Parameter optimization method of w-band wideband low-vswr resonator filter network

Abstract

The application discloses a parameter optimization method of a W-waveband wideband low standing wave ratio resonant filter network and relates to the field of resonant filter parameter optimization. The application adopts a variation method to optimize design variables, and utilizes the influence of the partial derivative of an objective function on design parameters to realize a more efficient optimization process, thereby reducing the standing wave ratio and the insertion loss.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of resonant filter parameter optimization, in particular to a parameter optimization method of a W-band wideband low standing wave ratio resonant filter network. BACKGROUND

[0002] The existing filter design technology is increasingly widely used in the millimeter wave band, especially in the W-band (75-110GHz) frequency band. With the development of wireless communication and radar systems, the performance requirements for filters are also increasingly high. In the W-band, due to the high frequency range, the traditional filter design technology is usually difficult to meet the requirements of low standing wave ratio and low insertion loss, so it is usually necessary to use electromagnetic simulation software to analyze and optimize the frequency response of the filter. In the design of the filter, the center frequency of the filter is usually first set by determining the frequency range and bandwidth target, which may involve overall analysis of the bandwidth performance to meet the system operating frequency band requirements. In order to achieve stable filtering effect in high frequency environment, the existing technology usually initializes the geometric parameters (such as the length, width and thickness of the resonator) of the filter according to the selected center frequency, and selects appropriate materials to ensure the controllability of electromagnetic performance. For example, in the selection of resonators, based on the characteristics of the dielectric constant and conductivity of the material, the electromagnetic response can be optimized to achieve lower loss in the target frequency band. However, these steps are usually only preliminary adjustments in the design process, which need to be further optimized.

[0003] While existing filter design techniques have achieved some success, particularly in lower-frequency bands, they still face several insurmountable drawbacks in the high-frequency W-band. First, due to the high frequency range of the W-band, electromagnetic waves experience significant losses during propagation, making it difficult to maintain a low insertion loss. Traditional filter design methods often struggle to optimize the standing wave ratio while maintaining low insertion loss, resulting in designs that lack good frequency selectivity and stability at high frequencies. Second, in existing techniques, the geometric parameters and material properties of the filter are often fixed at initialization. Although adjustments are made during the optimization process, the overall optimization accuracy and efficiency are low. Furthermore, due to the short wavelength of electromagnetic waves in the W-band, the filter design process requires precise geometric parameters. Conventional optimization methods used in existing techniques typically only achieve linear or simple nonlinear optimization, which limits optimization effectiveness and makes it difficult to achieve high-precision frequency response adjustment. Third, the weight allocation of the objective function in existing techniques is not flexible enough, making it impossible to allocate weights for specific frequency points. This makes it difficult to simultaneously meet performance requirements at multiple frequency points during the design process. Existing electromagnetic simulation and optimization methods for W-band filter design lack a systematic approach to the relationship between standing wave ratio (SWR) and insertion loss. This is particularly true for high-frequency designs, where existing technologies fail to provide effective optimization methods for the relationship between SWR and smoothness. Traditional optimization processes lack more advanced mathematical tools, such as the calculus of variations and functional design, making it difficult to effectively control the smoothness of the frequency response curve, resulting in unstable filter performance. Summary of the Invention

[0004] The present invention effectively solves the problem of limited optimization effect of the prior art in the high frequency band by optimizing the distribution setting of frequency points, introducing the variational method to design functionals, and establishing a weight distribution model for the relationship between standing wave ratio and insertion loss.

[0005] The parameter optimization method of the W-band wide-band low standing wave ratio resonant filter network includes the following steps:

[0006] S1. Set the center frequency and target bandwidth of the filter according to the W-band frequency requirements, select the optimized frequency points for distribution setting, set the initial geometric parameters as design variables based on the frequency points, and configure the material properties;

[0007] S2. Perform preliminary simulation analysis on the initialized filter structure using electromagnetic simulation software to obtain VSWR and insertion loss data within the bandwidth, and calculate VSWR and insertion loss curves at each frequency point.

