CN114781201A

CN114781201A - Method, system, device and medium for calculating temperature field of PCB in radiator

Info

Publication number: CN114781201A
Application number: CN202210223135.9A
Authority: CN
Inventors: 谢颖熙; 吴小华; 陆龙生; 梁庆铧; 梁澜之; 钟宜淋
Original assignee: South China University of Technology SCUT
Current assignee: South China University of Technology SCUT
Priority date: 2022-03-07
Filing date: 2022-03-07
Publication date: 2022-07-22

Abstract

The invention discloses a method, a system, a device and a medium for calculating a PCB temperature field in a radiator, wherein the method comprises the following steps: acquiring training data, wherein the training data comprises radiator fin distribution, chip layout and temperature field distribution; preprocessing the training data, and combining corresponding temperature field distribution as a label to form matching data; generating a countering neural network based on deep convolution, training to obtain a strong mapping relation between radiator fin distribution and chip layout, and constructing an agent model for quickly calculating a PCB (printed Circuit Board) steady-state temperature field under the heat dissipation condition according to the radiator fin distribution and the chip layout; inputting input data consisting of radiator fin distribution and chip layout into the trained proxy model, and rapidly calculating the steady-state PCB temperature cloud chart. According to the invention, through history acquisition, a proxy model is obtained through training and is used for helping a thermal design engineer to improve the design efficiency under the condition of ensuring the prediction precision; can be widely applied to the technical field of numerical heat transfer science calculation.

Description

Method, system, device and medium for calculating temperature field of PCB in radiator

Technical Field

The invention relates to the technical field of numerical heat transfer calculation, in particular to a method, a system, a device and a medium for calculating a PCB temperature field in a radiator.

Background

The current degree of automobile autopilot increases, and the quantity of the electronic control unit of car increases fast, when providing the challenge to integrated circuit integrated capability, raises higher requirement simultaneously to electronic equipment's such as controller heat-sinking capability. Generally speaking, the temperature field calculation of electronic equipment is a complex process of multi-physical field coupling, and when finite element simulation is performed, massive grids need to be divided for entities such as fins and chips, and boundary conditions and iteration conditions are set by combining hydromechanics and heat transfer content to perform solution.

When the heat flow coupling problem is solved, the occupied computing resources are high, the consumed time is long, and the requirement on a user is high. Therefore, how to simplify the numerical simulation process and reduce the design cost of the thermal simulation scheme is a problem to be solved urgently in the design and research of the new generation.

Disclosure of Invention

To solve at least one of the technical problems in the prior art to a certain extent, the present invention provides a method, a system, a device and a medium for calculating a temperature field of a PCB in a heat sink.

The technical scheme adopted by the invention is as follows:

a method for calculating a PCB temperature field in a radiator comprises the following steps:

acquiring training data, wherein the training data comprises radiator fin distribution, chip layout and temperature field distribution;

preprocessing the training data, and combining corresponding temperature field distribution as a label to form matching data;

generating an antagonistic neural network based on deep convolution, training to obtain a strong mapping relation between radiator fin distribution and chip layout, and constructing an agent model for quickly calculating a PCB steady-state temperature field under the heat dissipation condition according to the radiator fin distribution and the chip layout;

inputting input data consisting of radiator fin distribution and chip layout into the trained proxy model, and rapidly calculating the steady-state PCB temperature cloud chart.

Further, the agent model is trained in a mode of generating an antagonistic network, the training process consists of a generator and a discriminator, the generator is used for converting input radiator fin distribution and chip distribution into a PCB stable temperature field cloud picture, the discriminator is used for distinguishing the temperature field cloud picture generated by the generator from an actual label cloud picture, the balance point of final training is that when the temperature cloud picture generated by the generator is consistent with or very close to labels corresponding to a data set, and the probability that the discriminator can distinguish the cloud picture generated by the generator is 50%.

