WO2024120159A2 - 降低单元矩阵维度的仿真方法、系统及相关设备 - Google Patents

降低单元矩阵维度的仿真方法、系统及相关设备 Download PDF

Info

Publication number
WO2024120159A2
WO2024120159A2 PCT/CN2023/132522 CN2023132522W WO2024120159A2 WO 2024120159 A2 WO2024120159 A2 WO 2024120159A2 CN 2023132522 W CN2023132522 W CN 2023132522W WO 2024120159 A2 WO2024120159 A2 WO 2024120159A2
Authority
WO
WIPO (PCT)
Prior art keywords
simulation
unit
matrix
acoustic wave
surface acoustic
Prior art date
Application number
PCT/CN2023/132522
Other languages
English (en)
French (fr)
Inventor
朱玉泉
关鹏
杨睿智
胡锦钊
常林森
郭嘉帅
Original Assignee
深圳飞骧科技股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳飞骧科技股份有限公司 filed Critical 深圳飞骧科技股份有限公司
Publication of WO2024120159A2 publication Critical patent/WO2024120159A2/zh

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00Computer-aided design [CAD]
    • G06F30/20Design optimisation, verification or simulation
    • G06F30/23Design optimisation, verification or simulation using finite element methods [FEM] or finite difference methods [FDM]
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10Complex mathematical operations
    • G06F17/16Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10Complex mathematical operations
    • G06F17/18Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00Computer-aided design [CAD]
    • G06F30/10Geometric CAD
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T17/00Three dimensional [3D] modelling, e.g. data description of 3D objects
    • G06T17/20Finite element generation, e.g. wire-frame surface description, tesselation
    • G06T17/205Re-meshing
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02TCLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T90/00Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation

Definitions

  • the present invention belongs to the technical field of piezoelectric material electromechanical coupling, and in particular relates to a simulation method, system and related equipment for reducing the dimension of a unit matrix.
  • Surface acoustic wave devices are a kind of electromechanically coupled acoustic components, such as resonators, which are usually accurately simulated by the finite element method.
  • the finite element method (ISBN: 7-80159-853-9, 2015) was proposed in the 1950s. It is a mathematical calculation method that converts complex structural calculation problems into analysis and set problems of simple units.
  • surface acoustic wave devices are not fully simulated at full size. Instead, two-dimensional simulations are usually performed after the plane strain assumption is adopted.
  • the two-dimensional model will also consume a lot of computing resources and time. According to the characteristics of periodic changes in the interdigitated finger structure of the surface acoustic wave device, the hierarchical cascade technology has been widely used.
  • Hierarchical cascade technology eliminates the operation of internal degrees of freedom through Schur complement operations, which greatly reduces the demand for computing resources and makes full-scale simulation of surface acoustic wave devices possible.
  • the problem is that during the application of hierarchical cascade technology, multiple Schur complement operations are required for the basic structures in the surface acoustic wave device, such as a single finger, GAP, and left and right PML layers.
  • the basic structure is formed by splicing the unit matrix calculated by the finite element method, the overall matrix of the basic structure is still large, which will generate huge performance consumption during the Schur complement operation.
  • the computational efficiency is usually improved by reducing the number of grids and the order of the units. However, on the one hand, the reduction in the number of grids will affect the simulation results.
  • the simulation of the electromagnetic field also requires second-order units to ensure the calculation accuracy.
  • the embodiments of the present invention provide a simulation method, system and related equipment for reducing the dimension of a unit matrix, aiming to solve the problem that the unit matrix obtained by using the hierarchical cascade technology in the existing surface acoustic wave device simulation process is large in dimension and low in simulation efficiency.
  • an embodiment of the present invention provides a simulation method for reducing the dimension of a unit matrix, the simulation method is used to simulate a surface acoustic wave device, and the simulation method comprises the following steps:
  • the simulation unit matrices obtained from different basic structures are spliced and cascaded to obtain a simulation overall matrix, and the frequency response of the simulation overall matrix at a preset simulation frequency is calculated to obtain a simulation frequency response curve of the surface acoustic wave device.
  • the weight function is defined as N ⁇
  • the gradient function is defined as B ⁇
  • s and t are the horizontal and vertical coordinates in the local coordinate system respectively;
  • the six-node quadrilateral unit is a second-order interpolation in the horizontal direction and a first-order interpolation in the vertical direction.
  • the basic structure includes an interdigitation structure and a GAP structure.
  • the step of splicing and cascading the simulation unit matrices obtained from different basic structures to obtain a simulation overall matrix also includes:
  • the Schur complement operation is used to eliminate the common matrix nodes generated when the simulation unit matrices are spliced and cascaded.
  • an embodiment of the present invention further provides a simulation system for reducing the dimension of a unit matrix, wherein the simulation system is used to simulate a surface acoustic wave device, and comprises:
  • a simulation parameter acquisition module used for acquiring the geometric structure of the surface acoustic wave device, and dividing the surface acoustic wave device into a plurality of basic structures according to the geometric structure;
  • a node unit construction module is used to construct a weight function and a gradient function of a finite element method, and to construct a six-node quadrilateral unit according to the weight function and the gradient function;
  • a finite element modeling module used for meshing the base structure according to the finite element method, and calculating a simulation unit matrix corresponding to the base structure using the six-node quadrilateral unit;
  • the cascade simulation module is used to splice and cascade the simulation unit matrices obtained from different basic structures to obtain a simulation overall matrix, and calculate the frequency response of the simulation overall matrix at a preset simulation frequency to obtain a simulation frequency response curve of the surface acoustic wave device.
  • an embodiment of the present invention further provides a computer device, comprising: 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 simulation method for reducing the dimension of a unit matrix as described in any one of the above embodiments are implemented.
  • an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored.
  • the computer program is executed by a processor, the steps in the simulation method for reducing the dimension of a unit matrix as described in any one of the above embodiments are implemented.
  • the beneficial effects achieved by the present invention are compared with the method of calculating the unit matrix using the nine-node Lagrangian unit in the finite element theory for simulating the surface acoustic wave device in the prior art.
  • a new six-node quadrilateral unit is constructed and used in the unit matrix calculation, thereby reducing the dimension of the unit matrix, reducing the calculation requirements for the surface acoustic wave device simulation using the finite element method, and improving the calculation efficiency.
  • FIG1 is a flowchart of simulation steps of a surface acoustic wave device using a finite element method in the prior art
  • FIG2 is a schematic diagram showing the basic structure of the surface acoustic wave device provided by the present invention.
  • FIG3 is a schematic diagram of a second-order Lagrangian unit commonly used in the finite element method provided in an embodiment of the present invention.
  • FIG4 is a schematic diagram of a process of splicing unit matrices into an overall matrix according to an embodiment of the present invention
  • FIG5 is a flowchart of the steps of a simulation method for reducing the dimension of a unit matrix provided by an embodiment of the present invention
  • FIG6 is a schematic diagram of a six-node quadrilateral unit constructed according to an embodiment of the present invention.
  • FIG7 is a schematic diagram showing a comparison of units with different numbers of nodes provided by an embodiment of the present invention.
  • FIG8 is a schematic diagram showing simulation comparison results between a simulation method for reducing the dimension of a unit matrix provided by an embodiment of the present invention and the prior art;
  • FIG. 9 is a schematic diagram showing simulation comparison results of another simulation method for reducing the dimension of a unit matrix provided by an embodiment of the present invention and the prior art;
  • FIG10 is a schematic diagram of the structure of a simulation system 200 for reducing the dimension of a unit matrix provided in an embodiment of the present invention
  • FIG. 11 is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention.
  • the embodiment of the present invention first describes the simulation process of the surface acoustic wave device using the finite element method in the prior art.
  • the simulation process described in the embodiment of the present invention uses the geometric parameters of the surface acoustic wave device as initial parameters, and the ultimate goal of the simulation is to obtain a simulated frequency response curve of the surface acoustic wave device at a certain frequency.
  • FIG1 a flow chart of the simulation steps of a surface acoustic wave device using the finite element method in the prior art is shown in FIG1 , which includes the following steps:
  • the geometric structure information of the surface acoustic wave device includes the film thickness, metallization rate, pitch and other information of the interdigitated structure.
  • the upper part of FIG2 is the acquired geometric structure information, and the basic structure after dividing it is shown in the lower part of FIG2 , wherein the structure with electrode structure is the interdigitated structure, and the structure without electrode structure is the GAP structure.
  • step S3 includes the following sub-steps:
  • the role of the weight function can be understood as the fitting of the degrees of freedom within a range.
  • the more common interpolation method in the prior art is Lagrange interpolation. If a known curve f(x) has three known points (x 1 , y 1 ) (x 2 , y 2 ), (x 3 , y 3 ), then the curve f(x) can be represented by interpolation, and the interpolation result is:
  • the second-order interpolation function can be expressed as a second-order Lagrangian unit commonly used in the finite element method as shown in FIG3 .
  • the interpolation function of each node is obtained by using the above Lagrange interpolation form, and is expressed as:
  • the gradient function B ⁇ needed in the finite element method is the matrix composed of the partial derivatives of the weight function in different directions in the local coordinate system, that is,
  • the matrix B ⁇ can be fully expanded as:
  • the finite element method is to multiply both sides of the equation by a trial function And solve it by making the integral equal to 0 in a small area, that is:
  • the degree of freedom displacement u i and the potential ⁇ are still required to be continuous solutions.
  • FIG4 Exemplarily, the process of splicing unit matrices into an overall matrix is shown in FIG4 .
  • the Schur complement operation is used to eliminate the intermediate degrees of freedom of the basic structure, and only the left and right boundary degrees of freedom of the structure and the potential degrees of freedom at the junction of the electrode and the substrate are retained to obtain the overall matrix A Schur after Schur complementation, and the overall matrix A Schur after Schur complementation is used to replace the original overall matrix A.
  • the overall matrix A Schur of the basic structure after the Schur complement operation is spliced and cascaded.
  • the cascading principle is similar to the overall matrix splicing in the finite element method, which is to combine the overall matrices A1 and A2 of two adjacent basic structures in the surface acoustic wave device.
  • the combination principle in the finite element method is that for the M-dimensional overall matrix A 1 (M*M) and the N-dimensional overall matrix A 2 (N*N), when the two basic structures have K degrees of freedom in common, a new M+NK-dimensional overall matrix A 3 ((M+NK)*(M+NK)) can be constructed, in which the matrix information corresponding to the unshared degrees of freedom is directly used in the concatenated matrix, while the matrix information corresponding to the shared degrees of freedom is added before use.
