CN117807929A - Model simulation prediction method and system based on deep learning - Google Patents
Model simulation prediction method and system based on deep learning Download PDFInfo
- Publication number
- CN117807929A CN117807929A CN202311710768.3A CN202311710768A CN117807929A CN 117807929 A CN117807929 A CN 117807929A CN 202311710768 A CN202311710768 A CN 202311710768A CN 117807929 A CN117807929 A CN 117807929A
- Authority
- CN
- China
- Prior art keywords
- model
- parameters
- deep learning
- parameter
- simulation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004088 simulation Methods 0.000 title claims abstract description 73
- 238000000034 method Methods 0.000 title claims abstract description 48
- 238000013135 deep learning Methods 0.000 title claims abstract description 28
- 238000013136 deep learning model Methods 0.000 claims abstract description 37
- 238000012549 training Methods 0.000 claims abstract description 21
- 238000012545 processing Methods 0.000 claims abstract description 16
- 238000010606 normalization Methods 0.000 claims abstract description 13
- 238000007781 pre-processing Methods 0.000 claims description 6
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 4
- 230000006870 function Effects 0.000 claims description 4
- 238000012216 screening Methods 0.000 claims description 4
- 238000012795 verification Methods 0.000 claims description 3
- 238000013528 artificial neural network Methods 0.000 abstract description 6
- 239000003990 capacitor Substances 0.000 description 12
- 238000005259 measurement Methods 0.000 description 8
- 238000004364 calculation method Methods 0.000 description 6
- 238000013461 design Methods 0.000 description 4
- 238000010801 machine learning Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a model simulation prediction method based on deep learning, which comprises the following steps: determining a plurality of parameters required by the model simulation, and setting an adjustable boundary for each parameter; invoking a nanosalice simulator to acquire multiple groups of data of the multiple parameters, and carrying out normalization processing on the multiple parameters based on respective adjustable boundaries; inputting the normalized data of a plurality of parameters into a pre-configured deep learning model to train the deep learning model; and obtaining the deep learning model after the parameter input training of the current required simulation prediction model to predict so as to obtain a prediction result. The resistance and capacitance data of the model can be accurately predicted, and the simulation prediction efficiency is improved by training and predicting the current and capacitance change curves along with bias under different parameters through the neural network.
Description
Technical Field
The invention belongs to the technical field of chip design, and particularly relates to a model simulation prediction method and system based on deep learning.
Background
In the technical field of chip design, model parameter calculation aiming at a complex inherited circuit always works as a core, a group of model parameters are input in model simulation calculation, and current and capacitance values are calculated to be used as output. The traditional method uses a quasi-Newton method to calculate through a complex calculation formula, and has the defects of low calculation speed, difficult calculation of multiple outputs corresponding to multiple groups of parameters in parallel, incapability of directly obtaining the partial derivatives of simulation pair parameters and incapability of differentiating derivation.
Along with the development of technology, integrated circuit technology has been greatly advanced, the integration level is continuously improved, and the power consumption is continuously reduced. This variety of complex application scenarios places higher demands on the performance and power consumption of integrated circuits. Conventional hardware design methods have failed to meet these requirements and require optimization simulation by means of algorithmic design.
Disclosure of Invention
In order to solve the problems, the invention aims to provide a model simulation prediction method and a system based on deep learning, which can accurately predict resistance and capacitance data of a model, train and predict current and capacitance change curves along with bias under different parameters through a neural network.
The technical scheme of the invention is as follows: a model simulation prediction method based on deep learning comprises the following steps: determining a plurality of parameters required by the model simulation, and setting an adjustable boundary for each parameter; invoking a nanosalice simulator to acquire multiple groups of data of the multiple parameters, and carrying out normalization processing on the multiple parameters based on respective adjustable boundaries; inputting the normalized data of a plurality of parameters into a pre-configured deep learning model to train the deep learning model; and obtaining the deep learning model after the parameter input training of the current required simulation prediction model to predict so as to obtain a prediction result. Preferably, the machine learning model in this embodiment may be a deconvolution neural network, a fully connected network, a GAN, an AE, etc., and the core of the present technical solution is how to obtain parameters and how to preprocess the parameters based on the above machine learning model, so that the model can be effectively trained finally, and finally, a machine learning model capable of accurately predicting the resistance and capacitance of the model (model) is obtained.