[0008] S3. According to the standing wave ratio and the insertion loss of each frequency point obtained by calculation, a target function model with a weight is established, the target function contains numerical description of the standing wave ratio and the insertion loss, and the influence of each frequency point is calculated according to the weight;

[0009] S4. A functional is designed by the variational method, the functional contains design variables corresponding to the frequency response, and the relationship between the response function and the standing wave ratio and the smoothness is established by setting the response function;

[0010] S5. The design variables of the target function are analyzed by the variational method, the influence of each design parameter on the standing wave ratio and the insertion loss is analyzed, each parameter is optimized, and the key size and material properties of the initial geometric structure are adjusted according to the optimization result;

[0011] S6. Simulation feedback is performed, when the change of the target function after optimization iteration is less than a set threshold, it is judged that the current optimization has converged, and the result after adjustment is confirmed by full-band simulation; when the change of the target function after optimization iteration is greater than or equal to the set threshold, the optimization is continued, and step S5 is iteratively executed.

[0012] Further, in step S1, the initial geometric parameters include the length, width and thickness of the resonator.

[0013] Further, in step S1, the material properties include the dielectric constant and the conductivity.

[0014] Further, in step S3, the following sub-steps are specifically included:

[0015] S301. According to the standing wave ratio and the insertion loss of each frequency point obtained by calculation, the deviation of the actual standing wave ratio from the target standing wave ratio is calculated respectively;

[0016] S302. A penalty term is constructed in a square form, when the standing wave ratio deviation exceeds the target range, the penalty term increases with the increase of the deviation;

[0017] S303. An optimization target function containing the penalty term is constructed.

[0018] Further, in step S302, the penalty term is specifically a nonlinear penalty term, which is expressed as:

[0019]

[0020] Wherein, the ω i represents the i-th frequency point, the x i represents the i-th design variable, the P(ω i , x i ) represents the penalty term, and the VSWR(ω i , x iVSWR(ωi, xi) represents the actual VSWR of the ith frequency point at the ith design variable, the VSWR target represents the target VSWR.

[0021] Further, in the step S303, the objective function is specifically represented as:

[0022]

[0023] wherein, the J represents the objective function, the n represents the total number of frequency points, the i represents the index of the frequency point, the γ represents the weight coefficient of the influence of balancing the VSWR and the insertion loss, the L(ω i ) represents the actual insertion loss of the ith frequency point, the L target represents the target value of the insertion loss, the β i represents the weight of each frequency point.

[0024] Further, the step S4 specifically comprises the following sub-steps:

[0025] S401. Constructing a functional for describing the VSWR and the frequency response smoothness;

[0026] S402. Taking the partial derivative of each design variable, assuming that the partial derivative of the functional is and solving this partial derivative;

[0027] S403. Setting a response function and associating the response function with the frequency response.

[0028] Further, in the step S401, the functional is specifically represented as:

[0029]

[0030] wherein, the J(x i ) represents the functional for describing the VSWR and the frequency response smoothness, the f1 and f2 are used to represent the frequency range [f1, f2], the k1 represents the weight coefficient of the VSWR, the VSWR(ω i , x i ) represents the actual VSWR of the ith frequency point at the ith design variable, the VSWR target represents the target VSWR, the k2 represents the weight coefficient of the smoothness term, the f represents the frequency point, the x i represents the ith design variable, and the ω i represents the ith frequency point.

[0031] Further, in the step S402, the specific flow of solving this partial derivative is represented as:

[0032]

[0033] Wherein, the J represents a target function, the d represents derivation, and the Sensitivity of standing wave ratio to design variable is represented.

[0034] Further, the response function is specifically represented as:

[0035]

[0036] Wherein, the R(x i ) represents a response function.

[0037] The beneficial effects of the application are:

[0038] The application adopts variational method to optimize design variable, and utilizes partial derivative of target function to calculate influence on design parameter, so that more efficient optimization process is realized, thereby reducing standing wave ratio and insertion loss. BRIEF DESCRIPTION OF DRAWINGS

[0039] Figure 1 The method flowchart of the parameter optimization method of the W-band wideband low standing wave ratio resonant filter network provided by the embodiment of the application. DETAILED DESCRIPTION

[0040] The technical solutions of the application will be further described in detail below with reference to the drawings, but the protection scope of the application is not limited to the following description.

[0041] In order to make the purpose, technical solutions and advantages of the application clearer and more understandable, the application is further described in detail in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the application, and are not used to limit the application, that is, the described embodiments are only a part of the embodiments of the application, but not all the embodiments. The components of the embodiments of the application described and shown in the drawings can be arranged and designed in various different configurations.

