Automatic parameter calibration method and device, equipment, storage medium and product

Abstract

The invention discloses an automatic parameter calibration method and device, equipment, a storage medium and a product, and relates to the technical field of parameter calibration, the automatic parameter calibration method comprises the steps that initial parameters of a semiconductor device are acquired, and the initial parameters comprise fixed parameters and to-be-optimized parameters; based on the initial parameters, parameter calibration is carried out through a preset parameter calibration model to obtain calibrated output parameters, and the parameter calibration model is obtained by carrying out iterative training on a preset to-be-trained model based on an initial parameter sample and a parameter label of the initial parameter sample. According to the method, the machine learning model, namely the parameter calibration model, is used for replacing the traditional manual debugging process, the calibrated parameters are rapidly and accurately output through the data training model generated through simulation, and the method not only improves the calibration efficiency, but also improves the accuracy of parameter calibration.

Description

Technical Field

[0001] This application relates to the technical field of parameter calibration, and particularly to an automated parameter calibration method, device, equipment, storage medium, and product. Background Art

[0002] In the field of semiconductor device simulation, parameter calibration is a crucial step in ensuring the accuracy and reliability of simulation results. Through parameter calibration, the theoretical model of the device can be matched with experimental data, providing an important basis for device design, optimization, and performance prediction. However, for amorphous and polycrystalline materials (such as those widely used in thin-film transistors TFT and SiC devices), the parameter calibration process faces numerous challenges, which severely restrict the efficiency and accuracy of semiconductor device simulation.

[0003] In related technologies, parameter calibration usually adopts a manual adjustment method. However, this method relies on the experience of engineers, not only has low efficiency, but is also interfered by human factors, resulting in low accuracy of parameter calibration results.

[0004] The above content is only used to assist in understanding the technical solution of this application, and does not represent an admission that the above content is prior art. Summary of the Invention

[0005] The main purpose of this application is to provide an automated parameter calibration method, aiming to solve the technical problem of low accuracy of parameter calibration.

[0006] To achieve the above purpose, this application proposes an automated parameter calibration method, and the method of the automated parameter calibration includes:

[0007] Obtain the initial parameters of the semiconductor device, where the initial parameters include fixed parameters and parameters to be optimized;

[0008] Based on the initial parameters, perform parameter calibration through a preset parameter calibration model to obtain the calibrated output parameters, where the parameter calibration model is obtained by iteratively training a preset model to be trained based on the initial parameter samples and the parameter labels of the initial parameter samples.

[0009] Optionally, before the step of obtaining the initial parameters of the semiconductor device, the method includes:

[0010] Obtain the initial parameter samples and the parameter labels of the initial parameter samples;

[0011] Based on the initial parameter samples and the parameter labels, perform iterative training on a preset model to be trained to obtain a parameter calibration model.

[0012] Optionally, the step of obtaining the initial parameter samples includes:

[0013] Obtain a fixed parameter sample and a parameter sample to be optimized;

[0014] Perform unified processing on the file formats of the fixed parameter sample and the parameter sample to be optimized to obtain a fixed parameter sample and a parameter sample to be optimized with unified file formats;

[0015] Based on the fixed parameter sample and the parameter sample to be optimized with unified file formats, form an initial parameter sample.

[0016] Optionally, the step of obtaining the initial parameter sample and the parameter label of the initial parameter sample includes:

[0017] Obtain the parameter sample space of the semiconductor device;

[0018] Determine the value ranges of the parameters in the parameter sample space, and based on the value ranges of the parameters, sample the parameter sample space to obtain parameter combinations;

[0019] Generate simulation tasks corresponding to the parameter combinations, execute a preset number of simulation tasks in parallel, and obtain the initial parameter sample and the parameter label of the initial parameter sample.