Preferably, determining a plurality of parameters required for the model simulation, setting an adjustable boundary for each parameter further comprises: determining the plurality of parameters based on historical real engineering experience includes: k2, u0, ua, eu, uc, vsat, a1, a2, nfactor, eta0, etab, pclm, pdiblc, ndep; adopting standard average distribution random or taking logarithm to carry out random on the average distribution so as to obtain a plurality of groups of numerical values of the parameters; and replacing corresponding values in a preset model card based on the multiple groups of values of the multiple parameters.
Preferably, the method further comprises, prior to normalizing the plurality of parameters based on respective adjustable boundaries: writing a plurality of groups of numerical data of the parameters into a netlist file based on the pre-constructed netlist file, wherein the netlist file is pre-constructed according to real measuring points of a corresponding model; dividing according to the measuring points, wherein each size comprises a plurality of groups of different curves.
Preferably, normalizing the plurality of parameters based on respective adjustable boundaries further comprises: when the parameters are normalized by adopting a standard average distribution random method, the normalization is carried out based on the upper limit and the lower limit of the adjustable boundary of each parameter, and the processing result is between 0 and 1; when the parameter is normalized by adopting a random method for average distribution after taking the logarithms, the logarithm is taken firstly, the upper limit and the lower limit of the adjustable boundary of each parameter are also taken as the logarithm, and further the normalization is carried out based on the logarithm of the value, the logarithm of the upper limit and the logarithm of the lower limit, and the processing result is between 0 and 1.
Preferably, the method further comprises: after taking a standard average distribution random or taking a log to randomly obtain a plurality of groups of values of the plurality of parameters, preliminary screening is performed based on the set adjustable boundaries and physical facts to prevent errors of the simulator.
Preferably, before inputting the normalized data of the plurality of parameters into the pre-configured deep learning model to train the deep learning model, the method further includes: the results of the nanosalice simulator simulation are translated by the parsing module into tfrecord format for storage in order to train the network.
Preferably, inputting the normalized data of the plurality of parameters into a pre-configured simulator to train the deep learning model further comprises: the fit method of Model class in tensorflow. Keras library is adopted, the ratio of training set to verification set is 6:1, and the number of samples in each batch is 600; adam is adopted as an optimizer, the learning rate is 1e-3, current and capacitance are adopted as output, and a loss function is Mean Squared Logarithmic Error.
Based on the same conception, the invention also provides a model simulation prediction system based on deep learning, which comprises the following steps: the parameter determining module is used for determining a plurality of parameters required by the model simulation, and setting an adjustable boundary for each parameter; the preprocessing module is used for calling the nanosalice simulator to acquire multiple groups of data of the multiple parameters and carrying out normalization processing on the multiple parameters based on respective adjustable boundaries; the model training module is used for inputting the normalized data of the parameters into a pre-configured deep learning model to train the deep learning model; and the result prediction module is used for obtaining the current parameters of the simulation prediction model to input the parameters to the deep learning model after training to predict so as to obtain a prediction result.
Based on the same conception, the invention also provides an electronic device characterized by comprising: a memory for storing a processing program; and the processor is used for realizing the model simulation prediction method based on the deep learning when executing the processing program.
Based on the same conception, the invention also provides a readable storage medium, which is characterized in that a processing program is stored on the readable storage medium, and the processing program realizes the model simulation prediction method based on the deep learning when being executed by a processor.
By adopting the technical scheme, the invention has the following advantages and positive effects compared with the prior art:
1. according to the technical scheme, multiple groups of data of multiple parameters are randomly and automatically generated, preprocessing is carried out, the resistor and capacitor simulation results obtained by invoking the nanosalice simulator through simulation are input into the deep learning model to train the resistor and capacitor simulation results, simulation prediction of the resistor and capacitor data based on the deep learning model is achieved, the current and capacitor variation curves under different parameters are trained and predicted through a neural network, and simulation prediction efficiency is improved.
2. In the invention, a nanosalice simulator is adopted, and specific circuit conditions and specific measurement conditions are defined through a netlist. When the simulation netlist is generated, a corresponding netlist file is constructed according to the real measurement points of a group of foundry, and physical significance under the application scene is given to the machine learning model.
