CN113946991A - Semiconductor device temperature distribution prediction method based on GRNN model - Google Patents
Semiconductor device temperature distribution prediction method based on GRNN model Download PDFInfo
- Publication number
- CN113946991A CN113946991A CN202111007778.1A CN202111007778A CN113946991A CN 113946991 A CN113946991 A CN 113946991A CN 202111007778 A CN202111007778 A CN 202111007778A CN 113946991 A CN113946991 A CN 113946991A
- Authority
- CN
- China
- Prior art keywords
- data
- temperature distribution
- model
- semiconductor device
- test
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/23—Design optimisation, verification or simulation using finite element methods [FEM] or finite difference methods [FDM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/30—Circuit design
- G06F30/39—Circuit design at the physical level
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/08—Thermal analysis or thermal optimisation
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Geometry (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Design And Manufacture Of Integrated Circuits (AREA)
Abstract
The invention discloses a semiconductor device temperature distribution prediction method based on a GRNN model, which comprises the following steps: establishing a semiconductor device model based on the corresponding parameters of the target semiconductor device; acquiring a plurality of data sets of the semiconductor device model under a plurality of preset environments; training a GRNN neural network model based on the training data set to obtain an initial temperature distribution prediction model; and verifying the initial temperature distribution prediction model based on the test data set, and adjusting the initial temperature distribution prediction model according to a verification result to obtain a target temperature distribution prediction model. The method can quickly, efficiently and accurately obtain the target temperature distribution prediction model, so that the temperature distribution of the semiconductor device can be predicted based on the target temperature distribution prediction model.
Description
Technical Field
The invention belongs to the technical field of integrated circuits, and particularly relates to a semiconductor device temperature distribution prediction method based on a GRNN model.
Background
With the development of microelectronic technology, the size of semiconductor devices is continuously reduced, and the power density of integrated circuits and systems is multiplied, so that the self-heating effect of the integrated circuits is continuously intensified, wherein the self-heating effect affects the characteristics of the devices and also causes electromagnetic thermal coupling effect between the devices, and the electromagnetic thermal coupling effect causes the devices to seriously affect the performance of surrounding devices.
Therefore, in large-scale integrated circuit design, a circuit designer needs to predict the temperature characteristics of a semiconductor device at the initial stage of circuit design and evaluate the influence of the temperature distribution on the performance of the integrated circuit, so that reasonable layout and wiring optimization is performed at the initial stage of design, and the method has important significance for ensuring and improving the performance of the integrated circuit.
In the prior art, in order to accurately predict the temperature distribution data of the semiconductor device, two methods are mainly adopted: one is a finite element analysis method, but the finite element single simulation calculation cost in the method is higher, and the time consumption is longer; the other method is based on the temperature distribution result of the device, and selects a proper mathematical function for representation through the expression relationship among data to obtain the temperature distribution result of the device.
Disclosure of Invention
In order to solve the problems in the prior art, the invention provides a semiconductor device temperature distribution prediction method based on a GRNN model. The technical problem to be solved by the invention is realized by the following technical scheme:
a semiconductor device temperature distribution prediction method based on a GRNN model comprises the following steps: step 1: establishing a semiconductor device model based on the corresponding parameters of the target semiconductor device; step 2: acquiring a plurality of data sets of the semiconductor device model under a plurality of preset environments; wherein the plurality of data sets include a test data set and a plurality of training data sets, the data sets including ambient temperature data, power consumption data, distance to a heat source data, and temperature distribution data; and step 3: training a GRNN neural network model based on the training data set to obtain an initial temperature distribution prediction model; and 4, step 4: and verifying the initial temperature distribution prediction model based on the test data set, and adjusting the initial temperature distribution prediction model according to a verification result to obtain a target temperature distribution prediction model.
In one embodiment of the present invention, the parameters corresponding to the target semiconductor device include structural parameters and physical parameters.
