MEMS modeling method combining vector fitting with neural network transfer function

Abstract

The invention provides an MEMS modeling method combining vector fitting with a neural network transfer function, and relates to the technical field of micro electro mechanical systems. According to the embodiment of the invention, the fitting parameters corresponding to the performance curves of the MEMS devices with different geometric sizes and material characteristics can be predicted on the basis of vector fitting in combination with an artificial neural network transfer function algorithm, so that the performance curves of the MEMS devices with different geometric sizes and material characteristics are obtained; the modeling process of the MEMS device is quickly and accurately simulated and analyzed, and an MEMS designer can be guided to optimize the device.

Description

Technical Field

[0001] The invention relates to the technical field of micro-electromechanical systems, and in particular to a MEMS modeling method combining vector fitting with a neural network transfer function. Background Art

[0002] In recent years, with the continuous development of MEMS technology, the size of MEMS devices has gradually become smaller and more complex. When designing MEMS, it is necessary to use multiple geometric parameters as design variables and repeat finite element simulation iterative optimization. As the structure of MEMS devices becomes more complex, their finite element simulation becomes more time-consuming. In order to meet the design index specifications, the finite element method may take days or even months to obtain the global optimal solution. In order to optimize the efficiency of device modeling and realize the automated design of MEMS devices, it is necessary to develop an efficient MEMS device modeling method. Summary of the invention

[0003] The present invention provides a MEMS modeling method combining vector fitting with neural network transfer function to at least partially solve the above problems.

[0004] A first aspect of the present invention provides a MEMS modeling method combining vector fitting with a neural network transfer function, the method comprising: Determining target performance parameters of a target device and device parameters corresponding to the target performance parameters; Obtaining samples corresponding to the device parameters and sample device performance curves corresponding to the samples; Using a vector fitting method, the performance curve of the sample device is fitted, a transfer function is constructed, and sample fitting parameters are extracted; The artificial neural network is trained with the device parameters of the sample as input and the corresponding sample fitting parameters as output to obtain a fitting parameter prediction model; Inputting the device parameters of the sample to be tested corresponding to the target performance parameters of the target device into the fitting parameter prediction model to obtain predicted fitting parameters; Based on the predicted fitting parameters, a predicted device performance curve of the sample to be tested is obtained.

[0005] Optionally, the method further comprises: Based on the predicted device performance curve, it is determined whether the device parameter values ​​of the sample to be tested are qualified.

[0006] Optionally, the device parameters include multiple parameter dimensions, and the device parameters of the sample to be tested corresponding to the target performance parameters of the target device are input into the fitting parameter prediction model to obtain a predicted device performance curve of the sample to be tested, including: Acquire multiple samples to be tested, each sample to be tested having a different combination of device parameter values; Inputting the device parameters of the plurality of samples to be tested into the fitting parameter prediction model respectively to obtain a plurality of predicted fitting parameters; Based on the multiple predicted fitting parameters, obtaining predicted device performance curves of the multiple samples to be tested; The method further comprises: According to the predicted device performance curves of the multiple samples to be tested, a device parameter value combination corresponding to the best sample to be tested is determined.

[0007] Optionally, the method further comprises: The tolerance range of the device parameter is determined according to the predicted device performance curves of each of the plurality of samples to be tested.

[0008] Optionally, determining a target performance parameter of a target device and a device parameter corresponding to the target performance parameter includes: determining target performance parameters and multiple device parameters of a target device; Based on the principal component analysis method, main parameters corresponding to the target performance parameters are determined, and the main parameters are used as device parameters.

[0009] Optionally, obtaining a sample corresponding to the device parameter and a sample performance parameter corresponding to the sample includes: Determine parameter distribution intervals corresponding to device parameters; Sampling within the parameter distribution interval to obtain a sample; The finite element simulation method is used to determine the sample device performance curve corresponding to the sample.

