coupling compensation value based on each input sample and its corresponding antenna model response, wherein the coupling compensation value is used to measure the degree of mutual coupling between two working frequency bands of the dual-frequency Yagi antenna; A prediction model training module, used to train a preset neural network model using the input sample as input and the corresponding antenna model response and coupling compensation value as output to obtain an antenna performance prediction model; The antenna design parameter determination module is used to design an objective function based on the antenna model response and the coupling compensation value, search for antenna design parameters and input them into the antenna performance prediction model, calculate the objective function value according to the model output, judge whether the preset optimization standard is met based on the objective function value, and determine the optimal antenna design parameters.

9. An electronic device, characterized in that: The device includes a processor and a memory; a program is stored in the memory, and the program is loaded and executed by the processor to implement a design method for a dual-frequency Yagi antenna as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The storage medium stores a program, and when the program is executed by the processor, it is used to implement the design method of a dual-frequency Yagi antenna as described in any one of claims 1 to 7.

Images