Surface acoustic wave device model coefficient extraction and simulation method, device and related equipment

Through filtering and noise reduction and feature information search methods, the model coefficients of surface acoustic wave devices are optimized and neural network models are used for simulation, which solves the problem of low simulation accuracy in the existing technology and achieves more accurate simulation results.

CN118607439BActive Publication Date: 2024-11-01LANSUS TECH INC
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411041340.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-31
Publication Date
2024-11-01
Estimated Expiration
2044-07-31

AI Technical Summary

Technical Problem

The simulation accuracy of existing surface acoustic wave devices is low and are affected by random noise and clutter, resulting in the contamination of the truth value, affecting the accuracy of the simulation results.

Method used

The filtering and noise reduction algorithm removes abnormal waveforms and random noise in the measured curve, combines the moving window and limited conditions to search feature information, establish a magnificent model and optimize the model coefficients, and use a neural network model for simulation.

Benefits of technology

The simulation accuracy of surface acoustic wave devices is improved, accurate model coefficients are obtained, and the versatility and accuracy of simulation are enhanced.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118607439B_ABST
    Figure CN118607439B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of wireless communication technology, and discloses a method, device and related equipment for extracting and simulating model coefficients of surface acoustic wave devices. The method comprises the following steps: step S1, obtaining the admittance parameters of a batch of surface acoustic wave devices; step S2, removing abnormal waveforms and random noises in the measured curve by filtering and denoising algorithm, and obtaining the true value of the measured curve; step S3, searching for characteristic information in the measured curve by moving a window of preset size in combination with preset limiting conditions; step S4, optimizing all model coefficients step by step according to the characteristic change relationship between the coefficients of the phenomenological model and the conductance of the conductance curve and the susceptance of the susceptance curve; step S5, using a neural network model to develop a simulation method for surface acoustic wave devices within a preset structural size range and the obtained simulation results. The method for extracting and simulating model coefficients of surface acoustic wave devices of the present invention can improve the simulation accuracy of surface acoustic wave devices.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a method, device and related equipment for extracting and simulating model coefficients of a surface acoustic wave device. Background Art

[0002] At present, surface acoustic wave filters have good temperature stability, working stability, design flexibility, small size, light weight, low power consumption and other advantages, and are widely used in communications, radar, television, radio and other fields. Surface acoustic wave devices are composed of an upper metal structure and a lower piezoelectric substrate. Different metal structure designs correspond to different impedance characteristics. With the development of communication technology, surface acoustic wave filters are constantly developing in the direction of high frequency, low loss, and high power handling. In the process of new product development, high-quality iteration can not only gain an advantage but also reduce costs.

[0003] In related technologies, there are characteristic waveforms in the admittance curve of surface acoustic wave devices, which means that if these characteristics can be accurately described, the electrical characteristics of the device can be accurately simulated; the impedance characteristics of devices with the same structure from different manufacturers are slightly different, but they are enough to affect the performance of the device. Therefore, using measured data from different vendors to obtain different model coefficients can make the simulation more accurate.

[0004] However, since simulation development requires a large amount of measured data of surface acoustic wave filters of different structural sizes, random noise is inevitably introduced, and clutter may also be introduced to cause the true value to be contaminated. Therefore, the industry often uses HCT (hierarchical cascade technology) to first fit and then obtain the phenomenological model coefficients based on the uncontaminated data of HCT. However, this method has undergone two data fittings, which inevitably leads to a loss of accuracy. Summary of the invention

[0005] The purpose of the embodiment of the present invention is to provide a method for extracting and simulating a surface acoustic wave device model coefficient, so as to solve the problem of low simulation accuracy of existing surface acoustic wave devices.