[0010] A second aspect of the present invention provides a MEMS modeling device combining vector fitting with a neural network transfer function, wherein the MEMS modeling device combining vector fitting with a neural network transfer function comprises: A first determination module, used to determine target performance parameters of a target device and device parameters corresponding to the target performance parameters; An acquisition module, used to acquire samples corresponding to the device parameters and sample device performance curves corresponding to the samples; A fitting module, used to fit the performance curve of the sample device using a vector fitting method, construct a transfer function, and extract sample fitting parameters; A training module, used to train an artificial neural network with the device parameters of the sample as input and the corresponding sample fitting parameters as output to obtain a fitting parameter prediction model; A prediction module, used for inputting device parameters of the sample to be tested corresponding to the target performance parameters of the target device into the fitting parameter prediction model to obtain predicted fitting parameters; The curve determination module is used to obtain the predicted device performance curve of the sample to be tested based on the predicted fitting parameters.

[0011] Optionally, the device further comprises: The second determination module is used to determine whether the device parameter value of the sample to be tested is qualified based on the predicted device performance curve.

[0012] Optionally, the device parameter includes a plurality of device parameters, and the prediction module is used to: Acquire multiple samples to be tested, each sample to be tested having a different combination of device parameter values; Inputting the device parameters of the plurality of samples to be tested into the fitting parameter prediction model respectively to obtain a plurality of predicted fitting parameters; The curve determination module is used to obtain the predicted device performance curve of each of the multiple samples to be tested based on the multiple predicted fitting parameters; The device also includes: The third determination module is used to determine the device parameter value combination corresponding to the best sample to be tested according to the predicted device performance curves of each of the multiple samples to be tested.

[0013] Optionally, the device further comprises: The fourth determination module is used to determine the tolerance range of the device parameter according to the predicted device performance curves of each of the multiple samples to be tested.

[0014] Optionally, the first determining module is used to: determining target performance parameters and multiple device parameters of a target device; Based on the principal component analysis method, main parameters corresponding to the target performance parameters are determined, and the main parameters are used as device parameters.

[0015] Optionally, the acquisition module is used to: Determine parameter distribution intervals corresponding to device parameters; Sampling within the parameter distribution interval to obtain a sample; The finite element simulation method is used to determine the sample device performance curve corresponding to the sample.

[0016] The third aspect of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when executed, the processor implements the MEMS modeling method combining vector fitting with neural network transfer function as described in the first aspect of the present invention.

[0017] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the MEMS modeling method combining vector fitting with neural network transfer function as described in the first aspect of the present invention.

[0018] A fifth aspect of the present invention provides a computer program product, comprising a computer program / instruction, which is used by a processor to implement the steps in the MEMS modeling method of vector fitting combined with neural network transfer function as described in the first aspect of the present invention.

[0019] The embodiment of the present invention is based on vector fitting combined with an artificial neural network transfer function algorithm, and can predict the fitting parameters corresponding to the performance curves of MEMS devices with different geometric dimensions and material properties, thereby obtaining the performance curves of MEMS devices with different geometric dimensions and material properties, and quickly and accurately simulate and analyze the modeling process of MEMS devices, and can guide MEMS designers to optimize devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solution of the present invention, the accompanying drawings required for use in the description of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0021] Figure 1 It is a flow chart of the steps of the MEMS modeling method of vector fitting combined with neural network transfer function provided by the present invention; Figure 2 It is a model principle diagram of the MEMS modeling method combining vector fitting with neural network transfer function provided by the present invention; Figure 3 It is a principle diagram of an artificial neural network of a MEMS modeling method combining vector fitting with a neural network transfer function provided by the present invention; Figure 4 It is a structural schematic diagram of a MEMS modeling device combining vector fitting with a neural network transfer function provided by the present invention. DETAILED DESCRIPTION

[0022] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0023] Traditional MEMS device modeling methods usually use empirical formulas, equivalent models, and finite element simulation techniques. However, when optimizing the design of MEMS devices, multiple modeling and optimization iterations are usually required, which results in a long time required for each iteration and a large number of iterations, resulting in low computational efficiency. In order to overcome this problem, a new method has been proposed in recent years, which is a substitute model combining a neural network transfer function. This method can be used for parametric modeling and design optimization of microwave components, thereby significantly shortening the modeling and optimization time. Since MEMS devices have similar system characteristics to microwave components, neural network transfer functions can also be used to model the electrical, mechanical and other responses of MEMS devices. This new method provides a more efficient solution for the design of MEMS devices, and also provides new ideas and possibilities for future research and applications.