  • the final cascaded matrix is processed by preset electrical conditions combined with mathematical formulas, and the frequency response of the surface acoustic wave device at that frequency is obtained.
  • a matrix M describing the entire surface acoustic wave device can be obtained.
  • the matrix M describes the degrees of freedom at the left and right boundaries and the interface between the electrode and the substrate. It can be divided into 2*2 generalized block matrices to describe the electrical degrees of freedom E and other degrees of freedom B to be analyzed, as shown below:
  • the prior art uses a hierarchical cascade technology, which divides the surface acoustic wave device into multiple interdigitated finger structures, performs a Schur complement operation on each interdigitated finger structure to eliminate internal degrees of freedom, and then sequentially cascades the interdigitated finger structures after Schur complementation, thereby reducing the size of the overall matrix of the surface acoustic wave device.
  • step S4 involves a Schur complement operation, there is one matrix inversion operation and two matrix multiplication operations in the Schur complement operation. Since the time complexity of the matrix inversion and the matrix multiplication is approximately O(n 3 ), that is, when the overall matrix A is large, the time to obtain the Schur complement operation result A Schur will also increase significantly. Therefore, the Schur complement operation will occupy a lot of computing time, thereby slowing down the computing efficiency, and eventually becoming the computing efficiency bottleneck of the prior art.
  • step S3 in the prior art is specifically improved.
  • FIG. 5 is a flowchart of the steps of a simulation method for reducing the dimension of a unit matrix provided by an embodiment of the present invention.
  • the simulation method is used to simulate a surface acoustic wave device, and specifically includes the following steps:
  • the basic structure includes an interdigitation structure and a GAP structure.
  • S102 construct a weight function and a gradient function for a finite element method, and construct a six-node quadrilateral element according to the weight function and the gradient function.
  • the six-node quadrilateral unit is a second-order interpolation in the horizontal direction and a first-order interpolation in the vertical direction.
  • the six-node quadrilateral unit proposed in the embodiment of the present invention utilizes the characteristics of the surface acoustic wave device that the displacement field and the electric potential field mainly show nonlinear changes in the propagation direction, while the displacement field and the electric potential field perpendicular to the propagation direction change almost linearly and decay rapidly.
  • the six-node quadrilateral unit is a second-order interpolation in the horizontal direction and a first-order interpolation in the vertical direction.
  • FIG6 A schematic diagram of the six-node quadrilateral unit constructed in an embodiment of the present invention is shown in FIG6 .
  • the weight function and gradient function of the six-node element are constructed.
  • the weight function is defined as N ⁇
  • the gradient function is defined as B ⁇
  • s and t are the horizontal and vertical coordinates in the local coordinate system respectively;
  • the weight function and the gradient function of the six-node quadrilateral unit need to be determined in advance, and then the unit is constructed based on the weight function and the gradient function.
  • the present invention also provides another implementation mode.
  • COMSOL-Matlab joint simulation hierarchical cascade algorithm is often used. That is, steps S102 and S103 call COMSOL finite element method software and specify the geometric parameters of the basic structure in the COMSOL software. Thus modeling, automatically partitioning the network, and outputting the overall matrix;
  • FIG. 7 is a schematic diagram of comparing units with different numbers of nodes provided in an embodiment of the present invention.
  • the weight functions are N1 to N9, respectively.
  • N 1 6 N 1 9 + 1/2*N 8 9
  • N 2 6 N 2 9 + 1/2*N 6 9
  • N 3 6 N 3 9 + 1/2*N 6 9
  • N 4 6 N 4 9 + 1/2*N 8 9
  • the weight functions are N 1 to N 8 respectively.
  • the step of splicing and cascading the simulation unit matrices obtained from different basic structures to obtain a simulation overall matrix also includes:
  • the Schur complement operation is used to eliminate the common matrix nodes generated when the simulation unit matrices are spliced and cascaded.
  • the pitch of a single interpolated finger is 0.9635 microns
  • the height of the metal electrode is 0.146 microns
  • the metallization rate is 0.6
  • the thickness of the model is 32.759 microns.
  • the interpolation is used to construct a reflective grating with 100 interpolated fingers on both sides of the model. 50 pairs of 100 interpolated fingers are used as IDTs inside the model. The number of calculated frequency points is 400, and each frequency point is 1Mhz.
  • a substrate depth of 6 times the pitch is set in the experiment, and a PML layer of 2 times the pitch is set at the bottom of the substrate to prevent boundary reflection.
  • the dimension of the generated overall matrix is (3292 ⁇ 9292), and the total time spent on calculating the unit matrix and completing the overall matrix is about 700ms; while the six-node quadrilateral unit used in the embodiment of the present invention generates an overall matrix dimension of (1688 ⁇ 1688), and the total time spent on calculating the unit matrix and completing the overall matrix is about 330ms.
  • the beneficial effect achieved by the present invention is that compared with the simulation method of the prior art of surface acoustic wave device simulation using the nine-node Lagrangian unit in the finite element theory to calculate the unit matrix, by using the surface acoustic wave device Based on the characteristics that the displacement field and electric potential field of the device mainly show nonlinear changes in the propagation direction, while the displacement field and electric potential field perpendicular to the propagation direction change almost linearly and decay rapidly, a new six-node quadrilateral unit is constructed and used in the unit matrix calculation, thereby reducing the dimension of the unit matrix, reducing the computational requirements of surface acoustic wave device simulation using the finite element method, and improving the computational efficiency.
  • the embodiment of the present invention further provides a simulation system for reducing the dimension of a unit matrix, and the simulation system is used to simulate a surface acoustic wave device.
  • FIG. 10 is a structural schematic diagram of a simulation system 200 for reducing the dimension of a unit matrix provided by an embodiment of the present invention, and includes:
  • a simulation parameter acquisition module 201 is used to acquire the geometric structure of the surface acoustic wave device and divide the surface acoustic wave device into a plurality of basic structures according to the geometric structure;
  • a node unit construction module 202 is used to construct a weight function and a gradient function of a finite element method, and to construct a six-node quadrilateral unit according to the weight function and the gradient function;
  • a finite element modeling module 203 used for meshing the foundation structure according to the finite element method, and calculating a simulation element matrix corresponding to the foundation structure using the six-node quadrilateral element;
  • the cascade simulation module 204 is used to splice and cascade the simulation unit matrices obtained from different basic structures to obtain a simulation overall matrix, and calculate the frequency response of the simulation overall matrix at a preset simulation frequency to obtain a simulation frequency response curve of the surface acoustic wave device.
  • the simulation system 200 for reducing the dimension of a unit matrix can implement the steps in the simulation method for reducing the dimension of a unit matrix in the above embodiment, and can achieve the same technical effects. Please refer to the description in the above embodiment, which will not be repeated here.
  • An embodiment of the present invention further provides a computer device.
  • 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.
  • the processor 301 calls the computer program stored in the memory 302 to execute the steps in the simulation method for reducing the dimension of the unit matrix provided in the embodiment of the present invention, which specifically includes the following steps in conjunction with FIG. 5 :
  • the weight function is defined as N ⁇
  • the gradient function is defined as B ⁇
  • the weight function N ⁇ satisfies:
  • N ⁇ ⁇ N1 , N2 , N3 , N4 , N5 , N6 ⁇ ;
  • s and t are the horizontal and vertical coordinates in the local coordinate system respectively;
  • the six-node quadrilateral unit is a second-order interpolation in the horizontal direction and a first-order interpolation in the vertical direction.
  • the basic structure includes an interdigitation structure and a GAP structure.
  • the step of splicing and cascading the simulation unit matrices obtained from different basic structures to obtain a simulation overall matrix also includes:
  • the Schur complement operation is used to eliminate the common matrix nodes generated when the simulation unit matrices are spliced and cascaded.
  • the computer device 300 provided in the embodiment of the present invention can implement the steps in the simulation method of reducing the dimension of the unit matrix in the above embodiment, and can achieve the same technical effect. Please refer to the description in the above embodiment and will not be repeated here.
  • An embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored.
  • the computer program is executed by a processor, the various processes and steps in the simulation method for reducing the dimension of a unit matrix provided in an embodiment of the present invention are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.
  • the storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM).
  • the technical solution of the present invention can be embodied in the form of a software product.
  • the computer software product 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 may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Physics (AREA)
  • General Engineering & Computer Science (AREA)
  • Geometry (AREA)
  • Software Systems (AREA)
  • Databases & Information Systems (AREA)
  • Evolutionary Computation (AREA)
  • Computer Hardware Design (AREA)
  • Algebra (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Computer Graphics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computing Systems (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Operations Research (AREA)
  • Probability & Statistics with Applications (AREA)
  • Surface Acoustic Wave Elements And Circuit Networks Thereof (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

