Data processing method, device, apparatus, readable storage medium and computer program product

Abstract

The present application relates to a data processing method, device, computer equipment, computer-readable storage medium and computer program product. The method comprises: obtaining each overall modeling result of the target object and the local modeling result corresponding to the overall modeling result, and integrating the local modeling result into the overall modeling result to obtain multiple target finite element models; performing model analysis on the target finite element model to obtain the structural response index corresponding to each target finite element model; training the initial neural network model based on the structural response index, the overall structural parameters and the local structural parameters in the target finite element model to obtain the trained target neural network model; obtaining the target result according to the target neural network model, the initialization structural parameters of the target object and the preset calculation step size. The use of this method can improve the accuracy of the data processing results.

Description

Technical Field

[0001] The present application relates to the technical field of bridge engineering, and in particular to a data processing method, device, computer equipment, computer-readable storage medium and computer program product. Background Art

[0002] Before the construction of a cable-stayed bridge, it is necessary to design a reasonable side-to-mid-span ratio and the location of the steel-concrete joint. Different side-to-mid-span ratios and the location of the steel-concrete joint may lead to large differences in the overall force mechanism of the bridge. Therefore, it is necessary to determine a reasonable side-to-mid-span ratio and the location of the steel-concrete joint.

[0003] In traditional technology, for the determination of the side-to-mid-span ratio and the position of the steel-concrete junction, designers generally roughly determine the numerical range of the overall modeling results based on their experience, and determine multiple sets of parameters within the numerical range according to the designers' prior experience. A finite element model is constructed based on the multiple sets of parameters, and a better side-to-mid-span ratio and the position of the steel-concrete junction are determined as the target result based on the finite element models under different parameters.

[0004] However, in traditional technology, the method of manually determining the side-to-midspan ratio and the location of the steel-concrete joint based on experience is highly subjective, resulting in poor accuracy in determining the target results. Summary of the invention

[0005] Based on this, it is necessary to provide a data processing method, apparatus, computer device, computer-readable storage medium and computer program product to address the above technical issues.

[0006] In a first aspect, the present application provides a data processing method, comprising:

[0007] Acquire each overall modeling result of the target object and a local modeling result corresponding to the overall modeling result, and integrate the local modeling results into the overall modeling result to obtain a plurality of target finite element models;

[0008] Performing model analysis on the target finite element models to obtain structural response indicators corresponding to each of the target finite element models;

[0009] Training the initial neural network model based on the structural response index, the overall structural parameters and the local structural parameters in the target finite element model to obtain a trained target neural network model;

[0010] The target result is obtained according to the target neural network model, the initialization structure parameters of the target object and the preset calculation step size.

[0011] In one embodiment, before obtaining each overall modeling result of the target object and the local modeling result corresponding to the overall modeling result, the method further includes:

[0012] Acquire multiple overall structural parameters of the target object, and perform modeling analysis on the overall structural parameters based on the first finite element analysis model to obtain multiple overall modeling results; the overall structural parameters correspond to multiple different preset parameter values;

[0013] Determine the local structural parameters corresponding to each of the overall structural parameters according to the overall modeling result;

[0014] Modeling and analysis are performed on the local structural parameters based on the second finite element analysis model to obtain local modeling results corresponding to the overall modeling results.

[0015] In one embodiment, the integrating the local modeling results into the overall modeling results to obtain multiple target finite element models includes:

[0016] Unifying the formats of the overall modeling result and the local modeling result to obtain a first target modeling result and a second target modeling result after the formats are unified;

[0017] Each of the second target modeling results is integrated into the first target modeling result corresponding to the second target modeling result to obtain a target finite element model corresponding to each target parameter.

[0018] In one embodiment, the training of the initial neural network model based on the structural response index, the overall structural parameters and the local structural parameters in the target finite element model to obtain the trained target neural network model includes:

[0019] Predicting the overall structural parameters and local structural parameters in each of the target finite element models according to the initial neural network model to obtain prediction values ​​corresponding to each of the target finite element models;

[0020] Performing mean square error calculation based on the predicted value and the structural response index to obtain a loss value of the initial neural network model;

[0021] The model parameters of the initial neural network model are adjusted according to Bayesian regularization back propagation until the loss value meets the preset loss condition, thereby obtaining a trained target neural network model.

[0022] In one embodiment, obtaining a target result according to the target neural network model, the initialization structure parameters of the target object and the preset calculation step size includes:

[0023] Obtain static structural parameters of the target object, initialize target parameters and preset calculation step size;

[0024] Iteratively calculate the initialized target parameters according to a preset calculation step length to obtain multiple candidate target parameters; the static structure parameters include overall static structure parameters and local static structure parameters;

[0025] The candidate target parameters are screened according to the target neural network model to obtain a target result.

[0026] In one embodiment, the screening of the candidate target parameters according to the target neural network model to obtain the target result includes:

[0027] Predicting each of the candidate target parameters according to the target neural network model to obtain a structural response index corresponding to each of the candidate target parameters;

[0028] The structural response index corresponding to each of the candidate target parameters is screened according to a preset screening condition to determine a target result in which the structural response index meets the preset screening condition.

