Multi-physics computation method and system for digital twin online simulation

Abstract

A multi-physics computation method and system for digital twin online simulation, relating to the technical field of physics simulation. A multi-physics coupling simulation computation model of a simulated object is established, and the multi-physics coupling simulation computation model is simplified, and the order of a temperature field simulation model is reduced, thereby greatly reducing the amount of computation, and improving computational efficiency; a low-precision dataset is obtained by means of temperature field analysis, so that the data size required for a model is reduced by means of low-precision data, thus reducing modeling costs; and a basic data-driven model is constructed by means of the low-precision dataset and a sample space corresponding to the low-precision dataset, so that a temperature field distribution result can be rapidly outputted, further improving computational efficiency and reducing computational errors.

Description

A multi-physical field calculation method and system for digital twin online simulation

[0001] The present application claims priority to the Chinese patent application No. 202411134601.1, filed on August 19, 2024, and entitled "A multi-physical field calculation method and system for digital twin online simulation", the whole content of which is incorporated herein by reference. TECHNICAL FIELD

[0002] The present application relates to the field of physical field simulation, in particular to a multi-physical field calculation method and system for digital twin online simulation. BACKGROUND

[0003] The digital twin technology of power transmission and transformation equipment is one of the focuses of the digital transformation of current power grid systems. It uses digital and information technology to build a digital model in a virtual space and interacts with sensor data to realize real-time updating of equipment state output, achieving the goal of "virtual and real coexistence" of equipment entities and twins, supporting equipment operation and maintenance and full life cycle state evaluation.

[0004] Due to the insulation safety problem of power equipment, sensors cannot directly measure physical quantities inside the equipment, so one of the technical routes for real-time and accurate characterization of equipment operating characteristics and comprehensive utilization of sensing data in current equipment digital twins is to combine with multi-physical field simulation technology. When the sensor obtains specific monitoring data from the equipment entity, multi-physical field simulation is applied to solve the inverse problem of the black box of multi-physical field with limited signals as boundary conditions to obtain the overall multi-physical field distribution of the equipment under the condition, realizing the visualization of the internal state of the equipment.

[0005] Although multi-physical field simulation technology has been widely used in the design and manufacturing stage, due to the complex structure of the equipment, the characteristics of finite element and finite volume meshing, a large amount of time is required for simulation, and the real-time or quasi-real-time performance is weak. Developing a real-time simulation calculation method is a problem to be solved at present.

[0006] Using data-driven surrogate models can quickly obtain the multi-physical field distribution of the equipment, but the current technology needs to enumerate different working conditions that the limited group of equipment may occur, and perform high-precision finite element calculation, which has the following problems: the more working conditions, the more accurate model trained and constructed subsequently, but the more working conditions calculated, the higher the calculation cost; using less data can reduce the calculation cost but increase the error; under the limited group of working conditions, calculating multiple conditions will consume a lot of time in finite element simulation, resulting in high calculation cost. When conditions beyond the enumeration boundary occur, a new model needs to be built, which increases the data and the time required for reconfiguration.

[0007] In summary, the current multi-physics simulation technology modeling has high time cost and low efficiency; the traditional proxy model method needs a large number of accurate multi-physics simulation calculation results as input, has high modeling cost, and may have large errors.

[0008] SUMMARY

[0009] The application provides a multi-physics field calculation method and system for digital twin online simulation, which solves the technical problems that the current multi-physics simulation technology modeling has high time cost and low efficiency, the traditional proxy model method needs a large number of accurate multi-physics simulation calculation results as input, has high modeling cost, and may have large errors.

[0010] Therefore, the first aspect of the application provides a multi-physics field calculation method for digital twin online simulation, comprising:

[0011] establishing a multi-physics field coupling simulation calculation model of a simulated object, and simplifying the multi-physics field coupling simulation calculation model;

[0012] determining a temperature field simulation model according to the simplified multi-physics field coupling simulation calculation model, reducing the order of the temperature field simulation model, and obtaining a temperature field simulation reduced-order model;

[0013] performing temperature field analysis according to the temperature field simulation reduced-order model based on a preset sample space, and obtaining a low-precision data set;

[0014] constructing a basic data-driven model according to the low-precision data set and the sample space corresponding to the low-precision data set, wherein the basic data-driven model is used to input the sample space into the basic data-driven model and output the low-precision data set corresponding to the sample space;

[0015] inputting the current sample space into the basic data-driven model, and outputting the low-precision data set corresponding to the current sample space.