本发明适用于压电材料力电耦合技术领域,提供了一种降低单元矩阵维度的仿真方法、系统及相关设备,所述仿真方法包括:获取声表面波器件的几何结构,并根据几何结构将声表面波器件划分为多个基础结构;构建用于有限单元法的权函数和梯度函数,根据权函数和梯度函数构建六节点四边形单元;根据有限单元法对基础结构进行网格划分,并利用六节点四边形单元计算基础结构对应的仿真单元矩阵;将不同的基础结构得到的仿真单元矩阵进行拼接级联,得到仿真整体矩阵,并计算仿真整体矩阵在预设仿真频率下的频点频率响应,得到声表面波器件的仿真频率响应曲线。本发明通过在单元矩阵计算中使用六节点四边形单元,降低了单元矩阵的维度,提高了仿真计算效率。

Description

降低单元矩阵维度的仿真方法、系统及相关设备 技术领域
本发明属于压电材料力电耦合技术领域,尤其涉及一种降低单元矩阵维度的仿真方法、系统及相关设备。
背景技术
随着智能手机的发展,声表面波器件的需求量越来越大。声表面波器件是一种力电耦合的声学构件,例如谐振器,其通常通过有限单元法进行精准仿真。有限单元法(ISBN:7-80159-853-9,2015)在20世纪50年代被提出,是一种将复杂结构计算问题转化为简单单元的分析和集合问题的数学计算方法,但是由于有限单元法对计算资源的消耗巨大,因此在实际环境中,不会对声表面波器件进行全尺寸的完整仿真,而是通常采用平面应变假设后进行二维仿真。但是,当声表面波器件中的插指结构较多或模型精度要求较高时,二维模型也会存在较大的计算资源与时间消耗。根据声表面波器件插指结构周期性变化的特点,层次级联技术得到了广泛的运用。
层次级联技术通过Schur补(舒尔补)运算消除内部自由度的操作,大幅降低了对计算资源的需求,使得声表面波器件的全尺寸仿真变为可能。问题在于层次级联技术在应用过程中需要对声表面波器件中的基本结构,如单根插指、GAP、左右PML层进行多次的Schur补运算,在实际仿真过程中,由于基本结构是通过有限单元法计算的单元矩阵拼接形成,基本结构的整体矩阵仍然较大,在Schur补运算时会产生巨大的性能消耗。传统的有限单元法的仿真中通常会通过减少网格数量以及降低单元阶数来提升计算效率,但一方面,网格数量的降低会影响模拟结果,另一方面,电磁场的仿真也需要二阶的单元来保证计算精度。
因此,仿真中的基本结构的整体矩阵尺寸较难减小。现有技术中,一般通过并行技术、GPU加速等方法对基本结构的Schur补运算进行加速,但在个人计算机与没有GPU的集群上仍主要通过CPU进行串行计算,因此该技术仍存在优化空间。
发明内容
本发明实施例提供一种降低单元矩阵维度的仿真方法、系统及相关设备,旨在解决现有的声表面波器件仿真过程使用层次级联技术得到的单元矩阵维度大、仿真效率低的问题。
第一方面,本发明实施例提供一种降低单元矩阵维度的仿真方法,所述仿真方法用于对声表面波器件进行仿真,所述仿真方法包括以下步骤:
获取所述声表面波器件的几何结构,并根据所述几何结构将所述声表面波器件划分为多个基础结构;
构建用于有限单元法的权函数和梯度函数,根据所述权函数和所述梯度函数构建六节点四边形单元;
根据所述有限单元法对所述基础结构进行网格划分,并利用所述六节点四边形单元计算所述基础结构对应的仿真单元矩阵;
将不同的所述基础结构得到的所述仿真单元矩阵进行拼接级联,得到仿真整体矩阵,并计算所述仿真整体矩阵在预设仿真频率下的频点频率响应,得到所述声表面波器件的仿真频率响应曲线。
更进一步地,定义所述权函数为Nφ,所述梯度函数为Bφ,所述权函数Nφ满足:
Nφ={N1,N2,N3,N4,N5,N6};





其中,s、t分别为局部坐标系中的横坐标、纵坐标;
所述梯度函数Bφ满足:
更进一步地,所述六节点四边形单元在水平方向为二阶插值,在竖直方向为一阶插值。
更进一步地,所述基础结构包括插指结构和GAP结构。
更进一步地,将不同的所述基础结构得到的所述仿真单元矩阵进行拼接级联,得到仿真整体矩阵的步骤,还包括:
使用Schur补运算消去所述仿真单元矩阵进行拼接级联时产生的共用矩阵节点。
第二方面,本发明实施例还提供一种降低单元矩阵维度的仿真系统,所述仿真系统用于对声表面波器件进行仿真,包括:
仿真参数获取模块,用于获取所述声表面波器件的几何结构,并根据所述几何结构将所述声表面波器件划分为多个基础结构;
节点单元构建模块,用于构建有限单元法的权函数和梯度函数,根据所述权函数和所述梯度函数构建六节点四边形单元;
有限单元建模模块,用于根据所述有限单元法对所述基础结构进行网格划分,并利用所述六节点四边形单元计算所述基础结构对应的仿真单元矩阵;
级联仿真模块,用于将不同的所述基础结构得到的所述仿真单元矩阵进行拼接级联,得到仿真整体矩阵,并计算所述仿真整体矩阵在预设仿真频率下的频点频率响应,得到所述声表面波器件的仿真频率响应曲线。
第三方面,本发明实施例还提供一种计算机设备,包括:存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机程序,所述处理器执行所述计算机程序时实现如上述实施例中任意一项所述的降低单元矩阵维度的仿真方法中的步骤。
第四方面,本发明实施例还提供一种计算机可读存储介质,所述计算机可读存储介质上存储有计算机程序,所述计算机程序被处理器执行时实现如上述实施例中任意一项所述的降低单元矩阵维度的仿真方法中的步骤。
本发明所达到的有益效果,相对于现有技术的声表面波器件仿真使用有限元理论中九节点拉格朗日单元计算单元矩阵的方法,通过利用声表面波器件主要在传播方向的位移场、电势场呈现非线性变化,而在垂直于传播方向位移场、电势场近乎线性变化并快速衰减的特点,构建了一种新的六节点四边形单元并在单元矩阵计算中使用,从而降低了单元矩阵的维度,使得使用有限单元法的声表面波器件仿真的计算要求降低,提高了计算效率。
附图说明
图1是现有技术的使用有限单元法的声表面波器件的仿真步骤流程框图;
图2是本发明提供的声表面波器件的基础结构划分示意图;
图3是本发明实施例提供的有限单元法中常用的二阶拉格朗日单元示意图;
图4是本发明实施例提供的单元矩阵拼接为整体矩阵的过程示意图;
图5是本发明实施例提供的降低单元矩阵维度的仿真方法的步骤流程框图;
图6是本发明实施例构建的六节点四边形单元示意图;
图7是本发明实施例提供的不同节点数单元对比示意图;
图8是本发明实施例提供的降低单元矩阵维度的仿真方法与现有技术的仿真对比结果示意图;
图9是本发明实施例提供的另一种降低单元矩阵维度的仿真方法与现有技术的仿真对比结果示意图;
图10是本发明实施例提供的降低单元矩阵维度的仿真系统200的结构示意图;
图11是本发明实施例提供的计算机设备的结构示意图。
具体实施方式
为了使本发明的目的、技术方案及优点更加清楚明白,以下结合附图及实施例,对本发明进行进一步详细说明。应当理解,此处所描述的具体实施例仅仅用以解释本发明,并不用于限定本发明。
为便于理解,本发明实施例首先对现有技术的使用有限单元法的声表面波器件的仿真过程进行说明,本发明实施例所述的仿真过程,以声表面波器件的几何参数为初始参数,仿真的最终目的是获取声表面波器件在某一频率下的仿真频率响应曲线。
具体的,现有技术的使用有限单元法的声表面波器件的仿真步骤流程框图如图1所示,其包括以下步骤:
S1、获取需要仿真的声表面波器件的几何结构。
一般的,声表面波器件的几何结构信息,包括插指结构的膜厚、金属化率、pitch等信息。
S2、根据所述几何结构划分为基础结构。
如图2所示,图2中的上部为获取的几何结构信息,将其划分为后的基础结构如图2的下部所示,其中,具有电极结构的为插指结构,没有电极结构的为GAP结构。
S3、对所述基础结构根据有限单元法进行建模。
对基本结构使用有限单元法进行网格划分,并构建出特定频率下基础结构的整体矩阵A,一般的,步骤S3包括以下子步骤:
S31、网格划分。
S32、确定权函数与梯度函数类型。
权函数的作用可以理解为对一个范围内自由度的拟合,现有技术中较为常见的插值方法就是拉格朗日插值,假如一条已知的曲线f(x)中有三个已知点(x1,y1)(x2,y2),(x3,y3),则该段曲线f(x)可以通过插值进行表示,插值结果为:
上式可以优化地表示为P(x)=NxyN,其中:
yN={y1,y2,y3}T
而对于一个多维问题z(x,y),可以通过两个正交的插值函数Nx与Ny正交获得,即:
z(x,y)=NxNyz(x,y)N
综上所述,可以得到结论:对于拉格朗日插值函数,其拟合曲线的阶数取决于已知点的个数。
对于复杂的物理模型,其自由度的变化趋势往往是非线性的,所以一般需要至少二阶的插值函数进行拟合,二阶的插值函数可以表示为图3所示的有限单元法中常用的二阶拉格朗日单元。
一般地,使用有限单元法进行声表面波器件的仿真过程,需要对传统的二阶拉格朗日单元的权函数进行推导,其过程如下:
定义在一个网格中的自由度随空间变化为φ(x,y),而每个节点的自由度值可以表示为φN={φ1,φ2,φ3,φ4,φ5,φ6,φ7,φ8,φ9}T,根据各个节点的自由度可以构建一个权函数Nφ矩阵,从而对φ(x,y)进行插值,即φ(x,y)=NφφN
在有限单元法中,权函数的推导是在局部坐标系(s,t)中进行的,将九节点的坐标分别表示为:
p1=(-1,-1),p2=(1,-1),p3=(1,1),p4=(-1,1),p5=(0,-1),p6=(1,0),p7
(0,1),p8=(-1,,),p9=(0,0);
根据坐标,将上述的拉格朗日插值形式获得每个节点的插值函数,并分别表示为:








