Automobile plate selection method and device, electronic equipment and storage medium

Abstract

The invention discloses an automobile plate selection method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining the target type information and target performance range of a to-be-selected target automobile plate; obtaining a plurality of candidate automobile plates conforming to the target type information, and obtaining mechanical index parameters corresponding to the candidate automobile plates; for each candidate automobile sheet material, predicting a prediction performance range of the candidate automobile sheet material through the trained performance prediction model based on the target type information and the mechanical index parameters; and based on the target performance range and the predicted performance ranges corresponding to the candidate automobile plates, selecting a target automobile plate from the plurality of candidate automobile plates. The accuracy of automobile plate selection can be improved.

Description

Technical Field

[0001] The present application belongs to the field of artificial intelligence technology, and in particular relates to a method, device, electronic device and storage medium for selecting automobile panels. Background Art

[0002] When the steel industry provides sheet materials to automobile manufacturers, the sheet materials need to have sufficient strength to meet the safety requirements of the vehicle body structure. For example, high-strength sheet materials can effectively resist deformation during a collision and protect the passengers in the vehicle. However, at the same time, the sheet materials also need good formability during forming processes such as stamping. If the strength is too high, the elongation of the material may decrease, and cracking may occur easily when stamping complex-shaped parts (such as automobile hoods, doors, etc.).

[0003] Currently, automobile plates corresponding to different needs are usually selected according to set rules, and the accuracy of plate selection is low. Summary of the invention

[0004] The embodiments of the present application provide a method, device, electronic device and computer storage medium for selecting automobile plate materials, thereby improving the accuracy of automobile plate material selection at least to a certain extent.

[0005] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by the practice of the present application.

[0006] According to a first aspect of an embodiment of the present application, a method for selecting an automobile sheet material is provided, comprising: obtaining target type information and target performance range of a target automobile sheet material to be selected;

[0007] Acquire multiple candidate automobile plates that meet the target type information, and acquire mechanical index parameters corresponding to each candidate automobile plate;

[0008] For each candidate automotive sheet, based on the target type information and mechanical index parameters, the predicted performance range of the candidate automotive sheet is predicted by the trained performance prediction model;

[0009] A target automobile sheet material is selected from multiple candidate automobile sheets based on the target performance range and the predicted performance ranges corresponding to each candidate automobile sheet material.

[0010] In some possible implementations, the performance prediction model is obtained in the following manner:

[0011] Acquire a sample data set; the sample data set includes a plurality of sample automobile plates, each sample automobile plate includes corresponding sample type information, sample mechanical index parameters and sample reference performance range;

[0012] The initial prediction model is optimized at least once based on the sample data set until a preset end condition is met, and a prediction model is obtained based on the initial prediction model that meets the preset end condition.

[0013] In some possible implementations, the sample data set includes a training data set, and performing at least one optimization operation on the initial prediction model based on the sample data set includes:

[0014] Performing at least one training operation on the initial prediction model based on the training data set, wherein the preset end condition includes that at least one training operation satisfies the training end condition;

[0015] The training operation includes:

[0016] Inputting sample type information and sample mechanical index parameters of each sample automobile sheet in the training data set into the initial prediction model to obtain sample prediction information, wherein the sample prediction information includes a first sample prediction performance range;

[0017] For each sample automobile sheet, determining a first training loss of the sample automobile sheet based on a difference between a first sample predicted performance range and a sample reference performance range;

[0018] Determine the total training loss based on the first training loss of each sample automobile plate;

[0019] The parameters of the initial prediction model are adjusted based on the total training loss, and the initial prediction model after the adjustment of the parameters is used as the initial prediction model corresponding to the next training operation.

[0020] In some possible implementations, each sample automobile sheet further includes sample forming limit data, the sample prediction information further includes predicted forming limit data, and the training operation further includes:

[0021] For each sample automobile sheet, determining a second training loss of the sample automobile sheet based on a difference between the sample forming limit data and the predicted forming limit data;

[0022] Based on the first training loss of each sample automobile plate, the total training loss is determined, including:

[0023] The total training loss is determined based on the first training loss of each sample automobile sheet and the second training loss of each sample automobile sheet.

[0024] In some possible implementations, the sample data set further includes a validation data set, and performing at least one optimization operation on the initial prediction model based on the sample data set further includes:

[0025] Performing at least one verification operation on the initial prediction model that meets the training end condition based on the verification data set, wherein the preset end condition also includes at least one verification operation meeting the verification end condition;

[0026] The verification operations include:

[0027] Input the sample type information and sample mechanical index parameters of each sample automobile sheet in the validation data set into the first prediction model to obtain the second sample prediction performance range, wherein the first prediction model is an initial prediction model that meets the training end condition;

[0028] Determining a validation result based on a difference between a second sample predicted performance range and a sample reference performance range for each sample automotive sheet in the validation data set;

[0029] The hyperparameters of the initial prediction model are adjusted based on the verification results, and the initial prediction model after the hyperparameters are adjusted is used as the initial prediction model corresponding to the next optimization operation.

[0030] In some possible implementations, the initial prediction model includes an output layer, a hidden layer, and an output layer, and the hyperparameters of the initial prediction model include the number of nodes in the hidden layer;

[0031] Based on the validation results, the hyperparameters of the initial prediction model are adjusted, including:

[0032] Based on the verification result, the number of nodes in the hidden layer is adjusted within a preset range, wherein the preset range is determined based on the number of nodes in the input layer and the number of nodes in the output layer.

[0033] In some possible embodiments, the number of nodes in the input layer is determined based on the number of parameters corresponding to the mechanical index parameters, the middle value of the preset range is positively correlated with the number of nodes in the input layer, and the middle value of the preset range is positively correlated with the number of nodes in the output layer.

[0034] According to a second aspect of an embodiment of the present application, there is provided a device for selecting an automobile plate, comprising: an information acquisition module, for acquiring target type information and target performance range of a target automobile plate to be selected;

[0035] A plate acquisition module is used to acquire multiple candidate automobile plates that meet the target type information, and acquire mechanical index parameters corresponding to each candidate automobile plate;

[0036] A prediction module, for predicting the predicted performance range of each candidate automobile sheet material through a trained performance prediction model based on target type information and mechanical index parameters;

[0037] The selection module is used to select a target automobile sheet material from a plurality of candidate automobile sheets based on the target performance range and the predicted performance ranges corresponding to each candidate automobile sheet material.

[0038] In some possible implementations, the device further includes a training module for:

[0039] Acquire a sample data set; the sample data set includes a plurality of sample automobile plates, each sample automobile plate includes corresponding sample type information, sample mechanical index parameters and sample reference performance range;

[0040] The initial prediction model is optimized at least once based on the sample data set until a preset end condition is met, and a prediction model is obtained based on the initial prediction model that meets the preset end condition.

[0041] In some possible implementations, the sample data set includes a training data set, and the training module is specifically used to:

[0042] Performing at least one training operation on the initial prediction model based on the training data set, wherein the preset end condition includes that at least one training operation satisfies the training end condition;

[0043] The training operation includes:

[0044] Inputting sample type information and sample mechanical index parameters of each sample automobile sheet in the training data set into the initial prediction model to obtain sample prediction information, wherein the sample prediction information includes a first sample prediction performance range;

[0045] For each sample automobile sheet, determining a first training loss of the sample automobile sheet based on a difference between a first sample predicted performance range and a sample reference performance range;

[0046] Determine the total training loss based on the first training loss of each sample automobile plate;

[0047] The parameters of the initial prediction model are adjusted based on the total training loss, and the initial prediction model after the adjustment of the parameters is used as the initial prediction model corresponding to the next training operation.

[0048] In some possible implementations, each sample automobile sheet further includes sample forming limit data, the sample prediction information further includes predicted forming limit data, and the training module, when performing the training operation, is further used to: for each sample automobile sheet, determine a second training loss of the sample automobile sheet based on a difference between the sample forming limit data and the predicted forming limit data;

[0049] When the training module determines the total training loss based on the first training loss of each sample automobile sheet, it is specifically used to:

[0050] The total training loss is determined based on the first training loss of each sample automobile sheet and the second training loss of each sample automobile sheet.

[0051] In some possible implementations, the sample data set further includes a verification data set, and the training module is further used to: perform at least one verification operation on the initial prediction model that meets the training end condition based on the verification data set, and the preset end condition further includes at least one verification operation meeting the verification end condition;

[0052] When performing verification operations, the training module is specifically used to:

[0053] Input the sample type information and sample mechanical index parameters of each sample automobile sheet in the validation data set into the first prediction model to obtain the second sample prediction performance range, wherein the first prediction model is an initial prediction model that meets the training end condition;

[0054] Determining a validation result based on a difference between a second sample predicted performance range and a sample reference performance range for each sample automotive sheet in the validation data set;

[0055] The hyperparameters of the initial prediction model are adjusted based on the verification results, and the initial prediction model after the hyperparameters are adjusted is used as the initial prediction model corresponding to the next optimization operation.

[0056] In some possible implementations, the initial prediction model includes an output layer, a hidden layer, and an output layer, and the hyperparameters of the initial prediction model include the number of nodes in the hidden layer;

[0057] The training module adjusts the hyperparameters of the initial prediction model based on the validation results. Specifically, it is used to:

[0058] Based on the verification result, the number of nodes in the hidden layer is adjusted within a preset range, wherein the preset range is determined based on the number of nodes in the input layer and the number of nodes in the output layer.

[0059] In some possible embodiments, the number of nodes in the input layer is determined based on the number of parameters corresponding to the mechanical index parameters, the middle value of the preset range is positively correlated with the number of nodes in the input layer, and the middle value of the preset range is positively correlated with the number of nodes in the output layer.

[0060] According to a third aspect of an embodiment of the present application, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method of the above embodiment.

[0061] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, on which a computer program is stored, characterized in that when the computer program is executed by a processor, the steps of the method of the above embodiment are implemented.

[0062] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application.

[0063] The beneficial effects of the technical solution provided by the embodiment of the present application are:

[0064] If the target type information of the target automobile sheet required by the user is obtained, multiple candidate automobile sheets that meet the target type information can be obtained first, and then the trained performance prediction model can be used to combine the target type information and mechanical index parameters of each candidate automobile sheet to predict the predicted performance range corresponding to each mechanical performance index of each candidate automobile sheet. Then, based on the target performance range required by the user and the predicted performance range of each candidate automobile sheet, the target automobile sheet is selected from multiple candidate automobile sheets, so that the target automobile sheet that meets the target type information required by the user and meets the target performance range required by the user can be selected more accurately. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The drawings herein are incorporated into the specification and constitute a part of the specification, showing embodiments consistent with the present application, and together with the specification, are used to explain the principles of the present application. Obviously, the drawings described below are only some embodiments of the present application, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0066] Figure 1 A schematic flow chart of a method for selecting an automobile plate provided in an embodiment of the present application;

[0067] Figure 2 A schematic diagram of a data normalization scheme provided for an example of the present application;

[0068] Figure 3 A schematic diagram of a data normalization scheme provided for an example of the present application;

[0069] Figure 4 A schematic diagram of the structure of a device for selecting automobile plates provided in an embodiment of the present application;

[0070] Figure 5 A schematic diagram of the structure of an electronic device for selecting an automobile panel provided in an embodiment of the present application. DETAILED DESCRIPTION

[0071] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0072] In addition, described feature, structure or characteristic can be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present application. However, those skilled in the art will appreciate that the technical scheme of the present application can be put into practice without one or more of the specific details, or other methods, components, devices, steps, etc. can be adopted. In other cases, known methods, devices, realizations or operations are not shown or described in detail to avoid blurring the various aspects of the application.

[0073] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities may be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0074] The flowcharts shown in the accompanying drawings are only exemplary and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps can be decomposed, and some operations / steps can be combined or partially combined, so the actual execution order may change according to actual conditions.

[0075] It should also be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the objects used in this way can be interchanged where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those shown or described.

[0076] The following describes several exemplary embodiments to illustrate the technical solutions of the embodiments of the present application and the technical effects produced by the technical solutions of the present application. It should be noted that the following embodiments can refer to, draw on or combine with each other, and the same terms, similar features and similar implementation steps in different embodiments will not be described repeatedly.

[0077] The automobile plate selection method of the present application can be executed by any electronic device, and the electronic device may include a server or a terminal.

[0078] It can be understood by technicians in this technical field that the server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server or server cluster that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. The terminal can be a smart phone (such as Android phones, iOS phones, etc.), a tablet computer, a laptop computer, a digital broadcast receiver, a MID (Mobile Internet Devices), a PDA (personal digital assistant), a desktop computer, a smart home appliance, a vehicle-mounted terminal (such as a vehicle-mounted navigation terminal, a vehicle-mounted computer, etc.), a smart speaker, a smart watch, etc. The terminal and the server can be directly or indirectly connected by wired or wireless communication, but are not limited to this. The embodiments of the present invention can be applied to various scenarios, including but not limited to cloud technology, artificial intelligence, smart transportation, assisted driving, etc. It can also be determined based on the actual application scenario requirements, and is not limited here. The terminal (also referred to as a user terminal or user device) can be a smart phone, a tablet computer, a laptop computer, a desktop computer, an intelligent voice interaction device (such as a smart speaker), a wearable electronic device (such as a smart watch), a vehicle terminal, a smart home appliance (such as a smart TV), an AR / VR device, an aircraft, etc., but is not limited to these.

[0079] like Figure 1 As shown, in some possible implementations, the present application embodiment provides a method for selecting an automobile plate. Taking the execution subject as a server as an example, the method may include the following steps:

[0080] Step S101, obtaining target type information and target performance range of the target automobile sheet to be selected. The type information can be used to characterize the brand, model, component, grade, thickness, etc. corresponding to the automobile sheet; the performance range can include the applicable range corresponding to at least one mechanical performance index of the automobile sheet. Specifically, the target type information is the type information required by the automobile sheet to be selected, and the target performance range can include the performance range required by the automobile sheet to be selected.

[0081] Among them, the mechanical performance indicators may include yield strength, tensile strength, elongation, R value (work hardening index), N value (plastic strain ratio), etc.

[0082] Specifically, yield strength is the stress when the material begins to produce obvious plastic deformation; the yield strength of automobile plates cannot be too high, otherwise a large force will be required in the initial stage of stamping, which will easily cause increased die wear. Tensile strength reflects the maximum stress that the material can withstand during the stretching process. The appropriate yield strength and tensile strength range are crucial to ensure the smooth progress of the stamping process and the final performance of the stamped parts. Elongation is the ratio of the elongation of the material at the time of tensile fracture to the original length, which directly reflects the plastic deformation ability of the material; in the stamping process of automobile plates, especially for parts with complex shapes and large deformation (such as the curved part of the car body), high elongation materials can better adapt to the shape changes of the stamping die and avoid cracking during the stamping process.

[0083] The N value indicates the work hardening characteristics of the material during the plastic deformation process. A higher N value means that the strength of the material increases relatively slowly as the degree of deformation increases during the stamping process, which is conducive to uniform deformation of the material and improves the shape accuracy of the stamped parts.

[0084] The R value reflects the difference in plastic strain capacity of the material in the plane of the plate and in the direction perpendicular to the plane of the plate. A larger R value means that the material is more inclined to deform in the plane of the plate during the stamping process, which helps prevent the stamping parts from becoming thinner in the thickness direction and improves the quality of the stamping parts.

[0085] Step S102, obtaining a plurality of candidate automobile sheet materials that meet the target type information, and obtaining mechanical index parameters corresponding to each candidate automobile sheet material.

[0086] Specifically, multiple candidate automobile sheet materials that meet the target type information can be first screened out from the candidate automobile sheet materials corresponding to the multiple different types of information included in advance, wherein each candidate automobile sheet material has corresponding mechanical index parameters.

[0087] Step S103 : for each candidate automobile sheet material, based on the target type information and the mechanical index parameters, the predicted performance range of the candidate automobile sheet material is predicted by the trained performance prediction model.

[0088] Specifically, a performance prediction model can be pre-trained, and the performance prediction model can predict the predicted performance range of the automobile sheet material based on the type information and mechanical index parameters of the automobile sheet material.

[0089] That is, for the performance prediction model, the input data includes the type information and mechanical index parameters of the automobile sheet, and the output data includes the predicted performance range corresponding to at least one mechanical index parameter of the automobile sheet.

[0090] For example, input data of the performance prediction model may include: brand A, vehicle model B, front door inner panel (component), thickness 0.8, yield 153.53, tensile strength 3.9.45, elongation 46.35, N value 2.65, and R value 0.21.

[0091] The input data of the performance prediction model may include: the yield strength corresponds to a range of 162.4 to 169.7, the tensile strength ranges from 3.5.72 to 313.7, the elongation is 46.18 to 50.7, the N value is 2.59 to 2.78, and the R value is 0.24 to 0.25.

[0092] The specific training process for the performance prediction model will be further elaborated below.

[0093] Step S104 : selecting a target automobile sheet material from a plurality of candidate automobile sheet materials based on the target performance range and the predicted performance ranges corresponding to each candidate automobile sheet material.

[0094] Specifically, the predicted performance ranges corresponding to the candidate automobile sheet materials may be matched with the target performance range, and the candidate automobile sheet material corresponding to the predicted performance range that has the highest degree of matching with the target performance range may be used as the target automobile sheet material.

[0095] The above-mentioned method for selecting automobile plates, when the target type information of the target automobile plates required by the user is obtained, can first obtain multiple candidate automobile plates that meet the target type information, and then through the trained performance prediction model, it can combine the target type information and mechanical index parameters of each candidate automobile plate to predict the predicted performance range corresponding to each mechanical performance index of each candidate automobile plate, and then select the target automobile plate from multiple candidate automobile plates according to the target performance range required by the user and the predicted performance range of each candidate automobile plate, so that the target automobile plate that meets the target type information required by the user and meets the target performance range required by the user can be selected more accurately.

[0096] The specific training process of the performance prediction model will be further described below in conjunction with the embodiments.

[0097] In some possible implementations, the performance prediction model is obtained in the following manner:

[0098] (1) Obtain a sample dataset.

[0099] The sample data set includes a plurality of sample automobile panels, and each sample automobile panel includes corresponding sample type information, sample mechanical index parameters and sample reference performance range.

[0100] In a specific implementation process, the sample reference performance range of the sample automobile sheet material can be obtained by testing the sample automobile sheet material.

[0101] Specifically, obtaining a sample data set may include:

[0102] Get the initial sample data set;

[0103] The initial sample data set is standardized to obtain a sample data set.

[0104] Specifically, in the collected initial sample data set, the differences between the initial sample mechanical index parameters of each sample automobile sheet may vary greatly, for example, the tensile strength value is about 200 to 350, and the R value is about 0.1 to 0.3.

[0105] In order to improve the training efficiency and stability of the initial prediction model, the data can be standardized. Specifically, the standardization formula can refer to the following formula:

[0106]

[0107] Where Xnew is the sample mechanical index parameter of the sample automobile sheet after standardization; x is the initial sample mechanical index parameter of the sample automobile sheet, μ is the mean, and σ is the standard deviation.

[0108] like Figure 2 As shown, the initial yield strength of the sample automotive panels ranges from 140 to 180, and the yield strength after normalization ranges from -0.15 to 0.1.

[0109] (2) performing at least one optimization operation on the initial prediction model based on the sample data set until a preset end condition is met, and obtaining a prediction model based on the initial prediction model that meets the preset end condition.

[0110] Specifically, the optimization operation may include a training operation, that is, after the training operation, a prediction model is obtained; the optimization operation may also include a verification operation, that is, after at least one training operation, a verification operation is performed. In some possible implementations, the sample data set includes a training data set, and at least one optimization operation is performed on the initial prediction model based on the sample data set, including:

[0111] Performing at least one training operation on the initial prediction model based on the training data set, wherein the preset end condition includes that at least one training operation satisfies the training end condition;

[0112] The training operation includes:

[0113] ① Input the sample type information and sample mechanical index parameters of each sample automobile sheet in the training data set into the initial prediction model to obtain the sample prediction information.

[0114] ② For each sample automobile sheet, the first training loss of the sample automobile sheet is determined based on the difference between the first sample predicted performance range and the sample reference performance range.

[0115] ③ Based on the first training loss of each sample automobile plate, determine the total training loss.

[0116] ④ Adjust the parameters of the initial prediction model based on the total training loss, and use the initial prediction model after adjusting the parameters as the initial prediction model corresponding to the next training operation.

[0117] The sample prediction information includes a first sample prediction performance range.

[0118] The initial prediction model may include a BP (Back-Propagation Network) model, which is a multi-layer feedforward network trained by an error back-propagation algorithm.

[0119] Specifically, the initial prediction model includes an input layer, a hidden layer, and an output layer.

[0120] In the specific implementation process, the neural network can be trained through the back propagation algorithm, that is, according to the error of the output layer, it is back-propagated to the hidden layer, and the weights and biases of each layer are adjusted to reduce the error.

[0121] The total training loss can be calculated using the following formula:

[0122]

[0123] Among them, yi is the actual value, is the predicted value, n is the number of nodes in the output layer, and ε is the total training loss.

[0124] The parameters of the initial prediction model may include the weights of each layer of the initial prediction model and the gradient of the bias. In the specific implementation process, the parameters of the initial prediction model may be updated using the following formula:

[0125]

[0126] Among them, w iy is the weight from the i-th layer to the j-th layer, η is the learning rate, and E is the total training loss.

[0127] During the training process, the choice of learning rate is very critical. If the learning rate is too large, the model may diverge and fail to converge during the training process. If the learning rate is too small, the training speed will be very slow. Initially choose a small value of 0.01, and then adjust it according to the performance of the model on the validation set. The final learning rate is ideal at 0.03.

[0128] Specifically, the training end condition may be that the number of training times reaches a preset number of times, or that the total training loss converges, or that the total training loss is less than or equal to a first threshold.

[0129] In some possible implementations, each sample automobile sheet further includes sample forming limit data, the sample prediction information further includes predicted forming limit data, and the training operation further includes:

[0130] For each sample automobile sheet, a second training loss of the sample automobile sheet is determined based on a difference between the sample forming limit data and the predicted forming limit data.

[0131] In a specific implementation process, the process of determining the difference between the sample forming limit data and the predicted forming limit data can refer to the above formula (2), which will not be elaborated here.

[0132] When obtaining the sample forming limit data, the initial sample forming limit data may be obtained first, and then the initial sample forming limit data is standardized, such as Figure 3 As shown, the initial sample forming limit data of the sample automobile sheet material ranges from -0.1 to 0.1, and the sample forming limit data after normalization processing ranges from -2 to 2.

[0133] Based on the first training loss of each sample automobile plate, the total training loss is determined, including:

[0134] The total training loss is determined based on the first training loss of each sample automobile sheet and the second training loss of each sample automobile sheet.

[0135] Specifically, in addition to determining the total training loss based on the difference between the first sample predicted performance range and the sample reference performance range, the total training loss may also be calculated in combination with the difference between the sample forming limit data and the predicted forming limit data.

[0136] Specifically, the sum of the first training loss and the second training loss can be used as the total training loss.

[0137] In the above embodiment, based on the difference between the first sample predicted performance range and the sample reference performance range, the first training loss of the sample automobile sheet is determined, based on the difference between the sample forming limit data and the predicted forming limit data, the second training loss of the sample automobile sheet is determined, and then based on the first training loss of each sample automobile sheet and the second training loss of each sample automobile sheet, the total training loss is determined, so that the performance prediction model obtained by training can take into account the prediction ability of both the performance range and the forming limit data.

[0138] In the above embodiment, the initial prediction model is trained by using a training data set until the training end condition is met, and a trained performance prediction model can be obtained.

[0139] In other implementations, the initial prediction model may be trained using a training data set until a training end condition is met, and then the initial prediction model that meets the training end condition is verified using a verification data set to obtain a performance prediction model.

[0140] In some possible implementations, the sample data set further includes a validation data set, and performing at least one optimization operation on the initial prediction model based on the sample data set further includes:

[0141] Based on the verification data set, at least one verification operation is performed on the initial prediction model that meets the training end condition, and the preset end condition also includes that at least one verification operation meets the verification end condition.

[0142] That is to say, the preset end conditions include: at least one training operation meets the training end conditions, and the initial prediction model that meets the training end conditions passes the verification of the verification data set, that is, at least one verification operation meets the verification end conditions.

[0143] The verification end condition may be that the verification result is less than or equal to the second threshold, or that the verification result converges.

[0144] The verification operations include:

[0145] (1) The sample type information and sample mechanical index parameters of each sample automobile sheet in the validation data set are input into the first prediction model to obtain the second sample prediction performance range.

[0146] The first prediction model is an initial prediction model that meets the training end condition.

[0147] That is, when the initial prediction model has undergone at least one training operation and meets the training end conditions, the verification operation begins.

[0148] (2) Determine the verification result based on the difference between the second sample predicted performance range and the sample reference performance range of each sample automobile sheet in the verification data set.

[0149] Specifically, the process of determining the difference between the second sample predicted performance range and the sample reference performance range of each sample automobile sheet material may also refer to the above formula (2).

[0150] (3) Based on the verification results, the hyperparameters of the initial prediction model are adjusted, and the initial prediction model after the hyperparameter adjustment is used as the initial prediction model corresponding to the next optimization operation.

[0151] The initial prediction model includes an output layer, a hidden layer and an output layer, and the hyperparameters of the initial prediction model include the number of nodes in the hidden layer.

[0152] It should be noted that what is adjusted based on the verification results are the hyperparameters of the initial prediction model. That is, each time the hyperparameters of the initial prediction model are adjusted, the steps of training the initial prediction model at least once and verifying the initial prediction model that meets the training end conditions are repeated.

[0153] Specifically, the hyperparameters of the initial prediction model are adjusted based on the verification results, including:

[0154] Based on the validation results, the number of nodes in the hidden layer is adjusted within a preset range.

[0155] The preset range is determined based on the number of nodes in the input layer and the number of nodes in the output layer.

[0156] Specifically, the number of nodes in the input layer is determined based on the number of parameters corresponding to the mechanical index parameters, the middle value of the preset range is positively correlated with the number of nodes in the input layer, and the middle value of the preset range is positively correlated with the number of nodes in the output layer.

[0157] Specifically, the middle value of the preset range can be determined by referring to the following formula:

[0158]

[0159] Among them, n is the number of hidden layer nodes, m is the number of input layer nodes, l is the number of output layer nodes, and α is a constant between 1 and 10.

[0160] In the specific implementation process, the number of nodes in the hidden layer of the initial prediction model can be determined based on formula (4) and a preset range can be determined. After training and verification operations, the number of nodes in the hidden layer can be adjusted within the preset range, and training and verification operations can be performed again.

[0161] The above-mentioned method for selecting automobile sheet materials will be described below with reference to specific examples.

[0162] In an example, taking the target type information including brand A, vehicle model B, component being the front door inner panel, brand number C, and material performance index of thickness 0.8 as an example, a detailed description of this application may include the following process:

[0163] Sample data acquisition: obtain the sample mechanical properties of multiple sample automobile plates in the database, including yield strength, tensile strength, elongation, N value and R value;

[0164] Standardize the acquired data;

[0165] The standardized data is randomly divided into a training data set and a validation data set in a ratio of 8:2; the initial prediction model is trained based on the training data set; wherein the initial prediction model includes an input layer, a hidden layer and an output layer, and the input layer of the neural network structure adopted in this patent includes 5 neurons, each neuron corresponds to the yield strength, tensile strength, elongation, N value and R value; the output layer contains 1 neuron, namely the forming limit;

[0166] The training set data is imported into the initial prediction model, and the initial prediction model is trained. When the training end conditions are met, the validation data set is used to verify the initial prediction model that meets the training end conditions. In the specific training process, the hidden layer is set to have 9 neurons initially, and the learning rate is set to 0.2. The layers are connected by weights w and biases b, and the neurons in the same layer are not connected. The error back propagation algorithm is used to solve the update amount of w and b for each sample; the weight is updated using the following formula:

[0167] w′(ij)=w(ij)+l*Err j *O i (5)

[0169] b′(ij)=b(ij)+l*Err j (6)

[0171] Where unit i is connected to unit j, where l is the learning rate, Err j is the error of unit j, O i is the output of unit i, w′ is the new weight, and b′ is the new bias.

[0172] If the verification result meets the verification end conditions, then the optimal model, that is, the final performance prediction model, can be obtained; if the verification result does not meet the verification end conditions, then the steps of adjusting the hyperparameters of the initial prediction model (adjusting the number of hidden layer nodes), training the initial prediction model, and verifying the initial prediction model that meets the training end conditions are repeated until the training end conditions and the verification end conditions are met to obtain the performance prediction model.

[0173] The following describes an embodiment of the device of the present application, which can be used to execute the method in the above embodiment of the present application. For details not disclosed in the embodiment of the device of the present application, please refer to the embodiment of the method in the above embodiment of the present application. In some possible implementations of the present application, such as Figure 4 As shown, a device 40 for selecting automobile plates is provided, comprising:

[0174] The information acquisition module 401 is used to acquire the target type information and target performance range of the target automobile plate to be selected;

[0175] The plate acquisition module 402 is used to acquire multiple candidate automobile plates that meet the target type information, and acquire mechanical index parameters corresponding to each candidate automobile plate;

[0176] A prediction module 403 is used to predict the predicted performance range of each candidate automobile sheet material through a trained performance prediction model based on target type information and mechanical index parameters;

[0177] The selection module 404 is used to select a target automobile sheet material from a plurality of candidate automobile sheets based on the target performance range and the predicted performance ranges corresponding to each candidate automobile sheet material.

[0178] In some possible implementations, the device further includes a training module for:

[0179] Acquire a sample data set; the sample data set includes a plurality of sample automobile plates, each sample automobile plate includes corresponding sample type information, sample mechanical index parameters and sample reference performance range;

[0180] The initial prediction model is optimized at least once based on the sample data set until a preset end condition is met, and a prediction model is obtained based on the initial prediction model that meets the preset end condition.

[0181] In some possible implementations, the sample data set includes a training data set, and the training module is specifically used to:

[0182] Performing at least one training operation on the initial prediction model based on the training data set, wherein the preset end condition includes that at least one training operation satisfies the training end condition;

[0183] The training operation includes:

[0184] Inputting sample type information and sample mechanical index parameters of each sample automobile sheet in the training data set into the initial prediction model to obtain sample prediction information, wherein the sample prediction information includes a first sample prediction performance range;

[0185] For each sample automobile sheet, determining a first training loss of the sample automobile sheet based on a difference between a first sample predicted performance range and a sample reference performance range;

[0186] Determine the total training loss based on the first training loss of each sample automobile plate;

[0187] The parameters of the initial prediction model are adjusted based on the total training loss, and the initial prediction model after the adjustment of the parameters is used as the initial prediction model corresponding to the next training operation.

[0188] In some possible implementations, each sample automobile sheet further includes sample forming limit data, the sample prediction information further includes predicted forming limit data, and the training module, when performing the training operation, is further used to: for each sample automobile sheet, determine a second training loss of the sample automobile sheet based on a difference between the sample forming limit data and the predicted forming limit data;

[0189] When the training module determines the total training loss based on the first training loss of each sample automobile sheet, it is specifically used to:

[0190] The total training loss is determined based on the first training loss of each sample automobile sheet and the second training loss of each sample automobile sheet.

[0191] In some possible implementations, the sample data set further includes a verification data set, and the training module is further used to: perform at least one verification operation on the initial prediction model that meets the training end condition based on the verification data set, and the preset end condition further includes at least one verification operation meeting the verification end condition;

[0192] When performing verification operations, the training module is specifically used to:

[0193] Input the sample type information and sample mechanical index parameters of each sample automobile sheet in the validation data set into the first prediction model to obtain the second sample prediction performance range, wherein the first prediction model is an initial prediction model that meets the training end condition;

[0194] Determining a validation result based on a difference between a second sample predicted performance range and a sample reference performance range for each sample automotive sheet in the validation data set;

[0195] The hyperparameters of the initial prediction model are adjusted based on the verification results, and the initial prediction model after the hyperparameters are adjusted is used as the initial prediction model corresponding to the next optimization operation.

[0196] In some possible implementations, the initial prediction model includes an output layer, a hidden layer, and an output layer, and the hyperparameters of the initial prediction model include the number of nodes in the hidden layer;

[0197] The training module adjusts the hyperparameters of the initial prediction model based on the validation results. Specifically, it is used to:

[0198] Based on the verification result, the number of nodes in the hidden layer is adjusted within a preset range, wherein the preset range is determined based on the number of nodes in the input layer and the number of nodes in the output layer.

[0199] In some possible embodiments, the number of nodes in the input layer is determined based on the number of parameters corresponding to the mechanical index parameters, the middle value of the preset range is positively correlated with the number of nodes in the input layer, and the middle value of the preset range is positively correlated with the number of nodes in the output layer.

[0200] The above-mentioned automobile plate selection device, if obtaining the target type information of the target automobile plate required by the user, can first obtain multiple candidate automobile plates that meet the target type information, and then, through the trained performance prediction model, can combine the target type information and mechanical index parameters of each candidate automobile plate to predict the predicted performance range corresponding to each mechanical performance index of each candidate automobile plate, and then select the target automobile plate from multiple candidate automobile plates according to the target performance range required by the user and the predicted performance range of each candidate automobile plate, so that the target automobile plate that meets the target type information required by the user and meets the target performance range required by the user can be selected more accurately.

[0201] In an alternative embodiment, an electronic device is provided, such as Figure 5 As shown, Figure 5 The electronic device 4000 shown includes: a processor 4001 and a memory 4003. The processor 4001 and the memory 4003 are connected, such as through a bus 4002. Optionally, the electronic device 4000 may also include a transceiver 4004, which may be used for data interaction between the electronic device and other electronic devices, such as data transmission and / or data reception. It should be noted that in actual applications, the transceiver 4004 is not limited to one, and the structure of the electronic device 4000 does not constitute a limitation on the embodiments of the present application.

[0202] Processor 4001 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It may implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the disclosure of this application. Processor 4001 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0203] The bus 4002 may include a path for transmitting information between the above components. The bus 4002 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus 4002 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0204] The memory 4003 may be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, optical disk storage (including compressed optical disk, laser disk, optical disk, digital versatile disk, Blu-ray disk, etc.), magnetic disk storage media, other magnetic storage devices, or any other medium that can be used to carry or store computer programs and can be read by a computer, without limitation herein.

[0205] The memory 4003 is used to store the computer program for executing the embodiment of the present application, and the execution is controlled by the processor 4001. The processor 4001 is used to execute the computer program stored in the memory 4003 to implement the steps shown in the above method embodiment.

[0206] An embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps and corresponding contents of the aforementioned method embodiment can be implemented.

[0207] The embodiment of the present application also provides a computer program product, including a computer program, which can implement the steps and corresponding contents of the aforementioned method embodiment when executed by a processor.

[0208] It should be understood that, although each operation step is indicated by arrows in the flowchart of the embodiment of the present application, the implementation order of these steps is not limited to the order indicated by the arrows. Unless clearly stated herein, in some implementation scenarios of the embodiment of the present application, the implementation steps in each flowchart can be performed in other orders according to demand. In addition, some or all of the steps in each flowchart may include multiple sub-steps or multiple stages based on actual implementation scenarios. Some or all of these sub-steps or stages may be executed at the same time, and each sub-step or stage in these sub-steps or stages may also be executed at different times respectively. In different scenarios of execution time, the execution order of these sub-steps or stages may be flexibly configured according to demand, and the embodiment of the present application does not limit this.

[0209] The above are only optional implementation methods for some implementation scenarios of the present application. It should be pointed out that for ordinary technicians in this technical field, without departing from the technical concept of the scheme of the present application, other similar implementation methods based on the technical ideas of the present application are also within the protection scope of the embodiments of the present application.

Claims

1. A method for selecting automobile sheet materials, characterized in that: include: Obtain target type information and target performance range of target automobile sheet to be selected; Acquire multiple candidate automobile sheet materials that meet the target type information, and acquire mechanical index parameters corresponding to each candidate automobile sheet material; For each candidate automobile sheet material, based on the target type information and the mechanical index parameter, predicting the predicted performance range of the candidate automobile sheet material through the trained performance prediction model; Based on the target performance range and the predicted performance ranges corresponding to the candidate automobile sheet materials, a target automobile sheet material is selected from a plurality of candidate automobile sheet materials.

2. The method according to claim 1, characterized in that The performance prediction model is trained in the following way: Acquire a sample data set; the sample data set includes a plurality of sample automobile plates, each sample automobile plate includes corresponding sample type information, sample mechanical index parameters and sample reference performance range; The initial prediction model is optimized at least once based on the sample data set until a preset end condition is met, and the prediction model is obtained based on the initial prediction model that meets the preset end condition.

3. The method according to claim 2, characterized in that The sample data set includes a training data set, and performing at least one optimization operation on the initial prediction model based on the sample data set includes: performing at least one training operation on the initial prediction model based on the training data set, wherein the preset end condition includes that the at least one training operation satisfies the training end condition; The training operation includes: Inputting sample type information and sample mechanical index parameters of each sample automobile sheet in the training data set into the initial prediction model to obtain sample prediction information, wherein the sample prediction information includes a first sample prediction performance range; For each sample automobile plate, determining a first training loss of the sample automobile plate based on a difference between the first sample predicted performance range and the sample reference performance range; Determine the total training loss based on the first training loss of each sample automobile plate; The parameters of the initial prediction model are adjusted based on the total training loss, and the initial prediction model after the adjustment of the parameters is used as the initial prediction model corresponding to the next training operation.

4. The method according to claim 3, characterized in that Each of the sample automobile sheet materials further includes sample forming limit data, the sample prediction information further includes predicted forming limit data, and the training operation further includes: For each sample automobile sheet, determining a second training loss of the sample automobile sheet based on a difference between the sample forming limit data and the predicted forming limit data; The first training loss based on each sample automobile plate determines the total training loss, including: The total training loss is determined based on the first training loss of each sample automobile sheet and the second training loss of each sample automobile sheet.

5. The method according to claim 3, characterized in that: The sample data set also includes a validation data set, and performing at least one optimization operation on the initial prediction model based on the sample data set further includes: performing at least one validation operation on the initial prediction model that meets the training end condition based on the validation data set, and the preset end condition further includes that the at least one validation operation meets the validation end condition; The verification operation includes: Input the sample type information and sample mechanical index parameters of each sample automobile sheet in the validation data set into the first prediction model to obtain the second sample prediction performance range, wherein the first prediction model is an initial prediction model that meets the training end condition; Determining a validation result based on a difference between a second sample predicted performance range and a sample reference performance range for each sample automotive sheet in the validation data set; The hyperparameters of the initial prediction model are adjusted based on the verification result, and the initial prediction model after the hyperparameters are adjusted is used as the initial prediction model corresponding to the next optimization operation.

6. The method according to claim 5, characterized in that The initial prediction model includes an output layer, a hidden layer and an output layer, and the hyperparameters of the initial prediction model include the number of nodes in the hidden layer; and adjusting the hyperparameters of the initial prediction model based on the verification result includes: Based on the verification result, the number of nodes in the hidden layer is adjusted within a preset range, wherein the preset range is determined based on the number of nodes in the input layer and the number of nodes in the output layer.

7. The method according to claim 6, characterized in that The number of nodes in the input layer is determined based on the number of parameters corresponding to the mechanical index parameters, the middle value of the preset range is positively correlated with the number of nodes in the input layer, and the middle value of the preset range is positively correlated with the number of nodes in the output layer.

8. A device for selecting automobile plates, characterized in that: include: An information acquisition module, used to acquire target type information and target performance range of a target automobile sheet to be selected; A plate acquisition module is used to acquire multiple candidate automobile plates that meet the target type information, and acquire mechanical index parameters corresponding to each candidate automobile plate; A prediction module, configured to predict, for each candidate automobile sheet material, a predicted performance range of the candidate automobile sheet material through a trained performance prediction model based on the target type information and the mechanical index parameter; The selection module is used to select a target automobile sheet material from a plurality of candidate automobile sheets based on the target performance range and the predicted performance ranges corresponding to each candidate automobile sheet material.

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

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images