[0006] In order to solve the above technical problems, in a first aspect, an embodiment of the present invention provides a method for extracting and simulating model coefficients of a surface acoustic wave device, the method comprising the following steps:

[0007] Step S1, obtaining admittance parameters of a batch of surface acoustic wave devices;

[0008] Step S2, processing a corresponding measured curve according to the admittance parameter, and removing abnormal waveforms and random noise in the measured curve by a filtering and noise reduction algorithm to obtain a true value of the measured curve;

[0009] Step S3, searching for characteristic information in the measured curve by using a moving window of a preset size in combination with preset limiting conditions; wherein the measured curve includes a conductance curve and a susceptance curve;

[0010] Step S4, establishing a phenomenological model, and optimizing all model coefficients step by step according to the characteristic variation relationship between the coefficients of the phenomenological model and the conductance of the conductance curve and the susceptance of the susceptance curve;

[0011] Step S5, after obtaining the model coefficients of a batch, using the preset geometric dimensions as input and the model coefficients as output, a neural network model is obtained through training, and the neural network model is used to develop a simulation method for surface acoustic wave devices within a preset structural dimension range and obtain simulation results.

[0012] Preferably, in step S2, the filtering noise reduction algorithm includes any one or more combinations of median filtering, wavelet analysis, upper envelope algorithm, Fourier transform, least squares and B-spline fitting.

[0013] Preferably, in step S3, the moving window of the preset size is a moving window of 2 MHz-10 MHz.

[0014] Preferably, in step S3, the preset restriction condition is that the difference between the maximum value and the minimum value of the conductivity in the moving window satisfies a preset value.

[0015] Preferably, in step S4, the model coefficients include propagation loss, relative dielectric constant, two parameters, electromechanical coupling coefficient and speed adjustment coefficient of the surface acoustic wave device.

[0016] Preferably, the preset geometric dimensions are the film thickness ratio and metallization rate of the surface acoustic wave device.

[0017] In a second aspect, an embodiment of the present invention provides a surface acoustic wave device model coefficient extraction and simulation device, comprising:

[0018] An acquisition module, used to acquire the admittance parameters of a batch of surface acoustic wave devices;

[0019] A removal module, used to process the corresponding measured curve according to the admittance parameter, and remove abnormal waveforms and random noise in the measured curve by a filtering and noise reduction algorithm to obtain the true value of the measured curve;

[0020] A search module, used to search for characteristic information in the measured curve through a moving window of a preset size and preset limiting conditions; wherein the measured curve includes a conductance curve and a susceptance curve;

[0021] An optimization module is used to establish a phenomenological model, and optimize all model coefficients step by step according to the characteristic change relationship between the coefficients of the phenomenological model and the conductance of the conductance curve and the susceptance of the susceptance curve;

[0022] A training module is used to obtain a batch of model coefficients, take preset geometric dimensions as input and the model coefficients as output, obtain a neural network model through training, and use the neural network model to develop a simulation method for surface acoustic wave devices within a preset structural size range and obtain simulation results.

[0023] In a third aspect, an embodiment of the present invention provides a computer device, comprising: a memory, a processor, and a surface acoustic wave device model coefficient extraction and simulation program stored in the memory and executable on the processor, wherein the processor implements the steps in the surface acoustic wave device model coefficient extraction and simulation method as described above when executing the surface acoustic wave device model coefficient extraction and simulation program.

[0024] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a surface acoustic wave device model coefficient extraction and simulation program is stored. When the surface acoustic wave device model coefficient extraction and simulation program is executed by a processor, the steps in the surface acoustic wave device model coefficient extraction and simulation method as described above are implemented.

[0025] Compared with the prior art, the method for extracting and simulating the model coefficients of the surface acoustic wave device in the present invention uses a filtering noise reduction algorithm to remove noise and smooth the measured data through the above-mentioned steps S1 to S5, thereby achieving the purpose of obtaining a quasi-true value from the original measured data. At the same time, an algorithm for accurately obtaining feature points by using a moving window of appropriate size in combination with restriction conditions is used to effectively improve the simulation accuracy of the surface acoustic wave device; and the versatility of the simulation is greatly improved by the application of a neural network model. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work, among which:

[0027] Figure 1 A flow chart of a method for extracting and simulating surface acoustic wave device model coefficients provided by an embodiment of the present invention;

