A simulation effect prediction method, device, electronic device and storage medium

Abstract

The present application provides a simulation effect prediction method, device, electronic device and storage medium. By obtaining the values ​​of modeling parameters and inputting the values ​​of modeling parameters into a trained simulation effect prediction model, the target parameter simulation calculation error and simulation calculation time of the vehicle finite element simulation model under the values ​​of the modeling parameters are obtained, so that designers can weigh the simulation accuracy and efficiency of the vehicle finite element simulation model, and can provide a basis for establishing the optimal finite element model.

Description

Technical Field

[0001] The present application relates to the field of finite element simulation technology, and in particular to a simulation effect prediction method, device, electronic device and storage medium. Background Art

[0002] The finite element method is a very effective numerical analysis method for solving differential equations. Its basic idea is to use piecewise functions to approximate the original function, that is, to transform the infinite degree of freedom problem into a finite degree of freedom problem, and then solve a set of linear equations to obtain an approximate solution to the original equation. In finite element analysis, model establishment is the key to the finite element analysis process, which is related to the accuracy of the calculation results and the efficiency of the simulation. In actual engineering, the optimization design time takes up about 40% of the entire project process. In order to obtain an optimal solution, it is often necessary to verify several to dozens of solutions.

[0003] Different application fields have different requirements for simulation accuracy and efficiency. When the project requires high accuracy, it often takes a lot of computing and analysis time. When the project only focuses on the trend of results, the requirements for simulation accuracy will be reduced. Therefore, it is necessary to weigh the simulation accuracy and efficiency to select different analysis solutions under different requirements. Summary of the invention

[0004] The purpose of the embodiments of the present application is to provide a simulation effect prediction method, device, electronic device and storage medium to solve the above-mentioned technical problems.

[0005] On the one hand, a simulation effect prediction method is provided, comprising:

[0006] Obtaining values ​​of modeling parameters; the modeling parameters are parameters used to establish a vehicle finite element simulation model;

[0007] The values ​​of the modeling parameters are input into the trained simulation effect prediction model to obtain the target parameter simulation calculation error and simulation calculation time of the vehicle finite element simulation model under the values ​​of the modeling parameters.

[0008] In one embodiment, before inputting the values ​​of the modeling parameters into the trained simulation effect prediction model, the method includes:

[0009] Acquire a simulation effect training data set; the simulation effect training data set includes sample data of multiple value combinations of the modeling parameters, sample data of simulation calculation errors of target parameters of the vehicle finite element simulation model under each of the sample data of the value combinations, and sample data of simulation calculation time;

[0010] Training is performed based on the simulation effect training data set and a preset initial model to obtain the simulation effect prediction model.

[0011] In one embodiment, the step of obtaining a simulation effect training data set includes:

[0012] Obtaining the recommended value range of each modeling parameter;

[0013] Based on the recommended value range of each of the modeling parameters, a plurality of recommended values ​​of each of the modeling parameters are obtained;

[0014] Combining the recommended values ​​of the modeling parameters to obtain a plurality of value combination sample data of the modeling parameters;

[0015] Establishing a corresponding vehicle finite element simulation test model based on each of the value combination sample data;

[0016] Based on the target parameter simulation calculation value output by each of the vehicle finite element simulation test models and the corresponding target parameter expected value, determine the target parameter simulation calculation error sample data corresponding to the vehicle finite element simulation test model, and use the actual calculation time of the vehicle finite element simulation test model calculating the target parameter simulation calculation value as the corresponding simulation calculation time sample data;

[0017] The set of each of the value combination sample data, the target parameter simulation calculation error sample data corresponding to each of the value combination sample data, and the simulation calculation time sample data is used as a simulation effect training data set.

[0018] In one embodiment, the obtaining of multiple recommended values ​​of each modeling parameter based on the recommended value range of each modeling parameter includes:

[0019] Divide each of the modeling parameters into N equal parts within its corresponding recommended value interval to obtain a recommended value for each of the modeling parameters;

[0020] Combining the recommended values ​​of the modeling parameters to obtain a plurality of value combination sample data of the modeling parameters includes:

[0021] The recommended values ​​of the modeling parameters are combined by an optimal Latin square sampling method to obtain sample data of multiple value combinations of the modeling parameters.

[0022] In one embodiment, the initial model is a neural network model, and the training is performed based on the simulation effect training data set and the preset initial model to obtain the simulation effect prediction model, including:

[0023] In the process of training based on the simulation effect training data set and the preset initial model, determining whether a model training termination condition is met;

[0024] If so, stop model training and obtain the simulation effect prediction model;

[0025] If not, the neuron connection weights and / or neuron thresholds in the neural network model are updated, and model training is performed based on the updated neuron connection weights and / or neuron thresholds.

[0026] In one embodiment, the updating of the neuron connection weights and / or neuron thresholds in the neural network model includes:

[0027] The neuron connection weights and / or neuron thresholds in the neural network model are updated based on a genetic algorithm.

[0028] In one embodiment, the updating of the neuron connection weights and / or neuron thresholds in the neural network model based on a genetic algorithm includes:

[0029] Acquire a current population; the current population is composed of a plurality of combined coding data, each of which includes coding data obtained by encoding the previous generation of neuron connection weights and coding data obtained by encoding the previous generation of neuron thresholds;

[0030] Performing genetic processing on the current population to obtain a new generation population;

[0031] Calculating the fitness values ​​corresponding to each of the combined encoding data in the new generation population;

[0032] Determining an optimal combination of neuron connection weights and neuron thresholds based on the fitness value and the new generation population;

[0033] The neuron connection weights and neuron thresholds in the neural network model are updated based on the optimal neuron connection weights and neuron thresholds combination.

[0034] On the other hand, a simulation effect prediction device is provided, comprising:

[0035] An acquisition module, used to acquire the value of a modeling parameter; the modeling parameter is a parameter used to establish a vehicle finite element simulation model;

[0036] The prediction module is used to input the values ​​of the modeling parameters into the trained simulation effect prediction model to obtain the target parameter simulation calculation error and simulation calculation time of the vehicle finite element simulation model under the values ​​of the modeling parameters.

[0037] On the other hand, an electronic device is provided, including a processor and a memory, wherein a computer program is stored in the memory, and the processor executes the computer program to implement any one of the above methods.

[0038] On the other hand, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by at least one processor, any one of the above methods is implemented.

[0039] The simulation effect prediction method, device, electronic device and storage medium provided in the present application obtain the values ​​of modeling parameters, input the values ​​of modeling parameters into the trained simulation effect prediction model, and obtain the target parameter simulation calculation error and simulation calculation time of the vehicle finite element simulation model under the values ​​of the modeling parameters, so as to facilitate designers to balance the simulation accuracy and efficiency of the vehicle finite element simulation model, and provide a basis for establishing the optimal finite element model. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 A schematic diagram of a flow chart of a simulation effect prediction method provided in Example 1 of the present application;

[0041] Figure 2 A schematic diagram of a process for obtaining a simulation effect prediction model provided in Example 1 of the present application;

[0042] Figure 3 A comparison chart of prediction results provided in Example 1 of the present application;

[0043] Figure 4 A schematic diagram of the relative error between the actual value and the predicted value provided in Example 1 of the present application;

[0044] Figure 5 A schematic diagram of the structure of a neural network model provided in Example 1 of the present application;

[0045] Figure 6 A flowchart of a simulation effect prediction method provided in Example 1 of the present application;

[0046] Figure 7 A schematic diagram of the structure of a simulation effect prediction device provided in Example 2 of the present application;

[0047] Figure 8 This is a schematic diagram of the structure of an electronic device provided in Example 3 of the present application. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0049] Embodiment 1:

[0050] Finite element analysis is a numerical calculation method that approximates the physical behavior of an entity by discretizing a continuous physical structure into a finite number of small parts, namely "elements". When designing a vehicle, the finite element analysis method is introduced into the vehicle design in order to obtain the vehicle design parameters that meet the performance requirements. The vehicle finite element simulation model can be used to simulate the corresponding performance of the vehicle when the corresponding design parameters are selected for the relevant components of the vehicle.

[0051] Specifically, the vehicle finite element simulation model can simulate the values ​​of the target parameters of the vehicle under the corresponding design parameters. Designers can evaluate the performance of the vehicle under the design parameters based on the values ​​of the target parameters, thereby determining the design parameters that can achieve the best performance requirements.

[0052] The target parameter in the embodiment of the present application refers to the parameter used to evaluate the target to be optimized when optimizing the target to be optimized of the vehicle. It should be noted that in different application scenarios, the target parameters may also be different depending on the optimization target. For example, when it is necessary to optimize the NVH performance of the vehicle, the target parameter can be the modal frequency, and the NVH performance of the vehicle can be evaluated by establishing a vehicle finite element simulation model to simulate the value of the modal frequency of the vehicle under the corresponding design parameters.

[0053] It is understandable that the vehicle finite element simulation model can be established based on Hypermesh software, and of course, it can also be established based on other software. In different application scenarios, designers have different requirements for the simulation accuracy and efficiency of the vehicle finite element simulation model, so it is necessary to evaluate the simulation accuracy and efficiency of the vehicle finite element simulation model so that designers can select the best vehicle finite element simulation model to evaluate the target parameters of the vehicle according to the application scenario.

[0054] In view of this, the present application embodiment provides a simulation effect prediction method, see Figure 1 As shown, the following steps are included:

[0055] S11: Obtaining values ​​of modeling parameters, where the modeling parameters are parameters used to establish a vehicle finite element simulation model.

[0056] S12: Inputting the values ​​of the modeling parameters into the trained simulation effect prediction model to obtain the target parameter simulation calculation error and simulation calculation time of the vehicle finite element simulation model under the values ​​of the modeling parameters.

[0057] The following is a detailed introduction to the contents of the above steps.

[0058] The modeling parameters in the embodiment of the present application are parameters used to establish a vehicle finite element simulation model and have an impact on the simulation calculation error and simulation calculation time of the simulation target parameters of the vehicle finite element simulation model.

[0059] Specifically, the modeling parameters in the embodiments of the present application include but are not limited to at least one of mesh quality control parameters, material property parameters, and pipe boundary condition parameters.

[0060] The mesh quality control parameters in the embodiments of the present application include but are not limited to at least one of mesh size, aspect ratio, warping, maximum inner angle, minimum inner angle, surface distortion, Jacobian, taper, and triangular unit ratio.

[0061] The material property parameters in the embodiments of the present application include but are not limited to at least one of Young's modulus and Poisson's ratio. Among them, the Young's modulus of carbon steel is generally 200GPa to 220GPa, the Young's modulus of low alloy steel is generally 200GPa, the Young's modulus of high alloy steel is generally 210GPa, and the Young's modulus of ductile iron is generally 150GPa to 180GPa. The Poisson's ratio of carbon steel is generally 0.24 to 0.28, the Poisson's ratio of alloy steel is generally 0.25 to 0.30, and the Poisson's ratio of cast steel is generally 0.25 to 0.29.

[0062] The pipe boundary condition parameters in the embodiments of the present application include but are not limited to the dynamic stiffness in various directions of the installation point, for example, the dynamic stiffness in the X, Y, Z, RX, RY, and RZ directions. During use, the rubber bushing has a wide range of deformation, with a maximum deformation of up to 100%. Therefore, the bushing stiffness value range is defined as 20% to 200% of its original stiffness value.

[0063] In some embodiments, the modeling parameters may also include unit type, for example, determining whether the unit type is shell unit or solid unit based on the ratio of part thickness to length; and may also include analysis frequency and sampling frequency, wherein the sampling frequency is at least twice the upper limit of the reference frequency.

[0064] See also Figure 2 As shown, before step S12, the following sub-steps may also be included:

[0065] S121: Acquire a simulation effect training data set; the simulation effect training data set includes sample data of multiple value combinations of the modeling parameters, sample data of simulation calculation errors of target parameters of the vehicle finite element simulation model under each of the value combination sample data, and sample data of simulation calculation time;

[0066] S122: Perform training based on the simulation effect training data set and a preset initial model to obtain the simulation effect prediction model.

[0067] Specifically, step S121 may include the following sub-steps:

[0068] Sub-step 1: obtaining the recommended value range of each modeling parameter;

[0069] Sub-step 2: based on the recommended value range of each modeling parameter, obtaining a plurality of recommended values ​​of each modeling parameter;

[0070] Sub-step three: combining the recommended values ​​of the modeling parameters to obtain sample data of multiple value combinations of the modeling parameters;

[0071] Sub-step 4: establishing a corresponding vehicle finite element simulation test model based on each of the value combination sample data;

[0072] Sub-step 5: based on the target parameter simulation calculation value output by each of the vehicle finite element simulation test models and the corresponding target parameter expected value, determine the target parameter simulation calculation error sample data corresponding to the vehicle finite element simulation test model, and use the actual calculation time of the vehicle finite element simulation test model calculating the target parameter simulation calculation value as the corresponding simulation calculation time sample data;

[0073] Sub-step six: taking the set of each of the value combination sample data, the target parameter simulation calculation error sample data corresponding to each of the value combination sample data, and the simulation calculation time sample data as a simulation effect training data set.

[0074] Please refer to Table 1 below, which shows a recommended value range for each modeling parameter:

[0075] Table 1

[0076] Mesh Quality Recommended value range Grid size / mm 2-10 Aspect Ratio 2-8 Warpage / ° 2-8 Surface distortion / ° 20-60 Jacobi 0.3-0.9 Taper / ° 0.3-0.9 Triangular unit ratio / % 5-30 Minimum interior angle of a triangle / ° 10-40 Maximum interior angle of a triangle / ° 100-160 Minimum interior angle of a quadrilateral / ° 20-80 Maximum interior angle of a quadrilateral / ° 100-160

[0077] It is understandable that the recommended value range of each modeling parameter can be determined by the designer based on experience.

[0078] In the above sub-step 2, each of the modeling parameters may be divided into N equal parts within its corresponding recommended value interval to obtain a recommended value of each of the modeling parameters.

[0079] The larger N is, the more data sets there are, and the more accurate the simulation prediction model is. However, the amount of calculation will be larger. For the convenience of calculation, taking N as 51 as an example, the step size is (maximum value - minimum value) / 50. The recommended values ​​of the modeling parameters are shown in Table 2 below:

[0080] Table 2

[0081]

[0082] In the above sub-step three, the recommended values ​​of the modeling parameters can be combined by using an optimal Latin square sampling method to obtain sample data of multiple value combinations of the modeling parameters.

[0083] The Optimal Latin Hypercube Sampling (OLHS) method is a statistical method used in experimental design and simulation research, especially when sampling a multidimensional input space. Its purpose is to evenly distribute sample points within a given value range, thereby improving the accuracy and efficiency of the simulation or prediction model. Randomly select N points to ensure that each level of each factor is studied only once. Calculate the corresponding calculation error and calculation time, that is, determine the target parameter simulation calculation error sample data and simulation calculation time sample data corresponding to the vehicle finite element simulation test model. The embodiment of the present application takes the target parameter as the modal frequency as an example, and the results of the simulation calculation are shown in Table 3 below.

[0084] Table 3

[0085]

[0086] When calculating the error of the corresponding modal frequency using the established vehicle finite element simulation test model, the simulation calculation error of the modal frequency can be calculated according to the following formula:

[0087]

[0088] Where q is the modal order; A i The modal frequencies calculated from the vehicle finite element simulation test model; is the reference modal frequency.

[0089] Each column in Table 3 is a set of simulation effect training data. 80% of the data in Table 3 is used as the training set and 20% as the test set. Then, the true values ​​of 10 test samples in the test set are compared with the predicted values. Only efficiency is taken as an example. The results are as follows: Figure 3 As shown in the figure, the relative error between the predicted value and the true value of the sample is as follows Figure 4 shown.

[0090] The preset initial model in the embodiment of the present application may be a neural network model, and of course may also be other models. Specifically, step S122 may include the following sub-steps:

[0091] Sub-step 1: In the process of training based on the simulation effect training data set and the preset initial model, determine whether the model training termination condition is met; if so, go to the following sub-step 2; if not, go to the following sub-step 3;

[0092] Sub-step 2: stop model training and obtain the simulation effect prediction model;

[0093] Sub-step three: updating the neuron connection weights and / or neuron thresholds in the neural network model, and performing model training based on the updated neuron connection weights and / or neuron thresholds.

[0094] The neural network model in the embodiment of the present application can be a BP neural network model, and its model structure is as follows: Figure 5 As shown, it includes input layer, hidden layer and output layer.

[0095] In the embodiment of the present application, the number of neurons in the hidden layer can be determined based on the following formula:

[0096]

[0097] In the formula, n is the number of neurons in the input layer, m is the number of neurons in the output layer, and a is a constant between 0 and 10.

[0098] The calculation formula between the input layer neurons and the hidden layer neurons can be:

[0099] b μ =f(∑ u ω mu a m +k u );

[0100] The calculation formula between the hidden layer neurons and the output layer neurons can be:

[0101] c n =f(∑ u v nu b u +p n );

[0102] Among them, a m is the input layer neuron, b u is the hidden layer neuron, ω mu is the connection weight between the mth input layer neuron and the uth hidden layer neuron, f is the activation function, k u is the hidden layer neuron threshold, c n is the output layer neuron, v nu is the connection weight between the nth input layer neuron and the uth hidden layer neuron, p n is the output layer neuron threshold.

[0103] The model training termination condition in the embodiment of the present application can be flexibly set by the developer. For example, the actual output value of the model can be compared with the expected output value each time. If the mean square error does not meet the preset requirements, back propagation is performed, and the mean square error can be returned to the neurons assigned to each layer in the form of a gradient to update the neuron connection weights and / or the neuron thresholds. In some other embodiments, the model training can also be stopped when the number of iterations of the model training meets the preset requirements.

[0104] It should also be noted that before model training, the data in the simulation effect training data set can be preprocessed. Specifically, the data in the simulation effect training data set can be normalized.

[0105] Normalization can effectively avoid the influence of dimensions and units, prevent small values ​​in the data from being swallowed up, and improve the convergence speed of the model. The formula is as follows:

[0106]

[0107] X and Y are the values ​​before and after the conversion, MAX and MIN are the maximum and minimum values ​​of the samples, and the final result interval falls in the interval [0,1].

[0108] In the above sub-step three, the neuron connection weights and / or neuron thresholds in the neural network model may be updated based on a genetic algorithm, and of course, the neuron connection weights and / or neuron thresholds in the neural network model may also be updated based on other algorithms.

[0109] Genetic algorithm is based on the theory of biological evolution. It changes the search process through selection, crossover and mutation to find the optimal solution. It simulates the natural selection process of biological evolution in nature. Based on the fitness function value of individuals (similar to survival ability), individuals with better adaptation to the environment are selected as parents for crossover reproduction. The algorithm decodes the problem to be optimized into the form of "chromosomes", imitating the DNA molecular structure that carries genetic information in organisms. The genes on the chromosomes correspond to the traits of the organisms.

[0110] The following is a specific introduction using the example of updating the neuron connection weights and neuron thresholds in a neural network model based on a genetic algorithm.

[0111] Specifically, it includes the following sub-steps:

[0112] Sub-step 1: obtaining a current population; the current population is composed of a plurality of combined coding data, each of which includes coding data obtained by encoding the previous generation of neuron connection weights and coding data obtained by encoding the previous generation of neuron thresholds;

[0113] Sub-step 2: performing genetic processing on the current population to obtain a new generation population;

[0114] Sub-step three: calculating the fitness values ​​corresponding to each of the combined coding data in the new generation population;

[0115] Sub-step 4: determining an optimal combination of neuron connection weights and neuron thresholds based on the fitness value and the new generation population;

[0116] Sub-step five: updating the neuron connection weights and neuron thresholds in the neural network model based on the optimal combination of neuron connection weights and neuron thresholds.

[0117] The encoding in the above sub-step 1 refers to encoding possible solutions to the problem to be optimized (such as neuron connection weights and neuron thresholds) into "chromosome" representation using data structures such as binary strings and real number strings.

[0118] It is understandable that when the genetic algorithm is executed for the first time, an initial population needs to be generated. At this time, the initial population is also the current population.

[0119] In the above sub-step 2, reproductive reproduction, genetic manipulation, and natural selection may be performed based on the current population. Specifically, selection operations, crossover operations, and mutation operations may be performed based on the current population.

[0120] Among them, the selection operation refers to the process of selecting excellent individuals with strong survival ability in the population. Whether the chromosome can be inherited is determined according to the fitness value of the individual. The greater the proportion of the fitness of the chromosome in the population, the greater the possibility of inheritance.

[0121] After the selection operation, some excellent individuals will be saved, and the selected parent individuals will be cross-operated to combine beneficial genes in the same individual, thereby enhancing the global search ability of the genetic algorithm. The cross-operation is as follows:

[0122] a qj =a qj (1-b)+a lj b

[0123] a lj =a lj (1-b)+a qj b

[0124] In the formula, a qj and a lj Sample a q and a l Crossover operation at j; b is a random number in [0,1].

[0125] Performing mutation operations on new individuals can introduce new genetic changes. By randomly changing the genes of individuals, the diversity of the population can be increased. Non-uniform mutation methods are more common in actual operations. The mutation operation of the jth gene of the i-th individual is as follows:

[0126]

[0127] In the formula, a max is the maximum value of the gene, a min is the minimum value of the gene, μ is a random number between [0,1], g is the number of iterations, G max is the maximum number of evolutions.

[0128] After selection, crossover and mutation operations, a new generation of population can be generated.

[0129] The above process is repeated until the preset stopping condition is reached, such as reaching the maximum number of iterations or converging to the optimal solution. The genetic algorithm terminates and returns the optimal solution found, that is, the optimal combination of neuron connection weights and neuron thresholds.

[0130] In the embodiment of the present application, the fitness values ​​corresponding to each combination of coded data in the new generation population can be calculated based on the following fitness function:

[0131]

[0132] Among them, F(x) represents the fitness function, w i Represents the neuron connection weight corresponding to the i-th modeling parameter, Accuracy i (x) represents the contribution of the i-th modeling parameter to the simulation accuracy, AcComputationalTime i (x) represents the computational time consumption of the i-th modeling parameter, δ i represents the neuron threshold corresponding to the i-th modeling parameter, Penalty i (x) indicates that when the i-th modeling parameter is changed, the accuracy loss exceeds the threshold δ i The penalty item generated when .

[0133] The flowchart of a simulation effect prediction method provided in the embodiment of the present application can be found in Figure 6 As shown, Figure 6 The left part of the figure is the execution process of the genetic algorithm. Figure 6 The right part of the figure is the execution process for training based on the neural network model. By combining and executing these two sets of processes, the final simulation effect prediction model can be obtained, so that the simulation error and efficiency can be predicted based on the simulation effect prediction model.

[0134] In the embodiment of the present application, the weights and thresholds of the neural network model are optimized by using a genetic algorithm to solve the problem of being trapped in a local optimal solution. At the same time, the training speed of the simulation effect prediction model can be accelerated, the adaptability and convergence of the model can be enhanced, and thus the recognition accuracy can be improved.

[0135] It should be understood that, although the various steps in the above-mentioned flow chart are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the above-mentioned flow chart may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0136] Embodiment 2:

[0137] See also Figure 7 As shown, the embodiment of the present application also provides a simulation effect prediction device, including:

[0138] The acquisition module 701 is used to acquire the value of the modeling parameter; the modeling parameter is a parameter used to establish a vehicle finite element simulation model;

[0139] The prediction module 702 is used to input the values ​​of the modeling parameters into the trained simulation effect prediction model to obtain the target parameter simulation calculation error and simulation calculation time of the vehicle finite element simulation model under the values ​​of the modeling parameters.

[0140] It should be noted that, for the sake of brevity, the contents described in the above embodiments will not be repeated in this embodiment.

[0141] Embodiment three:

[0142] Based on the same inventive concept, the present application embodiment provides an electronic device, see Figure 8 As shown, it includes a processor 801 and a memory 802. The memory 802 stores a computer program, the processor 801 and the memory 802 communicate via a communication bus, and the processor 801 executes the computer program to implement the steps of the method described above, which will not be repeated here.

[0143] It should be noted that the electronic device in the embodiment of the present application can be a terminal or a server. Among them, the terminal can be a smart phone, a tablet computer, a laptop, a personal computer (PC), a smart home, a wearable electronic device, a VR / AR device, a car computer, etc. The server can be an intercommunication server or a background server between multiple heterogeneous systems, or an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, and basic cloud computing services such as big data and artificial intelligence platforms, etc.

[0144] The processor 801 may be an integrated circuit chip with signal processing capabilities. The processor 801 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. It may implement or execute various methods, steps and logic block diagrams disclosed in the embodiments of the present application. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0145] The memory 802 may include, but is not limited to, random access memory (RAM), read-only memory (ROM), programmable read-only memory (PROM), erasable read-only memory (EPROM), electrically erasable read-only memory (EEPROM), etc.

[0146] This embodiment also provides a computer-readable storage medium, such as a floppy disk, a CD, a hard disk, a flash memory, a USB flash disk, an SD card, an MMC card, etc., in which one or more programs for implementing the above steps are stored. These one or more programs can be executed by one or more processors to implement the steps of the method in the above embodiment 1, which will not be repeated here.

[0147] It should be noted that the diagram provided in the present embodiment only illustrates the basic concept of the present invention in a schematic manner, so the diagram only shows the components related to the present invention rather than drawing according to the number, shape and size of the components during actual implementation. The type, quantity and ratio of each component during actual implementation can be a random change, and the component layout type may also be more complicated. The structure, ratio, size, etc. illustrated in the drawings of the present specification are only used to match the content disclosed in the specification for people familiar with this technology to understand and read, and are not used to limit the limiting conditions that the present invention can implement, so they have no technical substantive significance. Any modification of the structure, change of the proportional relationship or adjustment of the size should still fall within the scope of the technical content disclosed by the present invention without affecting the effect that the present invention can produce and the purpose that can be achieved. At the same time, the terms such as "upper", "lower", "left", "right", "middle" and "one" quoted in this specification are only for the convenience of narration, and are not used to limit the scope of the present invention. The change or adjustment of its relative relationship should also be regarded as the scope of the present invention without substantially changing the technical content.

[0148] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0149] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A simulation effect prediction method, characterized in that: include: Obtaining values ​​of modeling parameters; the modeling parameters are parameters used to establish a vehicle finite element simulation model; The values ​​of the modeling parameters are input into the trained finite element simulation effect prediction model to obtain the target parameter simulation calculation error and simulation calculation time of the vehicle finite element simulation model under the values ​​of the modeling parameters.

2. The simulation effect prediction method according to claim 1, characterized in that: Before inputting the values ​​of the modeling parameters into the trained finite element simulation effect prediction model, the method includes: Acquire a simulation effect training data set; the simulation effect training data set includes sample data of multiple value combinations of the modeling parameters, sample data of simulation calculation errors of target parameters of the vehicle finite element simulation model under each of the sample data of the value combinations, and sample data of simulation calculation time; Training is performed based on the simulation effect training data set and a preset initial model to obtain the simulation effect prediction model.

3. The simulation effect prediction method according to claim 2, characterized in that: The step of obtaining a simulation effect training data set includes: Obtaining the recommended value range of each modeling parameter; Based on the recommended value range of each of the modeling parameters, a plurality of recommended values ​​of each of the modeling parameters are obtained; Combining the recommended values ​​of the modeling parameters to obtain a plurality of value combination sample data of the modeling parameters; Establishing a corresponding vehicle finite element simulation test model based on each of the value combination sample data; Based on the target parameter simulation calculation value output by each of the vehicle finite element simulation test models and the corresponding target parameter expected value, determine the target parameter simulation calculation error sample data corresponding to the vehicle finite element simulation test model, and use the actual calculation time of the vehicle finite element simulation test model calculating the target parameter simulation calculation value as the corresponding simulation calculation time sample data; The set of each of the value combination sample data, the target parameter simulation calculation error sample data corresponding to each of the value combination sample data, and the simulation calculation time sample data is used as a simulation effect training data set.

4. The simulation effect prediction method according to claim 3, characterized in that: The obtaining of a plurality of recommended values ​​of each of the modeling parameters based on the recommended value interval of each of the modeling parameters comprises: Divide each of the modeling parameters into N equal parts within its corresponding recommended value interval to obtain a recommended value for each of the modeling parameters; Combining the recommended values ​​of the modeling parameters to obtain a plurality of value combination sample data of the modeling parameters includes: The recommended values ​​of the modeling parameters are combined by an optimal Latin square sampling method to obtain sample data of multiple value combinations of the modeling parameters.

5. The simulation effect prediction method according to claim 2, characterized in that: The initial model is a neural network model, and the simulation effect prediction model is obtained by training based on the simulation effect training data set and the preset initial model, including: In the process of training based on the simulation effect training data set and the preset initial model, determining whether a model training termination condition is met; If so, stop model training and obtain the simulation effect prediction model; If not, the neuron connection weights and / or neuron thresholds in the neural network model are updated, and model training is performed based on the updated neuron connection weights and / or neuron thresholds.

6. The simulation effect prediction method according to claim 5, characterized in that: The updating of the neuron connection weights and / or neuron thresholds in the neural network model includes: The neuron connection weights and / or neuron thresholds in the neural network model are updated based on a genetic algorithm.

7. The simulation effect prediction method according to claim 6, characterized in that: The updating of the neuron connection weights and / or neuron thresholds in the neural network model based on the genetic algorithm includes: Acquire a current population; the current population is composed of a plurality of combined coding data, each of which includes coding data obtained by encoding the previous generation of neuron connection weights and coding data obtained by encoding the previous generation of neuron thresholds; Performing genetic processing on the current population to obtain a new generation population; Calculating the fitness values ​​corresponding to each of the combined encoding data in the new generation population; Determining an optimal combination of neuron connection weights and neuron thresholds based on the fitness value and the new generation population; The neuron connection weights and neuron thresholds in the neural network model are updated based on the optimal neuron connection weights and neuron thresholds combination.

8. A simulation effect prediction device, characterized in that: include: An acquisition module, used to acquire the value of a modeling parameter; the modeling parameter is a parameter used to establish a vehicle finite element simulation model; The prediction module is used to input the values ​​of the modeling parameters into the trained simulation effect prediction model to obtain the target parameter simulation calculation error and simulation calculation time of the vehicle finite element simulation model under the values ​​of the modeling parameters.

9. An electronic device, characterized in that: The method comprises a processor and a memory, wherein a computer program is stored in the memory, and the processor executes the computer program to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by at least one processor, the method according to any one of claims 1 to 7 is implemented.

Images