In one embodiment of the present invention, the step 2 comprises: step 2-1: carrying out finite element mesh division on a plurality of semiconductor device models in a preset environment; step 2-2: performing steady state solution on the semiconductor device model after the finite element meshing to obtain temperature distribution data of the target semiconductor device; step 2-3: and determining environmental temperature data, power consumption data and distance data from a heat source corresponding to a preset environment and temperature distribution data obtained by solving based on the preset environment into a data set so as to obtain a plurality of data sets.
The invention has the beneficial effects that:
the invention obtains a temperature distribution prediction model based on a GRNN model, the GRNN is based on a radial basis neural network, and has good nonlinear approximation performance. The data set comprises environment temperature data, power consumption data and distance data from a heat source, and can accurately represent the temperature distribution characteristics of the device.
The invention can avoid the problems of large occupation of computer resources and time consumption of analysis of the finite element analysis method. Compared with a characteristic function method, the method has the advantages that the accuracy of the prediction result obtained based on the GRNN model is higher, and the temperature distribution information of the device can be accurately represented.
The present invention will be described in further detail with reference to the accompanying drawings and examples.
Drawings
Fig. 1 is a schematic flowchart of a method for predicting temperature distribution of a semiconductor device based on a GRNN model according to an embodiment of the present invention;
fig. 2 is a schematic diagram of a topological structure of a GRNN model according to an embodiment of the present invention;
fig. 3 is a GRNN training error diagram of device temperature distribution as a function of position of a heat source and power consumption of a semiconductor device, obtained based on the method of the present invention.
Detailed Description
The present invention will be described in further detail with reference to specific examples, but the embodiments of the present invention are not limited thereto.
Examples
Referring to fig. 1, fig. 1 is a method for predicting temperature distribution of a semiconductor device based on a GRNN model according to an embodiment of the present invention, where the method includes:
step 1: and establishing a semiconductor device model based on the corresponding parameters of the target semiconductor device.
Optionally, the parameters corresponding to the target semiconductor device include a structural parameter and a physical parameter.
The invention obtains the structure parameter and the physical parameter of the semiconductor device in advance according to the process information of the semiconductor device. The invention is not limited to the semiconductor device, such as an InP HBT device, and obtains the information of the device in the circuit, including the structure parameters (active region, substrate size, etc.) and physical parameters (device material property parameters, doping, etc.) of the device, according to the InP HBT process library file. Device materials and dimensions are shown in table 1, and each material in the device uses a temperature dependent thermal conductivity:
TABLE 1 InP HBT device Structure and Material parameters
The method of fabricating the semiconductor device is not particularly limited, and may be, for example,
the method comprises the steps of constructing a semiconductor device model by using COMSOL finite element analysis software, applying a heat source to the established solid model, setting boundary conditions (setting the back surface temperature of a substrate as ambient temperature and insulating other surfaces of a chip), and meshing the established three-dimensional geometric semiconductor device model. And setting the power consumption of the active device and the solving range of the ambient temperature, performing steady-state thermal analysis, and obtaining the distribution condition of the temperature along with the change of the distance.
Step 2: acquiring a plurality of data sets of the semiconductor device model under a plurality of preset environments; wherein the plurality of data sets includes a test data set and a plurality of training data sets, the data sets including ambient temperature data, power consumption data, distance to heat source data, and temperature distribution data.
The method can change the power consumption (heat source) of the semiconductor device and the ambient temperature (boundary condition) of the semiconductor device, and perform steady-state thermal analysis, thereby obtaining the temperature distribution results of the semiconductor device under different ambient temperatures and different power consumptions, and obtaining the variation distribution curve of the surface temperature of the device along with the distance to generate a data set required by GRNN training.
The environment temperature data is 300K, 320K, 340K, 360K and 380K, and the power consumption is 5-15 mW. The training data set comprises power consumption data, distance data from a heat source and temperature distribution data corresponding to 300K, 320K, 340K and 360K; the test data set is mainly used for evaluating the prediction and generalization capability of the GRNN model, such as power consumption data corresponding to 380K, distance data from a heat source and temperature distribution data.
Optionally, step 2 includes:
step 2-1: carrying out finite element mesh division on a plurality of semiconductor device models in a preset environment;
examples are as follows: and meshing the semiconductor device model by using a free subdivision tetrahedron option.