[0029] In one embodiment, obtaining a target result according to the target neural network model, the initialization structure parameters of the target object and the preset calculation step size includes:

[0030] Obtaining static structural parameters of the target object, initializing target parameters and preset calculation steps, and constructing an integrated impact model according to the target neural network model;

[0031] Iteratively calculating the initialization target parameter based on the integrated impact model according to a preset calculation step length to obtain a plurality of candidate target parameters;

[0032] The candidate target parameters are screened based on the integrated impact model to obtain the target result. In a second aspect, the present application also provides a data processing device, including:

[0033] An acquisition module, used to acquire each overall modeling result of the target object and a local modeling result corresponding to the overall modeling result, and integrate the local modeling results into the overall modeling result to obtain a plurality of target finite element models;

[0034] An analysis module, used to perform model analysis on the target finite element models to obtain structural response indicators corresponding to each of the target finite element models;

[0035] A training module, used to train the initial neural network model based on the structural response index, the overall structural parameters and the local structural parameters in the target finite element model, to obtain a trained target neural network model;

[0036] The calculation module is used to obtain the target result according to the target neural network model, the initialization structure parameters of the target object and the preset calculation step size.

[0037] In one embodiment, the device further comprises:

[0038] A first modeling module is used to obtain a plurality of overall structural parameters of the target object, and to perform modeling and analysis on the overall structural parameters based on a first finite element analysis model to obtain a plurality of overall modeling results; the overall structural parameters correspond to a plurality of different preset parameter values;

[0039] A determination module, used to determine the local structural parameters corresponding to each of the overall structural parameters according to the overall modeling result;

[0040] The second modeling module is used to perform modeling analysis on the local structural parameters based on a second finite element analysis model to obtain local modeling results corresponding to each of the overall modeling results.

[0041] In one embodiment, the acquisition module is specifically used to unify the formats of the overall modeling result and the local modeling result to obtain the first target modeling result and the second target modeling result after the formats are unified;

[0042] Each of the second target modeling results is integrated into the first target modeling result corresponding to the second target modeling result to obtain a target finite element model corresponding to each target parameter.

[0043] In one embodiment, the training module is specifically used to predict the overall structural parameters and local structural parameters in each of the target finite element models according to the initial neural network model to obtain the prediction values ​​corresponding to each of the target finite element models;

[0044] Performing mean square error calculation based on the predicted value and the structural response index to obtain a loss value of the initial neural network model;

[0045] The model parameters of the initial neural network model are adjusted according to Bayesian regularization back propagation until the loss value meets the preset loss condition, thereby obtaining a trained target neural network model.

[0046] In one of the embodiments, the calculation module is specifically used to obtain static structural parameters of the target object, initialize target parameters and preset calculation steps;

[0047] Iteratively calculate the initialized target parameters according to a preset calculation step length to obtain multiple candidate target parameters; the static structure parameters include overall static structure parameters and local static structure parameters;

[0048] The candidate target parameters are screened according to the target neural network model to obtain a target result.

[0049] In one embodiment, the calculation module is specifically used to perform prediction processing on each of the candidate target parameters according to the target neural network model to obtain a structural response index corresponding to each of the candidate target parameters;

[0050] The structural response index corresponding to each of the candidate target parameters is screened according to a preset screening condition to determine a target result in which the structural response index meets the preset screening condition.

[0051] In one embodiment, the calculation module is specifically used to obtain static structural parameters of the target object, initialize target parameters and preset calculation steps, and construct an integrated influence model according to the target neural network model;

[0052] Iteratively calculating the initialization target parameter based on the integrated impact model according to a preset calculation step length to obtain a plurality of candidate target parameters;

[0053] The candidate target parameters are screened based on the integrated impact model to obtain target results.

[0054] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0055] Acquire each overall modeling result of the target object and a local modeling result corresponding to the overall modeling result, and integrate the local modeling results into the overall modeling result to obtain a plurality of target finite element models;

[0056] Performing model analysis on the target finite element models to obtain structural response indicators corresponding to each of the target finite element models;

[0057] Training the initial neural network model based on the structural response index, the overall structural parameters and the local structural parameters in the target finite element model to obtain a trained target neural network model;

[0058] The target result is obtained according to the target neural network model, the initialization structure parameters of the target object and the preset calculation step size.

[0059] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0060] Acquire each overall modeling result of the target object and a local modeling result corresponding to the overall modeling result, and integrate the local modeling results into the overall modeling result to obtain a plurality of target finite element models;

[0061] Performing model analysis on the target finite element models to obtain structural response indicators corresponding to each of the target finite element models;

[0062] Training the initial neural network model based on the structural response index, the overall structural parameters and the local structural parameters in the target finite element model to obtain a trained target neural network model;

[0063] The target result is obtained according to the target neural network model, the initialization structure parameters of the target object and the preset calculation step size.

[0064] In a fifth aspect, the present application further provides a computer program product, including a computer program, which implements the following steps when executed by a processor:

[0065] Acquire each overall modeling result of the target object and a local modeling result corresponding to the overall modeling result, and integrate the local modeling results into the overall modeling result to obtain a plurality of target finite element models;

[0066] Performing model analysis on the target finite element models to obtain structural response indicators corresponding to each of the target finite element models;

[0067] Training the initial neural network model based on the structural response index, the overall structural parameters and the local structural parameters in the target finite element model to obtain a trained target neural network model;

[0068] The target result is obtained according to the target neural network model, the initialization structure parameters of the target object and the preset calculation step size.