Further, the input data of the proxy model is obtained by:

step 1: determining fin distribution and fin height, chip distribution and chip heating power, and bump distribution and bump height for connecting the chip and the radiator;

and 2, step: creating a three-channel image with the same proportion as the researched radiator, and respectively performing linear mapping on the fin height, the chip heating power and the bump height and an image channel numerical value to obtain model training data;

and 3, step 3: and (3) creating a corresponding thermal model in ANSYS software according to the height of the fin, the heating power of the chip and the height of the bump, carrying out numerical simulation, obtaining the distribution of the PCB temperature field as training labels, and matching the training labels with the model training data in the step (2) one by one.

Further, the calculation method further comprises a fine tuning step:

and migrating the data training weight of the historical application scene to the current specific application scene weight.

The invention adopts another technical scheme that:

a computing system for a PCB temperature field in a heat sink, comprising:

the data acquisition module is used for acquiring training data, wherein the training data comprises radiator fin distribution, chip layout and temperature field distribution;

the label generation module is used for preprocessing the training data and combining the corresponding temperature field distribution as a label to form matching data;

the data training module is used for generating a confrontation neural network based on deep convolution, training to obtain a strong mapping relation between radiator fin distribution and chip layout, and constructing an agent model for quickly calculating a PCB steady-state temperature field under the heat dissipation condition according to the radiator fin distribution and the chip layout;

and the temperature calculation module is used for inputting input data consisting of radiator fin distribution and chip layout into the trained proxy model and quickly calculating the steady-state PCB temperature cloud chart.

Further, the input data of the proxy model is obtained by:

determining fin distribution and fin height, chip distribution and chip heating power, and bump distribution and bump height for connecting the chip and the radiator;

creating a three-channel image with the same proportion as the researched radiator, and respectively performing linear mapping on the fin height, the chip heating power and the bump height and the image channel numerical value to obtain model training data;

and (3) creating a corresponding thermal model in ANSYS software according to the height of the fin, the heating power of the chip and the height of the bump, performing numerical simulation, obtaining the distribution of the PCB temperature field as training labels, and pairing the training labels with model training data one by one.

Further, the computing system further comprises a fine-tuning module for migrating the data training weights of the historical application scenarios to the current specific application scenario weights.

The invention adopts another technical scheme that:

a computing device for a PCB temperature field within a heat sink, comprising:

at least one processor;

at least one memory for storing at least one program;

when executed by the at least one processor, cause the at least one processor to implement the method described above.

The other technical scheme adopted by the invention is as follows:

a computer readable storage medium in which a processor executable program is stored, which when executed by a processor is for performing the method as described above.

The beneficial effects of the invention are: according to the method, historical related thermal design scene structure data and corresponding PCB temperature field distribution data are collected to serve as training data for generating the countermeasure network, so that a proxy model capable of rapidly calculating the PCB temperature field corresponding to the corresponding scene structure data is obtained, and thermal design engineers can be helped to improve design efficiency under the condition of ensuring prediction accuracy through the proxy model.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following description is made on the drawings of the embodiments of the present invention or the related technical solutions in the prior art, and it should be understood that the drawings in the following description are only for convenience and clarity of describing some embodiments in the technical solutions of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to these drawings without creative efforts.

FIG. 1 is a schematic diagram of a system for rapidly calculating a temperature field of a PCB in a heat sink according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of a thermal simulation model used to generate a tag in an embodiment of the present invention;

FIG. 3 is an initial label graph obtained in the step of generating labels in an embodiment of the present invention;

FIG. 4 is a diagram illustrating pairs of matching data obtained in an embodiment of the present invention;

FIG. 5 is a diagram illustrating a structure of a model for generating a confrontation network and a training process selected in an embodiment of the present invention;

FIG. 6 is a diagram comparing a model training result and a simulation result according to an embodiment of the present invention;

FIG. 7 is a flowchart illustrating steps of a method for calculating a temperature field of a PCB in a heat sink according to an embodiment of the present invention.

Detailed Description

Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are illustrative only for the purpose of explaining the present invention and are not to be construed as limiting the present invention. The step numbers in the following embodiments are provided only for convenience of illustration, the order between the steps is not limited at all, and the execution order of each step in the embodiments can be adapted according to the understanding of those skilled in the art.

In the description of the present invention, it should be understood that the orientation or positional relationship referred to in the description of the orientation, such as the upper, lower, front, rear, left, right, etc., is based on the orientation or positional relationship shown in the drawings, and is only for convenience of description and simplification of description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus, should not be construed as limiting the present invention.

In the description of the present invention, the meaning of a plurality of means is one or more, the meaning of a plurality of means is two or more, and larger, smaller, larger, etc. are understood as excluding the number, and larger, smaller, inner, etc. are understood as including the number. If the first and second are described for the purpose of distinguishing technical features, they are not to be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated or implicitly indicating the precedence of the technical features indicated.

In the description of the present invention, unless otherwise explicitly limited, terms such as arrangement, installation, connection and the like should be understood in a broad sense, and those skilled in the art can reasonably determine the specific meanings of the above terms in the present invention in combination with the specific contents of the technical solutions.

The embodiment provides a rapid calculation system of a PCB temperature field in a radiator, which comprises a model preparation module and a front-end use module, and the system architecture diagram is shown in FIG. 1. The model preparation module comprises the steps of data sampling, label generation, data training, weight storage and the like.

The data sampling step comprises model training data preparation and thermal model structure parameter preparation. A thermal model is first built for a common passive heat dissipation scenario, as shown in fig. 2. The upper cover 100 includes fins 110 and bumps 120 contacting with the chips, a plurality of chip groups 140 are distributed on the PCB module 130, and the lower cover 150 is used to form a closed link where the PCB module 130 is located. And then generating a plurality of parameters in a specified range, and randomly sampling a plurality of groups in a certain range, wherein the plurality of groups comprise fin types, fin thicknesses, fin numbers, fin heights, substrate heights, bump position parameters, bump heights, chip position parameters and chip power consumption. And (3) parameterizing the thermal model according to the random parameter values, calculating a steady-state temperature field by using ICEPAK, and deriving the temperature cloud map distribution on the PCB under the steady-state condition. In order to facilitate loading and processing of subsequent model training, each parameter value is expressed in an image mode with the same proportion as that of a PCB, wherein fin position distribution information such as fin types, fin thicknesses and the like is expressed by different pixel values R of a red channel of an image, the specific pixel value is obtained by linearly mapping a fin height h1 and a substrate height h2 according to the following formula (1); the bump position information is represented by a green channel pixel value G of the image, and the specific pixel value is obtained by linearly mapping the bump height h3 through the following formula (2); the chip position information is represented by the image blue channel pixel value B, and the specific pixel value is obtained by linearly mapping the chip power value P and the sizes r1, r2 according to the following formula (3), and it is noted that h1, h2, h3, P, r1, r2 are not constants and depend on the parameters of the image corresponding to the position of the heat sink.

Formula (1):

wherein

A set of points representing corresponding locations of fins in the image,

representing the set of points where no fin is present in the image, only the boss corresponds to the point, h1, h2 being in mm.

Formula (2):

wherein

And h3 is a set of points representing the corresponding positions of bumps in the image, and the unit is mm.

Formula (3):

wherein

Set of points representing the corresponding positions of the chips in the image, P being in WR1, r2 units are mm.

In the label generating step, an ICEPAK module in an ANSYS workbench platform is utilized to establish a parameterized simulation model method. In the embodiment, a temperature field of a radiator model in a natural cooling state is simulated by using a CFD solver and a Blxinesk approximate model in ANSYS, and meanwhile, the radiation heat exchange phenomenon in natural convection is considered. Considering both the calculation speed and the accuracy, a Discrete Ordinates (DO) radiation model is adopted. The fluid state in the calculation domain is assumed to be turbulent flow, and a zero equation turbulence model with higher cost performance is adopted because the fluid state simulates an electronic product radiator. In order to ensure the calculation accuracy, 1.5 times of model feature length L is set for the calculation domain except for the Ymax direction of the top of the heat sink, the Ymax direction is set to be 2 times of the model feature length L, the region attributes in the six directions are all open, and the ambient temperature in the calculation domain is set to be 85 ℃. The chip was simulated using two copper blocks (blocks) doped with a zero thickness heat source (source) in between. And carrying out parameterization setting on the shape parameter and the power parameter of the chip. The shell is built by adopting an enclosure cavity module, six boundary surfaces are set to be thin shell features with the thickness of 2mm (the thickness actually participates in the calculation of a temperature field, but does not participate in the grid division of a model), and the whole shell is made of an aluminum alloy material. Because the chip on the PCB is at a certain distance from the upper end of the cavity shell, bumps (blcok) with corresponding quantity and size are established to link the cavity shell and the chip, which is beneficial to effective heat dissipation. A plate structure with the thickness of 1mm is arranged at the contact position of the bump and the chip and used for simulating heat conduction silica gel in an actual object, and the heat conduction coefficient of the heat conduction silica gel is set to be 5W/m.k. After the basic modeling of the electronic device is completed, an aluminum profile extruded radiator is added at the upper end of the cavity shell. The radiator is made of an aluminum alloy material with the heat conductivity coefficient of 240W/m.k, parameters are set for the overall height of the radiator, the height of the substrate, the number of the fins and the thickness of the fins, and the simulation of temperature fields of various radiators can be realized by changing the parameter values.

After the integral model is built, the model needs to be subjected to grid division, and because the shapes contained in the model are regular, Hexa Cartesian structured grids are adopted, and all grids are perpendicular and orthogonal. And after the grid is divided, checking the grid quality of the grid, and judging according to the Face alignment rate, the volume value volume of the grid and the Skewness Skewness of the grid. The values of all three are close to the expected values, so that the grid quality is good for the model and the model can be conformal. In addition to the need for good meshing quality, good meshing accuracy is also required. The number of meshes directly affects the speed and accuracy of the solution to the final result. For the overall model, the maximum mesh size for the X, Y, Z direction is limited to 10mm, 5mm, 15 mm. This is limited by 1/20-1/40 in terms of the calculated domain feature size. However, because the part sizes of the chip model and the fin model are much smaller than the characteristic size of the computational domain, the grid encryption needs to be independently performed to achieve the required computational accuracy. In order to test the grid independence, the quantity of grids of the chip and the fin model is increased in stages, and the change condition of the model temperature field under different grid quantities is tested. According to the results of six different mesh encryption models, when the number of meshes is 1161576, the model can calculate the temperature field of the model with a smaller calculation time while ensuring better calculation accuracy, so that the mesh division mode is finally selected.

Because finite element analysis simulation is carried out, requirements are required on the accurate value of the result, the corresponding iteration step number and the residual error standard of the corresponding parameter need to be set, the standard of the coupling residual error, namely the Flow residual error is 1e-3, the energy residual error value is 1e-7, and the calculation result is considered to meet the accuracy requirement as long as the three residual error standards are met. And (3) performing calculation solution of a small number of parameters in the early stage, and finding that most of model calculation converges within 250 steps, so that the iteration step number is set to be 250.

The artwork of data obtained using ICEPAK is shown in FIG. 3, where the PCB area is cropped by digital image technology and saved for use as a label.

The fin types in the training data are selected from a plurality of fin types for data acquisition, wherein a plurality of pieces of matching data are selected as shown in fig. 4. Fig. 4 shows six pairs of matching data, the left half is the combined image calculated according to the above formula, and the right half is obtained by intercepting the cloud image result obtained by the simulation software.

The data training part starts the training process by putting the paired data composition data sets into the condition generation countermeasure network as shown in fig. 5. The model used is a Spatially Adaptive pixel-level network (ASAP-Net). The training modes are divided into two types: there is no pre-training data and no scene data trimming. The training-free data means that no relevant historical data is used for pre-training, and only the current scene data is trained. And when training is carried out without pre-training data, selecting 16-32 matched data to train the model, and testing the predicted quality of the model by using a leave-out method to obtain and store the weight. Scene data fine tuning refers to selecting part of historical data for training in the presence of similar historical databases, and the quantity of the historical data is not limited herein. Generally, the more the data is, the more accurate the data is, after the pre-training weight is obtained, fine tuning is performed by using a small learning rate under a small amount of scene data, the obtained accurate weight is stored, and a result obtained through training without the pre-training data is shown in fig. 6, in which input of 4 data randomly extracted from a test set, a model calculation result and a corresponding label are respectively shown.

The front-end use module comprises a deployment weight and display part. By deploying the stored weight to a cloud resource, a quick calculation function can be provided for a WeChat applet through services such as WeChat cloud hosting and the like, and a user can simply design a fin structure and a chip position on interfaces such as the WeChat applet and the like to realize quick calculation; similarly, in view of the lightweight model, the model can be placed in a personal computer for off-line use, and a large amount of computing power is not required to be consumed to complete the rapid computing process.

In the embodiment, historical related thermal design scene structure data and corresponding PCB temperature field distribution data are collected to serve as training data for generating the countermeasure network, so that the proxy model capable of rapidly calculating the PCB temperature field corresponding to the corresponding scene structure data is obtained. The proxy model can help a thermal design engineer to improve the design efficiency under the condition of ensuring the prediction precision, and in addition, the model weight migration of historical data can be realized according to a small amount of novel data so as to adapt to the thermal design precision requirement of a corresponding scene. And when the selection of the heat dissipation scheme peripheral equipment needs to be evaluated quickly, an engineer is helped to estimate the temperature field of the PCB quickly. Compared with the complete CFD numerical simulation process in the traditional design flow, the method has the advantages that the calculation cost is greatly reduced, the calculation time and the learning cost of a user are reduced, and manpower and material resources are saved.

As shown in fig. 7, this embodiment further provides a method for calculating a temperature field of a PCB in a heat sink, including the following steps:

s1, acquiring training data, wherein the training data comprises radiator fin distribution, chip layout and temperature field distribution;

s2, preprocessing the training data, and combining the corresponding temperature field distribution as a label to form matching data;

s3, generating an antagonistic neural network based on the deep convolution, training to obtain a strong mapping relation between the distribution of the radiator fins and the chip layout, and constructing a proxy model for quickly calculating the steady-state temperature field of the PCB under the heat dissipation condition according to the distribution of the radiator fins and the chip layout;

and S4, inputting input data consisting of radiator fin distribution and chip layout into the trained proxy model, and rapidly calculating the steady-state PCB temperature cloud chart.

The calculation method of the PCB temperature field in the heat sink of the present embodiment has a corresponding relationship with the calculation system of the PCB temperature field in the heat sink shown in fig. 1, and thus has the functions and advantages of the system.

The present embodiment further provides a device for calculating a PCB temperature field in a heat sink, including:

at least one processor;

at least one memory for storing at least one program;

when executed by the at least one processor, cause the at least one processor to implement the method of fig. 7.

The device for calculating the temperature field of the PCB in the heat sink according to the embodiment of the present invention can perform the method for calculating the temperature field of the PCB in the heat sink according to the embodiment of the method of the present invention, can perform any combination of the implementation steps of the method embodiment, and has corresponding functions and advantages of the method.

Embodiments of the present application also disclose a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The computer instructions may be read by a processor of a computer device from a computer-readable storage medium, and the computer instructions executed by the processor cause the computer device to perform the method illustrated in fig. 7.

The embodiment also provides a storage medium, which stores instructions or programs capable of executing the method for calculating the temperature field of the PCB in the heat sink provided by the embodiment of the method of the invention, and when the instructions or the programs are run, the steps can be implemented by any combination of the embodiments of the method, and the method has corresponding functions and beneficial effects.

In alternative embodiments, the functions/acts noted in the block diagrams may occur out of the order noted in the operational illustrations. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality/acts involved. Furthermore, the embodiments presented and described in the flow charts of the present invention are provided by way of example in order to provide a more thorough understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is changed and in which sub-operations described as part of larger operations are performed independently.

Furthermore, although the present invention is described in the context of functional modules, it should be understood that, unless otherwise stated to the contrary, one or more of the described functions and/or features may be integrated in a single physical device and/or software module, or one or more functions and/or features may be implemented in a separate physical device or software module. It will also be appreciated that a detailed discussion of the actual implementation of each module is not necessary for an understanding of the present invention. Rather, the actual implementation of the various functional modules in the apparatus disclosed herein will be understood within the ordinary skill of an engineer, given the nature, function, and internal relationship of the modules. Accordingly, those skilled in the art can, using ordinary skill, practice the invention as set forth in the claims without undue experimentation. It is also to be understood that the specific concepts disclosed are merely illustrative of and not intended to limit the scope of the invention, which is to be determined from the appended claims along with their full scope of equivalents.

The functions, if implemented in the form of software functional units 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 invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk, and various media capable of storing program codes.

The logic and/or steps represented in the flowcharts or otherwise described herein, e.g., an ordered listing of executable instructions that can be considered to implement logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.

More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Further, the computer-readable medium could even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.

It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.

In the foregoing description of the specification, reference to the description of "one embodiment/example," "another embodiment/example," or "certain embodiments/examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

While embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that: various changes, modifications, substitutions and alterations can be made to the embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

While the preferred embodiments of the present invention have been illustrated and described, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

Claims

1. A method for calculating a temperature field of a PCB in a radiator is characterized by comprising the following steps:

2. The method as claimed in claim 1, wherein the surrogate model is trained by a generative countermeasure network, the training process is composed of two different models, i.e. a generator and a discriminator, the generator is used for converting the input fin distribution and chip distribution of the heat sink into a steady-state temperature field cloud picture of the PCB, the discriminator is used for distinguishing the temperature field cloud picture generated by the generator from an actual label cloud picture, the balance point of the final training is when the temperature cloud picture generated by the generator is consistent with the labels corresponding to the data sets, and the probability that the discriminator can discriminate the cloud picture generated by the generator is 50%.

3. The method for calculating the PCB temperature field in the radiator according to claim 1, wherein the input data of the proxy model is obtained by the following steps:

and (3) creating a corresponding thermal model in ANSYS software according to the height of the fin, the heating power of the chip and the height of the bump, carrying out numerical simulation, obtaining the distribution of the PCB temperature field as training labels, and matching the training labels with model training data one by one.

4. The method of claim 1, further comprising the step of fine tuning:

5. A computing system for a PCB temperature field in a heat sink, comprising:

the label generation module is used for preprocessing the training data and combining corresponding temperature field distribution as a label to form matching data;

the data training module is used for generating a countering neural network based on deep convolution, training to obtain a strong mapping relation between radiator fin distribution and chip layout, and constructing a proxy model for quickly calculating a PCB steady-state temperature field under the heat dissipation condition according to the radiator fin distribution and the chip layout;

6. The system of claim 5, wherein the surrogate model is trained by a generative countermeasure network, the training process comprises a generator and a discriminator, the generator is used for converting the input fin distribution and the chip distribution of the heat sink into a steady-state temperature field cloud picture of the PCB, the discriminator is used for distinguishing the temperature field cloud picture generated by the generator from an actual label cloud picture, the balance point of the final training is when the temperature cloud picture generated by the generator is consistent with the labels corresponding to the data sets, and the probability that the discriminator can distinguish the cloud picture generated by the generator is 50%.

7. The system of claim 5, wherein the input data of the proxy model is obtained by:

8. The system according to claim 5, further comprising a fine tuning module for migrating the data training weights of historical application scenarios to current application scenario-specific weights.

9. A computing device for a PCB temperature field in a heat sink, comprising:

at least one processor;

at least one memory for storing at least one program;

when executed by the at least one processor, cause the at least one processor to implement the method of any one of claims 1-4.

10. A computer readable storage medium in which a program executable by a processor is stored, wherein the program executable by the processor is adapted to perform the method according to any one of claims 1 to 4 when executed by the processor.