Step 2-2: performing steady state solution on the semiconductor device model after the finite element meshing to obtain temperature distribution data of the target semiconductor device;
step 2-3: and determining environmental temperature data, power consumption data and distance data from a heat source corresponding to a preset environment and temperature distribution data obtained by solving based on the preset environment into a data set so as to obtain a plurality of data sets.
And step 3: and training the GRNN neural network model based on the training data set to obtain an initial temperature distribution prediction model.
The invention can train a data set by using a GRNN (Generalized Regression Neural Network) model in MATLAB software, for example, calling a newgrnn function in Matlab, writing a GRNN program for training, setting a spread interval to be [0.001, 0.03], setting the step length to be 0.001, and setting the initial default value of spread to be 1.0. Fig. 2 is a schematic view of a topological structure of a GRNN model according to an embodiment of the present invention.
Optionally, determining the ambient temperature data, the power consumption data, and the distance data from a heat source as input data of the GRNN model; and determining the temperature distribution data as output data of the GRNN model.
And 4, step 4: and verifying the initial temperature distribution prediction model based on the test data set, and adjusting the initial temperature distribution prediction model according to a verification result to obtain a target temperature distribution prediction model.
Such as: the method comprises the steps of importing a training set into a GRNN model to carry out normalization processing on data, carrying out training learning based on the GRNN model on the data with the environmental temperatures of 300K, 320K, 340K and 360K in the data set, and carrying out subsequent simulation prediction on the data with the environmental temperature of 380K.
Optionally, the step 4 includes:
step 4-1: inputting the environmental temperature data, the power consumption data and the distance data from the heat source in the test data set into the initial temperature distribution model to obtain test temperature distribution data;
step 4-2: comparing the test temperature distribution data with the temperature distribution data in the test data set;
step 4-3: and when the error between the test temperature distribution data and the temperature distribution data in the test data set is smaller than a preset threshold value, determining the initial temperature distribution model as a target temperature distribution prediction model.
Optionally, after the step 4-2, the method further includes:
step S1: when the error between the test temperature distribution data and the temperature distribution data in the test data set is larger than a preset threshold value, adjusting the range and the step length of a diffusion constant in the GRNN model;
the range and step size of the model diffusion constant spread are adjusted, the range of the spread can be expanded, and/or the step size can be reduced, and the updated spread is used as the spread needed to be set in the next training.
Step S2: repeatedly executing the step 3 based on the adjusted GRNN model to obtain an updated initial temperature distribution prediction model;
step S3: and (4) executing the step 4 based on the updated initial temperature distribution prediction model until the error between the test temperature distribution data and the temperature distribution data in the test data set is smaller than a preset threshold value, and determining the updated initial temperature distribution model as a target temperature distribution prediction model.
The temperature prediction method can predict the temperature values of the semiconductor device at different positions away from the heat source of the device based on the target temperature distribution prediction model, and realize temperature distribution prediction.
FIG. 3 is a GRNN training error diagram of device temperature distribution varying with the position and power consumption of a semiconductor device heat source obtained based on the method of the present invention, wherein the X coordinate is the position from the semiconductor device heat source, the Y coordinate is the semiconductor device temperature, the blue curve represents the true value of finite element simulation, the red circle connected curve represents the predicted value obtained by the method of the present invention, and it can be seen from the diagram that at different sampling points, the predicted result obtained by the semiconductor temperature distribution prediction method of the present invention is very close to the true value.
Neural networks are adaptive nonlinear dynamical systems formed by a large number of simple neurons interconnected, are very adept at recognizing linear and nonlinear relations between single/multiple inputs and outputs, and also have high generalization capability, requiring minimal data storage. After learning from the initialization inputs and their relationships, it can also infer unknown relationships between the unknown data, enabling the model to generalize and predict the unknown data. And unlike many other predictive techniques, ANNs (Artificial neural networks), which include GRNN, do not impose any restrictions (e.g., how they are distributed) on input variables, and because ANNs have the ability to learn hidden relationships in data without imposing any fixed relationships in the data.