[0016] Preferably, the step of establishing a multi-physics field coupling simulation calculation model of a simulated object and simplifying the multi-physics field coupling simulation calculation model comprises:

[0017] constructing a multi-physics field coupling simulation calculation model according to the geometric structure of the simulated object, wherein the physical field includes a temperature field, a stress field, and a fluid field that are coupled with each other;

[0018] simplifying the coupling grid division between the temperature field and the stress field, and equivalent the convective heat transfer process of the fluid field to a heat conduction process.

[0019] Preferably, the method further comprises:

[0020] The temperature field simulation model is reduced by using an intrinsic orthogonal decomposition method to obtain a temperature field simulation reduced model.

[0021] Preferably, the step of constructing a basic data-driven model according to the low-precision data set and the sample space corresponding to the low-precision data set, and inputting the sample space into the basic data-driven model to output a low-precision data set corresponding to the sample space, specifically comprises:

[0022] A first training data set is constructed according to the low-precision data set and the sample space corresponding to the low-precision data set.

[0023] An initial basic data-driven model is constructed, and the initial basic data-driven model is trained by using the first training data set, and after training convergence, a basic data-driven model is obtained.

[0024] Preferably, the method further comprises:

[0025] The basic data-driven model is corrected.

[0026] Preferably, the step of correcting the basic data-driven model specifically comprises:

[0027] A plurality of sample spaces and a plurality of tolerance factors corresponding to the sample spaces are determined according to the difference between the basic data-driven model and the high-precision data set corresponding to the low-precision data set.

[0028] A second training data set is constructed according to the plurality of sample spaces and the plurality of tolerance factors corresponding to the sample spaces.

[0029] The second training data set is trained based on a deep neural network to obtain a tolerance factor prediction regression model.

[0030] The basic data-driven model is corrected according to the tolerance factor prediction regression model to obtain a multi-precision proxy model.

[0031] Preferably, the method further comprises:

[0032] The output result of the multi-precision proxy model is tested for errors.

[0033] If the error test result of the output result of the multi-precision proxy model is qualified, the multi-precision proxy model is output.

[0034] If the error test result of the output result of the multi-precision surrogate model is unqualified, go to the step of temperature field analysis according to the temperature field simulation reduced-order model based on the preset sample space to obtain a low-precision data set, and iterate until the error test result of the output result of the multi-precision surrogate model is qualified, stop iteration and output the multi-precision surrogate model.

[0035] In a second aspect, the present application further provides a multi-physical field calculation system for digital twin online simulation, comprising:

[0036] A simulation modeling module is configured to establish a multi-physical field coupling simulation calculation model of a simulated object, and simplify the multi-physical field coupling simulation calculation model.

[0037] A model reduction module is configured to determine a temperature field simulation model according to the simplified multi-physical field coupling simulation calculation model, reduce the temperature field simulation model, and obtain a temperature field simulation reduced-order model.

[0038] A temperature field analysis module is configured to perform temperature field analysis according to the temperature field simulation reduced-order model based on a preset sample space, and obtain a low-precision data set.

[0039] A driving model construction module is configured to construct a basic data driving model according to the low-precision data set and a sample space corresponding to the low-precision data set, wherein the basic data driving model is configured to output a low-precision data set corresponding to the sample space after inputting the sample space into the basic data driving model.

[0040] A data driving module is configured to input a current sample space into the basic data driving model, and output a low-precision data set corresponding to the current sample space.

[0041] In a third aspect, the present application further provides an electronic device, comprising a memory and a processor.

[0042] The memory is configured to store a program.

[0043] The processor executes the program to implement the steps of the digital twin online simulation multi-physical field calculation method.

[0044] In a fourth aspect, the present application further provides a computer readable storage medium, wherein the storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of the digital twin online simulation multi-physical field calculation method.