[0069] The above-mentioned data processing method, device, computer equipment, computer-readable storage medium and computer program product, by combining local modeling results with overall modeling results, can not only compensate for the overall modeling results by local complex structural construction and force transmission mechanism, but also reduce the amount of calculation for detailed modeling of the whole, thereby reducing the consumption of computing resources for constructing the target finite element model; the initial neural network model is trained by the target finite element model and the structural response index, and the target parameters of the target object are processed by the target neural network model, which can replace the modeling process of the target finite element model using a large number of target parameters with different parameter values, and can reduce the computational complexity and improve the computational efficiency, and then perform structural response index analysis based on a large number of target parameters with different parameter values, thereby realizing parameter analysis for a larger range and preventing the omission of the optimal parameter value, and performing analysis of the structural response index through the target neural network model, thereby avoiding the subjectivity of parameter analysis based on human experience, and further improving the accuracy of the data processing results. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0071] Figure 1 is a flow chart of a data processing method in one embodiment;

[0072] Figure 2 A schematic diagram of a process for constructing overall modeling results and local modeling results in one embodiment;

[0073] Figure 3 A schematic diagram of a flow chart of steps for integrating overall modeling results and local modeling results in one embodiment;

[0074] Figure 4 A schematic diagram of a process for training an initial neural network model in one embodiment;

[0075] Figure 5 A schematic diagram of a process for generating candidate target parameters and determining target results in one embodiment;

[0076] Figure 6 A schematic diagram of a process for determining a target result according to a target neural network model in one embodiment;

[0077] Figure 7 A schematic diagram of a process for constructing and applying an integrated impact model in an embodiment;

[0078] Figure 8 A schematic diagram of a model layer of a target neural network model in another embodiment;

[0079] Fig. 9 is a structural block diagram of a data processing device in one embodiment;

[0080] Fig.10 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0081] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0082] In one embodiment, Figure 1 As shown, a data processing method is provided. This embodiment takes the method applied to a terminal as an example. It can be understood that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0083] Step 102 , obtaining the overall modeling results of the target object and the local modeling results corresponding to the overall modeling results, and integrating the local modeling results into the overall modeling results to obtain a plurality of target finite element models.

[0084] In the embodiments of the present application, a hybrid beam cable-stayed bridge is used as an example for explanation. During the design stage of the hybrid beam cable-stayed bridge, the side-to-mid span ratio and the position of the steel-concrete joint are key parameters that affect the overall force mechanism of the bridge. Not only are the required construction counterweights different, but the internal forces of the structure will also be greatly different during operation and in the completed bridge state. Therefore, in the design of the hybrid beam cable-stayed bridge, it is necessary to determine an appropriate side-to-mid span ratio and the position of the steel-concrete joint.

[0085] The overall modeling result is the full-bridge modeling result of the target hybrid beam cable-stayed bridge (hereinafter referred to as the target bridge), where the overall modeling result of the target bridge is a number of overall modeling results obtained when using different side-to-span ratios and steel-concrete joint locations. The overall modeling result refers to the complete finite element model of the overall structure of the target bridge, which includes the overall geometry, material properties, boundary conditions, and loading conditions of the structure, and is used to analyze the global response of the overall structure of the target bridge, such as overall displacement, overall force distribution, etc.

[0086] The local modeling result is a local refined model constructed for the local structure of the target bridge. The local structure includes the structure of the steel-concrete joint, the distribution of components, etc., which is used to refine the local structure into smaller components or areas to capture the impact of local details on the overall structural behavior. Among them, the side-to-midspan ratio and the position of the steel-concrete joint directly affect the stress distribution inside the local structure (for example, the steel-concrete joint); different side-to-midspan ratios will lead to different force paths and stress concentration points in the steel-concrete joint. Therefore, different overall modeling results correspond to different local modeling results.

[0087] For the target bridge (target object), a series of design parameter schemes covering the conventional value range of the parameters can be pre-set. The terminal obtains the above-mentioned multiple design parameter schemes and performs overall and local modeling to obtain the overall modeling results and the local modeling results corresponding to the overall modeling results. On this basis, the overall modeling results and the local modeling results are not isolated. The terminal can associate the overall modeling results and the local modeling results with each other through engineering methods and software tools. The local modeling results are usually integrated into the overall modeling results through appropriate methods (for example, boundary conditions or connection conditions at nodes) to ensure the consistency and accuracy of the overall model.

[0088] Step 104 , performing model analysis on the target finite element models to obtain structural response indicators corresponding to each target finite element model.

[0089] In an embodiment of the present application, the terminal performs model analysis on the target finite element model according to the finite element automatic modeling program. During this process, the structural response index mainly reflects the overall performance of the target bridge. Therefore, the terminal can perform model analysis according to the finite element automatic modeling program in the finite element model used for overall modeling to obtain structural response indicators corresponding to different design schemes, wherein the structural response indicators include internal forces and displacements of key parts in the construction and bridge completion states, counterweights required for construction, cable tensions in the bridge completion state, etc. The structural response indicators are used to train the following initial neural network model.

[0090] Step 106, training the initial neural network model based on the structural response index, the overall structural parameters and the local structural parameters in the target finite element model to obtain a trained target neural network model.

[0091] In the embodiment of the present application, the initial neural network model is an ANN (Artificial Neural Networks) model, the terminal uses the structural response index as the sample label of the training sample, uses the overall structural parameters and local structural parameters in the target finite element model as the training samples of the initial neural network model, trains the model parameters of the initial neural network, optimizes the weights and thresholds of the neural network, and obtains the trained target neural network model. The target neural network model is used to quickly determine the structural response index based on the input structural parameters.

[0092] In addition, since the structural parameters, material properties, load conditions, etc. of different bridges may vary greatly, their structural behaviors and responses may also vary. Therefore, personalized customized training is required for each specific hybrid beam cable-stayed bridge (target object) to ensure that the target neural network model can accurately predict the structural response of the bridge. In customized training of the target object, the training data sets of different bridges may have different characteristics and distributions. Training a single target neural network model can better adapt to the data distribution characteristics of the bridge and improve the prediction accuracy of the model.

[0093] Step 108, obtaining the target result according to the target neural network model, the initialization structure parameters of the target object and the preset calculation step size.

[0094] In the embodiment of the present application, the terminal predicts the structural response index for the initialization structural parameters and preset calculation steps input to the target neural network based on the nonlinear relationship between the overall structural parameters and local structural parameters and the structural response index learned by the target neural network model, thereby replacing the process of modeling using a finite element model. Then, based on the structural response index output by the target neural network model, the stress distribution of the target bridge under different design parameters is analyzed, especially the stress concentration in key parts such as the steel-concrete joint, so as to evaluate the bearing capacity and fatigue life of the target bridge, and evaluate the overall displacement of the bridge under different parameter configurations, including structural deformation, deflection, etc., to determine whether the design requirements and use conditions are met. Finally, the terminal determines the side-to-mid span ratio and the position of the steel-concrete joint that meet the relevant performance of the target bridge, and obtains the target result.

[0095] In the above data processing method, by combining the local modeling results with the overall modeling results, the local complex structural structure and force transmission mechanism can be used to compensate for the overall modeling results, and the calculation amount of detailed modeling of the whole can be reduced, thereby reducing the consumption of computing resources for constructing the target finite element model; the initial neural network model is trained by the target finite element model and the structural response index, and the target parameters of the target object are processed by the target neural network model, which can replace the modeling process of the target finite element model using a large number of target parameters with different parameter values, and can reduce the calculation complexity and improve the calculation efficiency. Then, the structural response index analysis is performed according to the target parameters with a large number of different parameter values, realizing parameter analysis for a larger range and preventing the omission of the optimal parameter value. The structural response index analysis is performed through the target neural network model, avoiding the subjectivity of parameter analysis based on experience, and further improving the accuracy of data processing.

[0096] In an exemplary embodiment, before the terminal obtains the overall modeling result and the local modeling result corresponding to the overall modeling result, it is necessary to model the overall structure and the local structure of the target object, such as Figure 2 As shown, before step 102, the method further includes steps 202 to 206. Among them:

[0097] Step 202 , obtaining a plurality of overall structural parameters of the target object, and performing modeling analysis on the overall structural parameters based on a first finite element analysis model to obtain a plurality of overall modeling results.

[0098] Among them, the overall structural parameters correspond to multiple different preset parameter values.

[0099] In the embodiment of the present application, the first finite element analysis model can be a Midas model (a finite element analysis software for global modeling), and the overall structural parameters include static structural parameters and variable structural parameters as variables. Static structural parameters refer to structural parameters that are not affected by the side-to-mid span ratio and the position of the steel-concrete joint. Variable structural parameters refer to structural parameters that change with the mid-span ratio and the position of the steel-concrete joint. The terminal can pre-set multiple variable structural parameters with conventional value ranges to obtain multiple overall structural parameters. When the overall structural parameters are determined, the terminal inputs each overall structural parameter into the Midas model, performs modeling processing through the Midas model, and constructs the overall modeling result of the target bridge.

[0100] Step 204: determine the local structural parameters corresponding to each overall structural parameter according to the overall modeling result.

[0101] In an embodiment of the present application, the terminal determines the influence of different design parameters on the stress of the local structure (such as the steel-concrete joint) based on the overall modeling results, including analyzing how the overall structure applies stress at each key node or position, for example, determining the stress concentration area and the load transfer path, and then obtaining the local structural parameters corresponding to each overall structural parameter.

[0102] Step 206 , modeling and analyzing the local structural parameters based on the second finite element analysis model to obtain local modeling results corresponding to each overall modeling result.

[0103] The second finite element analysis model may be an Abaqus model (a finite element analysis software used for detailed modeling).

[0104] In the embodiment of the present application, although the fine model can better reflect the structure and force transmission mechanism, the amount of calculation is large. If it is used for full-bridge modeling, the modeling cost and analysis efficiency will be very low. Therefore, the terminal inputs the local structural parameters (determined key nodes, parts and external force influences) into the Abaqus model, performs modeling and analysis on local details, obtains specific connection details, local load transfer paths and stress states at specific locations for fine modeling, and obtains local modeling results for the steel-concrete joint structure.

[0105] In this embodiment, overall modeling provides a basis for determining the overall impact on local forces, while local modeling further refines the analysis of key parts or nodes, thereby comprehensively evaluating the performance of the target object under different design parameters, reducing the difference between the calculation model and the actual situation, and reducing the computing resource consumption of constructing the target finite element model, and ensuring that the target finite element model can accurately reflect the performance of the target object.

[0106] In an exemplary embodiment, before the terminal imports and integrates the local modeling results with the overall modeling results, it is necessary to ensure the data compatibility and consistency of different finite element analysis models, such as Figure 3 As shown, step 102 includes steps 302 to 304. Among them:

[0107] Step 302: unify the formats of the overall modeling results and the local modeling results to obtain the first target modeling results and the second target modeling results after unification.

[0108] In the embodiment of the present application, first, when performing local analysis in the Abaqus model, it is necessary to select a suitable output file format, and select the output as a text file (such as text post-processing output) or a specific binary file format, so as to be subsequently imported and parsed in the Midas model. When exporting data, the terminal unifies the format of the overall modeling results and the local modeling results to ensure that the data format output by the Abaqus model matches the format that the Midas model can recognize and process, involving the naming of fields, the unification of units, and the structured organization of data, to obtain the first target modeling results and the second target modeling results after the format is unified.

[0109] In an optional embodiment, after importing the data, some parameter adjustments and data calculations are required to ensure that the local analysis results of the Abaqus model are compatible with the overall calculations of the Midas model. First, for the node coordinates and displacement data, it is necessary to ensure that the node coordinates and displacement data imported from Abaqus are converted to the node coordinate system in the Midas model or the coordinate system is adjusted to maintain the consistency of the node coordinate system.

[0110] Step 304: Integrate each second target modeling result into the first target modeling result corresponding to the second target modeling result to obtain a target finite element model corresponding to each target parameter.

[0111] In an embodiment of the present application, the terminal imports the boundary conditions and material properties of the nodes, units or directly integrated local models contained in the second target modeling results into the first target modeling results based on the correspondence between the second target modeling results and the first target modeling results, completes the integration of the first target modeling results and the second target modeling results, and obtains the target finite element model corresponding to different target parameters.

[0112] In this embodiment, the overall modeling results and the local modeling results are formatted in a unified manner to ensure that they can form a complete and reasonable target finite element model after integration, thereby effectively revealing the influence of different design parameters on the overall and local structural performance of the hybrid beam cable-stayed bridge, improving the accuracy of the training samples for constructing the initial neural network model, and further improving the output accuracy of the trained target neural network model.

[0113] In an exemplary embodiment, Figure 4 As shown, step 106 includes steps 402 to 406. Among them:

[0114] Step 402 , predicting the overall structural parameters and local structural parameters in each target finite element model according to the initial neural network model, and obtaining the prediction values ​​corresponding to each target finite element model.

[0115] In an embodiment of the present application, the terminal takes the overall structural parameters and local structural parameters in each target finite element model as input, performs prediction processing through the initial neural network model, and for each target finite element model, inputs its specific structural parameters into the target neural network model to obtain the predicted structural response index, and obtains the predicted value corresponding to the target finite element model under different design parameters.

[0116] Step 404, performing mean square error calculation based on the predicted value and the structural response index to obtain the loss value of the initial neural network model.

[0117] In the embodiment of the present application, the terminal compares the predicted structural response index with the actual target finite element model analysis result (i.e., the sample label corresponding to the training sample) to determine the loss value of the initial neural network, wherein the performance function (PF, i.e., the loss function) of the initial neural network model is the network mean square error (MSE, i.e., MSE) based on the training set, and the performance function is as follows:

[0118]

[0119] Where T is the number of samples in the training set; is the network error of the i-th training sample in the training sample set, which is equal to the target true value With network prediction value deviation.

[0120] Step 406, adjusting the model parameters of the initial neural network model according to Bayesian regularization back propagation until the loss value meets the preset loss condition, thereby obtaining a trained target neural network model.

[0121] In an embodiment of the present application, the terminal uses a Bayesian Regularization Backpropagation algorithm to adjust the model parameters of the initial neural network model, and optimizes the parameters of the neural network model so that the mean square error can reach a preset loss condition when predicting the structural response of each target finite element model.

[0122] After each Bayesian regularization back propagation of the initial neural network, the prediction operation of step 402 is performed until the loss value meets the preset loss condition, and the current initial neural network model training is determined to be completed, and the trained target neural network model is obtained. The preset loss condition can be that during the iterative training process, the change rate of the current loss value is less than a preset threshold, or the current loss value is less than a preset loss threshold, or the number of current iterative training meets the preset training number.

[0123] In this embodiment, Bayesian regularized back propagation can automatically adjust hyperparameters, and can automatically adjust parameters according to the distribution of data and the complexity of the model in the case of different bridge structures or smaller training sets, so as to better adapt to the specific engineering application environment and improve the accuracy of the target neural network model output.

[0124] In an exemplary embodiment, Figure 5 As shown, step 108 includes steps 502 to 506. Among them:

[0125] Step 502, obtaining static structural parameters of the target object, initializing target parameters and preset calculation steps.

[0126] In an embodiment of the present application, the terminal receives static structural parameters of the target object, initialization target parameters and a preset calculation step length input by a user. The initialization target parameter may be a starting value used to iteratively calculate candidate target parameters according to the preset calculation step length.

[0127] Step 504, iteratively calculate the initialized target parameters according to a preset calculation step size to obtain multiple candidate target parameters.

[0128] Among them, the static structural parameters include overall static structural parameters and local static structural parameters.

[0129] In an embodiment of the present application, the terminal uses the initialization target parameters input by the user as the starting values, the initialization target parameters include the initialization side-to-midspan ratio and the initialization steel-concrete joint position, and iteratively calculates the initialization target parameters according to a preset calculation step size, and obtains a candidate target parameter in each iterative calculation, and obtains multiple candidate target parameters.

[0130] Step 506, screening candidate target parameters according to the target neural network model to obtain the target result.

[0131] In the embodiment of the present application, the terminal uses the trained target neural network model to evaluate and screen multiple candidate target parameters to obtain the final target result. Specifically, the terminal inputs the candidate target parameters as input data into the target neural network model, predicts each candidate target parameter through the target neural network model, obtains the corresponding structural response index, and sets screening conditions according to the design requirements and optimization goals, evaluates and screens the prediction results, and finally determines the best target parameters that meet the conditions as the final target result.

[0132] In this embodiment, structural response index analysis is performed based on a large number of candidate target parameters with different parameter values, thereby achieving parameter analysis for a wider range, preventing omission of optimal parameter values, achieving effective control of candidate target parameters, and improving the accuracy of target results.

[0133] In an exemplary embodiment, Figure 6 As shown, step 506 includes steps 602 to 604. Among them:

[0134] Step 602: predict each candidate target parameter according to the target neural network model to obtain a structural response index corresponding to each candidate target parameter.

[0135] In an embodiment of the present application, the terminal inputs each candidate target parameter into the target neural network model respectively, and predicts the structural response index corresponding to each candidate target parameter based on the mapping relationship learned by the target neural network model.

[0136] Step 604 , screening the structural response index corresponding to each candidate target parameter according to the preset screening condition, and determining the target result whose structural response index meets the preset screening condition.

[0137] In the embodiment of the present application, a series of screening conditions determined according to the actual needs of the engineering design are pre-set in the terminal. After obtaining the structural response index, the terminal compares the structural response index corresponding to each candidate target parameter with the preset screening condition. If all structural response indicators of a candidate target parameter meet the screening condition, the candidate target parameter is regarded as a target result that meets the design requirements. Optionally, the terminal can select the best target result as the final design result according to the actual situation, wherein the analysis factors of the actual situation include cost, safety, constructability, etc.

[0138] In this embodiment, structural response index analysis is performed based on a large number of candidate target parameters with different parameter values, thereby achieving parameter analysis for a wider range, effectively utilizing the predictive ability of the neural network model, and quickly screening out the optimal solution that meets the design requirements from a large number of candidate target parameters, thereby greatly improving the efficiency and accuracy of engineering design.

[0139] In an exemplary embodiment, the target neural network model can also be integrated with the candidate parameter generation and screening process, such as Figure 7 As shown, step 108 includes steps 702 to 706. Among them:

[0140] Step 702, obtain the static structural parameters of the target object, initialize the target parameters and preset the calculation step size, and build an integrated influence model according to the target neural network model.

[0141] In the embodiment of the present application, the terminal can integrate the target neural network model with the functions in step 504 and step 602 to step 604, construct and obtain an integrated impact model, and obtain the static structural parameters of the target object, initialize the target parameters and preset calculation step size. Figure 8 As shown, Figure 8 This is a schematic diagram of the model layers of the target neural network model integrated in the integrated impact model, where the target neural network (ANN, artificial neural network) usually contains multiple layers, including an input layer, one or more hidden layers, and an output layer. The neurons in each layer are fully connected to the neurons in the next layer. This structure is called a feedforward neural network. During the training process, the target neural network minimizes the difference between the predicted output and the actual output by adjusting the connection weights. This process is usually achieved by the backpropagation algorithm combined with the gradient descent method. The backpropagation algorithm first calculates the output error, then propagates these errors back through the network, and updates the weights to reduce the error of future predictions.

[0142] Step 704 , iteratively calculate the initialized target parameters based on the integrated influence model according to the preset calculation step length to obtain multiple candidate target parameters.

[0143] In the embodiment of the present application, the terminal automatically performs iterative calculation on the initialization target parameters through the integrated influence model according to the same principle in step 504 to obtain multiple candidate target parameters. The detailed process of iterative calculation in the embodiment of the present application will not be repeated.

[0144] Step 706, screening candidate target parameters based on the integrated impact model to obtain target results.

[0145] In an embodiment of the present application, preset screening conditions of structural response indicators are integrated into the integrated impact model, and based on the preset screening conditions, candidate target parameters are automatically predicted and structural response indicators corresponding to different candidate target parameters are screened to obtain target results that meet the preset screening conditions.

[0146] In this embodiment, the efficiency of determining the target result can be improved by automatically processing the entire process from the static structural parameters of the target object to the final target result through the integrated impact model.

[0147] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0148] Based on the same inventive concept, the embodiment of the present application also provides a data processing device for implementing the data processing method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in the one or more data processing device embodiments provided below can refer to the limitations on the data processing method above, and will not be repeated here.

[0149] In an exemplary embodiment, Fig. 9 As shown, a data processing device 900 is provided, comprising: an acquisition module 901, an analysis module 902, a training module 903 and a calculation module 904, wherein:

[0150] An acquisition module 901 is used to acquire the overall modeling results of the target object and the local modeling results corresponding to the overall modeling results, and integrate the local modeling results into the overall modeling results to obtain multiple target finite element models;

[0151] An analysis module 902 is used to perform model analysis on the target finite element model to obtain a structural response index corresponding to each target finite element model;

[0152] A training module 903 is used to train the initial neural network model based on the structural response index, the overall structural parameters and the local structural parameters in the target finite element model to obtain a trained target neural network model;

[0153] The calculation module 904 is used to obtain the target result according to the target neural network model, the initialization structure parameters of the target object and the preset calculation step size.

[0154] In one embodiment, the 900 device further comprises:

[0155] A first modeling module is used to obtain a plurality of overall structural parameters of the target object, and to perform modeling and analysis on the overall structural parameters based on the first finite element analysis model to obtain a plurality of overall modeling results; the overall structural parameters correspond to a plurality of different preset parameter values;

[0156] A determination module, used to determine the local structural parameters corresponding to each overall structural parameter according to the overall modeling result;

[0157] The second modeling module is used to perform modeling analysis on local structural parameters based on the second finite element analysis model to obtain local modeling results corresponding to each overall modeling result.

[0158] In one embodiment, the acquisition module 901 is specifically used to unify the formats of the overall modeling result and the local modeling result to obtain the first target modeling result and the second target modeling result after the formats are unified;

[0159] Each second target modeling result is integrated into the first target modeling result corresponding to the second target modeling result to obtain a target finite element model corresponding to each target parameter.

[0160] In one embodiment, the training module 903 is specifically used to predict the overall structural parameters and local structural parameters in each target finite element model according to the initial neural network model to obtain the prediction value corresponding to each target finite element model;

[0161] The mean square error is calculated based on the predicted value and the structural response index to obtain the loss value of the initial neural network model;

[0162] The model parameters of the initial neural network model are adjusted according to Bayesian regularization back propagation until the loss value meets the preset loss condition, and the trained target neural network model is obtained.

[0163] In one embodiment, the calculation module 904 is specifically used to obtain static structural parameters of the target object, initialize target parameters and preset calculation steps;

[0164] Iteratively calculate the initialized target parameters according to the preset calculation step length to obtain multiple candidate target parameters; the static structure parameters include overall static structure parameters and local static structure parameters;

[0165] The candidate target parameters are screened according to the target neural network model to obtain the target result.

[0166] In one embodiment, the calculation module 904 is specifically used to perform prediction processing on each candidate target parameter according to the target neural network model to obtain a structural response index corresponding to each candidate target parameter;

[0167] The structural response index corresponding to each candidate target parameter is screened according to the preset screening conditions to determine the target result in which the structural response index meets the preset screening conditions.

[0168] In one embodiment, the calculation module 904 is specifically used to obtain static structural parameters of the target object, initialize target parameters and preset calculation steps, and build an integrated impact model according to the target neural network model;

[0169] According to the preset calculation step length, the initialization target parameters are iteratively calculated based on the integrated influence model to obtain multiple candidate target parameters;

[0170] Based on the integrated impact model, candidate target parameters are screened to obtain target results.

[0171] Each module in the above data processing device can be implemented in whole or in part by software, hardware or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.

[0172] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Fig.10 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store overall structural parameters, local structural parameters and structural response indicators. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a data processing method is implemented.

[0173] Those skilled in the art will understand that Fig.10 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0174] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0175] Obtaining each overall modeling result of the target object and the local modeling result corresponding to the overall modeling result, and integrating the local modeling results into the overall modeling result to obtain multiple target finite element models;

[0176] Perform model analysis on the target finite element model to obtain the structural response index corresponding to each target finite element model;

[0177] The initial neural network model is trained based on the structural response index, the overall structural parameters and the local structural parameters in the target finite element model to obtain a trained target neural network model;

[0178] The target result is obtained according to the target neural network model, the initialization structure parameters of the target object and the preset calculation step size.

[0179] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0180] Acquire multiple overall structural parameters of the target object, perform modeling analysis on the overall structural parameters based on the first finite element analysis model, and obtain multiple overall modeling results; the overall structural parameters correspond to multiple different preset parameter values;

[0181] Determine the local structural parameters corresponding to each overall structural parameter according to the overall modeling results;

[0182] Based on the second finite element analysis model, local structural parameters are modeled and analyzed to obtain local modeling results corresponding to each overall modeling result.

[0183] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0184] Unifying the formats of the overall modeling results and the local modeling results to obtain the first target modeling results and the second target modeling results after the formats are unified;

[0185] Each second target modeling result is integrated into the first target modeling result corresponding to the second target modeling result to obtain a target finite element model corresponding to each target parameter.

[0186] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0187] According to the initial neural network model, the overall structural parameters and local structural parameters in each target finite element model are predicted and processed to obtain the corresponding predicted values ​​of each target finite element model;

[0188] The mean square error is calculated based on the predicted value and the structural response index to obtain the loss value of the initial neural network model;

[0189] The model parameters of the initial neural network model are adjusted according to Bayesian regularization back propagation until the loss value meets the preset loss condition, and the trained target neural network model is obtained.

[0190] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0191] Obtain static structural parameters of the target object, initialize target parameters and preset calculation step size;

[0192] Iteratively calculate the initialized target parameters according to the preset calculation step length to obtain multiple candidate target parameters; the static structure parameters include overall static structure parameters and local static structure parameters;

[0193] The candidate target parameters are screened according to the target neural network model to obtain the target result.

[0194] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0195] Predict and process each candidate target parameter according to the target neural network model to obtain the structural response index corresponding to each candidate target parameter;

[0196] The structural response index corresponding to each candidate target parameter is screened according to the preset screening conditions to determine the target result in which the structural response index meets the preset screening conditions.

[0197] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0198] Obtain static structural parameters of the target object, initialize target parameters and preset calculation steps, and build an integrated impact model based on the target neural network model;

[0199] According to the preset calculation step length, the initialization target parameters are iteratively calculated based on the integrated influence model to obtain multiple candidate target parameters;

[0200] Based on the integrated impact model, candidate target parameters are screened to obtain target results.

[0201] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0202] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0203] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0204] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0205] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0206] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A data processing method, It is characterized in that The method comprises: Acquire multiple overall structural parameters of the target object, and perform modeling analysis on the overall structural parameters based on the first finite element analysis model to obtain multiple overall modeling results; the overall structural parameters correspond to multiple different preset parameter values; Determine the local structural parameters corresponding to each of the overall structural parameters according to the overall modeling result; Performing modeling analysis on the local structural parameters based on the second finite element analysis model to obtain local modeling results corresponding to the overall modeling results; Obtain each overall modeling result of the target object and the local modeling result corresponding to the overall modeling result, and integrate the local modeling results into the overall modeling result to obtain multiple target finite element models; the target object is a hybrid beam cable-stayed bridge; the overall modeling result is a complete finite element model of the overall structure obtained when using different side-to-span ratios and steel-concrete joint positions, and the complete finite element model includes the overall geometry, material properties, boundary conditions and loading conditions of the overall structure; the local modeling result is a local refined model constructed for the structure and component distribution of the steel-concrete joint of the target object, wherein the overall modeling results with different side-to-span ratios correspond to different force paths and stress concentration points in the steel-concrete joint, and different overall modeling results correspond to different local modeling results; Performing model analysis on the target finite element models to obtain structural response indicators corresponding to each of the target finite element models; the structural response indicators include internal forces and displacements of key parts in the construction and completed bridge states, counterweights required for construction, and cable forces of inclined cables in the completed bridge state; The initial neural network model is trained based on the structural response index, the overall structural parameters and the local structural parameters in the target finite element model to obtain a trained target neural network model; the overall structural parameters and the local structural parameters are used as training samples of the initial neural network model, and the structural response index is used as a sample label of the training sample; Obtaining a target result according to the target neural network model, the initialization structure parameters of the target object and a preset calculation step size; The integrating the local modeling results into the overall modeling results to obtain a plurality of target finite element models includes: The local modeling results are integrated into the overall modeling results through boundary conditions or connection conditions at the nodes to obtain multiple target finite element models.

2. The method according to claim 1, It is characterized in that The integrating the local modeling results into the overall modeling results to obtain a plurality of target finite element models includes: Unifying the formats of the overall modeling result and the local modeling result to obtain a first target modeling result and a second target modeling result after the formats are unified; Each of the second target modeling results is integrated into the first target modeling result corresponding to the second target modeling result to obtain a target finite element model corresponding to each target parameter.

3. The method according to claim 1, It is characterized in that The initial neural network model is trained based on the structural response index, the overall structural parameters and the local structural parameters in the target finite element model to obtain a trained target neural network model, including: Predicting the overall structural parameters and local structural parameters in each of the target finite element models according to the initial neural network model to obtain prediction values ​​corresponding to each of the target finite element models; Performing mean square error calculation based on the predicted value and the structural response index to obtain a loss value of the initial neural network model; The model parameters of the initial neural network model are adjusted according to Bayesian regularization back propagation until the loss value meets the preset loss condition, thereby obtaining a trained target neural network model.

4. The method according to claim 1, It is characterized in that The step of obtaining a target result according to the target neural network model, the initialization structure parameters of the target object and the preset calculation step size includes: Obtain static structural parameters of the target object, initialize target parameters and preset calculation step size; Iteratively calculate the initialized target parameters according to a preset calculation step length to obtain multiple candidate target parameters; the static structure parameters include overall static structure parameters and local static structure parameters; The candidate target parameters are screened according to the target neural network model to obtain a target result.

5. The method according to claim 4, It is characterized in that The step of screening the candidate target parameters according to the target neural network model to obtain a target result includes: Predicting each of the candidate target parameters according to the target neural network model to obtain a structural response index corresponding to each of the candidate target parameters; The structural response index corresponding to each of the candidate target parameters is screened according to a preset screening condition to determine a target result in which the structural response index meets the preset screening condition.

6. The method according to claim 1, It is characterized in that The step of obtaining a target result according to the target neural network model, the initialization structure parameters of the target object and the preset calculation step size includes: Obtaining static structural parameters of the target object, initializing target parameters and preset calculation steps, and constructing an integrated impact model according to the target neural network model; Iteratively calculating the initialization target parameter based on the integrated impact model according to a preset calculation step length to obtain a plurality of candidate target parameters; The candidate target parameters are screened based on the integrated impact model to obtain target results.

7. A data processing device, It is characterized in that The device comprises: A first modeling module is used to obtain a plurality of overall structural parameters of the target object, and to perform modeling and analysis on the overall structural parameters based on a first finite element analysis model to obtain a plurality of overall modeling results; the overall structural parameters correspond to a plurality of different preset parameter values; A determination module, used to determine the local structural parameters corresponding to each of the overall structural parameters according to the overall modeling result; A second modeling module is used to perform modeling analysis on the local structural parameters based on a second finite element analysis model to obtain local modeling results corresponding to each of the overall modeling results; An acquisition module is used to acquire each overall modeling result of a target object and a local modeling result corresponding to the overall modeling result, and integrate the local modeling results into the overall modeling result to obtain a plurality of target finite element models; the target object is a hybrid beam cable-stayed bridge; the overall modeling result is a complete finite element model of the overall structure obtained when using different side-to-span ratios and steel-concrete joint positions, and the complete finite element model includes the overall geometry, material properties, boundary conditions and loading conditions of the overall structure; the local modeling result is a local refined model constructed for the structure and component distribution of the steel-concrete joint of the target object, wherein the overall modeling results with different side-to-span ratios correspond to different force paths and stress concentration points in the steel-concrete joint, and different overall modeling results correspond to different local modeling results; An analysis module is used to perform model analysis on the target finite element model to obtain structural response indicators corresponding to each target finite element model; the structural response indicators include internal forces and displacements of key parts in the construction and completed bridge states, counterweights required for construction, and cable forces of inclined cables in the completed bridge state; A training module is used to train the initial neural network model based on the structural response index, the overall structural parameters and the local structural parameters in the target finite element model to obtain a trained target neural network model; the overall structural parameters and the local structural parameters are used as training samples of the initial neural network model, and the structural response index is used as a sample label of the training sample; A calculation module, used to obtain a target result according to the target neural network model, the initialization structure parameters of the target object and a preset calculation step size; The acquisition module is specifically used to integrate the local modeling results into the overall modeling results through boundary conditions or connection conditions at nodes to obtain multiple target finite element models.

8. The device according to claim 7, It is characterized in that The acquisition module is specifically used to unify the formats of the overall modeling result and the local modeling result to obtain the first target modeling result and the second target modeling result after the formats are unified; Each of the second target modeling results is integrated into the first target modeling result corresponding to the second target modeling result to obtain a target finite element model corresponding to each target parameter.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program. It is characterized in that When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, It is 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 6 are implemented.

Images