Drawings
The invention is described in further detail below with reference to the attached drawing figures, wherein:
FIG. 1 is a flow chart of a model simulation prediction method based on deep learning;
FIG. 2 is a schematic diagram of a deep learning neural network model in one embodiment of the invention;
FIG. 3 is a schematic diagram of a model simulation prediction result based on deep learning.
Detailed Description
The invention is described in further detail below with reference to the drawings and the specific examples. Advantages and features of the invention will become more apparent from the following description and from the claims. It is noted that the drawings are in a very simplified form and utilize non-precise ratios, and are intended to facilitate a convenient, clear, description of the embodiments of the invention.
It should be noted that all directional indicators (such as up, down, left, right, front, and rear … …) in the embodiments of the present invention are merely used to explain the relative positional relationship, movement, etc. between the components in a particular posture (as shown in the drawings), and if the particular posture is changed, the directional indicator is changed accordingly.
First embodiment
As shown in fig. 1, the present embodiment provides a model simulation prediction method based on deep learning, which includes the following steps:
s100: determining a plurality of parameters required by the model simulation, and setting an adjustable boundary for each parameter;
s200: invoking a nanosalice simulator to acquire multiple groups of data of the multiple parameters, and carrying out normalization processing on the multiple parameters based on respective adjustable boundaries;
s300: inputting the normalized data of a plurality of parameters into a pre-configured deep learning model to train the deep learning model;
s400: and obtaining the deep learning model after the parameter input training of the current required simulation prediction model to predict so as to obtain a prediction result.
According to the technical scheme, multiple groups of data of multiple parameters are randomly and automatically generated, preprocessing is performed, and the resistor and capacitor simulation results obtained by calling the nanosalice simulator through simulation are input into the deep learning model to train the resistor and capacitor simulation results based on the preprocessed data, so that simulation prediction of the resistor and capacitor data based on the deep learning model is realized, and simulation prediction efficiency is improved. After determining a plurality of parameters, setting an adjustable boundary for each parameter, preferably, in the scheme, modifying the parameter according to a fixed step length one by one, and simulating to obtain a current and capacitance variation curve along with bias. Parameters and curves are the data required for training.
Preferably, determining a plurality of parameters required for the model simulation, setting an adjustable boundary for each parameter further comprises: determining the plurality of parameters based on historical real engineering experience includes: k2, u0, ua, eu, uc, vsat, a1, a2, nfactor, eta0, etab, pclm, pdiblc, ndep; adopting standard average distribution random or taking logarithm to carry out random on the average distribution so as to obtain a plurality of groups of numerical values of the parameters; and replacing corresponding values in a preset model card based on the multiple groups of values of the multiple parameters.
In connection with the application scenario of the present invention, the parameters determined in the present embodiment include k2, u0, ua, eu, uc, vsat, a1, a2, nfactor, eta0, etab, pclm, pdiblc, ndep, and the meanings of these parameters are not redundantly explained here, and all the above parameters are parameters commonly used in the semiconductor industry, and the model (model) may be different versions of bsim4, bsim6, bsim cmg, and the above parameters are derived from different versions of bsim4, bsim6, bsim cmg, and the like.
For example, referring to table 1, to obtain parameters of model simulation training, parameter values are obtained randomly, a set of real parameter ranges is adopted to be closer to the real situation, and a total of 14 parameters are selected according to engineering experience. Because the distribution of different parameters has different characteristics, the standard average distribution is selected to be random or the average distribution after taking the logarithm is selected to be random from parameter to parameter according to engineering experience for simplicity:
table 1 parameter table
And replacing the corresponding parameter values in the pre-configured model card with the random values generated in the previous process to obtain a new model card. Since a nanosalice simulator is employed, a netlist is needed to define specific circuit cases and specific measurement cases.
Preferably, the method further comprises, prior to normalizing the plurality of parameters based on respective adjustable boundaries: writing a plurality of groups of numerical data of the parameters into a netlist file based on the pre-constructed netlist file, wherein the netlist file is pre-constructed according to real measuring points of a corresponding model; dividing according to the measuring points, wherein each size comprises a plurality of groups of different curves.