[0028] Figure 2 A schematic diagram of a measured curve containing bimodal abnormal characteristic points provided by an embodiment of the present invention;

[0029] Figure 3A schematic diagram of a curve of a correction structure provided by an embodiment of the present invention;

[0030] Figure 4 A schematic diagram of a key feature point search result provided by an embodiment of the present invention;

[0031] Figure 5a Schematic diagram of comparison of simulation results without step S2 provided in the embodiment of the present invention Figure 1 ;

[0032] Figure 5b Schematic diagram of comparison of simulation results without step S2 provided in the embodiment of the present invention Figure 2 ;

[0033] Figure 6a Schematic diagram of comparison of simulation results of step S2 provided in an embodiment of the present invention Figure 1 ;

[0034] Figure 6b Schematic diagram of comparison of simulation results of step S2 provided in an embodiment of the present invention Figure 2 ;

[0035] Figure 7 A module diagram of a surface acoustic wave device model coefficient extraction and simulation device provided by an embodiment of the present invention;

[0036] Figure 8 A module diagram of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0037] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] Embodiment 1

[0039] Please refer to the attached Figure 1-6b As shown, an embodiment of the present invention provides a method for extracting and simulating model coefficients of a surface acoustic wave device, the method comprising the following steps:

[0040] Step S1, obtaining the admittance parameters of a batch of surface acoustic wave devices.

[0041] This example includes several resonators of different sizes and a DMS filter, where the resonator is used to extract the model coefficients and the filter DMS is used for simulation verification. Generally, there is random noise in the measured data, and there may also be clutter, which affects the accuracy of the main resonance characteristics.

[0042] Specifically, the probe station is used to obtain the admittance parameters of a batch of surface acoustic wave devices. Figure 2 The figure shows the amplitude-frequency characteristic curve of a resonator obtained by the probe station test, where the admittance decibel and the conductance decibel are respectively. A certain feature of the conductance decibel curve (the maximum point near 780M) is interfered by clutter to form a double peak.

[0043] Step S2: Process the corresponding measured curve according to the admittance parameter, and remove abnormal waveforms and random noise in the measured curve through a filtering and noise reduction algorithm to obtain the true value of the measured curve.

[0044] Specifically, Figure 3 The blue curve is Figure 2 The unprocessed conductance decibel curve in Figure 3 The yellow curve in the figure is the conductance decibel curve processed by median filtering, Fourier transform, inverse Fourier transform, spline fitting and other algorithms. It can be seen that the double peaks in step S1 have been processed into a single peak, which is in line with the general situation. At the same time, according to the conductance measurement results and correction results, it can be seen that the curve is smoothed by removing random noise.

[0045] Step S3, searching for characteristic information in the measured curve by moving a window of a preset size in combination with preset limiting conditions; wherein the measured curve includes a conductance curve and a susceptance curve. The characteristic information is related to the positions of several maxima in the measured curve, and the characteristic information obtained in this step is used to obtain the model coefficients in the subsequent step S4.

[0046] Step S4, establishing a phenomenological model, and optimizing all model coefficients step by step according to the characteristic variation relationship between the coefficients of the phenomenological model and the conductance of the conductance curve and the susceptance of the susceptance curve.

[0047] Among them, the phenomenological model is a commonly used model of surface acoustic wave devices, such as the COM model.

[0048] Step S5, after obtaining the model coefficients of a batch, using the preset geometric dimensions as input and the model coefficients as output, a neural network model is obtained through training, and the neural network model is used to develop a simulation method for surface acoustic wave devices within a preset structural dimension range and obtain simulation results.

[0049] In specific implementation, through the above-mentioned steps S1 to S5, the measured data is denoised and smoothed by a filtering noise reduction algorithm, thereby achieving the purpose of obtaining quasi-true values ​​from the original measured data. At the same time, an algorithm for accurately obtaining feature points by using a moving window of appropriate size in combination with restriction conditions is used to effectively improve the simulation accuracy of the surface acoustic wave device; the versatility of the simulation is greatly improved by the application of the neural network model.

[0050] In this embodiment, in step S2, the filtering noise reduction algorithm includes any one or more combinations of median filtering, wavelet analysis, upper envelope algorithm, Fourier transform, least squares and B-spline fitting.

[0051] In this embodiment, in step S3, the moving window of the preset size is a moving window of 2 MHz-10 MHz.

[0052] In this embodiment, in step S3, the preset restriction condition is that the difference between the maximum value and the minimum value of the conductivity in the moving window satisfies a preset value. The preset value may be that the difference between the maximum value and the minimum value in the moving window is greater than or less than a preset value, which will not be described here.

[0053] Specifically, a 10MHz moving window is used, and the maximum and minimum differences in the window are limited to be less than the preset extreme points, to find the extreme points that meet the requirements. Each small window has a 10MHz window, and the difference between the maximum and minimum values ​​is less than the preset value during the moving process. Figure 4 , the characteristic points are the maximum extreme point and the extreme points on its left and right sides.

[0054] Among them, the preset value is 3-5dB, and optionally, the preset value is 3dB.

[0055] In this embodiment, in step S4, the model coefficients include propagation loss, relative dielectric constant, two parameters, electromechanical coupling coefficient and speed adjustment coefficient of the surface acoustic wave device.

[0056] Specifically, the phenomenological model is used and the characteristic information difference in simulation and step S2 is used as the objective function to optimize the model coefficients such as propagation loss, relative dielectric constant, two parameters, electromechanical coupling coefficient, speed adjustment coefficient, etc. A large number of resonator model coefficients of different sizes are further optimized and used for neural network model training in the subsequent step S5.

[0057] In this embodiment, the preset geometric dimensions are the film thickness ratio and metallization rate of the surface acoustic wave device.

[0058] Among them, the film thickness ratio usually refers to the ratio of the change in temperature and the change in film thickness in a system. It is one of the important physical quantities of the film material properties and is an expression of the filter performance. The metallization ratio has an important influence on the frequency characteristics and electrical characteristics of the interdigital transducer.

[0059] Specifically, a neural network model was established with the film thickness ratio and metallization rate of the surface acoustic wave device as input and the model coefficient as output, and the model coefficient and its size information in S4 were used for model training. The electrical characteristics prediction program of the surface acoustic wave device was developed using the neural network model. Figure 5a , Figure 5b , Figure 6a and Figure 6b , is a simulation comparison of the s11 and s21 curves of a 5th-order DMS. Through the simulation comparison, it can be seen that the overall predicted feature position is relatively accurate. After processing some abnormal measured curves in step S2, the simulation results are more accurate.

[0060] Embodiment 2

[0061] like Figure 7 As shown, an embodiment of the present invention provides a surface acoustic wave device model coefficient extraction and simulation device 200, comprising:

[0062] An acquisition module 201 is used to acquire the admittance parameters of a batch of surface acoustic wave devices;

[0063] A removal module 202 is used to process the corresponding measured curve according to the admittance parameter, and remove abnormal waveforms and random noise in the measured curve by a filtering and noise reduction algorithm to obtain the true value of the measured curve;

[0064] A search module 203, used to search for characteristic information in the measured curve by using a moving window of a preset size in combination with preset limiting conditions; wherein the measured curve includes a conductance curve and a susceptance curve;

[0065] The optimization module 204 is used to establish a phenomenological model, and optimize all model coefficients step by step according to the characteristic change relationship between the coefficients of the phenomenological model and the conductance of the conductance curve and the susceptance of the susceptance curve;

[0066] The training module 205 is used to obtain the model coefficients of the batch, take the preset geometric dimensions as input and the model coefficients as output, obtain the neural network model through training, and use the neural network model to develop a simulation method for the surface acoustic wave device within the preset structural dimension range and obtain the simulation results.

[0067] Specifically, the admittance parameters of a batch of surface acoustic wave devices are obtained through the acquisition module 201; the corresponding measured curves are processed according to the admittance parameters through the removal module 202, and the abnormal waveforms and random noise in the measured curves are removed through the filtering and noise reduction algorithm to obtain the true value of the measured curves; the search module 203 is used to search for characteristic information in the measured curves through a moving window of a preset size in combination with preset limiting conditions; the optimization module 204 is used to establish a phenomenological model, and all model coefficients are optimized step by step according to the characteristic change relationship between the coefficients of the phenomenological model and the conductance of the conductance curve and the susceptance of the susceptance curve; the training module 205 is used to obtain the model coefficients of the batch, take the preset geometric size as input, and the model coefficients as output, obtain the neural network model through training, and use the neural network model to develop the simulation results of the surface acoustic wave devices within the preset structural size range. In this way, the measured data is denoised and smoothed using a filtering and denoising algorithm, thereby achieving the purpose of obtaining quasi-true values ​​from the original measured data. At the same time, an algorithm that accurately obtains feature points using a moving window of appropriate size combined with restriction conditions effectively improves the simulation accuracy of surface acoustic wave devices. The application of neural network models greatly improves the versatility of the simulation.

[0068] The functions implemented by the above-mentioned surface acoustic wave device model coefficient extraction and simulation device 200 are the same as those of the above-mentioned surface acoustic wave device model coefficient extraction and simulation method, and the technical effects produced are the same, which will not be described one by one here.

[0069] In this embodiment, in the removal module 202, the filtering noise reduction algorithm includes any one or more combinations of median filtering, wavelet analysis, upper envelope algorithm, Fourier transform, least squares and B-spline fitting.

[0070] In this embodiment, in the search module 203, the moving window of the preset size is a moving window of 10 MHz.

[0071] In this embodiment, in the search module 203, the preset restriction condition is that the difference between the maximum value and the minimum value of the moving window is less than a preset value.

[0072] Specifically, a 10MHz moving window is used, and the maximum and minimum differences in the window are limited to be less than the preset extreme points, to find the extreme points that meet the requirements. Each small window has a 10MHz window, and the difference between the maximum and minimum values ​​is less than the preset value during the moving process. Figure 3 , the characteristic points are the maximum extreme point and the extreme points on its left and right sides.

[0073] Among them, the preset value is 3-5dB, and optionally, the preset value is 3dB.

[0074] In this embodiment, in the optimization module 204, the model coefficients include the propagation loss, relative dielectric constant, two parameters, electromechanical coupling coefficient and speed adjustment coefficient of the surface acoustic wave device.

[0075] In this embodiment, the preset geometric dimensions are the film thickness ratio and metallization rate of the surface acoustic wave device.

[0076] Embodiment 3

[0077] like Figure 8 As shown, an embodiment of the present invention provides a computer device 300, comprising: a memory 301, a processor 302, and a surface acoustic wave device model coefficient extraction and simulation program stored in the memory 301 and executable on the processor 302, wherein the processor 302 implements the steps in the surface acoustic wave device model coefficient extraction and simulation method as described above when executing the surface acoustic wave device model coefficient extraction and simulation program:

[0078] Step S1, obtaining admittance parameters of a batch of surface acoustic wave devices;

[0079] Step S2, processing a corresponding measured curve according to the admittance parameter, and removing abnormal waveforms and random noise in the measured curve by a filtering and noise reduction algorithm to obtain a true value of the measured curve;

[0080] Step S3, searching for characteristic information in the measured curve by using a moving window of a preset size in combination with preset limiting conditions; wherein the measured curve includes a conductance curve and a susceptance curve;

[0081] Step S4, establishing a phenomenological model, and optimizing all model coefficients step by step according to the characteristic variation relationship between the coefficients of the phenomenological model and the conductance of the conductance curve and the susceptance of the susceptance curve;

[0082] Step S5, after obtaining the model coefficients of a batch, using the preset geometric dimensions as input and the model coefficients as output, a neural network model is obtained through training, and the neural network model is used to develop a simulation method for surface acoustic wave devices within a preset structural dimension range and obtain simulation results.

[0083] Embodiment 4

[0084] An embodiment of the present invention provides a computer-readable storage medium, on which a surface acoustic wave device model coefficient extraction and simulation program is stored. When the surface acoustic wave device model coefficient extraction and simulation program is executed by a processor, the steps in the surface acoustic wave device model coefficient extraction and simulation method as described above are implemented.

[0085] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0086] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an 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, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0087] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for extracting and simulating model coefficients of a surface acoustic wave device, characterized in that: The method comprises the following steps: Step S1, obtaining admittance parameters of a batch of surface acoustic wave devices; Step S2, processing a corresponding measured curve according to the admittance parameter, and removing abnormal waveforms and random noise in the measured curve by a filtering and noise reduction algorithm to obtain a true value of the measured curve; Step S3, searching for characteristic information in the measured curve by using a moving window of a preset size in combination with preset limiting conditions; wherein the measured curve includes a conductance curve and a susceptance curve; and the moving window of the preset size is a moving window of 2 MHz-10 MHz; Step S4, establishing a phenomenological model, and optimizing all model coefficients step by step according to the characteristic variation relationship between the coefficients of the phenomenological model and the conductance of the conductance curve and the susceptance of the susceptance curve; Step S5, after obtaining the model coefficients of a batch, using the preset geometric dimensions as input and the model coefficients as output, a neural network model is obtained through training, and the neural network model is used to develop a simulation method for surface acoustic wave devices within a preset structural dimension range and obtain corresponding simulation results.

2. The method for extracting and simulating surface acoustic wave device model coefficients according to claim 1, characterized in that: In step S2, the filtering noise reduction algorithm includes any one or more combinations of median filtering, wavelet analysis, upper envelope algorithm, Fourier transform, least squares and B-spline fitting.

3. The method for extracting and simulating the model coefficients of a surface acoustic wave device according to claim 1, wherein: In step S3, the preset restriction condition is that the difference between the maximum value and the minimum value of the conductance in the moving window satisfies a preset value.

4. The method for extracting and simulating surface acoustic wave device model coefficients according to claim 1, characterized in that: In the step S4, the model coefficients include the propagation loss, relative dielectric constant, two parameters, electromechanical coupling coefficient and speed adjustment coefficient of the surface acoustic wave device.

5. The method for extracting and simulating the model coefficients of a surface acoustic wave device according to claim 1, wherein: The preset geometric dimensions are the film thickness ratio and metallization rate of the surface acoustic wave device.

6. A surface acoustic wave device model coefficient extraction and simulation device, characterized in that: include: An acquisition module, used to acquire the admittance parameters of a batch of surface acoustic wave devices; A removal module, used to process the corresponding measured curve according to the admittance parameter, and remove abnormal waveforms and random noise in the measured curve by a filtering and noise reduction algorithm to obtain the true value of the measured curve; A search module, used to search for characteristic information in the measured curve by using a moving window of a preset size in combination with preset limiting conditions; wherein the measured curve includes a conductance curve and a susceptance curve; and the moving window of the preset size is a moving window of 2 MHz to 10 MHz; An optimization module is used to establish a phenomenological model, and optimize all model coefficients step by step according to the characteristic change relationship between the coefficients of the phenomenological model and the conductance of the conductance curve and the susceptance of the susceptance curve; The training module is used to obtain the model coefficients of the batch, take the preset geometric dimensions as input and the model coefficients as output, obtain the neural network model through training, and use the neural network model to develop a simulation method for the surface acoustic wave device within the preset structural size range and obtain the corresponding simulation results.

7. A computer device, characterized in that: include: A memory, a processor, and a surface acoustic wave device model coefficient extraction and simulation program stored in the memory and executable on the processor, wherein the processor implements the steps of the surface acoustic wave device model coefficient extraction and simulation method as described in any one of claims 1 to 5 when executing the surface acoustic wave device model coefficient extraction and simulation program.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a surface acoustic wave device model coefficient extraction and simulation program, which, when executed by a processor, implements the steps in the surface acoustic wave device model coefficient extraction and simulation method as described in any one of claims 1 to 5.