[0020] Optionally, the step of iteratively training a preset model to be trained based on the initial parameter sample and the parameter label to obtain a parameter calibration model includes:

[0021] Extract features from the initial parameter sample to obtain the feature vector of the initial parameter sample;

[0022] Based on the feature vector of the initial parameter sample, perform parameter calibration through a preset model to be trained to obtain predicted parameters;

[0023] Calculate the difference between the predicted parameters and the parameter labels to obtain an error result;

[0024] Based on the error result, determine whether the error result meets the error standard indicated by the preset error threshold range;

[0025] If the error result does not meet the error standard indicated by the preset error threshold range, update the model parameters of the model to be trained, and return to the step of performing parameter calibration through a preset model to be trained based on the feature vector of the initial parameter sample to obtain predicted parameters, until the training error result meets the error standard indicated by the preset error threshold range, and then stop training to obtain a parameter calibration model that meets the accuracy condition.

[0026] Optionally, the step of calculating the difference between the predicted parameters and the parameter labels to obtain an error result includes:

[0027] Simulate the prediction parameters to obtain a first simulation curve at different temperatures;

[0028] Compare the first simulation curve with a second simulation curve corresponding to the parameter label to obtain corresponding error results.

[0029] In addition, to achieve the above object, the present application also proposes an automated parameter calibration device, which includes:

[0030] An acquisition module for acquiring initial parameters of a semiconductor device, where the initial parameters include fixed parameters and parameters to be optimized;

[0031] A calibration module for calibrating parameters based on the initial parameters through a preset parameter calibration model to obtain calibrated output parameters, where the parameter calibration model is obtained by iteratively training a preset model to be trained based on initial parameter samples and parameter labels of the initial parameter samples.

[0032] In addition, to achieve the above object, the present application also proposes an automated parameter calibration device, which includes: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the computer program is configured to implement the steps of the automated parameter calibration method as described above.

[0033] In addition, to achieve the above object, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium, and when the computer program is executed by a processor, it implements the steps of the automated parameter calibration method as described above.

[0034] In addition, to achieve the above object, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the automated parameter calibration method as described above.

[0035] One or more technical solutions proposed by the present application have at least the following technical effects:

[0036] In related technologies, parameter calibration usually adopts a manual adjustment method. However, this method relies on the experience of engineers, not only has low efficiency, but is also interfered by human factors, resulting in low accuracy of parameter calibration results. In contrast, the present application proposes to obtain the initial parameters of a semiconductor device, where the initial parameters include fixed parameters and parameters to be optimized; based on the initial parameters, parameter calibration is performed through a preset parameter calibration model to obtain calibrated output parameters, where the parameter calibration model is obtained by iteratively training a preset model to be trained based on initial parameter samples and the parameter labels of the initial parameter samples. It can be understood that the present application uses a machine learning model, that is, a parameter calibration model, to replace the traditional manual debugging process, and trains the model through simulated generated data, so as to quickly and accurately output calibrated parameters. This method not only improves the calibration efficiency, but also improves the accuracy of parameter calibration. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The drawings herein are incorporated into the specification and form a part of the specification, showing embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, other drawings can also be obtained based on these drawings without creative efforts.

[0039] Figure 1 It is a schematic flowchart provided for the first embodiment of the automatic parameter calibration method of the present application;

[0040] Figure 2 It is a schematic flowchart provided for the second embodiment of the automatic parameter calibration method of the present application;

[0041] Figure 3 It is a schematic diagram of an example of a unified file format in the automatic parameter calibration method of the present application;

[0042] Figure 4 It is a schematic diagram of the machine learning model framework in the automatic parameter calibration method of the present application;

[0043] Figure 5 It is a schematic diagram of the module structure of the automatic parameter calibration device in the embodiment of the present application;

[0044] Figure 6 It is a schematic diagram of the device structure of the hardware operating environment involved in the automatic parameter calibration method in the embodiment of the present application.

[0045] The implementation, functional features, and advantages of the object of the present application will be further described with reference to the embodiments and the drawings. Specific Embodiments

[0046] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0047] To better understand the technical solutions of the present application, the following will be described in detail in conjunction with the accompanying drawings of the specification and specific embodiments.

[0048] The main solution of the embodiment of the present application is: obtaining the initial parameters of a semiconductor device, where the initial parameters include fixed parameters and parameters to be optimized; based on the initial parameters, parameter calibration is performed through a preset parameter calibration model to obtain calibrated output parameters, where the parameter calibration model is obtained by iteratively training a preset model to be trained based on initial parameter samples and parameter labels of the initial parameter samples.

[0049] In this embodiment, an automated parameter calibration device is used as the execution subject. For the convenience of description, the following will be described simply as "device".

[0050] In the related art, parameter calibration usually adopts a manual adjustment method. However, this method relies on the experience of engineers, not only has low efficiency, but also is interfered by human factors, resulting in low accuracy of parameter calibration results.

[0051] The present application provides a solution to achieve automated parameter calibration of semiconductor devices, thereby improving the accuracy of automated parameter calibration.

[0052] As can be seen from the above embodiments, the present application uses a machine learning model, that is, a parameter calibration model, to replace the traditional manual debugging process, and trains the model through simulated generated data, so as to quickly and accurately output calibrated parameters. This method not only improves the calibration efficiency, but also improves the accuracy of parameter calibration.

[0053] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication, and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device or a terminal system that can implement the above functions. The following takes the automated parameter calibration device as an example to illustrate this embodiment and the following embodiments.

[0054] Based on this, the embodiment of the present application provides an automated parameter calibration method, referring to Figure 1 , Figure 1 is a schematic flowchart of the first embodiment of the automated parameter calibration method of the present application.

[0055] In this embodiment, the automated parameter calibration method includes steps S100 to S200:

[0056] Step S100: Obtain the initial parameters of the semiconductor device, where the initial parameters include fixed parameters and parameters to be optimized;

[0057] It should be noted that this application relates to the technical field of semiconductor device simulation and parameter calibration, especially an AI-based method for calibrating material and defect parameters, which is applicable to the automatic calibration of amorphous / polycrystalline materials in devices such as TFTs, SiC, and GaN.

[0058] In specific implementation, fixed parameters refer to those that remain unchanged during the design and simulation of semiconductor devices, usually determined according to the physical structure or material characteristics of the device, including but not limited to: the geometric dimensions of the device (such as channel length, channel width), the dielectric constant of the material, the thickness of the gate insulating layer, and process parameters (such as doping concentration), etc.; parameters to be optimized refer to those that need to be determined through the calibration process, usually those that have a significant impact on device performance but are difficult to directly measure, including but not limited to: carrier mobility, threshold voltage, defect state density, subthreshold slope, etc.

[0059] In specific implementation, the initial parameters include fixed parameters and parameters to be optimized. That is, before starting parameter calibration, the device needs to obtain the initial parameters of the semiconductor device. The obtaining method can be receiving the initial parameters uploaded by the user or other obtaining methods, which are not specifically limited here.

[0060] Step S200: Based on the initial parameters, perform parameter calibration through a preset parameter calibration model to obtain the calibrated output parameters, where the parameter calibration model is obtained by iteratively training a preset model to be trained based on the initial parameter samples and the parameter labels of the initial parameter samples.

[0061] It should be noted that the parameter calibration model is a pre-trained machine learning model used to map the initial parameters to more accurate parameter values, where the training of the model is based on the initial parameter samples and the corresponding parameter labels. Specifically, iterative training is to train the preset model using the initial parameter samples and the corresponding parameter labels. Specifically, the weights and biases of the model are adjusted through multiple iterations so that the model can learn the mapping relationship between the input parameters and the target output. The training goal is to minimize the error between the model prediction value and the actual label (such as mean square error, mean absolute error).

[0062] In specific implementation, the calibrated output parameters are obtained by calibrating the initial parameters using the trained model to obtain more accurate output parameters. These calibrated parameters can make the device characteristic curves (such as IV curves) generated by simulation closer to the experimental data.

[0063] In related technologies, parameter calibration usually adopts a manual adjustment method. However, this method relies on the experience of engineers, not only has low efficiency, but also is interfered by human factors, resulting in low accuracy of parameter calibration results. In contrast, the present application proposes to obtain the initial parameters of a semiconductor device, where the initial parameters include fixed parameters and parameters to be optimized; based on the initial parameters, parameter calibration is performed through a preset parameter calibration model to obtain calibrated output parameters, where the parameter calibration model is obtained by iteratively training a preset model to be trained based on an initial parameter sample and the parameter label of the initial parameter sample. It can be understood that the present application uses a machine learning model, that is, a parameter calibration model, to replace the traditional manual debugging process, and trains the model through simulated generated data, so as to quickly and accurately output the calibrated parameters. This method not only improves the calibration efficiency, but also improves the accuracy of parameter calibration.

[0064] Based on the above first embodiment, the present application also proposes another embodiment. Referring to Figure 2 , the automated parameter calibration method includes:

[0065] Before the step of the device obtaining the initial parameters of the semiconductor device in specific implementation, the method includes:

[0066] Step A100, obtaining an initial parameter sample and the parameter label of the initial parameter sample;

[0067] It should be noted that the initial parameter sample is the input data for training the model, including different combinations of fixed parameters and parameters to be optimized. For example, parameter combinations generated by parameter space sampling (such as random sampling, Latin hypercube sampling). The parameter label is the "target value" corresponding to the initial parameter sample, usually the device performance indicators obtained through experimental measurement or high-precision simulation. For example, the measured IV curve (Id-Vg, Id-Vd), key performance indicators (such as threshold voltage, subthreshold slope), etc.

[0068] In specific implementation, the step of the device obtaining the initial parameter sample includes:

[0069] Obtaining a fixed parameter sample and a parameter sample to be optimized; performing unified file format processing on the fixed parameter sample and the parameter sample to be optimized to obtain the fixed parameter sample and the parameter sample to be optimized with unified file format; based on the fixed parameter sample and the parameter sample to be optimized with unified file format, forming an initial parameter sample.

[0070] It should be noted that in practical applications, parameter data may come from different sources and there may be inconsistent formats. In order to facilitate subsequent processing, it is necessary to perform unified format processing on these parameter samples.

[0071] In a specific implementation, referring to Figure 3 , this application provides a single file format for uniformly defining fixed parameters and parameters to be optimized, simplifying the parameter passing and management processes. That is, a single file format for defining fixed parameters and parameters to be optimized simplifies parameter management and passing. Specifically, the processing process includes: 1. Data formatting: Ensure that the formats of all parameter data are consistent (such as numerical format, unit uniformity). 2. Data structuring: Organize the parameter data into a structured form, such as a table format (CSV file), JSON format, or database format. 3. Data cleaning: Remove duplicate data, correct incorrect data, fill in missing data, etc.

[0072] In a specific implementation, after formatting, the data formats of the fixed parameter samples are consistent and can be directly used for subsequent processing. Similarly, after formatting, the data formats of the parameters to be optimized samples are also consistent.

[0073] In a specific implementation, the steps for the device to obtain the initial parameter samples and the parameter labels of the initial parameter samples include:

[0074] Obtain the parameter sample space of the semiconductor device; determine the value ranges of the parameters in the parameter sample space, and based on the value ranges of the parameters, sample the parameter sample space to obtain parameter combinations; generate simulation tasks corresponding to the parameter combinations, and execute a preset number of simulation tasks in parallel to obtain the initial parameter samples and the parameter labels of the initial parameter samples.

[0075] It should be noted that the parameter sample space refers to the set of all parameter combinations, and these parameters include fixed parameters and parameters to be optimized, which jointly describe the characteristics of the semiconductor device.

[0076] In a specific implementation, referring to Figure 4 , in order to generate diverse parameter combinations, it is necessary to determine the value range of each parameter, where the parameter value range is determined based on physical meaning, experimental data, or empirical knowledge. For example, the channel length value range: 10 μm to 20 μm, the carrier mobility value range: 10 cm 2 / V·s to 20 cm 2 / V·s, the defect state density value range: 1015 cm-3 to 1016 cm-3, etc.

[0077] In a specific implementation, the device samples the parameter sample space based on the value ranges of the parameters to obtain parameter combinations. Among them, sampling is to select a specific set of parameter combinations from the parameter sample space. Specifically, the sampling method can be random sampling, Latin hypercube sampling, or grid sampling.

[0078] Further, the device generates multiple simulation tasks according to the sampled parameter combinations, where each task corresponds to a specific set of parameter values. And it uses parallel computing resources (such as a high-performance computing cluster) to run multiple simulation tasks simultaneously to improve the efficiency of data generation. Specifically, the device randomly samples the parameter space to generate a training data set, and realizes the parallelization of data generation through the simulation task management module.

[0079] In a specific implementation, finally, the device extracts the generated simulation results from each simulation task. These results include: the initial parameter samples, that is, the sampled parameter combinations. Parameter labels, that is, the performance metrics generated by the simulation, such as IV curves (Id-Vg, Id-Vd curves), threshold voltage, subthreshold slope, etc.

[0080] Step A200: Based on the initial parameter samples and the parameter labels, iteratively train a preset model to be trained to obtain a parameter calibration model.

[0081] It should be noted that the preset model to be trained is a pre-designed machine learning model architecture, such as a deep neural network (DNN), a convolutional neural network (CNN), a random forest (Random Forest), etc. The structure and hyperparameters of the model (such as the number of layers, the number of nodes, the learning rate, etc.) have been defined before training.

[0082] In a specific implementation, referring to Figure 4 , the device uses the initial parameter samples as inputs and the parameter labels as the target outputs to train the preset model. The goal of training is to minimize the error between the model prediction value and the actual label. After iterative training, the model can learn the mapping relationship between the input parameters and the target outputs. This trained model is called a parameter calibration model. The parameter calibration model can be used to predict the device performance metrics corresponding to new input parameters. That is, by adjusting the parameters predicted by the model, the simulation results are made closer to the experimental data, thereby completing the parameter calibration.

[0083] In a specific implementation, the machine learning model training of this application includes two stages. Specifically, the first stage: unsupervised learning, using an unsupervised learning model to extract features from the simulated IV curves and refine the core features related to the parameters. The second stage: deep neural network, using the extracted features to train a deep neural network to predict the target parameters.

[0084] In a specific implementation, the steps of the device iteratively training a preset model to be trained based on the initial parameter samples and the parameter labels to obtain a parameter calibration model include:

[0085] Feature extraction is performed on the initial parameter samples to obtain the feature vectors of the initial parameter samples; based on the feature vectors of the initial parameter samples, parameter calibration is performed through a preset model to be trained to obtain predicted parameters; the difference between the predicted parameters and the parameter labels is calculated to obtain an error result; based on the error result, it is determined whether the error result meets the error standard indicated by a preset error threshold range; if the error result does not meet the error standard indicated by the preset error threshold range, the model parameters of the model to be trained are updated, and the step of performing parameter calibration through the preset model to be trained based on the feature vectors of the initial parameter samples to obtain predicted parameters is returned, until the training error result meets the error standard indicated by the preset error threshold range and then the training is stopped to obtain a parameter calibration model that meets the accuracy condition.

[0086] In a specific implementation, first, the device performs feature extraction on the initial parameter samples to obtain the feature vectors of the initial parameter samples. Specifically, the device extracts feature vectors from the initial parameter samples that can effectively describe the device characteristics. These feature vectors are the inputs of the model and are usually the results of processing or transforming the original parameters. The extracted feature vectors are in a format that the model can understand and process, such as normalized parameter values, combined features, etc.

[0087] In a specific implementation, the device inputs the feature vectors into the model, and the model calculates according to the current weights and biases and outputs predicted parameters. Secondly, the device calculates the difference between the parameters predicted by the model and the true parameter labels, usually using metrics such as mean square error (MSE), mean absolute error (MAE), etc. Among them, the true parameter label is the parameter label, and the parameter label is the true value obtained through experimental measurement or high-precision simulation.

[0088] Furthermore, the device determines whether the error is within the preset threshold range, indicating that the parameters predicted by the model are accurate enough; otherwise, the model needs to be further adjusted. If the error result does not meet the preset error threshold range, the weights and biases of the model need to be adjusted to reduce the error. Specifically, parameter calibration is performed again using the feature vectors until the error result meets the preset standard.

[0089] In a specific implementation, when the error result meets the preset error threshold range, the training is stopped. The trained model can accurately predict the parameters of the device according to the input feature vectors and is used for subsequent parameter calibration.

[0090] In a specific implementation, the step of calculating the difference between the predicted parameters and the parameter labels by the device to obtain an error result includes:

[0091] Simulate the prediction parameters to obtain the first simulation curves at different temperatures; compare the first simulation curves with the second simulation curves corresponding to the parameter tags to obtain the corresponding error results.

[0092] In a specific implementation, the device uses the prediction parameters output by the model to generate current-voltage (IV) curves at different temperatures in a simulation software (TCAD). These curves (the first simulation curves) reflect the operating states of the device under different temperature conditions. That is, this application supports multiple curves as training targets, and the model can perform parameter calibration by combining working condition data such as different temperatures to meet the calibration requirements under complex conditions.

[0093] It should be noted that the first simulation curve is a simulation curve generated using prediction parameters. The second simulation curve is an IV curve measured experimentally or an IV curve generated by high-precision simulation, serving as a parameter tag.

[0094] In a specific implementation, the device compares the first simulation curve with the second simulation curve to evaluate the accuracy of the model prediction. Specifically, the contents of the comparison include but are not limited to: Shape matching: Whether the shapes of the simulation curve and the experimental curve are consistent. Key parameter matching: Whether key parameters such as threshold voltage, subthreshold slope, and drain current are close. Error quantification: Quantify the difference between the two using indicators such as mean square error (MSE) and mean absolute error (MAE).

[0095] In a specific implementation, the device obtains the error value by comparing the first simulation curve and the second simulation curve. Judge whether the parameters predicted by the model are accurate according to the error result. If the error is within an acceptable range, it indicates that the parameters predicted by the model are accurate; otherwise, the model needs to be further adjusted.

[0096] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the automatic parameter calibration method of this application. Based on this technical concept, more forms of simple transformations are within the protection scope of this application.

[0097] This application also provides an automatic parameter calibration device. Please refer to Figure 5 , the automatic parameter calibration device includes:

[0098] An acquisition module 10, configured to acquire the initial parameters of a semiconductor device, where the initial parameters include fixed parameters and parameters to be optimized;

[0099] A calibration module 20, configured to perform parameter calibration on the basis of the initial parameters through a preset parameter calibration model to obtain calibrated output parameters, where the parameter calibration model is obtained by iteratively training a preset model to be trained based on an initial parameter sample and the parameter tag of the initial parameter sample.

[0100] Optionally, the automated parameter calibration device further includes:

[0101] A sample acquisition module, configured to acquire an initial parameter sample and a parameter label of the initial parameter sample;

[0102] A training module, configured to perform iterative training on a preset model to be trained based on the initial parameter sample and the parameter label, and obtain a parameter calibration model.

[0103] Optionally, the sample acquisition module includes:

[0104] A parameter sample acquisition module, configured to acquire a fixed parameter sample and a parameter sample to be optimized;

[0105] A format unification module, configured to perform file format unification processing on the fixed parameter sample and the parameter sample to be optimized, and obtain a fixed parameter sample and a parameter sample to be optimized with unified file formats;

[0106] A composition module, configured to compose an initial parameter sample based on the fixed parameter sample and the parameter sample to be optimized with unified file formats.

[0107] Optionally, the sample acquisition module further includes:

[0108] A parameter sample space acquisition module, configured to acquire a parameter sample space of a semiconductor device;

[0109] A sampling module, configured to determine a value range of each parameter in the parameter sample space, and sample the parameter sample space based on the value range of each parameter to obtain a parameter combination;

[0110] A simulation task parallel execution module, configured to generate a simulation task corresponding to the parameter combination, and parallelly execute a preset number of simulation tasks to obtain an initial parameter sample and a parameter label of the initial parameter sample.

[0111] Optionally, the training module includes:

[0112] A feature extraction module, configured to extract features from the initial parameter sample to obtain a feature vector of the initial parameter sample;

[0113] A parameter calibration module, configured to perform parameter calibration on the feature vector of the initial parameter sample through a preset model to be trained to obtain a predicted parameter;

[0114] A difference calculation module, configured to calculate a difference between the predicted parameter and the parameter label to obtain an error result;

[0115] A judgment module, configured to judge, based on the error result, whether the error result meets the error standard indicated by a preset error threshold range;

[0116] An iterative training module, configured to, if the error result does not meet the error standard indicated by the preset error threshold range, update the model parameters of the model to be trained, and return the step of obtaining a predicted parameter by calibrating parameters through a preset model to be trained based on the feature vector of the initial parameter sample, until the training error result meets the error standard indicated by the preset error threshold range, and then stop training to obtain a parameter calibration model that meets the accuracy condition.

[0117] Optionally, the difference calculation module includes:

[0118] A simulation module, configured to simulate the predicted parameter to obtain a first simulation curve at different temperatures;

[0119] A comparison module, configured to compare the first simulation curve with a second simulation curve corresponding to the parameter label to obtain a corresponding error result.

[0120] The automatic parameter calibration device provided by the present application adopts the automatic parameter calibration method in the above embodiment, and can solve the technical problem of automatic parameter calibration. Compared with the prior art, the beneficial effects of the automatic parameter calibration device provided by the present application are the same as those of the automatic parameter calibration method provided by the above embodiment, and other technical features in the automatic parameter calibration device are the same as those disclosed in the method of the above embodiment, and will not be elaborated herein.

[0121] The present application provides an automatic parameter calibration device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the automatic parameter calibration method in Embodiment 1 above.

[0122] Next, refer to Figure 6, which shows a schematic structural diagram of an automated parameter calibration device suitable for implementing the embodiments of the present application. The automated parameter calibration device in the embodiments of the present application may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistant), PADs (Portable Application Description: tablet computers), PMPs (Portable Media Player), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 6 The shown automated parameter calibration device is merely an example and should not impose any limitations on the functions and usage scope of the embodiments of the present application.

[0123] As Figure 6 shown, the automated parameter calibration device may include a processing device 1001 (such as a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM: Read Only Memory) 1002 or the program loaded from the storage device 1003 into the random access memory (RAM: Random Access Memory) 1004. In the RAM 1004, various programs and data required for the operation of the automated parameter calibration device are also stored. The processing device 1001, the ROM 1002, and the RAM 1004 are connected to each other through a bus 1005. The input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems may be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the automated parameter calibration device to communicate with other devices wirelessly or wiredly to exchange data. Although the figure shows an automated parameter calibration device with various systems, it should be understood that it is not required to implement or have all the shown systems. Instead, more or fewer systems may be implemented or had.

[0124] In particular, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product that includes a computer program carried on a computer-readable medium, and the computer program contains program codes for executing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by a processing device 1001, the above-mentioned functions defined in the methods of the embodiments disclosed in the present application are executed.

[0125] The automated parameter calibration device provided by the present application adopts the automated parameter calibration method in the above-mentioned embodiments and can solve the technical problems of automated parameter calibration. Compared with the prior art, the beneficial effects of the automated parameter calibration device provided by the present application are the same as those of the automated parameter calibration method provided in the above-mentioned embodiments, and other technical features in the automated parameter calibration device are the same as the features disclosed in the method of the previous embodiment and will not be elaborated here.

[0126] It should be understood that the various parts disclosed in the present application can be implemented by hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in a suitable manner in any one or more embodiments or examples.

[0127] As described above, the above are only specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed in the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0128] The present application provides a computer-readable storage medium having computer-readable program instructions (i.e., computer programs) stored thereon, and the computer-readable program instructions are used to execute the automated parameter calibration method in the above-mentioned embodiments.

[0129] The computer-readable storage medium provided by this application can be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM) or flash memory, optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this embodiment, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, device, or component. The program code contained on the computer-readable storage medium can be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0130] The above computer-readable storage medium can be included in an automated parameter calibration device; or it can exist separately and not be assembled into the automated parameter calibration device.

[0131] The above computer-readable storage medium carries one or more programs. When the one or more programs are executed by the automated parameter calibration device, the automated parameter calibration device is enabled to perform automated parameter calibration.

[0132] Computer program code for performing the operations of this application can be written in one or more programming languages or combinations thereof. The above programming languages include object-oriented programming languages such as Java, Python, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, by using an Internet service provider to connect through the Internet).

[0133] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a portion of code that contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, as well as combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0134] The modules described in the embodiments of the present application can be implemented in software or in hardware. In some cases, the name of the module does not constitute a limitation on the unit itself.

[0135] The readable storage medium provided by the present application is a computer-readable storage medium that stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned automated parameter calibration method, and can solve the technical problem of automated parameter calibration. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by the present application are the same as those of the automated parameter calibration method provided by the above embodiments, and will not be elaborated here.

[0136] The present application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the automated parameter calibration method as described above.

[0137] The computer program product provided by the present application can solve the technical problem of automated parameter calibration. Compared with the prior art, the beneficial effects of the computer program product provided by the present application are the same as those of the automated parameter calibration method provided by the above embodiments, and will not be elaborated here.

[0138] The above are only some embodiments of the present application, and do not limit the patent scope of the present application. Any equivalent structural transformation made using the content of the specification and drawings of the present application under the technical concept of the present application, or any direct / indirect application in other related technical fields, is included in the patent protection scope of the present application.

Claims

1. An automated parameter calibration method, characterized in that: The method for automatic parameter calibration comprises: Acquiring initial parameters of the semiconductor device, wherein the initial parameters include fixed parameters and parameters to be optimized; Based on the initial parameters, parameter calibration is performed through a preset parameter calibration model to obtain calibrated output parameters, wherein the parameter calibration model is obtained by iteratively training a preset model to be trained based on initial parameter samples and parameter labels of the initial parameter samples.

2. The automated parameter calibration method according to claim 1, characterized in that: Before the step of obtaining the initial parameters of the semiconductor device, the method includes: Obtaining an initial parameter sample and a parameter label of the initial parameter sample; Based on the initial parameter samples and the parameter labels, the preset model to be trained is iteratively trained to obtain a parameter calibration model.

3. The automated parameter calibration method according to claim 2, wherein: The step of obtaining the initial parameter sample comprises: Obtain fixed parameter samples and parameter samples to be optimized; Performing file format unification processing on the fixed parameter samples and the parameter samples to be optimized to obtain the fixed parameter samples and the parameter samples to be optimized after the file formats are unified; Based on the fixed parameter samples and the parameter samples to be optimized after the file formats are unified, an initial parameter sample is formed.

4. The automated parameter calibration method according to claim 2, wherein: The step of obtaining an initial parameter sample and a parameter label of the initial parameter sample comprises: Obtain parameter sample space of semiconductor devices; Determining a value range of each parameter in the parameter sample space, and based on the value range of each parameter, sampling the parameter sample space to obtain a parameter combination; Generate simulation tasks corresponding to the parameter combination, execute a preset number of simulation tasks in parallel, and obtain initial parameter samples and parameter labels of the initial parameter samples.

5. The automated parameter calibration method according to claim 2, wherein: The step of iteratively training the preset model to be trained based on the initial parameter sample and the parameter label to obtain a parameter calibration model includes: Performing feature extraction on the initial parameter sample to obtain a feature vector of the initial parameter sample; Based on the feature vector of the initial parameter sample, parameter calibration is performed through a preset model to be trained to obtain a prediction parameter; Calculate the difference between the predicted parameter and the parameter label to obtain an error result; Based on the error result, determining whether the error result meets an error standard indicated by a preset error threshold range; If the error result does not meet the error standard indicated by the preset error threshold range, the model parameters of the model to be trained are updated, and the feature vector based on the initial parameter sample is returned. The parameter calibration is performed through the preset model to be trained to obtain the predicted parameter step. The training is stopped until the training error result meets the error standard indicated by the preset error threshold range, and a parameter calibration model that meets the accuracy conditions is obtained.

6. The automated parameter calibration method according to claim 5, characterized in that: The step of calculating the difference between the prediction parameter and the parameter label to obtain an error result includes: Simulating the predicted parameters to obtain first simulation curves at different temperatures; The first simulation curve is compared with a second simulation curve corresponding to the parameter label to obtain a corresponding error result.

7. An automated parameter calibration device, characterized in that: The device comprises: An acquisition module, used for acquiring initial parameters of the semiconductor device, wherein the initial parameters include fixed parameters and parameters to be optimized; A calibration module is used to perform parameter calibration based on the initial parameters through a preset parameter calibration model to obtain calibrated output parameters, wherein the parameter calibration model is obtained by iteratively training a preset model to be trained based on initial parameter samples and parameter labels of the initial parameter samples.

8. An automated parameter calibration device, characterized in that: The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the automated parameter calibration method according to any one of claims 1 to 6.

9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the automated parameter calibration method according to any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the automated parameter calibration method according to any one of claims 1 to 6 are implemented.

Images