[0024] The embodiment of the present invention is based on an artificial neural network transfer function algorithm, can predict the performance of MEMS devices with different geometric dimensions and material properties, quickly and accurately simulate and analyze the modeling process of MEMS devices, and can guide MEMS designers to optimize devices.

[0025] The embodiment of the present invention provides a MEMS modeling method combining vector fitting with neural network transfer function, specifically, Figure 1 As shown, it shows a flow chart of the steps of the MEMS modeling method of vector fitting combined with neural network transfer function provided by an embodiment of the present invention, and the method includes the following steps: S101, determining a target performance parameter of a target device and a target parameter corresponding to the target performance parameter.

[0026] In the embodiment of the present invention, the target device is the MEMS device that currently needs to be analyzed, which may specifically include: sensors, micromotors, micropumps, filters, oscillators, antennas, micromirror arrays, interferometers, microfluidic chips, biochips, drug delivery systems, energy harvesters, etc.

[0027] In the embodiment of the present invention, after the target device is determined, the industrial background of the device can be determined based on relevant production requirements and industrial application requirements of the target device, and then all performance parameters that affect the performance of the MEMS device can be determined.

[0028] In the embodiment of the present invention, the target performance parameter refers to the main performance parameter concerned in the manufacturing process of the target device. For example, for a microfluidic chip, its target performance parameter may include: fluid control performance and geometric accuracy.

[0029] In the embodiment of the present invention, the target performance parameter may include one or more performance parameters. In actual application, each performance parameter may be analyzed separately.

[0030] In the embodiment of the present invention, the target parameters may include: all parameters that affect the performance of the MEMS device, such as material parameters, structural parameters, morphology parameters, etc.

[0031] S102, obtaining samples corresponding to the device parameters and sample device performance curves corresponding to the samples.

[0032] In the embodiment of the present invention, random sampling may be performed based on the parameter range of the device parameters designed during the manufacturing process of the target device to obtain samples, and then determine the sample device performance curve corresponding to the sample.

[0033] In the embodiment of the present invention, a sample may be obtained based on parameter data of a sample device obtained in an actual manufacturing process, and a sample device performance curve corresponding to the sample may be determined.

[0034] In the embodiment of the present invention, the data corresponding to the sample includes the value corresponding to the device parameter. When the device parameter includes multiple parameter dimensions, a sample is obtained by taking the value of each device parameter dimension.

[0035] For example, if the device parameters include width, thickness and roughness, then the width value can be 1, the thickness value can be 100, and the roughness value can be 1000 as a sample. Accordingly, a sample performance curve can be obtained for the sample with the width value 1, the thickness value 100, and the roughness value 1000.

[0036] In the embodiment of the present invention, the performance curve describes the response of the sample under different conditions, such as frequency response, temperature response, etc.

[0037] S103, using a vector fitting method to fit the performance curve of the sample device, construct a transfer function, and extract sample fitting parameters.

[0038] Vector fitting approximates given frequency domain data by fitting a set of rational functions (transfer functions), thereby extracting system parameters such as poles, residues, and constant terms. It is mainly used for fitting frequency domain response data, such as in circuit design and electromagnetic field analysis, to obtain a rational function model from measurement or simulation data, usually in the form of a transfer function.

[0039] In the embodiment of the present invention, the sample device performance curve refers to the response of the sample at different frequencies, such as S parameters, impedance, admittance, etc. The purpose of constructing the transfer function is for system-level simulation or circuit analysis, so it is necessary to extract parameters such as poles, zeros, residues, etc. from the data.

[0040] Usually the process of vector fitting includes the following steps: initializing poles, constructing a linear equation system, iteratively optimizing poles, and forming a transfer function.

[0041] In the embodiment of the present invention, a vector fitting method is used to fit the device performance curve, construct a transfer function, and extract the fitting parameters. Taking the FBAR device as an example, since the FBAR performance curve has an obvious peak response, the vector fitting method can obtain a better fit. At this time, the transfer function is: ; Where T represents the order of the transfer function, P i represents the pole, Z i Indicates zero point; h and d is a constant. s is a function of frequency, H is the frequency response curve of the FBAR device, P i and Z i , h, d represent the fitting parameters obtained by vector fitting , is the value of the device parameter corresponding to the sample.

[0042] S104, training an artificial neural network with the device parameters of the sample as input and the corresponding sample fitting parameters as output to obtain a fitting parameter prediction model.

[0043] In the embodiment of the present invention, an artificial neural network method is used (specifically, the model principle diagram is as follows Figure 2 As shown in the figure), a mathematical substitution model is constructed, and the values ​​of the device parameters corresponding to the samples obtained above are As model input, the sample fitting parameters of the transfer function (P i and Z i , h, d) as model output to train and learn the artificial neural network.

[0044] like Figure 2 As shown, in the embodiment of the present invention, the device parameters include multiple device parameters (piezoelectric layer thickness, piezoelectric performance and surface roughness), and the values ​​of each group of device parameters can be used as a sample. As shown in the figure, k groups of device parameter values ​​can be obtained, that is, k samples. Finite element simulation is performed on each sample to obtain the sample device performance curve of each k sample. Then, the vector fitting method can be used to fit the performance curve of each sample device to obtain the fitting parameters. Subsequently, the values ​​of the device parameters corresponding to the above-obtained samples are As model input, the sample fitting parameters of the transfer function (P i and Z i, h, d) as model output to train and learn the artificial neural network.

[0045] In the embodiment of the present invention, the principle diagram of the artificial neural network is as follows Figure 3 As shown, the model is trained with n-dimensional device parameters as input and m-dimensional fitting parameters as output.

[0046] In the embodiment of the present invention, the values ​​of the device parameters of the test sample can also be obtained by sampling. , the finite element simulation method is used for the test sample to obtain the device performance curve corresponding to the test sample, and the vector fitting method is used to extract the fitting parameters of the device performance curve of each test sample, which is called the actual value of the test sample .

[0047] The values ​​of the device parameters of the test sample Input into the trained model to obtain the predicted value of the fitting parameters output by the model , compare the predicted value with the actual value of each test sample Compare and measure the model prediction performance. If the error between the model prediction result and the actual result is small, the model training is confirmed to be complete. If the prediction result error is large, the model hyperparameter adjustment method can be used to iteratively modify the model hyperparameters until the model prediction accuracy meets the requirements and obtain the fitting parameter prediction model.

[0048] S105, inputting the device parameters of the sample to be tested corresponding to the target performance parameters of the target device into the fitting parameter prediction model to obtain predicted fitting parameters.

[0049] S106, obtaining a predicted device performance curve of the sample to be tested based on the predicted fitting parameters.

[0050] In the embodiment of the present invention, after the fitting parameter prediction model is obtained through training, the values ​​of the device parameters of the sample to be tested can be , input the fitting parameter prediction model, obtain the corresponding prediction fitting parameters, and then obtain the corresponding prediction device performance curve, so as to complete the mathematical modeling of the test sample without using the finite element simulation method.

[0051] Therefore, the MEMS modeling method of vector fitting combined with neural network transfer function proposed in the embodiment of the present invention can solve the problems faced by the design process of MEMS devices, such as complex structure, long simulation time, and difficult optimization. Although the traditional finite element simulation method has high accuracy, its calculation cost is high and the iteration cycle is long, which makes it difficult to meet the needs of modern industry for rapid design and efficient optimization. The MEMS device modeling method based on the artificial neural network transfer function algorithm can effectively solve these problems.

[0052] Based on the MEMS modeling method of vector fitting combined with neural network transfer function provided by the embodiment of the present invention, it can provide high-speed and accurate MEMS device modeling methods for MEMS design units and MEMS manufacturing wafer-level foundries, and at the same time bring new device yield analysis methods to MEMS manufacturing wafer-level foundries. By using the method of the embodiment of the present invention, it is expected to reduce the design iteration cycle of typical MEMS products by at least one order of magnitude, and greatly save the hardware resources, computing time and energy consumption costs of MEMS design units for MEMS device modeling. In addition, the method is particularly suitable for MEMS frequency domain devices (such as radio frequency MEMS devices, etc.), and the fitting degree and modeling accuracy of the frequency domain characteristic curve of the frequency domain device are high, which can meet the design requirements of high-end frequency domain MEMS devices. Finally, by applying the method of the embodiment of the present invention, the design-manufacturing iteration times of MEMS design units and MEMS wafer foundries can be effectively reduced, and the design-manufacturing iteration times of MEMS product research and development are expected to be reduced by 30%-50%, thereby reducing the development time of MEMS devices.

[0053] Specifically, in the embodiment of the present invention, the method further includes the following steps: S21, based on the predicted device performance curve, determining whether the device parameter values ​​of the sample to be tested are qualified.

[0054] In the embodiment of the present invention, the reliability and life evaluation of MEMS devices can be realized based on the predicted device performance curve of the sample to be tested, for example: combining the stress simulation curve under periodic load to estimate the fatigue life of the device (such as the long-term swing reliability of the micromirror). Another example: simulating the performance drift of MEMS devices under temperature, humidity or vibration environment, optimizing the packaging design or compensation algorithm.

[0055] Specifically, in an embodiment of the present invention, when the device parameter includes multiple device parameters, the device parameters of the sample to be tested corresponding to the target performance parameters of the target device are input into the fitting parameter prediction model to obtain the predicted device performance curve of the sample to be tested, including: S1, obtaining a plurality of samples to be tested, each sample to be tested having a different combination of device parameter values.

[0056] S2, inputting the device parameters of the plurality of samples to be tested into the fitting parameter prediction model respectively to obtain a plurality of predicted fitting parameters.

[0057] S3, based on the multiple predicted fitting parameters, obtaining predicted device performance curves of the multiple samples to be tested.

[0058] In this case, the method further comprises: S22, determining a device parameter value combination corresponding to an optimal sample to be tested according to the predicted device performance curves of each of the multiple samples to be tested.

[0059] In the embodiment of the present invention, after completing the MEMS device modeling based on the artificial neural network transfer function algorithm, the design optimization and iteration of the MEMS device can be realized based on the predicted device performance curve of the sample to be tested, for example: observing the influence of different design parameters (such as material properties, structural dimensions, cantilever beam thickness, etc.) on the performance through the device performance curve, and quickly screening the optimal parameter combination. Another example: before manufacturing the MEMS device, the device performance curve is used to predict the key indicators of the device such as sensitivity, resonant frequency, displacement, power consumption, etc., to reduce the cost of trial and error.

[0060] Specifically, in the embodiment of the present invention, the method further includes the following steps: S23, determining the tolerance range of the device parameter according to the predicted device performance curves of each of the plurality of samples to be tested.

[0061] In the embodiment of the present invention, the predicted device performance curve of the sample to be tested can also provide guidance on the process and manufacturing of MEMS devices, for example, simulating the impact of different process deviations (such as photolithography errors and etching angle deviations) on performance and determining the tolerance range of key dimensions. Another example is identifying structural weaknesses (such as stress concentration areas) through stress / strain curves and optimizing geometry or material selection to avoid fracture.

[0062] In an optional implementation, step S101 includes the following sub-steps: S1011, determining a target performance parameter and multiple device parameters of a target device.

[0063] S1012, determining main parameters corresponding to the target performance parameters based on a principal component analysis method, and using the main parameters as device parameters.

[0064] In the embodiment of the present invention, the device parameters that affect the target performance parameters may include multiple parameters. In order to simplify the calculation process, the embodiment of the present invention proposes that data dimension reduction can be performed, specifically, the principal component analysis method can be used to achieve data dimension reduction, and the main parameters corresponding to the target performance parameters can be obtained as device parameters.

[0065] In the embodiment of the present invention, other alternative methods may be used to achieve data dimensionality reduction. Alternatively, data dimensionality reduction may not be performed, and all device parameters related to the target performance parameters may be directly used as device parameters for subsequent sampling and model training.

[0066] In an optional implementation, step S102 includes the following sub-steps: S1021, determining a parameter distribution interval corresponding to a device parameter.

[0067] S1022: Sampling within the parameter distribution interval to obtain samples.

[0068] S1023, using a finite element simulation method to determine a sample device performance curve corresponding to the sample.

[0069] In the embodiment of the present invention, the upper and lower bounds of the parameter distribution interval corresponding to the device parameter can be obtained by online data statistics in the manufacturing process, and samples are obtained within the interval by uniform sampling or other sampling methods.

[0070] In the embodiment of the present invention, finite element simulation can be performed based on the values ​​of device parameters corresponding to the sample to obtain a sample device performance curve corresponding to the sample.

[0071] Based on the same inventive concept, the present invention also provides a MEMS modeling device combining vector fitting with neural network transfer function, such as Figure 4 As shown, it shows a structural schematic diagram of a MEMS modeling device for vector fitting combined with a neural network transfer function provided by an embodiment of the present invention, wherein the MEMS modeling device 400 for vector fitting combined with a neural network transfer function includes: A first determination module 401 is used to determine a target performance parameter of a target device and a device parameter corresponding to the target performance parameter; An acquisition module 402 is used to acquire samples corresponding to the device parameters and sample device performance curves corresponding to the samples; A fitting module 403 is used to fit the sample device performance curve using a vector fitting method, construct a transfer function, and extract sample fitting parameters; A training module 404 is used to train an artificial neural network using the device parameters of the sample as input and the corresponding sample fitting parameters as output to obtain a fitting parameter prediction model; Prediction module 405, used for inputting device parameters of the sample to be tested corresponding to the target performance parameters of the target device into the fitting parameter prediction model to obtain predicted fitting parameters; The curve determination module 406 is used to obtain the predicted device performance curve of the sample to be tested based on the predicted fitting parameters.

[0072] Optionally, the device further comprises: The second determination module is used to determine whether the device parameter value of the sample to be tested is qualified based on the predicted device performance curve.

[0073] Optionally, the device parameter includes a plurality of device parameters, and the prediction module 405 is used to: Acquire multiple samples to be tested, each sample to be tested having a different combination of device parameter values; Inputting the device parameters of the plurality of samples to be tested into the fitting parameter prediction model respectively to obtain a plurality of predicted fitting parameters; The curve determination module 406 is used to obtain the predicted device performance curve of each of the multiple samples to be tested based on the multiple predicted fitting parameters; The device also includes: The third determination module is used to determine the device parameter value combination corresponding to the best sample to be tested according to the predicted device performance curves of each of the multiple samples to be tested.

[0074] Optionally, the device further comprises: The fourth determination module is used to determine the tolerance range of the device parameter according to the predicted device performance curves of each of the multiple samples to be tested.

[0075] Optionally, the first determining module 401 is configured to: determining target performance parameters and multiple device parameters of a target device; Based on the principal component analysis method, main parameters corresponding to the target performance parameters are determined, and the main parameters are used as device parameters.

[0076] Optionally, the acquisition module 402 is used to: Determine parameter distribution intervals corresponding to device parameters; Sampling within the parameter distribution interval to obtain a sample; The finite element simulation method is used to determine the sample device performance curve corresponding to the sample.

[0077] Based on the same inventive concept, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes, the steps in the MEMS modeling method combining vector fitting with neural network transfer function as described in any of the above embodiments are implemented.

[0078] Based on the same inventive concept, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the MEMS modeling method combining vector fitting with neural network transfer function described in any of the above embodiments.

[0079] Based on the same inventive concept, the present invention provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps in the MEMS modeling method of vector fitting combined with neural network transfer function described in any of the above embodiments.

[0080] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0081] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, devices, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0082] The present invention is described with reference to the flowcharts and / or block diagrams of the method, terminal device (apparatus), and computer program product according to the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0083] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0084] These computer program instructions can also be loaded onto a computer or other programmable terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0085] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0086] Finally, it should be noted that, in the present invention, relational terms such as first and second, etc., are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal 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 terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise one..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0087] The above is a detailed introduction to the MEMS modeling method of vector fitting combined with neural network transfer function provided by the present invention. The present invention uses specific examples to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A MEMS modeling method combining vector fitting with neural network transfer function, characterized in that: The method comprises: Determining target performance parameters of a target device and device parameters corresponding to the target performance parameters; Obtaining samples corresponding to the device parameters and sample device performance curves corresponding to the samples; Using a vector fitting method, the performance curve of the sample device is fitted, a transfer function is constructed, and sample fitting parameters are extracted; The artificial neural network is trained with the device parameters of the sample as input and the corresponding sample fitting parameters as output to obtain a fitting parameter prediction model; Inputting the device parameters of the sample to be tested corresponding to the target performance parameters of the target device into the fitting parameter prediction model to obtain predicted fitting parameters; Based on the predicted fitting parameters, a predicted device performance curve of the sample to be tested is obtained.

2. The MEMS modeling method combining vector fitting with neural network transfer function according to claim 1 is characterized in that: The method further comprises: Based on the predicted device performance curve, it is determined whether the device parameter values ​​of the sample to be tested are qualified.

3. The MEMS modeling method combining vector fitting with neural network transfer function according to claim 1 is characterized in that: The device parameters include multiple parameter dimensions, and the device parameters of the sample to be tested corresponding to the target performance parameters of the target device are input into the fitting parameter prediction model to obtain the predicted device performance curve of the sample to be tested, including: Acquire multiple samples to be tested, each sample to be tested having a different combination of device parameter values; Inputting the device parameters of the plurality of samples to be tested into the fitting parameter prediction model respectively to obtain a plurality of predicted fitting parameters; Based on the multiple predicted fitting parameters, obtaining predicted device performance curves of the multiple samples to be tested; The method further comprises: According to the predicted device performance curves of the multiple samples to be tested, a device parameter value combination corresponding to the best sample to be tested is determined.

4. The MEMS modeling method combining vector fitting with neural network transfer function according to claim 1 is characterized in that: The method further comprises: The tolerance range of the target parameter is determined according to the predicted device performance curves of each of the plurality of samples to be tested.

5. The MEMS modeling method combining vector fitting with neural network transfer function according to claim 1 is characterized in that: Determining target performance parameters of a target device and device parameters corresponding to the target performance parameters includes: determining target performance parameters and multiple device parameters of a target device; Based on the principal component analysis method, main parameters corresponding to the target performance parameters are determined, and the main parameters are used as device parameters.

6. The MEMS modeling method of vector fitting combined with neural network transfer function according to any one of claims 1 to 4, characterized in that: Acquiring a sample corresponding to the device parameter and a sample performance parameter corresponding to the sample, including: Determine the parameter distribution interval corresponding to the device parameter; Sampling within the parameter distribution interval to obtain a sample; The finite element simulation method is used to determine the sample device performance curve corresponding to the sample.

7. A MEMS modeling device combining vector fitting with neural network transfer function, characterized in that: The MEMS modeling device combining vector fitting with neural network transfer function includes: A first determination module, used to determine target performance parameters of a target device and device parameters corresponding to the target performance parameters; An acquisition module, used to acquire samples corresponding to the device parameters and sample device performance curves corresponding to the samples; A fitting module, used to fit the performance curve of the sample device using a vector fitting method, construct a transfer function, and extract sample fitting parameters; A training module, used to train an artificial neural network with the device parameters of the sample as input and the corresponding sample fitting parameters as output to obtain a fitting parameter prediction model; A prediction module, used for inputting device parameters of the sample to be tested corresponding to the target performance parameters of the target device into the fitting parameter prediction model to obtain predicted fitting parameters; The curve determination module is used to obtain the predicted device performance curve of the sample to be tested based on the predicted fitting parameters.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the MEMS modeling method combining vector fitting with neural network transfer function described in any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the MEMS modeling method combining vector fitting with neural network transfer function described in any one of claims 1 to 6 is implemented.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instruction is executed by a processor, the steps in the MEMS modeling method combining vector fitting with neural network transfer function described in any one of claims 1 to 6 are implemented.

Images