[0042] Therefore, the detailed description of the embodiments of the application provided in the drawings below is not intended to limit the scope of the claimed application, but only represents selected embodiments of the application. Based on the embodiments of the application, all other embodiments obtained by those skilled in the art without creative work belong to the scope of protection of the application. It should be noted that the relationship terms such as "first" and "second" and the like are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between the entities or operations.

[0043] Moreover, the term "include", "includes" or any other variation thereof, is intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the statement "including a" does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0044] The features and characteristics of the present application are further described in detail below with reference to the embodiments.

[0045] wherein, as Figure 1 The parameter optimization method of the W-band wideband low standing wave ratio resonant filter network comprises the following steps:

[0046] S1. According to the requirements of the W-band frequency range, the center frequency and target bandwidth range of the filter are set, the optimized frequency points are distributed and set, the initial geometric parameters are set as design variables based on the frequency points, and the material properties are configured;

[0047] S2. The initialized filter structure is analyzed by electromagnetic simulation software, the standing wave ratio and insertion loss data in the bandwidth range are obtained, and the standing wave ratio curve and loss curve of each frequency point are calculated;

[0048] S3. According to the calculated standing wave ratio and insertion loss of each frequency point, a target function model with weight is established, the target function contains numerical description of standing wave ratio and insertion loss, and the influence of each frequency point is calculated according to the weight;

[0049] S4. A functional is designed by variational method, the functional contains design variables corresponding to frequency response, and the relationship between standing wave ratio and smoothness is established by setting response function;

[0050] S5. The design variables of the target function are analyzed by variational method, the influence of each design parameter on the standing wave ratio and insertion loss is analyzed, each parameter is optimized, and the key size of the initial geometric structure and the material properties are adjusted according to the optimization result;

[0051] S6. Simulation feedback is performed, when the change of the target function after optimization iteration is less than the set threshold, it is judged that the current optimization has converged, and the adjusted result is confirmed by full-band simulation; when the change of the target function after optimization iteration is greater than or equal to the set threshold, the optimization is continued, and step S5 is iterated.

[0052] Further, in step S1, the initial geometric parameters include resonator length, width, and thickness.

[0053] Further, in the step S1, the material properties include dielectric constant and conductivity.

[0054] Specifically, in the step S1, based on the frequency requirement of the W-band, the center frequency and the target bandwidth range of the filter are selected to ensure the coverage of the required communication bandwidth. According to the distribution of different frequency points in the range, the initial design geometric parameters are determined and used as design variables. The selected frequency points can cover the key points in the frequency response curve, which can include the upper and lower limits of the bandwidth, the center frequency position, etc. The material properties are configured according to the high-frequency requirements of the filter, which include dielectric constant, conductivity, etc. The above parameters directly affect the VSWR and insertion loss performance of the filter at the W-band frequency. The above implementation steps select a more comprehensive frequency point for optimization at the initial design stage, making the subsequent optimization process more targeted and accurate.

[0055] Further, in the step S2, the initial structure is simulated and analyzed by electromagnetic simulation software to obtain the VSWR and IL data in the specified bandwidth range, and the simulation results are summarized into curves to analyze the VSWR and loss performance of each frequency point. This implementation can be used to identify defects in the initial design, such as frequency points with excessively high VSWR or abnormally high insertion loss. Compared with the traditional method of simulating and analyzing only at specific frequency points, this process can more comprehensively understand the response of the filter in the entire bandwidth range and provide basic data support for the subsequent optimization process.

[0056] Further, the step S3 specifically includes the following sub-steps:

[0057] S301. According to the calculated VSWR and IL of each frequency point, the deviation of the actual VSWR from the target VSWR is calculated respectively; specifically, according to the simulation results, the deviation between the actual VSWR of each frequency point and the preset target VSWR is calculated, and the part exceeding the target is identified. This implementation is used to identify the frequency range that needs to be optimized most, so as to more efficiently allocate optimization resources.

[0058] S302. A penalty term is constructed in the form of square, which increases with the increase of the deviation when the VSWR deviation exceeds the target range;

[0059] S303. An optimization objective function containing the penalty term is constructed.

[0060] Further, in the step S302, the penalty term is a nonlinear penalty term, which is represented as:

[0061]

[0062] wherein, ω i represents the i-th frequency point, and xi represents the i-th design variable, the P(ω i , x i ) represents a penalty term, the VSWR(ω i , x i ) represents an actual standing wave ratio of the i-th frequency point at the i-th design variable, and the VSWR target represents a target standing wave ratio.

[0063] Specifically, for the standing wave ratio deviation ΔVSWR(ω i , x i ), it is necessary to note that the specific calculation process is to calculate the deviation between the actual standing wave ratio VSWR(ω i , x i ) and the target standing wave ratio VSWR target , that is, ΔVSWR(ω i , x i ) = VSWR(ω i , x i )-VSWR target , so as to quantify the degree of deviation of the standing wave ratio from the target range. In addition, when ΔVSWR(ω i , x i ) exceeds VSWR target , that is, ΔVSWR(ω i , x i )>0, it indicates that the actual standing wave ratio is higher than the target standing wave ratio, that is, the filter performs worse than expected at this frequency point, and needs to be adjusted to prompt the optimization algorithm to adjust the design parameters to reduce the standing wave ratio to the target value; otherwise, it indicates that the standing wave ratio performs well, so it does not need to be punished, and the optimization algorithm can not be adjusted.

[0064] Further, the penalty term is constructed in a square form, and when the standing wave ratio deviation exceeds the target range, the priority of the frequency point is increased by increasing the penalty value of the deviation, so that the optimization process focuses on the standing wave ratio control of the key frequency point. The above embodiment ensures that the optimization algorithm can effectively reduce the excessively high standing wave ratio in subsequent iterations to obtain a smooth response curve.

[0065] Further, in the step S303, the optimization objective function is specifically represented as:

[0066]

[0067] Wherein, the J represents the objective function, the n represents the total number of frequency points, the i represents the index of the frequency point, the γ represents a weight coefficient balancing the influence between the standing wave ratio and the insertion loss, the L(ω i ) represents the actual insertion loss of the i-th frequency point, and the L targeta target value of insertion loss, the β i The weight of each frequency point is represented. Specifically, the penalty term is integrated into the optimization objective function, making it more targeted and constrained. Compared with the traditional objective function model, the control of the frequency points exceeding the standing wave ratio range is strengthened, and the optimization of the objective function in the frequency band is more balanced through the weighted penalty term adjustment.

[0068] Further, the step S4 specifically includes the following sub-steps:

[0069] S401. Construct a functional for describing the standing wave ratio and frequency response smoothness;

[0070] S402. Partially derive each design variable, and set the partial derivative of the functional as and solve this partial derivative;

[0071] S403. Set the response function, and associate the response function with the frequency response.

[0072] Further, in the step S401, the functional is specifically represented as:

[0073]

[0074] wherein the J(x i ) represents a functional for describing the standing wave ratio and frequency response smoothness, the f1 and f2 are used to represent the frequency range [f1, f2], the k1 represents a weight coefficient of the standing wave ratio, the VSWR(ω i , x i ) represents an actual standing wave ratio of the i-th frequency point at the i-th design variable, the VSWR target represents a target standing wave ratio, the k2 represents a weight coefficient of the smoothness term, the f represents a frequency point, the x i represents the i-th design variable, and the ω i represents the i-th frequency point. Specifically, the functional for describing the standing wave ratio and frequency response smoothness is constructed, and the changes of the standing wave ratio and insertion loss are integrated into the variational method model. Through the construction of the functional, the standing wave ratio and insertion loss change smoothly in the full frequency band, so as to achieve a smooth and undulating response.

[0075] Further, in the step S402, the specific flow of solving the partial derivative is represented as:

[0076]

[0077] wherein the J represents the objective function, the d represents the derivative, the sensitivity of the standing wave ratio to the design variables. Specifically, the partial derivative of the functional with respect to each design variable is calculated to determine the influence of each variable on the standing wave ratio and the insertion loss response. This process is used to subsequently adjust the optimization direction and provide specific optimization directions for different design parameters. The calculation of the partial derivative makes the response curve optimization more accurate and enables the optimal path of the standing wave ratio and insertion loss optimization to be found more quickly. In addition, the sensitivity of the standing wave ratio to the design variables is measured by calculating the rate of change of the standing wave ratio (VSWR) with respect to each design variable (such as the geometric parameters of the resonator, the dielectric constant of the material, etc.) to measure the degree of dependence of the standing wave ratio on these variables. The core of this process is to understand how the response of the standing wave ratio changes under different design parameter values, so as to find the optimal parameter configuration. The sensitivity value of the standing wave ratio to the design variables provides a quantitative direction of the influence of each design variable on the standing wave ratio. When the sensitivity of a variable is large, it indicates that the variable has a significant influence on the standing wave ratio; conversely, when the sensitivity is small, it indicates that the variable has little influence. Through this information, the variable that has a greater influence on the standing wave ratio can be adjusted first, thereby more effectively optimizing the filter performance. In order to minimize the functional, the partial derivative of the standing wave ratio with respect to each design variable (i.e. the sensitivity) needs to be calculated, and combined with a numerical optimization method to adjust the variable value. The sensitivity value will serve as the direction and size of the gradient descent guidance, and by gradually reducing the value of the functional, the optimization and smoothness improvement of the standing wave ratio within the bandwidth are achieved.

[0078] Further, the sensitivity of the standing wave ratio to the design variables is calculated as follows:

[0079]

[0080] where a small increment Δx is applied to each variable by the finite difference method i to recalculate the standing wave ratio change to estimate the sensitivity. Preferably, the parameter scanning function in the electromagnetic simulation software can also be used to adjust each variable to observe the change of the standing wave ratio, thereby obtaining the sensitivity of the standing wave ratio to the variable.

[0081] Further, in the step S403, the response function is specifically represented as:

[0082]

[0083] where R(x i ) represents the response function. Specifically, by optimizing the standing wave ratio value and the frequency response smoothness of each frequency point through the relationship between the response function and the frequency response, it is ensured that the final response curve meets the low standing wave ratio requirement of the W-band wide frequency band.

[0084] Further, as a preferred embodiment of the above embodiment, an exemplary implementation process is proposed:

[0085] First, according to the W-band wideband requirement, the main working frequency range of the filter is determined, which is between 75GHz and 110GHz, the center frequency is selected as the design core, and several key frequency points are set on both sides of the center frequency to ensure the performance optimization in the wideband coverage. For the initial parameters of the filter structure, including the size (length, width, thickness) of the resonator, the dielectric constant and conductivity of the configured material, the initial design is ensured to meet the basic requirements;

[0086] The initial filter model is simulated by electromagnetic simulation software to obtain the voltage standing wave ratio (VSWR) and insertion loss curve of each frequency point. By analyzing the VSWR and insertion loss values of the preliminary simulation, the performance of each frequency point is understood;

[0087] A target function with weights is constructed, and the VSWR and insertion loss of each frequency point in the target function are described according to the preset weights, and the weights are allocated according to the importance of the frequency points;

[0088] The target function is processed by variational method to design a corresponding functional, which includes variables related to the frequency response of the filter, such as geometric size parameters and material properties. The response function of the functional is established to integrate the relationship between VSWR and smoothness into the optimization framework to balance the VSWR optimization and frequency response smoothness;

[0089] The partial derivative is used to quantify the influence of each design variable on the VSWR and insertion loss, and the design variables, including key dimensions and material properties, are adjusted according to the direction and amplitude of the derivative calculation. After each adjustment, the simulation is repeated to obtain new VSWR and insertion loss data, and the frequency response is gradually optimized;

[0090] The change of the optimized target function is observed. If the change value of the target function is less than the preset convergence threshold, it means that the optimization process has basically stabilized, and the performance of the current parameters is confirmed by full-band simulation. If the change of the target function is greater than or equal to the set threshold, it means that the optimization is not complete, and the iteration optimization continues until the convergence condition is reached.

[0091] In addition, as a preferred embodiment of the above embodiment, in the design of a wideband filter, in order to make the VSWR and insertion loss of the edge frequency points have more significant influence on the overall filter performance, the weights of the edge frequency points in the target function are adjusted separately, so that the optimization process pays more attention to the performance of the edge frequencies. For example, according to the specific requirements of the W-band wideband, the frequency range of the bandwidth is divided into "edge frequency range" and "center frequency range", and the weights of the upper and lower edge frequency points are higher:

[0092]

[0093] Wherein, the a edgeωa center ωc min ωa max ωa

[0094] The above description is merely that of the preferred embodiments of the application and is not to be taken in a limiting sense but is made merely for the purpose of disclosure with the fullest possible scope of the application being apparent by the teachings herein, and it is understood that variations and modifications can be effected within the scope of the concepts described above, by a person of ordinary skill in the art, without departing from the scope of the application. It is therefore intended to cover in the appended claims all such changes and modifications that come within the scope of the application.

Claims

1. A method for parameter optimization of a W-band wideband low-vswr resonator filter network, characterized in that, The method comprises the following steps: S1. According to the requirements of the W-band frequency band, the center frequency and the target bandwidth range of the filter are set, the optimized frequency points are distributed, the initial geometric parameters are set as design variables based on the frequency points, and the material properties are configured; S2. The initialized filter structure is simulated and analyzed by an electromagnetic simulation software, the standing wave ratio and the insertion loss data in the bandwidth range are obtained, and the standing wave ratio curve and the loss curve of each frequency point are calculated; S3. According to the calculated standing wave ratio and insertion loss of each frequency point, a target function model with a weight value is established, the target function includes the numerical description of the standing wave ratio and the insertion loss, and the influence of each frequency point is calculated according to the weight; S4. A functional is designed by the variation method, the functional includes the design variables corresponding to the frequency response, and the relationship between the standing wave ratio and the smoothness is established by setting a response function; S5. The design variables of the target function are analyzed by the variation method, the influence of each design parameter on the standing wave ratio and the insertion loss is analyzed, each parameter is optimized, and the key size of the initial geometric structure and the material properties are adjusted according to the optimization result; S6. Simulation feedback is performed, when the change of the target function after optimization iteration is less than a set threshold, it is judged that the current optimization has converged, and the adjusted result is confirmed by full-band simulation; when the change of the target function after optimization iteration is greater than or equal to the set threshold, the optimization is continued, and step S5 is iterated; The step S4 specifically comprises the following sub-steps: S401. A functional for describing the standing wave ratio and the frequency response smoothness is constructed; S402. Take the partial derivative of each design variable, set the partial derivative of the functional as and solve this partial derivative; S403. A response function is set, and the response function is associated with the frequency response.

2. The method of claim 1, wherein the W-band wideband low- VSWR resonator filter network is optimized by, In the step S1, the initial geometric parameters include the resonator length, width and thickness.

3. The method of claim 1, wherein the W-band wideband low- VSWR resonator filter network is optimized by, In the step S1, the material properties include the dielectric constant and the conductivity.

4. The method of claim 1, wherein the W-band wideband low- VSWR resonator filter network is optimized by, The step S3 specifically comprises the following sub-steps: S301. According to the calculated standing wave ratio and insertion loss of each frequency point, the deviation of the actual standing wave ratio from the target standing wave ratio is calculated; S302. A penalty term is constructed in a square form, when the standing wave ratio deviation exceeds the target range, the penalty term increases with the increase of the deviation; S303. An optimization target function containing the penalty term is constructed.

5. The method of claim 4, wherein the W-band wideband low- VSWR resonator filter network is optimized by, In the step S302, the penalty term is specifically a nonlinear penalty term, which is represented as: ； Wherein the represents the i-th frequency point, and the represents the i-th design variable, and the represents a penalty term, and the represents the actual standing wave ratio of the i-th frequency point at the i-th design variable, and the represents a target standing wave ratio.

6. The method of claim 5, wherein the W-band wideband low- VSWR resonator filter network is optimized by: In the step S303, the optimization target function is specifically represented as: ； Wherein the represents a target function, the represents the total number of frequency points, the represents the index of the frequency point, the represents a weight coefficient of the influence of balancing the standing wave ratio and the insertion loss, the represents the actual insertion loss of the i-th frequency point, the represents the target value of the insertion loss, the represents the weight of each frequency point.

7. The method of claim 1, wherein the W-band wideband low- VSWR resonator filter network is optimized by, In the step S401, the functional is specifically represented as: ； wherein the denotes a functional for describing standing wave ratio and frequency response smoothness, the and is used to denote a frequency range , the denotes a weight coefficient of the standing wave ratio, the denotes an actual standing wave ratio of the i-th frequency point at the i-th design variable, the denotes a target standing wave ratio, the denotes a weight coefficient of the smoothness term, the denotes a frequency point, the denotes the i-th design variable, the denotes the i-th frequency point.

8. The method of claim 6, wherein the W-band wideband low- VSWR resonator filter network is optimized by, In the step S402, the specific flow of solving the partial derivative is represented as: ； Wherein, the Indicates the objective function, and the Indicates the derivative, and the Indicates the sensitivity of the standing wave ratio to the design variable.

9. The method of claim 8, wherein the W-band wideband low-vswr resonator filter network is optimized by, In the step S403, the response function is specifically represented as: ； wherein the represents a response function.

Images