[0045] As can be seen from the above technical solutions, the present application has the following advantages:

[0046] The application greatly reduces the amount of calculation and improves the calculation efficiency by establishing a multi-physics field coupling simulation calculation model of the simulated object, simplifying the multi-physics field coupling simulation calculation model, and reducing the order of the temperature field simulation model. The low-precision data set obtained through temperature field analysis is used to reduce the amount of data required by the model and reduce the cost of modeling. In addition, the basic data-driven model is constructed by using the low-precision data set and its corresponding sample space, so that the temperature field distribution result can be quickly output, further improving the calculation efficiency and reducing the calculation error. BRIEF DESCRIPTION OF DRAWINGS

[0047] Fig. 1 is a flowchart of a multi-physics field calculation method of digital twin online simulation provided by an embodiment of the application;

[0048] Fig. 2 is a flowchart of step S1 provided by an embodiment of the application;

[0049] Fig. 3 is a flowchart of correcting the basic data-driven model provided by an embodiment of the application;

[0050] Fig. 4 is a structural schematic diagram of a multi-physics field calculation system of digital twin online simulation provided by an embodiment of the application;

[0051] Fig. 5 is a structural schematic diagram of an electronic device provided by an embodiment of the application. DETAILED DESCRIPTION

[0052] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of the present application.

[0053] It should be noted that the terms "first", "second", and the like in the specification and claims of the present application and the above-described drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or a chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to the process, method, product or device.

[0054] The current multi-physical field simulation technology needs to enumerate different working conditions that a limited set of devices can occur, and perform high-precision finite element calculation, and the following problems exist: the more the working conditions, the more accurate the model trained and constructed subsequently, but the more the working conditions calculated, the higher the calculation cost; using less data has low calculation cost but large error; under a limited set of working conditions, calculating multiple conditions will consume a lot of time in finite element simulation itself, and the calculation cost is high. When conditions beyond the enumeration boundary occur, the model needs to be rebuilt, and the added data and the time required for the reconstruction of the configuration are relatively long.

[0055] In summary, the current multi-physical field simulation technology has high time cost, low efficiency and large calculation error in modeling.

[0056] Therefore, the embodiment of the present application provides a multi-physical field calculation method for digital twin online simulation, which can be applied to the case of multi-physical field calculation. The method can be executed by a multi-physical field calculation device, which can be realized in the form of hardware and / or software, and can be configured in a computer device.

[0057] Referring to FIG. 1, FIG. 1 illustrates a flow of a multi-physical field calculation method for digital twin online simulation provided by the embodiment of the present application.

[0058] The multi-physical field calculation method for digital twin online simulation provided by the embodiment of the present application comprises the following steps.

[0059] Step S1, a multi-physical field coupling simulation calculation model of a simulated object is established, and the multi-physical field coupling simulation calculation model is simplified.

[0060] In the field of power application, the simulated object can be a power device, such as a UHV converter valve side bushing.

[0061] Specifically, as shown in FIG. 2, step S1 specifically comprises the following steps.

[0062] Step S101, a multi-physical field coupling simulation calculation model is constructed according to the geometric structure of the simulated object, wherein the physical field includes a temperature field, a stress field and a fluid field that are coupled with each other.

[0063] Step S102, the coupling grid division between the temperature field and the stress field is simplified, and the convection heat transfer process of the fluid field is equivalent to a heat conduction process.

[0064] It should be noted that the stress characteristics of the elastic electrical connection structure will change after the temperature changes due to the influence of material thermal expansion. Considering this problem, increasing the grid division difficulty will lead to unnecessary increase in the number of grids and increase the amount of calculation. Therefore, the embodiment simplifies the coupling simulation process between the temperature field and the stress field by coupling grid division between the temperature field and the stress field, so as to not consider the influence of temperature field distribution on material thermal expansion, thereby reducing the difficulty of grid division and reducing the amount of calculation.

[0065] Wherein, the convection heat transfer process is not easy to obtain, the embodiment equivalent the convection heat transfer process of fluid field to heat conduction process, thereby reducing the amount of calculation.

[0066] Step S2, determine a temperature field simulation model according to the simplified multi-physical field coupling simulation calculation model, reduce the order of the temperature field simulation model, and obtain a temperature field simulation reduced-order model.

[0067] In some specific embodiments, the intrinsic orthogonal decomposition method is used to reduce the order of the temperature field simulation model to obtain the temperature field simulation reduced-order model.

[0068] Specifically, the temperature field simulation model can be characterized by a temperature field finite element discrete equation. The temperature field simulation reduced-order model can map the temperature field calculation nodes to a set of orthogonal basis eigenvectors to form a low-dimensional space, so as to further improve the calculation efficiency of the model. Through research, it is found that the single simulation calculation time of the temperature field simulation reduced-order model can be controlled within seconds.