Since a nanosalice simulator is employed, a netlist is needed to define specific circuit cases and specific measurement cases. When generating the simulated netlist, constructing a corresponding netlist file according to the real measurement points of a set of foundry. 9 devices of different sizes were selected:
table 2 size table
Dividing according to the measurement points, each size contains 7 different sets of curves:
table 3 measurement Point Condition Meter
The method writes randomly generated parameters and simulation resistor and capacitor results based on a nanosalice simulator into a pre-configured netlist file, and preferably the method is completed based on tensorf low, so that after simulation corresponding to a model card is completed completely, the simulation results are translated into a tfrecord format for storage by means of an analysis module so as to train a network. The simulator runs the CPU with a main frequency of 3.6GHz.
Preferably, normalizing the plurality of parameters based on respective adjustable boundaries further comprises: when the parameters are normalized by adopting a standard average distribution random method, the normalization is carried out based on the upper limit and the lower limit of the adjustable boundary of each parameter, and the processing result is between 0 and 1; when the parameter is normalized by adopting a random method for average distribution after taking the logarithms, the logarithm is taken firstly, the upper limit and the lower limit of the adjustable boundary of each parameter are also taken as the logarithm, and further the normalization is carried out based on the logarithm of the value, the logarithm of the upper limit and the logarithm of the lower limit, and the processing result is between 0 and 1.
According to the technical scheme, each parameter is normalized, so that the parameter processing of a subsequent model is more convenient, the model is prevented from additionally processing complex data, the accuracy of data processing is improved, and the possibility of failure of the model is reduced.
Preferably, the method further comprises: after taking a standard average distribution random or taking a log to randomly obtain a plurality of groups of values of the plurality of parameters, preliminary screening is performed based on the set adjustable boundaries and physical facts to prevent errors of the simulator.
Since a nanosalice simulator is employed, a netlist is needed to define specific circuit cases and specific measurement cases. Because the random parameters in combination may violate physical facts, such as the effective length cannot be less than 0. Therefore, one-step preliminary screening is needed, and a model card with a certain problem is skipped according to some criteria, so that errors of a simulator are reduced, and simulation data generation time is shortened.
Preferably, before inputting the normalized data of the plurality of parameters into the pre-configured deep learning model to train the deep learning model, the method further includes: the results of the nanosalice simulator simulation are translated by the parsing module into tfrecord format for storage in order to train the network.
Preferably, inputting the normalized data of the plurality of parameters into a pre-configured simulator to train the deep learning model further comprises: the fit method of Model class in tensorflow. Keras library is adopted, the ratio of training set to verification set is 6:1, and the number of samples in each batch is 600; adam is adopted as an optimizer, the learning rate is 1e-3, current and capacitance are adopted as output, and a loss function is Mean Squared Logarithmic Error.
Preferably, the deep learning model is built before training the model, and referring to fig. 2, a schematic diagram of the deep learning neural network model in one embodiment is shown. The neural network takes 14 model parameters as input; the network adopts a deconvolution network, namely, the network consists of a plurality of full-connection layers and a plurality of deconvolution layers; then 6 pieces of 61-point curve data are obtained through deconvolution calculation. The training samples were 50 ten thousand simulation curves. Training uses the fit method of Model class in tensorfilow. Keras library. The number of samples was 45 ten thousand, the ratio of training set to validation set was 6:1, and the number of samples per batch was 600. Adam is adopted as an optimizer, and the learning rate is 1e-3. The current capacitance is taken as output, and Mean Squared Logarithmic Error is selected as a loss function. After training, see fig. 3, the mean error of fit to the simulation curve is 1%.
Second embodiment
Based on the same conception, the invention also provides a model simulation prediction system based on deep learning, which comprises the following steps: the parameter determining module is used for determining a plurality of parameters required by the model simulation, and setting an adjustable boundary for each parameter; the preprocessing module is used for calling the nanosalice simulator to acquire multiple groups of data of the multiple parameters and carrying out normalization processing on the multiple parameters based on respective adjustable boundaries; the model training module is used for inputting the normalized data of the parameters into a pre-configured deep learning model to train the deep learning model; and the result prediction module is used for obtaining the current parameters of the simulation prediction model to input the parameters to the deep learning model after training to predict so as to obtain a prediction result.
According to the technical scheme, multiple groups of data of multiple parameters are randomly and automatically generated, preprocessing is performed, the resistor and capacitor simulation results obtained by invoking the nanosalice simulator through simulation are input into a deep learning model to train the resistor and capacitor simulation results, simulation prediction of the resistor and capacitor data based on the deep learning model is achieved, current and capacitor variation curves along with bias under different parameters are trained and predicted through a neural network, and simulation prediction efficiency is improved.
Third embodiment
Based on the same conception, the invention also provides an electronic device characterized by comprising: a memory for storing a processing program; and the processor is used for realizing the model simulation prediction method based on the deep learning when executing the processing program.
Based on the same conception, the invention also provides a readable storage medium, which is characterized in that a processing program is stored on the readable storage medium, and the processing program realizes the model simulation prediction method based on the deep learning when being executed by a processor.
The deep learning based model simulation prediction method, if implemented in the form of program instructions and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present embodiment may be essentially or a part contributing to the prior art or all or part of the technical solution may be embodied in the form of software, where the computer software is stored in a storage medium, and includes several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in the embodiments of the present disclosure. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-only memory (ROM), a random access memory (Random access memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
It will be apparent to those skilled in the art that, for convenience and brevity of description, reference may be made to the corresponding procedures in the foregoing method embodiments for identifying the specific implementation of the above-described system and apparatus.
The embodiments of the present invention have been described in detail with reference to the drawings, but the present invention is not limited to the above embodiments. Even if various changes are made to the present invention, it is within the scope of the appended claims and their equivalents to fall within the scope of the invention.
Claims (10)
1. The model simulation prediction method based on deep learning is characterized by comprising the following steps of:
determining a plurality of parameters required by the model simulation, and setting an adjustable boundary for each parameter;
invoking a nanosalice simulator to acquire multiple groups of data of the multiple parameters, and carrying out normalization processing on the multiple parameters based on respective adjustable boundaries;
inputting the normalized data of a plurality of parameters into a pre-configured deep learning model to train the deep learning model;
and obtaining the deep learning model after the parameter input training of the current required simulation prediction model to predict so as to obtain a prediction result.
2. The deep learning based model simulation prediction method of claim 1, wherein determining a plurality of parameters required for the model simulation, setting an adjustable boundary for each parameter further comprises:
determining the plurality of parameters based on historical real engineering experience includes: k2, u0, ua, eu, uc, vsat, a1, a2, nfactor, eta0, etab, pclm, pdiblc, ndep;
adopting standard average distribution random or taking logarithm to carry out random on the average distribution so as to obtain a plurality of groups of numerical values of the parameters;
and replacing corresponding values in a preset model card based on the multiple groups of values of the multiple parameters.
3. The deep learning based model simulation prediction method of claim 2, wherein prior to normalizing the plurality of parameters based on respective adjustable boundaries, the method further comprises:
writing a plurality of groups of numerical data of the parameters into a netlist file based on the pre-constructed netlist file, wherein the netlist file is pre-constructed according to real measuring points of a corresponding model;
dividing according to the measuring points, wherein each size comprises a plurality of groups of different curves.
4. The deep learning based model simulation prediction method of claim 1, wherein normalizing the plurality of parameters based on respective adjustable boundaries further comprises:
when the parameters are normalized by adopting a standard average distribution random method, the normalization is carried out based on the upper limit and the lower limit of the adjustable boundary of each parameter, and the processing result is between 0 and 1;
when the parameter is normalized by adopting a random method for average distribution after taking the logarithms, the logarithm is taken firstly, the upper limit and the lower limit of the adjustable boundary of each parameter are also taken as the logarithm, and further the normalization is carried out based on the logarithm of the value, the logarithm of the upper limit and the logarithm of the lower limit, and the processing result is between 0 and 1.
5. The deep learning based model simulation prediction method of claim 2, further comprising: after taking a standard average distribution random or taking a log to randomly obtain a plurality of groups of values of the plurality of parameters, preliminary screening is performed based on the set adjustable boundaries and physical facts to prevent errors of the simulator.
6. The deep learning based model simulation prediction method of claim 5, wherein before inputting the normalized data of the plurality of parameters into the pre-configured deep learning model to train the deep learning model, further comprising:
the results of the nanosalice simulator simulation are translated by the parsing module into tfrecord format for storage in order to train the network.
7. The deep learning based model simulation prediction method of claim 6, wherein inputting the normalized data of the plurality of parameters into a pre-configured simulator to train the deep learning model further comprises:
the fit method of Model class in tensorflow. Keras library is adopted, the ratio of training set to verification set is 6:1, and the number of samples in each batch is 600;
adam is adopted as an optimizer, the learning rate is 1e-3, current and capacitance are adopted as output, and a loss function is Mean Squared Logarithmic Error.
8. A model simulation prediction system based on deep learning, comprising:
the parameter determining module is used for determining a plurality of parameters required by the model simulation, and setting an adjustable boundary for each parameter;
the preprocessing module is used for calling the nanosalice simulator to acquire multiple groups of data of the multiple parameters and carrying out normalization processing on the multiple parameters based on respective adjustable boundaries;
the model training module is used for inputting the normalized data of the parameters into a pre-configured deep learning model to train the deep learning model;
and the result prediction module is used for obtaining the current parameters of the simulation prediction model to input the parameters to the deep learning model after training to predict so as to obtain a prediction result.
9. An electronic device, comprising:
a memory for storing a processing program;
a processor that implements the deep learning-based model simulation prediction method as claimed in any one of claims 1 to 7 when executing the processing program.
10. A readable storage medium, wherein a processing program is stored on the readable storage medium, and when executed by a processor, the processing program implements the model simulation prediction method based on deep learning according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311710768.3A CN117807929A (en) | 2023-12-13 | 2023-12-13 | Model simulation prediction method and system based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311710768.3A CN117807929A (en) | 2023-12-13 | 2023-12-13 | Model simulation prediction method and system based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117807929A true CN117807929A (en) | 2024-04-02 |
Family
ID=90428833
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311710768.3A Pending CN117807929A (en) | 2023-12-13 | 2023-12-13 | Model simulation prediction method and system based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117807929A (en) |
-
2023
- 2023-12-13 CN CN202311710768.3A patent/CN117807929A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111382255B (en) | Method, apparatus, device and medium for question-answering processing | |
| US10936781B1 (en) | Method for setting parameters in design of printed circuit board, device employing method, and non-transitory storage medium | |
| CN110728313B (en) | Classification model training method and device for intention classification recognition | |
| CN111144548A (en) | Method and device for identifying working condition of pumping well | |
| CN115082920A (en) | Deep learning model training method, image processing method and device | |
| CN116822383A (en) | Equipment life prediction model construction method and device, readable storage medium and equipment | |
| CN113904915A (en) | Intelligent power communication fault analysis method and system based on Internet of things | |
| CN116166967B (en) | Data processing method, equipment and storage medium based on meta learning and residual error network | |
| CN117807929A (en) | Model simulation prediction method and system based on deep learning | |
| CN108959415A (en) | A kind of exception dimension localization method, device and electronic equipment | |
| CN115237355A (en) | Directional exchange method and system based on flash memory data class identification | |
| CN111382757B (en) | Method and terminal equipment for dynamically adjusting training samples in deep learning classification algorithm | |
| US20160063149A1 (en) | Design tool apparatus, method and computer program for designing an integrated circuit | |
| CN112906386B (en) | Method and device for determining text characteristics | |
| CN115146596B (en) | Recall text generation method and device, electronic equipment and storage medium | |
| CN113449382B (en) | Simulation method and system for whole vehicle emission result | |
| CN117971357B (en) | Finite state automaton verification method and device, electronic equipment and storage medium | |
| CN117170686B (en) | Method and computing device for neural network compilation optimization | |
| CN114219307A (en) | Product determination method, device, equipment and computer storage medium | |
| CN117571321B (en) | Bearing fault detection method, device, equipment and storage medium | |
| CN114385894B (en) | Dictionary-based public opinion monitoring method and device | |
| CN110782942B (en) | Quality control method, device and storage medium in metabonomics | |
| CN110750548B (en) | Problem evaluation method based on neural network, storage medium and application server | |
| CN114676820A (en) | Network model tuning method and device, electronic equipment and storage medium | |
| CN115345637A (en) | Consumption prediction method, consumption prediction device, electronic equipment and computer readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |