Method and device for predicting uniaxial tensile stress of viscoelastic material, electronic equipment and storage medium

Abstract

The embodiment of the invention discloses a method, device and equipment for predicting the uniaxial tensile stress of a viscoelastic material and a medium, and the method comprises the steps: obtaining uniaxial tensile experimental data of a target type of viscoelastic material under different experimental conditions, and constructing a training set and a verification set according to the uniaxial tensile experimental data; based on the training set, training a deep neural network model based on a residual error through an L-BFGS algorithm to obtain an initial prediction model; evaluating the initial prediction model through the verification set to obtain an evaluation result, and updating the initial model according to the evaluation result to obtain a target prediction model; and according to the prediction model and the target experiment condition data of the to-be-predicted viscoelastic material during the uniaxial tensile experiment, predicting the uniaxial tensile stress of the to-be-predicted viscoelastic material during the uniaxial tensile experiment.

Description

Technical Field

[0001] The present invention relates to the field of mechanical behavior prediction, and particularly to a method, device, electronic device and storage medium for predicting uniaxial tensile stress of viscoelastic materials. Background Art

[0002] In modern engineering and scientific fields, the application of nonlinear viscoelastic materials is becoming increasingly widespread. These materials have unique mechanical properties. When subjected to external loads, they exhibit a nonlinear stress-strain relationship and time-dependent viscoelastic characteristics. For example, in the aerospace field, composite materials used to manufacture aircraft components often have nonlinear viscoelastic properties; in the automotive industry, the performance of tire rubber is also affected by nonlinear viscoelasticity; in the biomedical field, human tissues such as muscles and cartilage also exhibit nonlinear viscoelastic behavior.

[0003] Accurately predicting the uniaxial tensile mechanical properties of nonlinear viscoelastic materials is crucial for the design, manufacture and use of materials. However, due to the complex mechanical behavior of such materials, which is affected by a variety of factors, including temperature, strain rate, strain, etc., the prediction work is extremely challenging.

[0004] Traditional prediction methods are mainly based on theoretical models and numerical simulations. Theoretical models usually describe the mechanical behavior of materials based on some simplified assumptions and physical laws, but these models often cannot fully consider the complex nonlinearity and multi-factor interactions in reality, resulting in a large deviation between the prediction results and the actual situation.

[0005] Numerical simulation methods, such as finite element analysis, although able to simulate the mechanical response of materials to a certain extent, usually require a large amount of computing resources and time, and have extremely high requirements for the establishment of models and the selection of parameters. Summary of the Invention

[0006] In view of this, the present invention provides a method, device, electronic device and storage medium for predicting uniaxial tensile stress of viscoelastic materials, which are used to solve the problems of inaccurate prediction and difficulty in implementation in the prior art.

[0007] In a first aspect, an embodiment of the present invention provides a method for predicting uniaxial tensile stress of viscoelastic materials, the method comprising:

[0008] Obtaining a uniaxial tensile experimental data set of viscoelastic materials of a target type under different experimental conditions, and constructing a training set and a validation set according to the uniaxial tensile experimental data set;

[0009] Based on the training set, training a preset deep neural network model by the L-BFGS algorithm to obtain an initial prediction model;

[0010] Evaluate the initial prediction model using the validation set according to the preset evaluation index data to obtain an evaluation result, and adjust the initial prediction model according to the evaluation result to obtain a target prediction model, or use the initial model as the target prediction model;

[0011] Predict the uniaxial tensile stress on the viscoelastic material to be predicted during the uniaxial tensile experiment according to the prediction model and the target experimental condition data of the viscoelastic material to be predicted during the uniaxial tensile experiment, and the material type of the viscoelastic material to be predicted is the target type.

[0012] Optionally, the experimental conditions include temperature conditions, strain rate conditions, and strain conditions. The step of obtaining the uniaxial tensile experiment data set of the viscoelastic material of the target type under different experimental conditions includes:

[0013] Obtain the historical uniaxial tensile stress of the viscoelastic material of the target type under each experimental condition, determine the temperature value corresponding to the historical uniaxial tensile stress according to the temperature condition in the experimental condition, determine the strain rate data corresponding to the historical uniaxial tensile stress according to the strain rate condition in the experimental condition, and determine the strain data corresponding to the historical uniaxial tensile stress according to the strain condition in the experimental condition;

[0014] Integrate the temperature value, the strain rate data, the strain data, and the historical uniaxial tensile stress into the initial uniaxial tensile experiment data under each experimental condition to obtain an initial uniaxial tensile experiment data set;

[0015] Preprocess each initial uniaxial tensile experiment data in the initial uniaxial tensile experiment data set to obtain the uniaxial tensile experiment data set, and the data in the uniaxial tensile experiment data set is within the same dimension range. The preprocessing includes at least one of cleaning processing, screening processing, and normalization processing.

[0016] Optionally, the preset deep neural network model includes an input layer, an output layer, and a preset number of hidden layers. Among them, the preset number of hidden layers are connected in series in sequence, and there is a residual connection between adjacent hidden layers. The input layer includes a first neuron for receiving the temperature value, a second neuron for receiving the strain rate data, and a third neuron for receiving the strain data. The input layer inputs the temperature value, the strain rate data, and the strain data to the first hidden layer in the series through the first neuron, the second neuron, and the third neuron. The output end of the last hidden layer in the series is connected to the output layer.

[0017] Optionally, the hidden layer includes a linear layer and an activation layer. The output end of the linear layer is connected to the input end of the activation layer. The input end of the linear layer serves as the input end of the hidden layer, and the output end of the activation layer serves as the output end of the hidden layer. The input end of the hidden layer is connected in series with the input end of the next hidden layer through a residual connection.

[0018] Optionally, the activation function of the preset deep neural network model is the hyperbolic tangent function.

[0019] Optionally, the step of evaluating the initial prediction model through the validation set according to the preset evaluation index data to obtain an evaluation result includes:

[0020] Obtain preset evaluation index data for the prediction model. The preset evaluation index data includes at least the mean square error index, the root mean square error index, and the mean absolute error index;

[0021] Substitute the validation set into the initial prediction model to obtain a prediction result set for the validation set;

[0022] Calculate the mean square error, the root mean square error, and the mean absolute error of the initial prediction model according to the prediction result set and the actual uniaxial tensile stress corresponding to the validation set;

[0023] Compare the mean square error with a first threshold to obtain a first result;

[0024] Compare the root mean square error with a second threshold to obtain a second result;

[0025] Compare the mean absolute error with a third threshold to obtain a third result;

[0026] When the first result, the second result, and the third result all meet the preset rules, obtain the first evaluation result in the evaluation result;

[0027] When at least one of the first result, the second result, and the third result does not meet the preset rules, obtain the second evaluation result in the evaluation result.

[0028] Optionally, the step of adjusting the initial prediction model according to the evaluation result to obtain a target prediction model, or using the initial model as the target prediction model includes:

[0029] When obtaining the second evaluation result in the evaluation result, adjust the initial prediction model according to the second evaluation result to obtain a target prediction model;

[0030] Among them, the step of adjusting the initial prediction model according to the second evaluation result includes: adjusting the model structure of the initial prediction model according to the second evaluation result, adjusting the training process of the initial prediction model according to the second evaluation result, and adjusting the model structure of the initial prediction model by using a regularization technique according to the second evaluation result;

[0031] When the first evaluation result in the evaluation results is obtained, the initial model is used as the target prediction model.

[0032] In a second aspect, an embodiment of the present invention further provides a uniaxial tensile stress prediction device for viscoelastic materials, and the device includes:

[0033] A data acquisition module, configured to acquire a uniaxial tensile experiment data set of a viscoelastic material of a target type under different experimental conditions, and construct a training set and a validation set according to the uniaxial tensile experiment data set;

[0034] A training module, configured to train a preset deep neural network model based on the training set by using the L-BFGS algorithm to obtain an initial prediction model;

[0035] An evaluation module, configured to evaluate the initial prediction model through the validation set according to preset evaluation index data, obtain an evaluation result, and adjust the initial prediction model according to the evaluation result to obtain a target prediction model, or use the initial model as the target prediction model;

[0036] A prediction module, configured to predict the uniaxial tensile stress received by the viscoelastic material to be predicted during a uniaxial tensile experiment according to the prediction model and the target experimental condition data of the viscoelastic material to be predicted during the uniaxial tensile experiment, and the material type of the viscoelastic material to be predicted is the target type.

[0037] In a third aspect, an embodiment of the present invention further provides an electronic device, and the electronic device includes:

[0038] One or more processors;

[0039] A storage device, configured to store one or more programs;

[0040] When the one or more programs are executed by the one or more processors, the one or more processors implement the uniaxial tensile stress prediction method for viscoelastic materials in any embodiment of the present invention.

[0041] Fourthly, an embodiment of the present invention further provides a storage medium containing computer-executable instructions, and when the computer-executable instructions are executed by a computer processor, they are used to execute the uniaxial tensile stress prediction method for viscoelastic materials in any embodiment of the present invention.

[0042] The technical solution of the embodiment of the present invention is as follows: by obtaining the uniaxial tensile experiment data set of viscoelastic materials of the target type under different experimental conditions, and constructing a training set and a validation set according to the uniaxial tensile experiment data set; based on the training set, training a preset deep neural network model through the L-BFGS algorithm to obtain an initial prediction model; evaluating the initial prediction model according to the preset evaluation index data through the validation set to obtain an evaluation result, and adjusting the initial prediction model according to the evaluation result to obtain a target prediction model, or using the initial model as the target prediction model; according to the prediction model and the target experimental condition data of the viscoelastic material to be predicted during the uniaxial tensile experiment, predicting the uniaxial tensile stress received by the viscoelastic material to be predicted during the uniaxial tensile experiment, and the material type of the viscoelastic material to be predicted is the target type. Using a deep neural network can more accurately fit non-linear laws. Taking the target experimental condition data as the input can comprehensively consider the influence of the target experimental condition data on the output stress, reduce the calculation cost and time, improve the data processing efficiency, meet the rapid prediction requirements in practical applications, enhance the ability to capture non-linear relationships, more truly reflect the mechanical property change law of the material, improve the generality and adaptability of the model, and enable it to be widely applied to non-linear viscoelastic materials of different types and characteristics. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained according to these drawings.

[0044] Among them:

[0045] Figure 1 is a schematic flow chart of a method for predicting uniaxial tensile stress of a viscoelastic material in an embodiment;

[0046] Figure 2 is a network structure diagram of a preset deep neural network model in a method for predicting uniaxial tensile stress of a viscoelastic material in an embodiment;

[0047] Figure 3 is a structure diagram of a residual connection in a method for predicting uniaxial tensile stress of a viscoelastic material in an embodiment;

[0048] Figure 4 It is a schematic structural diagram of a uniaxial tensile stress prediction device for a viscoelastic material in an embodiment;

[0049] Figure 5 It is a schematic structural diagram of an electronic device provided by an embodiment of the present application;

[0050] Figure 6 It is a schematic structural diagram of a storage medium provided by an embodiment of the present application. Detailed implementation manners

[0051] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts belong to the scope of protection of the present invention.

[0052] Before elaborating on the technical solutions of the embodiments of the present invention, first, an exemplary description of the application scenarios of the embodiments of the present invention is given:

[0053] In modern engineering and scientific fields, the application of non-linear viscoelastic materials is becoming increasingly widespread. These materials have unique mechanical properties. When subjected to external loads, they exhibit non-linear stress-strain relationships and time-dependent viscoelastic characteristics. For example, in the aerospace field, composite materials used to manufacture aircraft components often have non-linear viscoelastic properties; in the automotive industry, the performance of tire rubber is also affected by non-linear viscoelasticity; in the biomedical field, human tissues such as muscles and cartilage also exhibit non-linear viscoelastic behavior.

[0054] Accurately predicting the uniaxial tensile mechanical properties of non-linear viscoelastic materials is crucial for the design, manufacture, and use of materials. However, due to the complex mechanical behavior of such materials, which is affected by a variety of factors, including temperature, strain rate, strain, etc., the prediction work is extremely challenging.

[0055] In view of this, the embodiments of the present invention provide a method for predicting the uniaxial tensile stress of viscoelastic materials. By using a deep neural network to more accurately fit non-linear laws, taking the target experimental condition data as input, it can comprehensively consider the influence of the target experimental condition data on the output stress, reduce the calculation cost and time, improve the data processing efficiency, to meet the rapid prediction requirements in practical applications, enhance the ability to capture non-linear relationships, more truly reflect the change law of the mechanical properties of materials, improve the generality and adaptability of the model, and enable it to be widely applied to different types and characteristics of non-linear viscoelastic materials.

[0056] In one embodiment, the embodiment of the present invention provides a method for predicting the uniaxial tensile stress of a viscoelastic material. The method for predicting the uniaxial tensile stress of the viscoelastic material in the embodiment of the present invention is applicable to the analysis of viscoelastic materials in different scenarios. The method for predicting the uniaxial tensile stress of the viscoelastic material in the embodiment of the present invention can be executed by a device for predicting the uniaxial tensile stress of the viscoelastic material. The device for predicting the uniaxial tensile stress of the viscoelastic material can be implemented by software and / or hardware.

[0057] As Figure 1 shown, the method for predicting the uniaxial tensile stress of the viscoelastic material in the embodiment of the present invention specifically includes the following steps:

[0058] S110. Obtain a uniaxial tensile experiment data set of a viscoelastic material of a target type under different experimental conditions, and construct a training set and a validation set according to the uniaxial tensile experiment data set;

[0059] Exemplarily, the target type is a non-linear type, that is, the viscoelastic material of the target type is a non-linear viscoelastic material, and the uniaxial tensile experiment data set includes different experimental condition data and the stress corresponding to the experimental condition data.

[0060] Exemplarily, the training set and the validation set are obtained by dividing the uniaxial tensile experiment data set according to a preset ratio.

[0061] S120. Based on the training set, train a preset deep neural network model by using the L-BFGS algorithm to obtain an initial prediction model;

[0062] Exemplarily, the L-BFGS algorithm is an optimization algorithm based on the quasi-Newton method, which can effectively handle large-scale optimization problems. The L-BFGS algorithm performs well in terms of convergence speed and accuracy. Compared with the Stochastic Gradient Descent (SGD) algorithm, the L-BFGS algorithm can converge to a better solution faster, especially when dealing with complex non-linear problems. Compared with algorithms such as Adaptive Moment Estimation (Adam), the L-BFGS algorithm can provide more stable and accurate optimization results.

[0063] Specifically, the L-BFGS algorithm belongs to the limited-memory quasi-Newton method. The L-BFGS algorithm only stores the information of the recent several iterations to approximate the inverse of the Hessian matrix. In each iteration, the L-BFGS algorithm constructs an approximation of the inverse of the Hessian matrix by using the gradient information of the previous iterations. Specifically, the L-BFGS algorithm updates this approximation by storing a certain number of vector pairs (displacement vectors and gradient difference vectors). When calculating the update direction, the L-BFGS algorithm uses these stored vector pairs and the current gradient to construct an approximate search direction. The constructed approximate search direction is more accurate than the simple gradient descent direction and can converge to the optimal solution faster.

[0064] S130. Evaluate the initial prediction model using the validation set according to the preset evaluation index data to obtain an evaluation result, and adjust the initial prediction model according to the evaluation result to obtain a target prediction model, or use the initial model as the target prediction model.

[0065] Exemplarily, different performance of the initial prediction model can be evaluated through different evaluation index data. For example, accuracy, stability, etc.; for different requirements, the initial prediction model can be evaluated according to different preset evaluation index data.

[0066] Exemplarily, when the evaluation result indicates that the initial prediction model does not meet the requirements, adjust the initial prediction model according to the evaluation result. For example, when the evaluation result indicates that the accuracy of the initial prediction model does not meet the requirements, the network structure of the initial prediction model can be adjusted until the accuracy of the adjusted initial prediction model meets the requirements to obtain a target prediction model.

[0067] Exemplarily, when the evaluation result indicates that the initial prediction model meets the requirements, use the initial model as the target prediction model.

[0068] S140. According to the prediction model and the target experimental condition data of the viscoelastic material to be predicted during the uniaxial tensile test, predict the uniaxial tensile stress received by the viscoelastic material to be predicted during the uniaxial tensile test, and the material type of the viscoelastic material to be predicted is the target type.

[0069] Exemplarily, take the target experimental condition data of the viscoelastic material to be predicted during the uniaxial tensile test as input data and input it into the target prediction model. The target prediction model outputs the uniaxial tensile stress received by the viscoelastic material to be predicted during the uniaxial tensile test to complete the prediction process.

[0070] By obtaining the uniaxial tensile test data sets of viscoelastic materials of the target type under different experimental conditions, and constructing a training set and a validation set according to the uniaxial tensile test data sets; based on the training set, training a preset deep neural network model through the L-BFGS algorithm to obtain an initial prediction model; evaluating the initial prediction model according to preset evaluation index data through the validation set to obtain an evaluation result, and adjusting the initial prediction model according to the evaluation result to obtain a target prediction model, or taking the initial model as the target prediction model; according to the prediction model and the target experimental condition data of the viscoelastic material to be predicted during the uniaxial tensile test, predicting the uniaxial tensile stress received by the viscoelastic material to be predicted during the uniaxial tensile test, and the material type of the viscoelastic material to be predicted is the target type. Using a deep neural network can more accurately fit non-linear laws. Taking the target experimental condition data as the input can comprehensively consider the influence of the target experimental condition data on the output stress, reduce the calculation cost and time, improve the data processing efficiency, meet the rapid prediction requirements in practical applications, enhance the ability to capture non-linear relationships, more truly reflect the mechanical property change law of the material, improve the generality and adaptability of the model, and enable it to be widely applied to non-linear viscoelastic materials of different types and characteristics.

[0071] In a possible implementation manner, the experimental conditions include temperature conditions, strain rate conditions, and strain conditions, and the step of obtaining the uniaxial tensile test data sets of viscoelastic materials of the target type under different experimental conditions includes:

[0072] Obtaining the historical uniaxial tensile stress of the viscoelastic material of the target type under each experimental condition, determining the temperature value corresponding to the historical uniaxial tensile stress according to the temperature condition in the experimental condition, determining the strain rate data corresponding to the historical uniaxial tensile stress according to the strain rate condition in the experimental condition, and determining the strain data corresponding to the historical uniaxial tensile stress according to the strain condition in the experimental condition;

[0073] Integrating the temperature value, the strain rate data, the strain data, and the historical uniaxial tensile stress into the initial uniaxial tensile test data under each experimental condition to obtain an initial uniaxial tensile test data set;

[0074] Performing preprocessing on each initial uniaxial tensile test data in the initial uniaxial tensile test data set to obtain the uniaxial tensile test data set, and the data in the uniaxial tensile test data set is within the same dimension range, and the preprocessing includes at least one of cleaning processing, screening processing, and normalization processing.

[0075] Exemplarily, the existing methods have insufficient integration ability for input factors and cannot fully utilize the information of factors such as temperature, strain rate, and strain. By taking temperature, strain rate, and strain as inputs, the influence of temperature, strain rate, and strain on the output stress can be comprehensively considered.

[0076] Exemplarily, the stress of the nonlinear viscoelastic material under different temperature conditions, strain rate conditions, and strain conditions is obtained, and the temperature data corresponding to the temperature condition, the strain rate data corresponding to the strain rate condition, the strain condition corresponding to the strain condition, and the stress are combined to form an initial uniaxial tensile experiment data set.

[0077] Exemplarily, the initial uniaxial tensile experiment data set is cleaned, screened, and normalized to eliminate noise and outliers, and to make the data in the initial uniaxial tensile experiment data set within the same dimension range, which is convenient for the subsequent learning and training of the neural network.

[0078] In a possible implementation manner, as Figure 2 shown, the preset deep neural network model includes an input layer, an output layer, and a preset number of hidden layers. Among them, the preset number of hidden layers are connected in series in sequence, and a residual connection is provided between adjacent two hidden layers. The input layer includes a first neuron for receiving the temperature value, a second neuron for receiving the strain rate data, and a third neuron for receiving the strain data. The input layer inputs the temperature value, the strain rate data, and the strain data to the first hidden layer in the series through the first neuron, the second neuron, and the third neuron. The output end of the last hidden layer in the series is connected to the output layer.

[0079] Exemplarily, a deep neural network (DNN for short) is a machine learning model with a multi-layer structure. The strength of the deep neural network lies in its ability to automatically learn complex patterns and features from a large amount of data. Through the interconnection and cooperation of neurons at multiple levels, the DNN can gradually abstract and represent the input data, so as to achieve accurate prediction of various tasks. The preset deep neural network model adopts a deep neural network architecture composed of multiple layers of neurons, including an input layer, a hidden layer, and an output layer. The input layer receives temperature data, strain rate data, and strain data. The hidden layer contains 60 layers and uses residual connections. The output layer outputs the predicted value of the stress.

[0080] In a possible implementation manner, as Figure 3As shown, the hidden layer includes a linear layer and an activation layer. The output end of the linear layer is connected to the input end of the activation layer. The input end of the linear layer serves as the input end of the hidden layer, and the output end of the activation layer serves as the output end of the hidden layer. The input end of the hidden layer is connected in series with the input end of the next hidden layer through a residual connection.

[0081] Exemplarily, a residual connection means skipping certain layers in the network and directly adding the input information to the output of the subsequent layer. Residual connections enable the network to be trained and optimized more easily. Through residual connections, the information flow can be transmitted more smoothly in the network, enabling deep networks to effectively learn useful features. Residual connections provide a shortcut for the network, enabling the gradient to be backpropagated more directly, enhancing the training effect and generalization ability of the model.

[0082] In a possible implementation manner, the activation function of the preset deep neural network model is the hyperbolic tangent function.

[0083] Exemplarily, the Tanh function is selected as the activation function of the neural network. The output value range of the Tanh function is between -1 and 1. Compared with the Sigmoid function (the output range is between 0 and 1), the Tanh function can handle a wider numerical range, which enables the Tanh function to better represent the characteristics of the data.

[0084] In a possible implementation manner, the step of evaluating the initial prediction model through the validation set according to the preset evaluation index data includes:

[0085] Obtain the preset evaluation index data for the prediction model. The preset evaluation index data at least includes the mean square error index, the root mean square error index, and the mean absolute error index;

[0086] Substitute the validation set into the initial prediction model to obtain a prediction result set for the validation set;

[0087] Calculate the mean square error, the root mean square error, and the mean absolute error of the initial prediction model according to the prediction result set and the actual uniaxial tensile stress corresponding to the validation set;

[0088] Compare the mean square error with a first threshold to obtain a first result;

[0089] Compare the root mean square error with a second threshold to obtain a second result;

[0090] Compare the mean absolute error with a third threshold to obtain a third result;

[0091] When the first result, the second result, and the third result all meet the preset rules, the first evaluation result in the evaluation results is obtained;

[0092] When at least one of the first result, the second result, and the third result does not meet the preset rules, the second evaluation result in the evaluation results is obtained.

[0093] Exemplarily, the initial prediction model is evaluated by the mean squared error, the root mean squared error, and the mean absolute error of the initial prediction model. The smaller the values of the mean squared error, the root mean squared error, and the mean absolute error, the better the performance of the initial prediction model. Therefore, a first threshold for the mean squared error, a second threshold for the root mean squared error, and a third threshold for the mean absolute error are set.

[0094] Exemplarily, when the evaluation result of any one of the mean squared error, the root mean squared error, and the mean absolute error of the initial prediction model does not meet the preset rules, the second evaluation result in the evaluation results is obtained.

[0095] In a possible implementation manner, the step of adjusting the initial prediction model according to the evaluation result to obtain a target prediction model, or using the initial model as the target prediction model includes:

[0096] When the second evaluation result in the evaluation results is obtained, the initial prediction model is adjusted according to the second evaluation result to obtain a target prediction model;

[0097] Among them, the step of adjusting the initial prediction model according to the second evaluation result includes: adjusting the model structure of the initial prediction model according to the second evaluation result, adjusting the training process of the initial prediction model according to the second evaluation result, and adjusting the model structure of the initial prediction model by using a regularization technique according to the second evaluation result;

[0098] When the first evaluation result in the evaluation results is obtained, the initial model is used as the target prediction model.

[0099] Exemplarily, the process of adjusting the model structure of the initial prediction model includes, but is not limited to, increasing the number of hidden layers of the initial prediction model; the process of adjusting the training process of the initial prediction model includes, but is not limited to, increasing the number of training sets for training the initial prediction model.

[0100] In another embodiment of the present invention, a device for predicting the uniaxial tensile stress of a viscoelastic material is provided. Figure 4The figure is a schematic structural diagram of a uniaxial tensile stress prediction device for viscoelastic materials provided by an embodiment of the present invention. The uniaxial tensile stress prediction device for viscoelastic materials provided by the embodiment of the present invention can execute the uniaxial tensile stress prediction method provided by any embodiment of the present invention, and has corresponding functional modules and beneficial effects for executing the method. The device includes:

[0101] A data acquisition module 201, configured to acquire a uniaxial tensile experiment data set of a viscoelastic material of a target type under different experimental conditions, and construct a training set and a validation set according to the uniaxial tensile experiment data set;

[0102] A training module 202, configured to train a preset deep neural network model based on the training set by using the L-BFGS algorithm to obtain an initial prediction model;

[0103] An evaluation module 203, configured to evaluate the initial prediction model according to preset evaluation index data through the validation set to obtain an evaluation result, and adjust the initial prediction model according to the evaluation result to obtain a target prediction model, or use the initial model as the target prediction model;

[0104] A prediction module 204, configured to predict the uniaxial tensile stress suffered by the viscoelastic material to be predicted during the uniaxial tensile experiment according to the prediction model and the target experimental condition data of the viscoelastic material to be predicted during the uniaxial tensile experiment, and the material type of the viscoelastic material to be predicted is the target type.

[0105] It should be noted that the various modules included in the above device are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be realized; in addition, the specific names of the functional modules are only for the convenience of mutual distinction and do not limit the protection scope of the embodiments of the present invention.

[0106] In another embodiment of the present invention, an electronic device is also provided. Figure 5 The block diagram of an exemplary electronic device 50 suitable for implementing the embodiment mode of the embodiment of the present invention is shown. Figure 5 The shown electronic device 50 is only an example and should not bring any limitation to the functions and usage scope of the embodiments of the present invention.

[0107] As Figure 5 shown, the electronic device 50 is presented in the form of a general computing device. The components of the electronic device 50 may include, but are not limited to: one or more processors or processing units 501, a system memory 502, and a bus 503 connecting different system components (including the system memory 502 and the processing unit 501).

[0108] The bus 503 represents one or more of several types of bus architectures, including a memory bus or memory controller, a peripheral bus, an Accelerated Graphics Port, a processor bus, or a local bus using any of the several bus architectures. By way of example, and not limitation, these architectures include the Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MAC) bus, Enhanced ISA bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.

[0109] The electronic device 50 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the electronic device 50, including both volatile and nonvolatile media, removable and non-removable media.

[0110] The system memory 502 can include computer system readable media in the form of volatile memory, such as random access memory (RAM) 504 and / or cache memory 505. The electronic device 50 may further include other removable / non-removable, volatile / nonvolatile computer system storage media. By way of example only, a storage system 506 can be provided for reading from and writing to non-removable, nonvolatile magnetic media ( Figure 5 not shown and typically called a "hard disk drive"). Although Figure 5 not shown in the figures, a disk drive for reading from and writing to a removable nonvolatile disk (e.g., a "floppy disk"), and an optical disk drive for reading from and writing to a removable nonvolatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) can be provided. In these cases, each drive can be connected to the bus 503 by one or more data media interfaces. The memory 502 can include at least one program product having a set (e.g., at least one) of program modules that are configured to carry out the functions of the embodiments of the present invention.

[0111] A program / utility 508 having a set (at least one) of program modules 507 can be stored, for example, in the memory 502, such program modules 507 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which examples or some combination thereof may include an implementation of a networking environment. The program modules 507 typically carry out the functions and / or methods of the embodiments described herein.

[0112] The electronic device 50 can also communicate with one or more external devices 509 (such as a keyboard, a pointing device, a display 510, etc.), and can also communicate with one or more devices that enable a user to interact with the electronic device 50, and / or communicate with any device that enables the electronic device 50 to communicate with one or more other computing devices (such as a network card, a modem, etc.). Such communication can be carried out through the input / output (I / O) interface 511. Moreover, the electronic device 50 can also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through the network adapter 512. As shown in the figure, the network adapter 512 communicates with other modules of the electronic device 50 through the bus 503. It should be understood that although Figure 5 not shown in the figure, other hardware and / or software modules can be used in combination with the electronic device 50, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0113] The processing unit 501 executes various functional applications and data processing by running the programs stored in the system memory 502, for example, implementing the uniaxial tensile stress prediction method for viscoelastic materials provided by the embodiments of the present invention.

[0114] In another embodiment of the present invention, a storage medium containing computer-executable instructions is also provided. As Figure 6 shown, an embodiment of the present application provides a computer-readable storage medium 400, on which a computer program 411 is stored. When the computer program 411 is executed by a processor, it realizes: obtaining a uniaxial tensile experiment data set of a viscoelastic material of a target type under different experimental conditions, and constructing a training set and a validation set according to the uniaxial tensile experiment data set; based on the training set, training a preset deep neural network model through the L-BFGS algorithm to obtain an initial prediction model; evaluating the initial prediction model through the validation set according to preset evaluation index data to obtain an evaluation result, and adjusting the initial prediction model according to the evaluation result to obtain a target prediction model, or using the initial model as the target prediction model; predicting the uniaxial tensile stress received by the viscoelastic material to be predicted during the uniaxial tensile experiment according to the prediction model and the target experimental condition data of the viscoelastic material to be predicted during the uniaxial tensile experiment, and the material type of the viscoelastic material to be predicted is the target type.

[0115] The computer storage medium of the embodiment of the present invention may adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive list) of the computer-readable storage medium include: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0116] The computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, which carries the computer-readable program code. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium can also be any computer-readable medium other than the computer-readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0117] The program code contained on the computer-readable medium can be transmitted by any appropriate medium, including but not limited to wireless, wire, optical fiber, RF, etc., or any suitable combination of the above.

[0118] The computer program code for performing the operations of the embodiment of the present invention can be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, by using an Internet service provider to connect through the Internet).

[0119] The above-disclosed is only the preferred embodiment of the present invention. Of course, the scope of the rights of the present invention cannot be limited thereby. Therefore, equivalent changes made according to the claims of the present invention still fall within the scope covered by the present invention.

Claims

1. A method for predicting uniaxial tensile stress of viscoelastic materials, characterized in that: include: Acquire a uniaxial tensile experimental data set of a target type of viscoelastic material under different experimental conditions, and construct a training set and a validation set based on the uniaxial tensile experimental data set; Based on the training set, the preset deep neural network model is trained by the L-BFGS algorithm to obtain an initial prediction model; According to preset evaluation index data, the initial prediction model is evaluated through the validation set to obtain an evaluation result, and the initial prediction model is adjusted according to the evaluation result to obtain a target prediction model, or the initial model is used as the target prediction model; The uniaxial tensile stress to which the viscoelastic material to be predicted is subjected during the uniaxial tensile test is predicted according to the prediction model and target experimental condition data of the viscoelastic material to be predicted during the uniaxial tensile test, wherein the material type of the viscoelastic material to be predicted is the target type.

2. The method according to claim 1, characterized in that: The experimental conditions include temperature conditions, strain rate conditions and strain conditions. The step of obtaining a uniaxial tensile experimental data set of a target type of viscoelastic material under different experimental conditions includes: Obtaining the historical uniaxial tensile stress of the target type of viscoelastic material under each experimental condition, and determining the temperature value corresponding to the historical uniaxial tensile stress according to the temperature condition in the experimental condition, determining the strain rate data corresponding to the historical uniaxial tensile stress according to the strain rate condition in the experimental condition, and determining the strain data corresponding to the historical uniaxial tensile stress according to the strain condition in the experimental condition; Integrate the temperature value, the strain rate data, the strain data and the historical uniaxial tensile stress into the initial uniaxial tensile test data under each experimental condition to obtain an initial uniaxial tensile test data set; Each initial uniaxial tensile experimental data in the initial uniaxial tensile experimental data set is preprocessed to obtain the uniaxial tensile experimental data set, wherein the data in the uniaxial tensile experimental data set are in the same dimensional range, and the preprocessing includes at least one of cleaning processing, screening processing and normalization processing.

3. The method according to claim 2, characterized in that The preset deep neural network model includes an input layer, an output layer and a preset number of hidden layers, wherein the preset number of hidden layers are connected in series in sequence, and a residual connection is set between two adjacent hidden layers, the input layer includes a first neuron for receiving the temperature value, a second neuron for receiving the strain rate data and a third neuron for receiving the strain data, and the input layer inputs the temperature value, the strain rate data and the strain data to the first hidden layer in the series through the first neuron, the second neuron and the third neuron, and the output end of the last hidden layer in the series is connected to the output layer.

4. The method according to claim 3, characterized in that The hidden layer includes a linear layer and an activation layer, the output end of the linear layer is connected to the input end of the activation layer, the input end of the linear layer serves as the input end of the hidden layer, the output end of the activation layer serves as the output end of the hidden layer, and the input end of the hidden layer is connected in series with the input end of the next hidden layer through a residual connection.

5. The method according to claim 1, characterized in that The activation function of the preset deep neural network model is a hyperbolic tangent function.

6. The method according to claim 1, characterized in that The step of evaluating the initial prediction model through the validation set according to the preset evaluation index data to obtain the evaluation result includes: Acquire preset evaluation index data for the prediction model, wherein the preset evaluation index data at least includes a mean square error index, a root mean square error index, and a mean absolute error index; Substituting the verification set into the initial prediction model to obtain a prediction result set for the verification set; Calculating the mean square error, root mean square error and mean absolute error of the initial prediction model according to the prediction result set and the actual uniaxial tensile stress corresponding to the verification set; Compare the mean square error with a first threshold to obtain a first result; Compare the root mean square error with a second threshold to obtain a second result; Compare the mean absolute error with a third threshold to obtain a third result; When the first result, the second result and the third result all satisfy the preset rule, a first evaluation result in the evaluation results is obtained; When at least one of the first result, the second result, and the third result does not satisfy the preset rule, a second evaluation result in the evaluation results is obtained.

7. The method according to claim 6, characterized in that The step of adjusting the initial prediction model according to the evaluation result to obtain a target prediction model, or taking the initial model as the target prediction model, comprises: When a second evaluation result in the evaluation results is obtained, adjusting the initial prediction model according to the second evaluation result to obtain a target prediction model; The step of adjusting the initial prediction model according to the second evaluation result includes: adjusting the model structure of the initial prediction model according to the second evaluation result, adjusting the training process of the initial prediction model according to the second evaluation result, and adjusting the model structure of the initial prediction model using a regularization technique according to the second evaluation result; When the first evaluation result in the evaluation results is obtained, the initial model is used as the target prediction model.

8. A device for predicting uniaxial tensile stress of viscoelastic materials, characterized in that: The device comprises: A data acquisition module, used to acquire a uniaxial tensile experimental data set of a target type of viscoelastic material under different experimental conditions, and to construct a training set and a validation set based on the uniaxial tensile experimental data set; A training module, used to train a preset deep neural network model based on the training set by using an L-BFGS algorithm to obtain an initial prediction model; An evaluation module, used to evaluate the initial prediction model through the validation set according to preset evaluation index data to obtain an evaluation result, and adjust the initial prediction model according to the evaluation result to obtain a target prediction model, or use the initial model as the target prediction model; The prediction module is used to predict the uniaxial tensile stress to which the viscoelastic material to be predicted is subjected during the uniaxial tensile experiment according to the prediction model and the target experimental condition data of the viscoelastic material to be predicted during the uniaxial tensile experiment, wherein the material type of the viscoelastic material to be predicted is the target type.

9. An electronic device, characterized in that: The electronic device comprises: One or more processors; A storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the uniaxial tensile stress prediction method for viscoelastic materials as described in any one of claims 1-7.

10. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions are used to execute the uniaxial tensile stress prediction method of viscoelastic materials as described in any one of claims 1 to 7 when executed by a computer processor.

Images