[0069] Step S3, based on the preset sample space, performing temperature field analysis according to the temperature field simulation reduced-order model to obtain a low-precision data set.

[0070] Wherein, the sample space refers to the boundary conditions and loads in the multi-physical field simulation.

[0071] The sample space is obtained by optimized nested Latin hypercube. That is, first, the low-dimensional sample space value U L is obtained by sampling, and then the high-precision sample space U H is selected from the low-dimensional sample space.

[0072] After setting the sample space, the temperature field analysis distribution result, i.e. the grid node data, is obtained through temperature field analysis, and constitutes a low-precision data set. Wherein, a set of sample spaces corresponds to a set of grid node data.

[0073] Step S4, constructing a basic data driven model according to the low-precision data set and the sample space corresponding to the low-precision data set. The basic data driven model is used to input the sample space into the basic data driven model, and output the low-precision data set corresponding to the sample space.

[0074] The basic data-driven model can be constructed by using one of a neural network, a response surface method, and a Kriging method, respectively establishing the basic data-driven model by using the above methods, and selecting the method with the minimum error to construct the basic data-driven model through error testing on the basic data-driven model.

[0075] In one specific embodiment, step S4 comprises:

[0076] Step S401, constructing a first training data set according to the low-precision data set and the sample space corresponding to the low-precision data set.

[0077] Step S402, constructing an initial basic data-driven model, training the initial basic data-driven model through the first training data set, and obtaining the basic data-driven model after training convergence.

[0078] The basic data-driven model is constructed by taking the sample space as input and the low-precision data set corresponding to the sample space as output. That is, the basic data-driven model can quickly output the grid node data under the condition of inputting any simulation boundary condition.

[0079] Step S5, inputting the current sample space into the basic data-driven model to output the low-precision data set corresponding to the current sample space.

[0080] It should be noted that the embodiments of the present application greatly reduce the calculation amount and improve the calculation efficiency by establishing a multi-physical field coupling simulation calculation model of the simulated object, simplifying the multi-physical field coupling simulation calculation model, and reducing the order of the temperature field simulation model. The low-precision data set obtained through temperature field analysis reduces the data amount required by the model and reduces the modeling cost. The basic data-driven model is constructed by using the low-precision data set and the corresponding sample space, so that the temperature field distribution result can be quickly output, and the calculation efficiency is further improved and the calculation error is reduced.

[0081] In some specific embodiments, in order to further reduce the error of the basic data-driven model, the embodiments of the present application also modify the basic data-driven model.

[0082] Specifically, as shown in FIG. 3, the step of modifying the basic data-driven model comprises:

[0083] Step S601, determining a plurality of sample spaces and a plurality of tolerance factors corresponding to the plurality of sample spaces according to the difference between the basic data-driven model and the high-precision data set corresponding to the low-precision data set.

[0084] The process of obtaining the high-precision data set is as follows:

[0085] By refining the multi-physical field coupling simulation calculation model, taking the multi-physical field coupling simulation calculation model of the valve side bushing of the ultra-high voltage converter as an example, the elastic electrical connection structure is finely modeled according to the actual design to consider the influence of the current-carrying connection structure assembly on the compression amount and contact stress of the elastic electrical connection, and the relationship between the contact stress and the contact resistance is obtained according to the Holm electrical contact theory.

[0086] The contact resistance value is taken as a parameter condition, the overall loss distribution of the converter valve side bushing is obtained as a heat source through electromagnetic field calculation, and is substituted into the heat-fluid coupling calculation equation to obtain the temperature field distribution.

[0087] Due to the influence of material thermal expansion, the stress characteristics of the elastic electrical connection structure will change after the temperature changes, the contact resistance of the electrical connection structure is updated according to the calculated temperature field distribution, and the electrical conductivity of the metal material is considered to change with the temperature field in the electromagnetic field calculation, the electromagnetic field equation is solved again to obtain a new loss distribution, and the heat-fluid coupling calculation result is updated until the calculation result error of the two temperature field distributions is less than a set value, it can be judged that the convergence state is reached, at this time, the high-precision calculation result of the temperature field distribution of the converter valve side bushing is obtained, that is, the high-precision data set.

[0088] Since the process of obtaining the high-precision data set is relatively complex, it is generally measured in days, so the calculation time of the low-precision data is very fast compared with the high-precision data set, and the combination of high-precision and low-precision data can shorten the data set acquisition time, thereby reducing the modeling cost.

[0089] The tolerance factor is represented as:

[0090] δ(x i )=y i (x H )-f i (x base )

[0091] In the formula, x i represents a sample space, δ(x i ) represents a tolerance factor, y H (x H ) represents high-precision data, and f (x

[0001] ) represents a basic data-driven model, wherein x ∈U .

[0092] Step S602, constructing a second training data set according to a plurality of sample spaces and a plurality of tolerance factors corresponding to the plurality of sample spaces.

[0093] The mapping relationship between the sample space and the tolerance factor is obtained through step S601, and a second training data set is constructed based on the mapping relationship between the sample space and the tolerance factor.

[0094] In step S603, the second training data set is trained based on the deep neural network to obtain a tolerance factor prediction regression model.

[0095] In the training process, the sample space is input and the tolerance factor is output to train the deep neural network to obtain the tolerance factor prediction regression model.

[0096] In step S604, the base data-driven model is corrected according to the tolerance factor prediction regression model to obtain a multi-precision surrogate model.

[0097] The multi-precision surrogate model is:

[0098] In the formula, f H (x) is the output of the multi-precision surrogate model, is the tolerance factor prediction regression model.

[0099] In order to verify the accuracy of the output result of the multi-precision surrogate model, the embodiment further includes:

[0100] In step S701, the output result of the multi-precision surrogate model is tested for error.

[0101] In step S702, if the error test result of the output result of the multi-precision surrogate model is qualified, the multi-precision surrogate model is output.

[0102] In step S703, if the error test result of the output result of the multi-precision surrogate model is not qualified, the iteration is performed until the error test result of the output result of the multi-precision surrogate model is qualified, and the iteration is stopped and the multi-precision surrogate model is output.

[0103] For example, a set of data other than the constructed surrogate model is taken as a sample space a, a set of grid node data b is obtained through finite element calculation, and another set of grid node data c is obtained through the multi-precision surrogate model. The difference between the grid node data b and the grid node data c is calculated, and it is determined whether the difference is less than a preset error value. If the difference is less than the preset error value, it is determined that the error test result of the output result of the multi-precision surrogate model is qualified, otherwise, it is determined that the error test result of the output result of the multi-precision surrogate model is not qualified.

[0104] If it is determined that the error test result of the output result of the multi-precision surrogate model is unqualified, the network parameter optimization of the tolerance factor prediction regression model can also be performed. That is, according to the convergence index, the differential calculator is used to perform differential calculation on the temperature field distribution result obtained by the multi-precision surrogate model, and the error value of the corresponding simulation calculation model convergence index, that is, the physical information error, is obtained.

[0105] The tolerance factor is taken as the observation data error, the physical information error and the sample space are combined to construct a loss function, the weight and structure of the tolerance factor prediction regression model are analyzed and adjusted by using the loss function, so that the loss function is minimized or meets the convergence condition requirement, the tolerance factor considering the physical information constraint is obtained, and the multi-precision surrogate model with higher precision is obtained.

[0106] The above is a detailed description of an embodiment of the multi-physical field calculation method of digital twin online simulation provided by the present application, and the following is a detailed description of an embodiment of a multi-physical field calculation system of digital twin online simulation provided by the present application.

[0107] As shown in FIG. 4, FIG. 4 illustrates the structure of a multi-physical field calculation system of digital twin online simulation provided by an embodiment of the present application.

[0108] Some embodiments of the present application also provide a multi-physical field calculation system of digital twin online simulation, comprising:

[0109] The simulation modeling module 100 is used to establish a multi-physical field coupling simulation calculation model of the simulated object and simplify the multi-physical field coupling simulation calculation model.

[0110] Specifically, the multi-physical field coupling simulation calculation model is constructed according to the geometric structure of the simulated object, wherein the physical field includes a temperature field, a stress field and a fluid field that are coupled with each other.

[0111] The coupling grid division between the temperature field and the stress field is simplified, and the convection heat transfer process of the fluid field is equivalent to a heat conduction process.

[0112] It should be noted that due to the influence of material thermal expansion, the stress characteristics of the elastic electrical connection structure will change after the temperature changes, and considering this problem will increase the difficulty of grid division and cause unnecessary increase in the number of grids and increase in the calculation amount. Therefore, the embodiment simplifies the coupling simulation process between the temperature field and the stress field by dividing the coupling grid between the temperature field and the stress field, so as to not consider the influence of the temperature field distribution on the material thermal expansion, thereby reducing the difficulty of grid division and reducing the calculation amount.

[0113] Among them, the convection heat transfer process is not easy to obtain, and the embodiment equivalent to the heat conduction process of the fluid field, thereby reducing the calculation amount.

[0114] The model reduction module 200 is configured to determine a temperature field simulation model according to the simplified multi-physical field coupling simulation calculation model, reduce the temperature field simulation model, and obtain a temperature field simulation reduced model.

[0115] In some embodiments, the temperature field simulation model is reduced by using an intrinsic orthogonal decomposition method to obtain the temperature field simulation reduced model.

[0116] Specifically, the temperature field simulation model can be represented by a temperature field finite element discrete equation. The temperature field simulation reduced model can map the temperature field calculation nodes to a set of orthogonal basis eigenvectors to form a low-dimensional space, so as to further improve the calculation efficiency of the model. It is found through research that the single simulation calculation time of the temperature field simulation reduced model can be controlled within seconds.

[0117] The temperature field analysis module 300 is configured to perform temperature field analysis based on the preset sample space and the temperature field simulation reduced model, and obtain a low-precision data set.

[0118] The driving model construction module 400 is configured to construct a basic data driving model according to the low-precision data set and the sample space corresponding to the low-precision data set. The basic data driving model is configured to input the sample space into the basic data driving model, and output the low-precision data set corresponding to the sample space.

[0119] The basic data driving model can be constructed by using one of a neural network, a response surface method, and a Kriging method. The basic data driving model is constructed by using the above methods respectively, and error testing is performed on the basic data driving model. The method with the smallest error is selected to construct the basic data driving model.

[0120] The data driving module 500 is configured to input the current sample space into the basic data driving model, and output the low-precision data set corresponding to the current sample space.

[0121] As shown in FIG. 5, some embodiments of the present application further provide an electronic device 10, which includes a memory 20 and a processor 30.

[0122] The memory 20 is configured to store a program.

[0123] The processor 30 executes the program to implement the steps of the multi-physical field calculation method of digital twin online simulation in any of the above embodiments.

[0124] Some embodiments of the present application further provide a computer readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the multi-physical field calculation method of digital twin online simulation in any of the above embodiments are implemented.

[0125] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the system, the electronic device and the computer storage medium described above can refer to the corresponding process in the foregoing method embodiment, and will not be repeated here.

[0126] In several embodiments provided by the present application, it can be understood that each block in the flowchart or block diagram can represent a module, a program segment or a part of code, and the module, the program segment or the part of code contain one or more executable instructions for implementing a specified logic function. It should also be noted that, in some alternative implementation manners, the functions annotated in the blocks can also occur in an order different from that annotated in the drawings. For example, two continuous blocks can actually be executed substantially in parallel, and they can also be executed in reverse order in some cases, depending on the functions involved.

[0127] In several embodiments provided by the present application, it can be understood that the disclosed system, electronic device, computer storage medium and method can be implemented in other manners. For example, the above-described device embodiments are merely schematic, and the division of the units is merely a logical function division, and there can be another division manner in actual implementation. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed coupling or direct coupling or communication connection between the units can be indirect coupling or communication connection through some interface, device or unit, and can be electrical, mechanical or in other forms.

[0128] The units described as separated components can or can not be physically separated, and the components displayed as units can or can not be physical units, i.e., can be located in one place, or can be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purposes of the embodiments.

[0129] In addition, each functional unit in the embodiments of the present application can be integrated in one processing unit, or each unit can exist physically as a separate unit, or two or more units can be integrated in one unit. The integrated unit can be implemented in the form of hardware, or in the form of a software functional unit.

[0130] If the integrated unit is implemented in the form of a software function unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for executing all or part of the steps of the method described in various embodiments of the present application by a computer device (which can be a personal computer, a server, or a network device, etc.). The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (English full name: Read-Only Memory, English abbreviation: ROM), a random access memory (English full name: Random Access Memory, English abbreviation: RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0131] The above embodiments are only used to illustrate the technical solutions of the present application, not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A multi-physics field computing method of digital twin online simulation, characterized in that, The method comprises the following steps: establishing a multi-physical field coupling simulation calculation model of a simulated object, and simplifying the multi-physical field coupling simulation calculation model; determining a temperature field simulation model according to the simplified multi-physical field coupling simulation calculation model, and reducing the order of the temperature field simulation model to obtain a temperature field simulation reduced-order model; performing temperature field analysis based on a preset sample space and according to the temperature field simulation reduced-order model to obtain a low-precision data set; constructing a basic data-driven model according to the low-precision data set and a sample space corresponding to the low-precision data set, wherein the basic data-driven model is used to output a low-precision data set corresponding to a sample space after the sample space is input into the basic data-driven model; inputting a current sample space into the basic data-driven model to output a low-precision data set corresponding to the current sample space.

2. The multi-physics field calculation method of digital twin online simulation according to claim 1, characterized in that, The step of establishing a multi-physical field coupling simulation calculation model of a simulated object and simplifying the multi-physical field coupling simulation calculation model comprises the following steps: constructing a multi-physical field coupling simulation calculation model according to the geometric structure of the simulated object, wherein the physical fields include a temperature field, a stress field and a fluid field that are coupled with each other; simplifying the coupling grid division between the temperature field and the stress field, and equivalently converting the heat transfer process of the fluid field into a heat conduction process.

3. The multi-physics computing method of digital twin online simulation according to claim 1, wherein, The method further comprises the following steps: reducing the order of the temperature field simulation model by using an intrinsic orthogonal decomposition method to obtain a temperature field simulation reduced-order model.

4. The multi-physics computing method of digital twin online simulation according to claim 1, wherein, The step of constructing a basic data-driven model according to the low-precision data set and a sample space corresponding to the low-precision data set, wherein the basic data-driven model is used to output a low-precision data set corresponding to a sample space after the sample space is input into the basic data-driven model, comprises the following steps: constructing a first training data set according to the low-precision data set and a sample space corresponding to the low-precision data set; constructing an initial basic data-driven model, training the initial basic data-driven model by using the first training data set, and obtaining a basic data-driven model after the training converges.

5. The multi-physics computing method of digital twin online simulation according to claim 1, wherein, The method further comprises the following step: modifying the basic data-driven model.

6. The multi-physics computing method of digital twin online simulation according to claim 5, wherein, The step of modifying the basic data-driven model comprises the following steps: determining a plurality of sample spaces and a plurality of tolerance factors corresponding to the sample spaces according to the difference between the basic data-driven model and a high-precision data set corresponding to the low-precision data set; constructing a second training data set according to the plurality of sample spaces and the plurality of tolerance factors corresponding to the sample spaces; training the second training data set based on a deep neural network to obtain a tolerance factor prediction regression model; modifying the basic data-driven model according to the tolerance factor prediction regression model to obtain a multi-precision surrogate model.

7. The multi-physics computational method of digital twin online simulation according to claim 6, wherein, The method further comprises the following steps: performing error testing on the output result of the multi-precision surrogate model; if the error testing result of the output result of the multi-precision surrogate model is qualified, outputting the multi-precision surrogate model. If the error test result of the output result of the multi-precision surrogate model is unqualified, go to the step of temperature field analysis based on the preset sample space according to the temperature field simulation reduced-order model to perform iteration until the error test result of the output result of the multi-precision surrogate model is qualified, stop iteration and output the multi-precision surrogate model.

8. A multi-physics computing system for digital twin online simulation, characterized in that, Comprise: The simulation modeling module is used for establishing a multi-physics field coupling simulation calculation model of a simulated object, and simplifying the multi-physics field coupling simulation calculation model; The model reduction module is used for determining a temperature field simulation model according to the simplified multi-physics field coupling simulation calculation model, reducing the temperature field simulation model, and obtaining a temperature field simulation reduced-order model; The temperature field analysis module is used for performing temperature field analysis based on a preset sample space according to the temperature field simulation reduced-order model to obtain a low-precision data set; The driving model construction module is used for constructing a basic data driving model according to the low-precision data set and a sample space corresponding to the low-precision data set, the basic data driving model being used for inputting the sample space into the basic data driving model to output a low-precision data set corresponding to the sample space; The data driving module is used for inputting a current sample space into the basic data driving model to output a low-precision data set corresponding to the current sample space.

9. An electronic device, comprising: The electronic device comprises a memory and a processor; The memory is used for storing a program; The processor executes the program to realize the steps of the multi-physics field calculation method of digital twin online simulation according to any one of claims 1 to 7.

10. A computer readable storage medium, the storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to realize the steps of the multi-physics field calculation method of digital twin online simulation according to any one of claims 1 to 7.

Images