根据以上插值函数,权函数矩阵表示为:
Nφ={N1,N2,N3,N4,N5,N6,N7,N8,N9};
而有限单元法中需要用到的梯度函数Bφ则是权函数在局部坐标系对不同方向的偏导数构成的矩阵,即可以将矩阵Bφ完全展开为:
S33、计算单元矩阵。
声表面波器件的数值求解就是对动力学方程与麦克斯韦方程耦合的方程组进行求解,动力学方程与麦克斯韦方程组表达式为:
而有限单元法九是通过将方程两侧乘一个试函数并在一个小区域内使积分等于0来进行求解,即:
其中的应力σij与电位移Di均为中间变量,实际求解的自由度是位移ui与电势φ,因此有应力与电位移相对位移与电势的关系σi=ijkIuk,I-eijkφ,i、Di=eijkuj,k+eijφ,j,其中,uk,I=εkI表示应变,cijkI为弹性常数,eijk为压电常数,eij为介电常数;
上述方程组使用自由度位移ui与电势φ进行表示时,其需要对位移与电势进行二阶求导,对解的连续性要求较高,使用散度定理将上述方程组进行转化得到:
此时仍然要求自由度位移ui与电势φ是连续解,而在有限单元法中,目标是计算的是每个节点的自由度,因此,需要使用权函数Nu与Nφ通过节点位移与电势φN对位移ui与电势φ进行插值,即以及φ=NφφN
而对于应变εij与电位移Di,其是位移ui与电势φ的梯度,可以对权函数求导获得梯度函数Bu=Nu,i与Bφ=Nφ,i,并且使用梯度函数Bu与Bφ通过节点位移与电势φN对应变εij与电位移Di进行插值,即以及
此时,原动力学方程与麦克斯韦方程组可以表示为:
接着使用伽辽金方法(Galerkin method),选择与权函数Nu与Nφ相同的试函数原动力学方程与麦克斯韦方程组进一步转换为:
将上述方程组用矩阵表述为:
此刻的动力学方程与麦克斯韦方程组虽然可以用矩阵表述,但其仍然为积分形式,而不是传统数值方法中的代数形式,因此有限单元法中对于该积分转换到局部坐标系下使用高斯积分法进行处理,将全局坐标系下的积分dV=dxdy转化为局部坐标系下的积分dsdt,即:
上式中的就是单元矩阵,其具体表达式如下所示:


S34、单元矩阵拼接为整体矩阵。
以两个基础结构为例,其各自的单元矩阵都简化表示为4×4的矩阵,分别为k1与k2,如下所示,其中,k1的第3、4个自由度(右下四个)与k2的第1、2个自由度(左上四个)表示的是同一个节点。
因此将这两个单元进行拼接时,需要将这两个自由度的信息进行相加,从而获得这两个单元的整体矩阵A,如下所示:
示例性的,将单元矩阵拼接为整体矩阵的过程示意如图4所示。
S4、对所述基础结构消去内部自由度。
利用Schur补运算消去基础结构的中间自由度,只保留结构的左右边界自由度以及电极与基底交界处的电势自由度从而获得Schur补后的整体矩阵ASchur,并使用Schur补之后的整体矩阵ASchur替代其原有的整体矩阵A。
S5、将所述基础结构级联。
将Schur补运算后基础结构的整体矩阵ASchur进行拼接级联,其级联原理类似于有限单元法中的整体矩阵拼接,是将声表面波器件中相邻的两个基础结构的整体矩阵A1与A2进行组合。
有限单元法中的组合原理,是对于M维的整体矩阵A1(M*M),与N维的整体矩阵A2(N*N),当两个基础结构有K个自由度是共用的,就可以构建一个新的M+N-K维的整体矩阵A3((M+N-K)*(M+N-K)),其中,不共用自由度对应的矩阵信息在拼接后的矩阵中直接使用,而共用自由度对应的矩阵信息则相加后使用。
S6、获得声表面波器件频率响应。
将最终级联后的矩阵通过预设的电学条件结合数学公式进行处理,并获得该频率下声表面波器件的频率响应。
例如,通过对基础结构的两个单元矩阵进行级联后,可以得到一个描述整个声表面波器件的矩阵M,矩阵M描述了左右边界及电极与基底交界处的自由度,可以对其进行分块为2*2的广义分块矩阵,分别描述要分析的电自由度E与其他自由度B,如下所示:
对于仿真目标需要计算的频率响应,如导纳参数,可以使用如下的公式进行处理即可获得导纳参数:
Y=jwWMEE
至此,现有技术的使用有限单元法的进行声表面波器件的仿真,并得到某一频率下的仿真频率响应曲线的过程结束。
现有技术使用层次级联技术,通过将声表面波器件划分成多个插指结构,对每个插指结构进行Schur补运算消去内部自由度,之后将Schur补后的插指结构按照顺序进行级联,从而减小了声表面波器件整体矩阵的尺寸,但由于步骤S4中涉及Schur补运算,在Schur补运算中存在一次矩阵求逆操作与两次矩阵乘法操作,由于并且矩阵求逆与矩阵乘法的时间复杂度约为O(n3),也就是当整体矩阵A较大时,获得Schur补运算结果ASchur的时间也会显著的增大,因此,Schur补运算会占据较多的计算时间从而拖慢计算效率,最终成为现有技术的计算效率瓶颈。
在本发明实施例具体是对现有技术中的步骤S3进行改进,具体的,请参照图5,图5是本发明实施例提供的降低单元矩阵维度的仿真方法的步骤流程框图,所述仿真方法用于对声表面波器件进行仿真,具体包括以下步骤:
S101、获取所述声表面波器件的几何结构,并根据所述几何结构将所述声表面波器件划分为多个基础结构。
更进一步地,所述基础结构包括插指结构和GAP结构。
S102、构建用于有限单元法的权函数和梯度函数,根据所述权函数和所述梯度函数构建六节点四边形单元。
更进一步地,所述六节点四边形单元在水平方向为二阶插值,在竖直方向为一阶插值。
具体的,本发明实施例提出的所述六节点四边形单元利用声表面波器件主要在传播方向的位移场、电势场呈现非线性变化,而在垂直于传播方向位移场、电势场近乎线性变化并快速衰减的特点,相较于原本的九节点拉格朗日单元, 所述六节点四边形单元在水平方向为二阶插值,在竖直方向为一阶插值,本发明实施例构建的六节点四边形单元示意图如图6所示。
之后,根据拉格朗日插值法,与传统的九节点拉格朗日单元类似,构建出该六节点单元的权函数与梯度函数。
更进一步地,定义所述权函数为Nφ,所述梯度函数为Bφ,所述权函数Nφ满足:
Nφ={N1,N2,N3,N4,N5,N6};





其中,s、t分别为局部坐标系中的横坐标、纵坐标;
所述梯度函数Bφ满足:
基于逆推导的关系,在实际过程中,所述六节点四边形单元的权函数与梯度函数需要提前确定,再基于所述权函数与所述梯度函数进行单元构建。
S103、根据所述有限单元法对所述基础结构进行网格划分,并利用所述六节点四边形单元计算所述基础结构对应的仿真单元矩阵。
本发明还提供另一种实施方式,在层次级联的应用中,经常也会采用COMSOL-Matlab联合仿真层次级联算法,即步骤S102、S103通过调用COMSOL有限单元法软件,并在COMSOL软件中指定基础结构的几何参数, 从而建模、自动划分网络、并输出整体矩阵;
此时,由于无法在COMSOL软件中确定所述六节点四边形单元的权函数与梯度函数,而只能调用COMSOL软件中内置的二阶拉格朗日单元或者巧凑边点元,因此需要采用一定的方法将COMSOL软件导出的整体矩阵进行处理,从而获得本发明实施例中的所述六节点四边形单元的整体矩阵。
具体的,请参照图7,图7是本发明实施例提供的不同节点数单元对比示意图,对于COMSOL软件中导出的二阶拉格朗日单元(9-node)构成的整体矩阵,其权函数分别为N1至N9,可以将编号为8、9、6的形函数平分叠加到1、5、2单元与4、7、3单元中,以获得本发明实施例中的所述六节点四边形单元的权函数,即:
N1 6=N1 9+1/2*N8 9
N2 6=N2 9+1/2*N6 9
N3 6=N3 9+1/2*N6 9
N4 6=N4 9+1/2*N8 9
N5 6=N5 9+1/2*N9 9
N6 6=N7 9+1/2*N9 9
而对于COMSOL软件中构建的巧凑边点元(8-node),其权函数分别为N1至N8,可以将编号为8、6的形函数平分叠加到1、2单元与4、3单元中以获得本发明实施例中的所述六节点四边形单元的权函数,即:
N1 6=N1 9+1/2*N8 8
N2 6=N2 9+1/2*N6 8
N3 6=N3 9+1/2*N6 8
N4 6=N4 9+1/2*N8 8
N5 6=N5 9
N6 6=N7 9
S104、将不同的所述基础结构得到的所述仿真单元矩阵进行拼接级联,得到仿真整体矩阵,并计算所述仿真整体矩阵在预设仿真频率下的频点频率响应, 得到所述声表面波器件的仿真频率响应曲线。
更进一步地,将不同的所述基础结构得到的所述仿真单元矩阵进行拼接级联,得到仿真整体矩阵的步骤,还包括:
使用Schur补运算消去所述仿真单元矩阵进行拼接级联时产生的共用矩阵节点。
示例性的,本发明实施例提供的降低单元矩阵维度的仿真方法与现有技术的仿真对比示例如下:
在单一谐振器频率下,单根插值的指条宽度pitch为0.9635微米,金属电极高度为0.146微米,金属化率0.6,模型的厚度为32.759微米,模型两侧使用该插值各构建100根插指的反射栅,模型内部使用50对共100根插指作为IDT,计算的频点数为400个频点,每个频点均为1Mhz,同时,为了进行仿真对比,在实验时设置了6倍pitch的基底深度,并且在基底底部设置了2倍pitch的PML层防止边界反射,在反射栅外侧同样设置了2倍pitch宽度的PML防止边界反射;之后,对其进行网格的划分,其中电极的网格数量为3*3=9,基底的网格数量为30*5=150,PML的网格数量为5*3=15,共174个网格。
作为理论对比,当使用传统的二阶拉格朗日单元,生成的整体矩阵维度为(3292×9292),在计算单元矩阵、并完整整体矩阵的计算共计花费时间约700ms;而本发明实施例使用的所述六节点四边形单元,生成的整体矩阵维度为(1688×1688),在计算单元矩阵、并完整整体矩阵的计算共计花费时间约330ms。
本发明实施例提供的降低单元矩阵维度的仿真方法与现有技术的仿真对比结果如图8、图9所示,可以看出,本发明实施例采用的六结点四边形单元的计算耗时远小于九节点单元与八结点单元,并且,六结点四边形单元的计算耗时为九节点单元的20%,并且频率响应也较为接近,体现了本发明实施例提供的降低单元矩阵维度的仿真方法在计算精度与计算效率上具有更好的性能。
本发明所达到的有益效果,相对于现有技术的声表面波器件仿真使用有限元理论中九节点拉格朗日单元计算单元矩阵的仿真方法,通过利用声表面波器 件主要在传播方向的位移场、电势场呈现非线性变化,而在垂直于传播方向位移场、电势场近乎线性变化并快速衰减的特点,构建了一种新的六节点四边形单元并在单元矩阵计算中使用,从而降低了单元矩阵的维度,使得使用有限单元法的声表面波器件仿真的计算要求降低,提高了计算效率。
本发明实施例还提供一种降低单元矩阵维度的仿真系统,所述仿真系统用于对声表面波器件进行仿真,请参照图10,图10是本发明实施例提供的降低单元矩阵维度的仿真系统200的结构示意图,其包括:
仿真参数获取模块201,用于获取所述声表面波器件的几何结构,并根据所述几何结构将所述声表面波器件划分为多个基础结构;
节点单元构建模块202,用于构建有限单元法的权函数和梯度函数,根据所述权函数和所述梯度函数构建六节点四边形单元;
有限单元建模模块203,用于根据所述有限单元法对所述基础结构进行网格划分,并利用所述六节点四边形单元计算所述基础结构对应的仿真单元矩阵;
级联仿真模块204,用于将不同的所述基础结构得到的所述仿真单元矩阵进行拼接级联,得到仿真整体矩阵,并计算所述仿真整体矩阵在预设仿真频率下的频点频率响应,得到所述声表面波器件的仿真频率响应曲线。
所述降低单元矩阵维度的仿真系统200能够实现如上述实施例中的降低单元矩阵维度的仿真方法中的步骤,且能实现同样的技术效果,参上述实施例中的描述,此处不再赘述。
本发明实施例还提供一种计算机设备,请参照图11,图11是本发明实施例提供的计算机设备的结构示意图,所述计算机设备300包括:存储器302、处理器301及存储在所述存储器302上并可在所述处理器301上运行的计算机程序。
所述处理器301调用所述存储器302存储的计算机程序,执行本发明实施例提供的降低单元矩阵维度的仿真方法中的步骤,请结合图5,具体包括:
S101、获取所述声表面波器件的几何结构,并根据所述几何结构将所述声 表面波器件划分为多个基础结构;
S102、构建用于有限单元法的权函数和梯度函数,根据所述权函数和所述梯度函数构建六节点四边形单元;
S103、根据所述有限单元法对所述基础结构进行网格划分,并利用所述六节点四边形单元计算所述基础结构对应的仿真单元矩阵;
S104、将不同的所述基础结构得到的所述仿真单元矩阵进行拼接级联,得到仿真整体矩阵,并计算所述仿真整体矩阵在预设仿真频率下的频点频率响应,得到所述声表面波器件的仿真频率响应曲线。
更进一步地,定义所述权函数为Nφ,所述梯度函数为Bφ,所述权函数Nφ满足:
Nφ={N1,N2,N3,N4,N5,N6};





其中,s、t分别为局部坐标系中的横坐标、纵坐标;
所述梯度函数Bφ满足:
更进一步地,所述六节点四边形单元在水平方向为二阶插值,在竖直方向为一阶插值。
更进一步地,所述基础结构包括插指结构和GAP结构。
更进一步地,将不同的所述基础结构得到的所述仿真单元矩阵进行拼接级联,得到仿真整体矩阵的步骤,还包括:
使用Schur补运算消去所述仿真单元矩阵进行拼接级联时产生的共用矩阵节点。
本发明实施例提供的计算机设备300能够实现如上述实施例中的降低单元矩阵维度的仿真方法中的步骤,且能实现同样的技术效果,参上述实施例中的描述,此处不再赘述。
本发明实施例还提供一种计算机可读存储介质,所述计算机可读存储介质上存储有计算机程序,该计算机程序被处理器执行时实现本发明实施例提供的降低单元矩阵维度的仿真方法中的各个过程及步骤,且能实现相同的技术效果,为避免重复,这里不再赘述。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程,是可以通过计算机程序来指令相关的硬件来完成,所述的程序可存储于一计算机可读取存储介质中,该程序在执行时,可包括如上述各方法的实施例的流程。其中,所述的存储介质可为磁碟、光盘、只读存储记忆体(Read-Only Memory,ROM)或随机存取存储器(Random Access Memory,简称RAM)等。
需要说明的是,在本文中,术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含,从而使得包括一系列要素的过程、方法、物品或者装置不仅包括那些要素,而且还包括没有明确列出的其他要素,或者是还包括为这种过程、方法、物品或者装置所固有的要素。在没有更多限制的情况下,由语句“包括一个……”限定的要素,并不排除在包括该要素的过程、方法、物品或者装置中还存在另外的相同要素。
通过以上的实施方式的描述,本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现,当然也可以通过硬件,但很多情况下前者是更佳的实施方式。基于这样的理解,本发明的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来, 该计算机软件产品存储在一个存储介质(如ROM/RAM、磁碟、光盘)中,包括若干指令用以使得一台终端(可以是手机,计算机,服务器,空调器,或者网络设备等)执行本发明各个实施例所述的方法。
上面结合附图对本发明的实施例进行了描述,所揭露的仅为本发明较佳实施例而已,但是本发明并不局限于上述的具体实施方式,上述的具体实施方式仅仅是示意性的,而不是限制性的,本领域的普通技术人员在本发明的启示下,在不脱离本发明宗旨和权利要求所保护的范围情况下,还可做出很多形式用等同变化,均属于本发明的保护之内。

Claims (8)

  1. 一种降低单元矩阵维度的仿真方法,所述仿真方法用于对声表面波器件进行仿真,其特征在于,所述仿真方法包括以下步骤:
    获取所述声表面波器件的几何结构,并根据所述几何结构将所述声表面波器件划分为多个基础结构;
    构建用于有限单元法的权函数和梯度函数,根据所述权函数和所述梯度函数构建六节点四边形单元;
    根据所述有限单元法对所述基础结构进行网格划分,并利用所述六节点四边形单元计算所述基础结构对应的仿真单元矩阵;
    将不同的所述基础结构得到的所述仿真单元矩阵进行拼接级联,得到仿真整体矩阵,并计算所述仿真整体矩阵在预设仿真频率下的频点频率响应,得到所述声表面波器件的仿真频率响应曲线。
  2. 如权利要求1所述的降低单元矩阵维度的仿真方法,其特征在于,定义所述权函数为Nφ,所述梯度函数为Bφ,所述权函数Nφ满足:
    Nφ={N1,N2,N3,N4,N5,N6};





    其中,s、t分别为局部坐标系中的横坐标、纵坐标;
    所述梯度函数Bφ满足:
  3. 如权利要求1所述的降低单元矩阵维度的仿真方法,其特征在于,所述六节点四边形单元在水平方向为二阶插值,在竖直方向为一阶插值。
  4. 如权利要求1所述的降低单元矩阵维度的仿真方法,其特征在于,所述基础结构包括插指结构和GAP结构。
  5. 如权利要求1所述的降低单元矩阵维度的仿真方法,其特征在于,将不同的所述基础结构得到的所述仿真单元矩阵进行拼接级联,得到仿真整体矩阵的步骤,还包括:
    使用Schur补运算消去所述仿真单元矩阵进行拼接级联时产生的共用矩阵节点。
  6. 一种降低单元矩阵维度的仿真系统,所述仿真系统用于对声表面波器件进行仿真,其特征在于,包括:
    仿真参数获取模块,用于获取所述声表面波器件的几何结构,并根据所述几何结构将所述声表面波器件划分为多个基础结构;
    节点单元构建模块,用于构建有限单元法的权函数和梯度函数,根据所述权函数和所述梯度函数构建六节点四边形单元;
    有限单元建模模块,用于根据所述有限单元法对所述基础结构进行网格划分,并利用所述六节点四边形单元计算所述基础结构对应的仿真单元矩阵;
    级联仿真模块,用于将不同的所述基础结构得到的所述仿真单元矩阵进行拼接级联,得到仿真整体矩阵,并计算所述仿真整体矩阵在预设仿真频率下的频点频率响应,得到所述声表面波器件的仿真频率响应曲线。
  7. 一种计算机设备,其特征在于,包括:存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机程序,所述处理器执行所述计算机程序时实现如权利要求1至5中任意一项所述的降低单元矩阵维度的仿真方法中的步骤。
  8. 一种计算机可读存储介质,其特征在于,所述计算机可读存储介质上存储有计算机程序,所述计算机程序被处理器执行时实现如权利要求1至5中任意一项所述的降低单元矩阵维度的仿真方法中的步骤。
PCT/CN2023/132522 2022-12-06 2023-11-20 降低单元矩阵维度的仿真方法、系统及相关设备 WO2024120159A2 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202211557508.2A CN115577603B (zh) 2022-12-06 2022-12-06 降低单元矩阵维度的仿真方法、系统及相关设备
CN202211557508.2 2022-12-06

Publications (1)

Publication Number Publication Date
WO2024120159A2 true WO2024120159A2 (zh) 2024-06-13

Family

ID=84590282

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/132522 WO2024120159A2 (zh) 2022-12-06 2023-11-20 降低单元矩阵维度的仿真方法、系统及相关设备

Country Status (2)

Country Link
CN (1) CN115577603B (zh)
WO (1) WO2024120159A2 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116384204B (zh) * 2023-05-31 2023-08-22 深圳飞骧科技股份有限公司 混合单元阶数的有限元仿真方法、系统及相关设备
CN116362093B (zh) * 2023-05-31 2023-09-01 深圳飞骧科技股份有限公司 基于多物理场分离自由度的仿真方法、系统及相关设备

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11182522B2 (en) * 2016-08-29 2021-11-23 Resonant, Inc. Hierarchical cascading in FEM simulations of SAW devices
US10797673B2 (en) * 2016-08-29 2020-10-06 Resonant Inc. Hierarchical cascading in two-dimensional finite element method simulation of acoustic wave filter devices
WO2020123698A1 (en) * 2018-12-11 2020-06-18 Resonant, Inc. Fast, highly accurate, full-fem surface acoustic wave simulation
CN110442907B (zh) * 2019-07-02 2023-04-28 浙江中科电声研发中心 压电式mems扬声器基本特性的数值仿真分析方法
CN113962084A (zh) * 2021-10-22 2022-01-21 中国电子科技集团公司第二十六研究所 基于降维pde模型的声表面波谐振器的频响特性分析方法
CN114117690B (zh) * 2022-01-27 2022-06-14 深圳飞骧科技股份有限公司 声表面波滤波器仿真方法、相关设备及存储介质

Also Published As

Publication number Publication date
CN115577603B (zh) 2023-03-17
CN115577603A (zh) 2023-01-06

Similar Documents

Publication Publication Date Title
WO2024120159A2 (zh) 降低单元矩阵维度的仿真方法、系统及相关设备
WO2024120165A1 (zh) 快速构建整体矩阵的仿真方法、系统及相关设备
WO2024120164A1 (zh) 减小单元矩阵的级联误差的仿真方法、系统及相关设备
WO2023142626A1 (zh) 声表面波滤波器仿真方法、相关设备及存储介质
Sundar et al. Low-constant parallel algorithms for finite element simulations using linear octrees
CN114861576A (zh) 超导量子芯片版图的仿真方法及装置、电子设备和介质
Khaitan et al. A class of new preconditioners for linear solvers used in power system time-domain simulation
Balewski et al. Step on it bringing fullwave finite-element microwave filter design up to speed
Tan Fundamental implicit FDTD schemes for computational electromagnetics and educational mobile apps
CN111859835A (zh) 一种电路互连网络模型的降阶方法、降阶装置及降阶设备
JP2023058606A (ja) 超伝導量子チップの設計方法及び装置、電子機器並びに媒体
Basermann et al. Dynamic load-balancing of finite element applications with the drama library
WO2024007652A1 (zh) 一种大型稀疏矩阵加速求解方法、系统及存储介质
Ghosh et al. A FETI‐preconditioned conjugate gradient method for large‐scale stochastic finite element problems
CN108763777B (zh) 基于泊松方程显式解的vlsi全局布局模型建立方法
Stavroulakis et al. Non-overlapping domain decomposition solution schemes for structural mechanics isogeometric analysis
Mittal et al. Design exploration and implementation of simplex algorithm over reconfigurable computing platforms
Ma et al. An explicit-implicit mixed staggered asynchronous step integration algorithm in structural dynamics
CN116384204B (zh) 混合单元阶数的有限元仿真方法、系统及相关设备
Majak et al. Higher order Haar wavelet method for solving differential equations
CN116362093B (zh) 基于多物理场分离自由度的仿真方法、系统及相关设备
Perko et al. Low-cost, high-performance CNN simulator implemented in FPGA
Xu et al. Accommodating linear and nonlinear boundary conditions in wave digital simulations of PDE systems
Wong et al. Towards Scalable Full-Device Simulation for Surface Acoustic Wave Devices
Garcia-Mallen et al. Towards an Accelerator for Differential and Algebraic Equations Useful to Scientists