Method and system for designing filter parameters on basis of bayesian optimization, and related device

Abstract

The present invention is applicable to the technical field of filter optimization design, and in particular relates to a method and system for designing filter parameters on the basis of Bayesian optimization, and a related device. The method comprises the following steps: constructing an optimization model on the basis of parameters of a surface acoustic wave filter that need to be optimized, and performing optimization calculation on the parameters on the basis of the optimization model, so as to obtain optimized parameters; constructing a preset optimization algorithm based on a Bayesian optimization method; and performing iterative calculation by taking the optimized parameters as inputs of the preset optimization algorithm, and obtaining a parameter optimization result corresponding to the parameters. Compared with the prior art, the present invention uses an HCT-FEM method to replace a COM to calculate the parameters of a filter, which improves the calculation precision of filter parameters when the number of IDTs is small; moreover, a Bayesian optimization method is used during parameter optimization, which avoids the problems of gradient optimization relying on initial values and the optimization efficiency of global optimization being low, and improves the iteration speed of the parameter optimization process.

Description

Method, system and related equipment for designing filter parameters based on Bayesian optimization Technical Field

[0001] The present invention is applicable to the technical field of filter optimization design, and in particular relates to a method, system and related equipment for designing filter parameters based on Bayesian optimization. Background Art

[0002] Surface acoustic wave (SAW) filters, due to their miniaturization, low cost, and high selectivity, have become indispensable components in electronic products such as mobile phones. Designing SAW filters requires simulation tools, typically using the MBVD (modified Butterworth-Van Dyke) equivalent circuit model, the COM (Couple Of Model) model, and the hierarchical cascade-based finite element method (HCT-FEM).

[0003] Low insertion loss and high out-of-band rejection filters are required in the RF Rx band. These filters include DMS (Double Mode SAW) resonators and ladder resonators. However, the equivalent circuit model cannot simulate DMS resonators. Therefore, the COM model must be combined with the COM model during the simulation process. The COM model usually simulates in seconds, but the COM model has poor simulation accuracy for IDT (Interdigital Transducer) resonators with a small number of elements, which greatly restricts the design tape-out speed.

[0004] To address this issue, related technologies use HCT-FEM to simulate filters. This method is usually completed in minutes, which improves simulation accuracy and increases design tape-out speed. However, even though the HCT-FEM method is much faster than the ordinary finite element (FEM) method, it still cannot meet the requirements of the gradient optimization algorithm. This is because the gradient optimization algorithm is very dependent on the initial values ​​of the DMS resonator's geometric parameters and has no analytical gradient. Therefore, finite differences are required to calculate the gradient. When finite differences are used to calculate the gradient, multiple HCT-FEM calls are required. Global optimization is very limited by the number of optimization variables and is usually combined with the HCT-FEM method. When the number of resonator variables is around 30, approximately 5,000 HCT-FEM calls are required for calculation. The multiple calls seriously restrict the overall efficiency of the simulation.

[0005] Summary of the Invention

[0006] The present invention provides a method, system and related equipment for designing filter parameters based on Bayesian optimization, aiming to solve the technical problems of poor simulation accuracy and slow iteration speed of filter parameters based on COM model in the prior art.

[0007] To solve the above technical problems, in a first aspect, the present invention provides a method for designing filter parameters based on Bayesian optimization, the method comprising the following steps:

[0008] S101, constructing an optimization model according to the parameters of the surface acoustic wave filter to be optimized, and performing optimization calculation on the parameters based on the optimization model to obtain optimized parameters;

[0009] S102. Constructing a preset optimization algorithm based on the Bayesian optimization method;

[0010] S103: performing iterative calculations using the optimization parameters as inputs of the preset optimization algorithm to obtain parameter optimization results corresponding to the parameters.

[0011] Furthermore, the optimization model satisfies the following relationship: Min IL(x); stlb≤x≤ub; G(x)≤0;

[0012] Wherein, IL represents the insertion loss of the surface acoustic wave filter, x represents the parameter, lb and ub represent the lower limit and upper limit of the parameter respectively, and G(x) represents the preset additional parameter of the surface acoustic wave filter.