In summary, the invention obtains the temperature distribution prediction model based on the GRNN model, the GRNN is based on the radial basis neural network, and has good nonlinear approximation performance, compared with the radial basis neural network, the training is more convenient, the learning speed can be greatly accelerated, and the problem of local minimum value can be avoided. The data set comprises environment temperature data, power consumption data and distance data from a heat source, and can accurately represent the temperature distribution characteristics of the device.
The invention can avoid the problems of large occupation of computer resources and time consumption of analysis of the finite element analysis method. Compared with a characteristic function method, the method has the advantages that the accuracy of the prediction result obtained based on the GRNN model is higher, and the temperature distribution information of the device can be accurately represented.
The foregoing is a more detailed description of the invention in connection with specific preferred embodiments and it is not intended that the invention be limited to these specific details. For those skilled in the art to which the invention pertains, several simple deductions or substitutions can be made without departing from the spirit of the invention, and all shall be considered as belonging to the protection scope of the invention.
Claims (6)
1. A semiconductor device temperature distribution prediction method based on a GRNN model is characterized by comprising the following steps:
step 1: establishing a semiconductor device model based on the corresponding parameters of the target semiconductor device;
step 2: acquiring a plurality of data sets of the semiconductor device model under a plurality of preset environments; wherein the plurality of data sets include a test data set and a plurality of training data sets, the data sets including ambient temperature data, power consumption data, distance to a heat source data, and temperature distribution data;
and step 3: training a GRNN neural network model based on the training data set to obtain an initial temperature distribution prediction model;
and 4, step 4: and verifying the initial temperature distribution prediction model based on the test data set, and adjusting the initial temperature distribution prediction model according to a verification result to obtain a target temperature distribution prediction model.
2. The method of claim 1, wherein the parameters corresponding to the target semiconductor device comprise structural parameters and physical parameters.
3. The method of claim 1, wherein the step 2 comprises:
step 2-1: carrying out finite element meshing on the semiconductor device models in a plurality of preset environments;
step 2-2: performing steady state solution on the semiconductor device model after the finite element meshing to obtain temperature distribution data of the target semiconductor device;
step 2-3: and determining environmental temperature data, power consumption data and distance data from a heat source corresponding to a preset environment and temperature distribution data obtained by solving based on the preset environment into a data set so as to obtain a plurality of data sets.
4. The method of claim 1, wherein the ambient temperature data, power consumption data, distance from heat source data are determined as input data to the GRNN model; and determining the temperature distribution data as output data of the GRNN model.
5. The method of claim 1, wherein the step 4 comprises:
step 4-1: inputting the environmental temperature data, the power consumption data and the distance data from the heat source in the test data set into the initial temperature distribution model to obtain test temperature distribution data;
step 4-2: comparing the test temperature distribution data with the temperature distribution data in the test data set;
step 4-3: and when the error between the test temperature distribution data and the temperature distribution data in the test data set is smaller than a preset threshold value, determining the initial temperature distribution model as a target temperature distribution prediction model.