[0013] Furthermore, the preset optimization algorithm includes the following sub-steps:

[0014] S1021. Determine a calculation sample for the parameters of the preset optimization algorithm;

[0015] S1022, using the HCT-FEM method to calculate parameter variables of the surface acoustic wave filter, where the parameter variables include the insertion loss IL and the preset additional parameter G(x);

[0016] S1023, establishing a proxy function using the calculation sample and the parameter variable as a function;

[0017] S1024. Calculate the minimum value of the calculation samples in the proxy function using a preset solution algorithm;

[0018] S1025: Determine whether the iteration of the preset optimization algorithm reaches a preset threshold:

[0019] If so, output the minimum value as the optimal solution of the parameter;

[0020] If not, the HCT-FEM method is called to calculate the parameter variable corresponding to the minimum value, the number of iterations is increased by 1, and then the process returns to step S1022.

[0021] Furthermore, step S1021 is specifically as follows:

[0022] Stratified sampling is performed between the lower limit lb and the upper limit ub of the parameter according to Latin hypercube sampling to generate the calculation sample.

[0023] Furthermore, step S1023 further includes the steps of:

[0024] The position of a point of a new sample added to the calculated sample is calculated using the EI acquisition function, and the proxy function is updated according to the position of the point.

[0025] Furthermore, the proxy function is a Gaussian process regression model.

[0026] Furthermore, the preset solution algorithm is a genetic algorithm.

[0027] In a second aspect, the present invention further provides a system for designing filter parameters based on Bayesian optimization, comprising:

[0028] A parameter modeling module is used to construct an optimization model according to the parameters of the surface acoustic wave filter that need to be optimized, and to optimize the parameters based on the optimization model to obtain the optimized parameters;

[0029] Bayesian modeling module, used to build preset optimization algorithms based on Bayesian optimization methods;

[0030] The Bayesian optimization module is used to perform iterative calculations using the optimization parameters as inputs of the preset optimization algorithm to obtain parameter optimization results corresponding to the parameters.

[0031] In a third aspect, the present invention also provides a computer device comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the method for designing filter parameters based on Bayesian optimization as described in any one of the above embodiments are implemented.

[0032] In a fourth aspect, the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps in the method for designing filter parameters based on Bayesian optimization as described in any one of the above embodiments are implemented.

[0033] The beneficial effect achieved by the present invention lies in proposing a method for designing filter parameters based on Bayesian optimization. Compared with the existing technology, this method uses the HCT-FEM method instead of the COM model to calculate the parameters of the filter, thereby improving the calculation accuracy of the filter parameters when the number of IDT roots is small. In addition, the Bayesian optimization method is used in the parameter optimization process to replace the gradient optimization and global optimization methods of the existing technology, thereby avoiding the problems of gradient optimization relying on initial values ​​and low optimization efficiency of global optimization, and improving the iteration speed of the parameter optimization process. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] 1 is a flowchart of the steps of a method for designing filter parameters based on Bayesian optimization according to an embodiment of the present invention;

[0035] FIG2 is a flow chart of a preset optimization algorithm based on the Bayesian optimization method constructed in an embodiment of the present invention;

[0036] FIG3 is a schematic diagram of parameter optimization results provided by an embodiment of the present invention;

[0037] 4 is a schematic diagram of the structure of a system for designing filter parameters based on Bayesian optimization according to an embodiment of the present invention;

[0038] FIG5 is a schematic diagram of the structure of a computer device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0039] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0040] Please refer to FIG1 , which is a flowchart of a method for designing filter parameters based on Bayesian optimization according to an embodiment of the present invention. The method includes the following steps:

[0041] S101 : Constructing an optimization model according to the parameters of the surface acoustic wave filter to be optimized, and performing optimization calculation on the parameters based on the optimization model to obtain optimized parameters.

[0042] Specifically, the optimization model constructed by the present invention is used to calculate the admittance of surface acoustic wave devices including ladder filters and DMS filters. The optimization model satisfies the following relationship: Min IL(x); stlb≤x≤ub; G(x)≤0;

[0043] Wherein, IL represents the insertion loss of the surface acoustic wave filter, x represents the parameter, lb and ub represent the lower limit and upper limit of the parameter respectively, and G(x) represents the preset additional parameter of the surface acoustic wave filter.

[0044] The parameters include at least one of the period length, metallization rate, aperture, spacing, and number of roots. For example, for a ladder filter, the specific parameters include n*Pitch (period length), Duty (metallization rate), W (aperture), Gap (spacing), and N_IDT (number of roots); and for a DMS filter, the specific parameters include the Pitch (period length), Duty (metallization rate) of each IDT (interdigitated transducer), and the gap (spacing) of the reflector between the IDTs. For some devices with special structures, such as a gradient DMS filter, geometric information such as the number of gradient roots and the pitch (spacing) of the gradient part will also be involved. The preset additional parameters include information such as the ripple (ripple), standing wave ratio (vswr), and out-of-band rejection (out-rejection) corresponding to the parameter x. In actual implementation, the variable x is uniformly restricted to less than or equal to 0 by multiplying by a negative sign.

[0045] S102: Construct a preset optimization algorithm based on the Bayesian optimization method.

[0046] Specifically, please refer to Figure 2, which is a flow chart of a preset optimization algorithm based on the Bayesian optimization method constructed in an embodiment of the present invention. Bayesian optimization is a method that uses Bayes' theorem to guide the search to find the minimum or maximum value of the objective function. It is mainly used to solve black box optimization problems. In an embodiment of the present invention, the calculation sample is used as the hyperparameter x of Bayesian optimization, and the corresponding preset additional parameter G(x) is the hyperparameter y. Its purpose is to calculate the unknown function of x and y, thereby calculating the minimum value of the hyperparameter x. The preset optimization algorithm includes the following sub-steps:

[0047] S1021. Determine a calculation sample of the parameters for the preset optimization algorithm.

[0048] Step S1021 is specifically as follows:

[0049] Stratified sampling is performed between the lower limit lb and the upper limit ub of the parameter based on Latin hypercube sampling to generate the calculation samples. Latin hypercube sampling is a sampling method with a low number of iterations. Compared to methods such as Monte Carlo sampling, Latin hypercube sampling can maintain the independence of variables during the sampling process and ensure that outlying events are accurately reflected in the output, thereby improving sampling efficiency. In an embodiment of the present invention, stratified sampling is performed between the lower limit lb and the upper limit ub of the parameter based on Latin hypercube sampling. This allows for the extraction of more representative data from the parameters of a pre-designed surface acoustic wave filter, facilitating subsequent parameter control in Bayesian optimization.

[0050] S1022 : Calling the HCT-FEM method to calculate parameter variables of the surface acoustic wave filter, where the parameter variables include the insertion loss IL and the preset additional parameter G(x).

[0051] Specifically, as shown in Figure 2, the calculation sample is used as the parameter x in the Bayesian optimization process, which is consistent with the parameter x of the surface acoustic wave filter in step S101. Correspondingly, the parameter y in the Bayesian optimization process is the preset additional parameter G(x) of the surface acoustic wave filter in step S101. In a single calculation process, IL(x) and G(x) can both be calculated by the HCT-FEM method. At this time, the HCT-FEM method optimizes and calculates each of the parameters separately, and transforms and cascades the results of the respective calculations to obtain the corresponding admittance parameters, which are then transformed into scattering parameters. Finally, the scattering parameters of all devices are cascaded to obtain the complete optimized parameters of the surface acoustic wave filter.

[0052] S1023: Establish a proxy function using the calculation sample and the parameter variable as a function.

[0053] S1023 also includes the steps of:

[0054] The position of a point of a new sample added to the calculated sample is calculated using the EI acquisition function, and the proxy function is updated according to the position of the point.

[0055] During Bayesian optimization, the objective function is unknown, so a surrogate function is constructed to pre-fit the desired curve. In this embodiment, the surrogate function is a Gaussian process regression model. Based on the surrogate function, the expected improvement (EI) acquisition function can be used to collect more points near possible minimum points during the calculation process, or to collect more points in previously unsampled areas. The surrogate function expression is then updated sequentially based on the positions of the collected points, bringing it closer to the actual expression of the objective function.

[0056] S1024: Calculate the minimum value (x_min) of the calculation samples in the proxy function using a preset solution algorithm.

[0057] To improve computational efficiency, the embodiment of the present invention uses the preset solving algorithm to improve computational efficiency during the proxy function update process. The preset solving algorithm is a genetic algorithm. In one possible implementation, step S1024 can also be performed using machine learning or other computing power-based methods to accelerate the updating of the proxy function.

[0058] S1025: Determine whether the iteration of the preset optimization algorithm reaches a preset threshold:

[0059] If so, output the minimum value as the optimal solution of the parameter;

[0060] If not, call the HCT-FEM method to calculate the parameter variable (IL(x_min), G(x_min)) corresponding to the minimum value, increase the number of iterations by 1, and then return to step S1022.

[0061] S103: performing iterative calculations using the optimization parameters as inputs of the preset optimization algorithm to obtain parameter optimization results corresponding to the parameters.

[0062] For example, according to the above method, an embodiment of the present invention uses a normal-saw process and a 42° YX LiTaO3 substrate to design a band 2028 filter. Three ladder resonators plus one DMS resonator are used, where the DMS resonator contains multiple IDTs with a smaller number of roots. The final parameter optimization results are shown in Figure 3. Figure 3 shows the optimization results through S parameters (Scatter parameters, scattering parameters), where S21 represents the forward transmission coefficient and S11 represents the reflection coefficient when the port in the filter is matched. Through continuous iterative calculation, the optimal parameter values ​​obtained according to the curve in Figure 3 are IL = 2.0db, vswr = 1.7, out-of-band suppression of 0.748, and a noise value of -30db at 0.832GHz. In other words, the optimization method proposed in the embodiment of the present invention can well obtain the required filter parameters, and in actual use, combined with the Bayesian optimization method, the number of calls to the HCT-FEM method is reduced from thousands to hundreds.

[0063] The beneficial effect achieved by the present invention lies in proposing a method for designing filter parameters based on Bayesian optimization. Compared with the existing technology, this method uses the HCT-FEM method instead of the COM model to calculate the parameters of the filter, thereby improving the calculation accuracy of the filter parameters when the number of IDT roots is small. In addition, the Bayesian optimization method is used in the parameter optimization process to replace the gradient optimization and global optimization methods of the existing technology, thereby avoiding the problems of gradient optimization relying on initial values ​​and low optimization efficiency of global optimization, and improving the iteration speed of the parameter optimization process.

[0064] The embodiment of the present invention further provides a system 200 for designing filter parameters based on Bayesian optimization. Please refer to FIG4 , which is a schematic structural diagram of the system for designing filter parameters based on Bayesian optimization provided by an embodiment of the present invention, which includes:

[0065] The parameter modeling module 201 is used to construct an optimization model according to the parameters of the surface acoustic wave filter to be optimized, and optimize the parameters based on the optimization model to obtain the optimized parameters;

[0066] The Bayesian modeling module 202 is used to construct a preset optimization algorithm based on the Bayesian optimization method;

[0067] The Bayesian optimization module 203 is configured to perform iterative calculations using the optimization parameters as inputs of the preset optimization algorithm to obtain parameter optimization results corresponding to the parameters.

[0068] The optimization model satisfies the following relationship: Min IL(x); stlb≤x≤ub; G(x)≤0;

[0069] IL represents the insertion loss of the surface acoustic wave filter, x represents the parameter, lb and ub represent the lower limit and upper limit of the parameter respectively, and G(x) represents the preset additional parameter of the surface acoustic wave filter

[0070] The preset optimization algorithm is specifically:

[0071] Determine the calculation sample of the parameter for the preset optimization algorithm; this step specifically performs stratified sampling between the upper limit lb and the lower limit ub of the parameter according to Latin hypercube sampling to generate the calculation sample

[0072] Invoking the HCT-FEM method to calculate parameter variables of the surface acoustic wave filter, the parameter variables including the insertion loss IL and the preset additional parameter G(x);

[0073] Establishing a proxy function using the calculation sample and the parameter variable as a function, wherein the proxy function is a Gaussian process regression model; calculating the position of a point of a new sample used to add to the calculation sample using an EI acquisition function, and updating the proxy function according to the position of the point;

[0074] A preset solving algorithm is used to calculate the minimum value of the calculation sample in the proxy function; the preset solving algorithm is a genetic algorithm;

[0075] Determine whether the iteration of the preset optimization algorithm reaches a preset threshold:

[0076] If so, output the minimum value as the optimal solution of the parameter;

[0077] If not, the HCT-FEM method is called to calculate the parameter variable corresponding to the minimum value, the number of iterations is increased by 1, and then the process returns to the step of calling the HCT-FEM method to calculate the parameter variable of the surface acoustic wave filter.

[0078] The system 200 for designing filter parameters based on Bayesian optimization can implement the steps in the method for designing filter parameters based on Bayesian optimization in the above embodiment, and can achieve the same technical effects. Please refer to the description in the above embodiment and will not be repeated here.

[0079] An embodiment of the present invention further provides a computer device. Please refer to Figure 5, which is a structural diagram of the computer device provided by an embodiment of the present invention. The computer device 300 includes: a memory 302, a processor 301, and a computer program stored in the memory 302 and executable on the processor 301.

[0080] The processor 301 calls the computer program stored in the memory 302 to execute the steps of the method for designing filter parameters based on Bayesian optimization provided by an embodiment of the present invention. Referring to FIG1 , the method specifically includes the following steps:

[0081] S101 : Constructing an optimization model according to the parameters of the surface acoustic wave filter to be optimized, and performing optimization calculation on the parameters based on the optimization model to obtain optimized parameters.

[0082] The optimization model satisfies the following relationship: Min IL(x); stlb≤x≤ub; G(x)≤0;

[0083] Wherein, IL represents the insertion loss of the surface acoustic wave filter, x represents the parameter, lb and ub represent the lower limit and upper limit of the parameter respectively, and G(x) represents the preset additional parameter of the surface acoustic wave filter.

[0084] S102: Construct a preset optimization algorithm based on the Bayesian optimization method.

[0085] The preset optimization algorithm includes the following sub-steps:

[0086] S1021. Determine a calculation sample of the parameters for the preset optimization algorithm.

[0087] S1022 : Calling the HCT-FEM method to calculate parameter variables of the surface acoustic wave filter, where the parameter variables include the insertion loss IL and the preset additional parameter G(x).

[0088] S1023: Establish a proxy function using the calculation sample and the parameter variable as a function.

[0089] The proxy function is a Gaussian process regression model.

[0090] S1024: Calculate the minimum value of the calculation samples in the proxy function using a preset solution algorithm.

[0091] The preset solution algorithm is a genetic algorithm.

[0092] S1025: Determine whether the iteration of the preset optimization algorithm reaches a preset threshold:

[0093] If so, output the minimum value as the optimal solution of the parameter;

[0094] If not, the HCT-FEM method is called to calculate the parameter variable corresponding to the minimum value, the number of iterations is increased by 1, and the process returns to step S1022.

[0095] Step S1021 is specifically as follows:

[0096] Stratified sampling is performed between the lower limit lb and the upper limit ub of the parameter according to Latin hypercube sampling to generate the calculation sample.

[0097] Step S1023 further includes the following steps:

[0098] The position of a point of a new sample added to the calculated sample is calculated using the EI acquisition function, and the proxy function is updated according to the position of the point.

[0099] S103: performing iterative calculations using the optimization parameters as inputs of the preset optimization algorithm to obtain parameter optimization results corresponding to the parameters.

[0100] The computer device 300 provided in the embodiment of the present invention can implement the steps in the method for designing filter parameters based on Bayesian optimization in the above embodiment, and can achieve the same technical effects. Please refer to the description in the above embodiment and will not be repeated here.

[0101] An embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the various processes and steps in the method for designing filter parameters based on Bayesian optimization provided in an embodiment of the present invention are implemented, and the same technical effects can be achieved. To avoid repetition, they will not be described here.

[0102] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware through a computer program. The program can be stored in a computer-readable storage medium, and when executed, the program can include the processes in the above-described method embodiments. The storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0103] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0104] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better embodiment. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present invention.

[0105] The embodiments of the present invention are described above in conjunction with the accompanying drawings. What is disclosed is only a preferred embodiment of the present invention. However, the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms and equivalent changes without departing from the scope of protection of the purpose of the present invention and the claims, which are all within the protection of the present invention.

Claims

1. A method for designing filter parameters based on Bayesian optimization, characterized in that: The method comprises the following steps: S101, constructing an optimization model according to the parameters of the surface acoustic wave filter to be optimized, and optimizing and calculating the parameters based on the optimization model to obtain optimized parameters; S102, constructing a preset optimization algorithm based on the Bayesian optimization method; S103: performing iterative calculations using the optimization parameters as inputs of the preset optimization algorithm to obtain parameter optimization results corresponding to the parameters.

2. The method for designing filter parameters based on Bayesian optimization as claimed in claim 1, characterized in that: The optimization model satisfies the following relationship: Min IL(x); stlb≤x≤ub; G(x)≤0; Among them, IL represents the insertion loss of the surface acoustic wave filter, x represents the parameter, lb and ub represent the lower limit and upper limit of the parameter respectively, and G(x) represents the preset additional parameter of the surface acoustic wave filter.

3. The method for designing filter parameters based on Bayesian optimization as claimed in claim 2, characterized in that: The preset optimization algorithm includes the following sub-steps: S1021, determining a calculation sample of the parameters used for the preset optimization algorithm; S1022, calling the HCT-FEM method to calculate the parameter variables of the surface acoustic wave filter, the parameter variables including the insertion loss IL and the preset additional parameter G(x); S1023, establishing a proxy function using the calculation sample and the parameter variable as a function; S1024, using a preset solution algorithm to calculate the minimum value of the calculation samples in the proxy function; S1025: Determine whether the iteration of the preset optimization algorithm reaches a preset threshold: If so, output the minimum value as the optimal solution of the parameter; If not, call the HCT-FEM method to calculate the parameter variable corresponding to the minimum value, increase the number of iterations by 1, and then return to step S1022.

4. The method for designing filter parameters based on Bayesian optimization as claimed in claim 3, characterized in that: Step S1021 is specifically as follows: Stratified sampling is performed between the lower limit lb and the upper limit ub of the parameter according to Latin hypercube sampling to generate the calculation sample.

5. The method for designing filter parameters based on Bayesian optimization as claimed in claim 3, characterized in that: Step S1023 also includes the steps of: The position of a point of a new sample added to the calculated sample is calculated using the EI acquisition function, and the proxy function is updated according to the position of the point.

6. The method for designing filter parameters based on Bayesian optimization as claimed in claim 3, characterized in that: The proxy function is a Gaussian process regression model.

7. The method for designing filter parameters based on Bayesian optimization as claimed in claim 3, characterized in that: The preset solution algorithm is a genetic algorithm.

8. A system for designing filter parameters based on Bayesian optimization, characterized in that: include: A parameter modeling module is used to construct an optimization model according to the parameters of the surface acoustic wave filter that need to be optimized, and optimize the parameters based on the optimization model to obtain the optimized parameters; Bayesian modeling module, used to build preset optimization algorithms based on Bayesian optimization methods; The Bayesian optimization module is used to perform iterative calculations using the optimization parameters as inputs of the preset optimization algorithm to obtain parameter optimization results corresponding to the parameters.

9. A computer device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps in the method for designing filter parameters based on Bayesian optimization as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the method for designing filter parameters based on Bayesian optimization as described in any one of claims 1 to 7 are implemented.

Images