6. The method of claim 5, wherein after the step 4-2, the method further comprises:
step S1: when the error between the test temperature distribution data and the temperature distribution data in the test data set is larger than a preset threshold value, adjusting the range and the step length of a diffusion constant in the GRNN model;
step S2: repeatedly executing the step 3 based on the adjusted GRNN model to obtain an updated initial temperature distribution prediction model;
step S3: and (4) executing the step 4 based on the updated initial temperature distribution prediction model until the error between the test temperature distribution data and the temperature distribution data in the test data set is smaller than a preset threshold value, and determining the updated initial temperature distribution model as a target temperature distribution prediction model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111007778.1A CN113946991B (en) | 2021-08-30 | 2021-08-30 | Semiconductor device temperature distribution prediction method based on GRNN model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111007778.1A CN113946991B (en) | 2021-08-30 | 2021-08-30 | Semiconductor device temperature distribution prediction method based on GRNN model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113946991A true CN113946991A (en) | 2022-01-18 |
| CN113946991B CN113946991B (en) | 2023-08-15 |
Family
ID=79327750
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111007778.1A Active CN113946991B (en) | 2021-08-30 | 2021-08-30 | Semiconductor device temperature distribution prediction method based on GRNN model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113946991B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100211536A1 (en) * | 2007-08-31 | 2010-08-19 | Al-Fattah Saud Mohammed A | Artificial neural net work models for determining relative permeability of hydrocarbon reservoirs |
| CN106991212A (en) * | 2017-03-07 | 2017-07-28 | 西安交通大学 | A kind of blade root Strength Safety characteristic prediction method for optimizing GRNN network algorithms based on GA_PSO |
-
2021
- 2021-08-30 CN CN202111007778.1A patent/CN113946991B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100211536A1 (en) * | 2007-08-31 | 2010-08-19 | Al-Fattah Saud Mohammed A | Artificial neural net work models for determining relative permeability of hydrocarbon reservoirs |
| CN106991212A (en) * | 2017-03-07 | 2017-07-28 | 西安交通大学 | A kind of blade root Strength Safety characteristic prediction method for optimizing GRNN network algorithms based on GA_PSO |
Non-Patent Citations (2)
| Title |
|---|
| 刘子英;朱琛磊;: "基于Elman神经网络模型的IGBT寿命预测", 半导体技术, no. 05 * |
| 朱正林;刘权;张欢;: "基于优化广义回归神经网络的碟式太阳能温度预测研究", 南京工程学院学报(自然科学版), no. 01 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113946991B (en) | 2023-08-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhang et al. | Advanced parallel space-mapping-based multiphysics optimization for high-power microwave filters | |
| Zhang et al. | A novel training approach for parametric modeling of microwave passive components using Padé via Lanczos and EM sensitivities | |
| Koziel et al. | Improved modeling of microwave structures using performance-driven fully-connected regression surrogate | |
| Yuan et al. | Solder joint reliability modeling by sequential artificial neural network for glass wafer level chip scale package | |
| CN113177290B (en) | Satellite component temperature field prediction method based on depth agent model normalization | |
| Xu et al. | Surrogate thermal model for power electronic modules using artificial neural network | |
| Na et al. | Automated neural network-based multiphysics parametric modeling of microwave components | |
| Spina et al. | Polynomial chaos-based macromodeling of general linear multiport systems for time-domain analysis | |
| CN110362881B (en) | Microwave power device nonlinear model method based on extreme learning machine | |
| Deng et al. | A 2D hyperchaotic discrete memristive map and application in reservoir computing | |
| Dolin et al. | Modified methods of circuit simulation of radio engineering devices in the time domain | |
| Fu et al. | Solving non-homogeneous linear ordinary differential equations using memristor-capacitor circuit | |
| CN103366033B (en) | method and system for statistical circuit simulation | |
| CN113946991B (en) | Semiconductor device temperature distribution prediction method based on GRNN model | |
| CN113947008B (en) | Semiconductor device temperature distribution prediction method based on BP neural network model | |
| Han et al. | Reservoir computing dissection and visualization based on directed network embedding | |
| Du et al. | A surrogate-based parallel optimization of analog circuits using multi-acquisition functions | |
| Rosa et al. | Using anns to size analog integrated circuits | |
| Jiang et al. | Performance prediction of large-scale 1S1R resistive memory array using machine learning | |
| CN112926259B (en) | Method for predicting junction temperature of semiconductor device based on RBF neural network model | |
| CN115061387A (en) | Component use reliability assessment method based on field-circuit coupling collaborative simulation model | |
| Sami et al. | Artificial neural network and dataset optimization for implementation of linear system models in resource‐constrained embedded systems | |
| US9582622B1 (en) | Evaluating on-chip voltage regulation | |
| Ghasemzadeh et al. | A new mixed-signal CMOS fuzzy logic controller in current mode | |
| JP2023513754A (en) | Design of Dynamic Random Access Memory Path Transistor with Statistical Variation in Leakage Current |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |