A complex superstructure multi-class strong nonlinear curve prediction method and system

Abstract

The application discloses a kind of complex superstructure multi-class strong nonlinear curve prediction method and system, and relates to superstructure technical field.For the problems, such as complex mechanical behavior, strong nonlinear characteristics, multiple types and excessive calculation cost, encountered in current complex superstructure mechanical property curve prediction, the application greatly improves the prediction accuracy of multi-class strong nonlinear curve by accurately describing the curve characteristics;And effectively avoids the problem of mixed variety of curve characteristics, only a small amount of data is needed to complete the construction of prediction model, reduces the cost of data generation and neural network training.The application provides a new idea for the prediction of complex superstructure multi-class strong nonlinear mechanical property curve, and has good application prospect in the field of complex superstructure research.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of mechanical superstructure, and particularly relates to a multi-category strong nonlinear curve prediction method and system for complex superstructure. BACKGROUND

[0002] In recent years, the research on mechanical superstructure has achieved outstanding results, and a series of unconventional and counterintuitive mechanical properties that are rare or even nonexistent in nature have been realized. Through reasonable design of size parameters, the superstructure can exhibit programmable mechanical response. For example, through reasonable design of the shape of the unit cell, the stiffness, Poisson's ratio and thermal expansion coefficient of the superstructure can be programmed to span from negative to positive values. Based on the simple geometric parameterization of the microstructure layout design method, the superstructure can realize the customized design of mechanical response, so as to meet the different functional requirements under actual working conditions. Therefore, it is an important task to study the influence of the structure size parameters of the complex supermaterial on the mechanical properties to realize accurate prediction of the mechanical properties.

[0003] Through reasonable arrangement of the unit, the negative stiffness superstructure can exhibit a snap-through behavior in which the force suddenly changes in an instant, which has been proved to be helpful to further improve the buffering performance of the superstructure. However, the snap-through behavior makes the mechanical property curve non-smooth, and the strong nonlinear mutation feature greatly increases the difficulty of accurate prediction. Secondly, the mechanical properties of the complex superstructure are influenced by the geometric parameters of the structure itself, and when the geometric parameters are more, the high-dimensional design space will increase the difficulty of accurate prediction. Therefore, the strong nonlinearity of the mechanical response of the complex superstructure and the high-dimensional design space greatly increase the complexity of performance prediction, which makes the prediction of the mechanical curve of the superstructure a very challenging task.

[0004] Machine learning can be used as a data curve processing method, which can derive logic and rules from the provided data set, and predict new data according to the derived logic and rules. In order to improve the prediction accuracy, the most commonly used method at present is to increase the sample size of the training set, but this will greatly increase the calculation cost of generating the data set. Therefore, an efficient and feasible method is needed to realize accurate prediction of the multi-category curve of the superstructure, especially the non-smooth curve with strong nonlinear characteristics. SUMMARY

[0005] The technical problem to be solved by the present application is to provide a multi-category strong nonlinear curve prediction method and system for complex superstructure to solve the technical problem that the traditional machine learning algorithm needs a large number of simulations and experiments, needs to provide a large number of training sets, and consumes a large amount of time and cost, in order to save time, cost and computing resources and improve the accuracy.

[0006] The technical scheme adopted by the present application is as follows:

[0007] A complex superstructure multi-class strong nonlinear curve prediction method, comprising the following steps:

[0008] S1, according to the influence factors of the mechanical properties of the complex superstructure, selecting the geometric parameters of the complex superstructure as the size parameters, changing the size parameters to obtain different structures, obtaining the correlation database of the structure size parameters and the mechanical response by carrying out the quasi-static compression simulation, and normalizing the curve and the size parameter to obtain the curve data set;

[0009] S2, using the monotonicity and smoothness of the curve to classify the curve data set obtained in step S1 to obtain the monotonic increasing positive stiffness curve, the smooth negative stiffness curve and the non-smooth snap negative stiffness curve;

[0010] S3, according to the monotonic increasing positive stiffness curve, the smooth negative stiffness curve and the non-smooth snap negative stiffness curve obtained in step S2, combining the machine learning algorithm to train the neural network, and constructing the curve type prediction model;

[0011] S4, according to the initial data set obtained in step S1, the interpolation calculation is carried out on the curve with strong nonlinearity, that is, the non-smooth snap negative stiffness curve obtained by the curve prediction model in step S3, the similar features of the curve are concentrated in a specific area, that is, some specific sections of the curve, so as to facilitate the feature prediction;

[0012] S5, for the curve after the reconstruction in step S4, combining the machine learning algorithm to train the neural network respectively, and constructing the curve feature prediction model, the curve feature prediction model taking the size parameter as the input parameter and outputting a predicted quasi-static compression mechanical property curve.

[0013] Preferably, according to the influence factors of the mechanical properties of the complex superstructure, 9 geometric parameters of the complex superstructure are selected as the size parameters, and the value ranges of the 9 parameters are as follows:

[0014]

[0015] Wherein, AB rod thickness and C rod thickness, AB rod boundary length and C rod boundary length, AB rod thickness and C rod thickness, C rod length L2, AB rod angle, C rod angle, and the specific parameters are shown in Figure 3 .

[0016] Preferably, the curve feature classification of the curve data set by using the monotonicity and smoothness of the curve is specifically:

[0017] When all the first derivatives are greater than 0, it is classified as the first type of curve;

[0018] When there is a derivative less than 0, it is considered smooth when it does not exceed a set threshold, and it is classified as the second type of curve;

[0019] Otherwise, it is a non-smooth snap behavior with a force mutation, and it is classified as the third type of curve.

[0020] Preferably, the monotonically increasing positive stiffness curve, the smooth negative stiffness curve and the non-smooth snap negative stiffness curve are specifically:

[0021]

[0022] wherein, F and u are the normalized force and displacement, represents the set threshold for judging whether the curve is smooth.

[0023] Preferably, the curve type prediction model is specifically constructed as:

[0024] The size parameter is taken as input, the monotonically increasing positive stiffness curve, the smooth negative stiffness curve and the non-smooth snap negative stiffness curve are marked as 1-3 respectively, the class number is taken as output, LSTM neural network is used for classification training, and the curve type prediction model is obtained.

[0025] Preferably, the forget gate of the curve type prediction model is:

[0026]

[0027] wherein, is a sigmoid function, is a weight matrix of the forget gate, is the hidden state of the previous time step, is the input of the current time step, is a bias term;

[0028] The input gate is:

[0029]

[0030]

[0031] wherein, is the activation value of the input gate, is the candidate memory cell, and are weight matrices of the input gate and the candidate memory cell, and are bias terms;

[0032] update memory cell:

[0033]

[0034] wherein, is the memory cell of the current time step, is the memory cell of the previous time step;

[0035] output gate:

[0036]

[0037]

[0038] wherein, is the activation value of the output gate, is the hidden state of the current time step, is the weight matrix of the output gate, is the bias term.

[0039] Preferably, the partition reconstruction is specifically:

[0040] For a positive stiffness curve, directly perform machine learning classification training,

[0041] For a smooth negative stiffness curve, when performing curve reconstruction, the curve is divided into 3 regions, including 2 positive stiffness regions and 1 negative stiffness region;

[0042] For a curve with sharp negative stiffness, the curve is reconstructed into 5 regions, including 2 smooth segments and 1 rapidly descending non-smooth segment, plus 2 positive stiffness regions, a total of 5 regions.

[0043] Preferably, an interpolation calculation method is used to make the same partition of the same type of curve contain the same number of feature points, so as to obtain a mechanical response vector describing the characteristics of the curve composed of the feature points on each partition and the respective points between different partitions.

[0044] Preferably, an error evaluation mechanism of minimum Euclidean distance is established to verify the constructed curve characteristic prediction model, the force-displacement curve is discretized into 10000 evaluation points, the Euclidean distance of each prediction point to the 10000 evaluation points is calculated, and the minimum distance is extracted to evaluate the deviation of the prediction point from the reference curve.

[0045] In a second aspect, the embodiments of the present application provide a multi-class strong nonlinear curve prediction system of complex superstructure, comprising

[0046] The data module selects geometric parameters of the complex superstructure as size parameters according to influencing factors of mechanical properties of the complex superstructure, obtains different structures by changing the size parameters, obtains a correlation database of the structure size parameters and mechanical responses by performing quasi-static compression simulation, and obtains a curve data set by normalizing the curve and the size parameters;

[0047] The classification module classifies the curve data set according to curve characteristics by using monotonicity and smoothness of the curve, and obtains a monotonically increasing positive stiffness curve, a smooth negative stiffness curve and a non-smooth snap negative stiffness curve.

[0048] The construction module constructs a curve type prediction model by training a neural network according to the monotonically increasing positive stiffness curve, the smooth negative stiffness curve and the non-smooth snap negative stiffness curve, and combining a machine learning algorithm.

[0049] The reconstruction module reconstructs the curve with strong nonlinearity by interpolation calculation according to the obtained initial data set, and concentrates similar features of the curve in a specific region, that is, a certain specific section of the curve, so as to facilitate feature prediction.

[0050] The prediction module trains a neural network by combining a machine learning algorithm, and constructs a curve feature prediction model for the reconstructed curve, wherein the curve feature prediction model takes the size parameter as an input parameter, and outputs a predicted mechanical property curve of quasi-static compression.

[0051] In a third aspect, a computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the above-mentioned complex superstructure multi-class strong nonlinear curve prediction method when executing the computer program.

[0052] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium including a computer program, and the computer program implements the steps of the above-mentioned complex superstructure multi-class strong nonlinear curve prediction method when executed by a processor.

[0053] Compared with the prior art, the present application has at least the following beneficial effects:

[0054] A complex superstructure multi-class strong nonlinear curve prediction method can greatly reduce the data of the training set and the cost of experiment and simulation calculation while ensuring the prediction accuracy by performing curve classification and curve reconstruction on the data curve set in advance, and the prediction method of the complex superstructure multi-class nonlinear curve based on machine learning can accurately predict the mechanical properties of the complex superstructure, with an accuracy of more than 95%, which provides an algorithm basis for accurate and efficient prediction of the mechanical properties of the complex superstructure, and has good social and economic benefits.

[0055] Further, 9 geometric parameters of complex superstructure are selected as size parameters, aiming to deeply study the influence of these parameters on the mechanical properties of superstructure, and to realize the precise control of mechanical response through reasonable geometric design. Selecting 9 geometric parameters as size parameters can comprehensively cover the geometric characteristics of superstructure, ensuring that the model can fully consider the influence of various design variables. By adjusting these 9 key geometric parameters, the optimization design of superstructure can be realized, so that its mechanical properties can meet the requirements of specific applications.

[0056] Further, the curve feature classification is performed on the curve data set by using the monotonicity and smoothness of the curve, aiming to efficiently and accurately identify curves of different categories, thereby improving the prediction accuracy of the model for the mechanical properties of complex superstructure. By analyzing the monotonicity of the curve, the curve can be effectively classified into two categories: positive stiffness curve and negative stiffness curve; the smoothness classification can help identify the mutation area in the curve, further subdividing the negative stiffness curve into smooth and non-smooth categories, which has important guiding significance for subsequent model training and prediction.

[0057] Further, the nonlinear curve is preliminarily classified to improve the adaptability and prediction accuracy of strong nonlinear and non-smooth curves.

[0058] Further, the complex nonlinear mechanical property curve is partitioned to concentrate similar features in a specific area, thereby simplifying the prediction process of complex curves, improving prediction accuracy, reducing training data requirements, and being applicable to various neural network models.

[0059] Further, the minimum Euclidean distance error evaluation mechanism is adopted, which greatly improves the adaptability of the model to non-smooth and strong nonlinear curve features when predicting the multi-category nonlinear mechanical property curve of complex superstructure by discrete sampling points

[0060] It can be understood that the beneficial effects of the above-mentioned second aspect can be referred to the related description in the above-mentioned first aspect, which will not be repeated here.

[0061] In summary, the present application greatly improves the prediction accuracy of multi-category strong nonlinear curves by accurately describing the curve features, and effectively avoids the problem of mixing various curve features, requiring only a small amount of data to complete the construction of the prediction model, reducing the cost of data generation and neural network training. It provides a new idea for predicting the multi-category strong nonlinear mechanical property curve of complex superstructure, and has good application prospect in the field of complex superstructure research.

[0062] The technical solutions of the present application will be further described in detail below with the help of drawings and examples. BRIEF DESCRIPTION OF DRAWINGS

[0063] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings used in the relative embodiment description are briefly introduced as follows. Obviously, the drawings in the following description only represent some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort based on these drawings.

[0064] Figure 1 Flow chart of the method of the present application;

[0065] Figure 2 Flow chart of the neural network for establishing the curve feature prediction model according to the present application;

[0066] Figure 3 Schematic diagram of the complex superstructure and its geometric parameters studied in the present application;

[0067] Figure 4 Schematic diagram of the curve classification and curve reconstruction method of the present application;

[0068] Figure 5 Error evaluation method, minimum Euclidean distance principle diagram;

[0069] Figure 6 Comparison diagram of the prediction accuracy of different neural networks under feature driving and non-feature driving;

[0070] Figure 7 Schematic diagram of the computer device provided by an embodiment of the present application;

[0071] Figure 8 Block diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0072] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without any creative effort fall within the scope of the present application.

[0073] In the description of the present application, it should be understood that the terms "comprise" and "include" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude one or more other features, integers, steps, operations, elements, components and / or sets thereof.

[0074] It should also be understood that the terms used in the specification and the following claims are for the purpose of describing particular embodiments only and do not intend to limit the present application. As used in the specification and the appended claims, the singular forms "a," "an" and "the" are intended to include plural forms as well, unless the context clearly indicates otherwise.

[0075] It should be further understood that the term "and / or" used in the specification and the appended claims means one or more of the associated listed items as well as all possible combinations of the items and includes these combinations, for example, A and / or B can mean A alone, A and B together, or B alone. In addition, the character " / " in the present application generally represents an "or" relationship between the objects before and after it.

[0076] It should be understood that although the terms first, second, third, etc. can be used in the embodiments of the present application to describe a certain range, etc., these ranges should not be limited to these terms. These terms are only used to distinguish the ranges from each other. For example, the first range can also be referred to as the second range, and similarly, the second range can also be referred to as the first range without departing from the scope of the embodiments of the present application.

[0077] Depending on the context, the word "if" as used herein can be interpreted to mean "when" or "while" or "in response to determining" or "in response to detecting." Similarly, the phrase "if determined" or "if detecting (a stated condition or event)" can be interpreted to mean "when determined" or "in response to determining" or "when detecting (a stated condition or event)" or "in response to detecting (a stated condition or event)", depending on the context.

[0078] Various structural diagrams according to the disclosed embodiments of the present application are shown in the accompanying drawings. These drawings are not drawn to scale, in which certain details are exaggerated for the purpose of clarity and certain details can be omitted. The shapes of various regions, layers and their relative sizes and positional relationships shown in the drawings are only exemplary, and in actuality, they can deviate due to manufacturing tolerances or technical limitations, and a person skilled in the art can additionally design regions / layers with different shapes, sizes and relative positions according to actual needs.

[0079] The present application provides a multi-class strong nonlinear curve prediction method of complex superstructure, based on classification, reconstruction and other curve preprocessing technologies, realizes fast and accurate prediction of multi-class strong nonlinear curve, can not design a certain parameter structure, does not need to be simulated, and directly predicts the mechanical properties by inputting the characteristic parameters of complex superstructure.

[0080] Please refer to Figure 1 andFigure 2 The application discloses a complex superstructure multi-class strong nonlinear curve prediction method, and comprises the following steps:

[0081] S1, data generation: according to the influence factors of the mechanical properties of the complex superstructure, 9 geometric parameters of the complex superstructure are selected as size parameters, the sizes of various parameters are changed in a reasonable range, and a quasi-static compression simulation is performed in a commercial finite element software Abaqus to obtain a stress-strain data curve set of the structure. The data curve set and the size parameters are subjected to data preprocessing;

[0082] Please refer to Figure 3 , the geometric structure of the complex superstructure is researched, and 12 geometric parameters are determined. The complex superstructure is composed of three rods A, B and C, and the 12 geometric parameters are: AB rod included angle α, C rod included angle β, AB rod thin rod thickness t1, AB rod thickness δ1, C rod thin rod thickness t2, C rod thickness δ2, AB rod boundary length l1, C rod boundary length l2, AB rod included angle length l3, AB rod included angle thickness t3, AB rod length L1 and C rod length L2. Then, 9 of the 12 geometric parameters are reasonably changed, Py code based on Python secondary development is used, finite element simulation experiments are performed in the Abaqus software, stress-strain curves under different parameter modeling are obtained, and the initial database is built. In this embodiment, 1000 groups of data are subjected to batch simulation.

[0083] The value ranges of the 9 parameters are as follows:

[0084]

[0085] In order to improve the efficiency of simulation, Abaqus is secondarily developed based on Python, Py files are used for finite element simulation, and a curve data set is obtained.

[0086] The initial data set is randomly divided into a training set and a test set according to 8:2, and all size parameters are subjected to normalization processing, so as to weaken the influence of different size parameters on the results due to different dimensions and units and eliminate the adverse effects of singular sample data.

[0087] The normalization formula is as follows:

[0088]

[0089] Wherein, X is a normalized result, X i is an original size parameter value, X min is a minimum value of a size parameter value range, X max is a maximum value of the value range.

[0090] S2, curve classification: according to the different characteristics of the curve, the monotonicity and smoothness of the curve are used to classify the obtained data set, and three different categories of curves are obtained: monotone increasing positive stiffness curve, smooth negative stiffness curve and non-smooth snap negative stiffness curve;

[0091] For the classification of the curve, when all the first derivatives are greater than 0, that is, the curve is monotonically increasing, it shows the positive stiffness characteristic, and it is classified as the first type of curve;

[0092] When there is a derivative less than 0, that is, there is a decreasing interval, that is, it shows the negative stiffness characteristic, and on this basis, the size of the slope is compared, when it does not exceed the set threshold, it is considered that the curve is smooth, and it is classified as the second type of curve;

[0093] Otherwise, it is considered that the non-smooth snap behavior occurs, and it is classified as the third type of curve.

[0094] Please refer to Figure 4 , the slope of each point on the curve and the difference between the adjacent two points of the curve are calculated, and the curve is divided into three different categories of curves according to the monotonicity and smoothness characteristics: monotone increasing positive stiffness curve, smooth negative stiffness curve and non-smooth snap negative stiffness curve:

[0095]

[0096] wherein, F and u are the normalized force and displacement, represents the threshold value set to judge whether the curve is smooth.

[0097] S3, establish a curve type prediction model: according to the curve classification result, combine the machine learning algorithm to build a curve type prediction model;

[0098] After curve classification, the size parameters are taken as input, the three types of curves are marked as 1-3, the nine designable size parameters are taken as input, and the category number is taken as output. LSTM neural network is used for classification training, which contains 100 neurons and iterates 1500 times. The prediction model LSTM of the curve type is built. Long short-term memory network is a special type of recurrent neural network (RNN), which consists of three gates (gate units): forget gate, input gate and output gate. The following is the basic formula of LSTM:

[0099] Forget gate:

[0100]

[0101] wherein, is a sigmoid function, is a weight matrix of the forget gate, is the hidden state of the previous time step, is the input of the current time step, is a bias term.

[0102] Input gate:

[0103]

[0104]

[0105] wherein, is the activation value of the input gate, is the candidate memory cell, and are weight matrices of the input gate and the candidate memory cell, and are bias terms.

[0106] Update memory cell:

[0107]

[0108] wherein, is the memory cell of the current time step, is the memory cell of the previous time step.

[0109] Output gate:

[0110]

[0111]

[0112] wherein, is the activation value of the output gate, is the hidden state of the current time step, is a weight matrix of the output gate, is a bias term.

[0113] S4, curve reconstruction: according to the obtained initial data set, the curve reconstruction is performed on the curve which is difficult to predict by interpolation calculation, so as to concentrate the similar features of the curve in a certain region, i.e. some specific sections of the curve, and facilitate the feature prediction;

[0114] According to the curve type prediction model in step S3, the initial database is divided into small databases of three different curve types; then, the curve containing negative stiffness is partitioned by calculating the slope, and for the second type of weakly nonlinear curve, it is divided into positive stiffness, negative stiffness and positive stiffness according to the positive and negative of the slope; for the third type of strongly nonlinear curve, it is first divided into positive stiffness, negative stiffness and positive stiffness according to the positive and negative of the slope, and then divided into weak negative stiffness, strong negative stiffness and weak negative stiffness according to whether the slope exceeds the threshold in the negative stiffness partition, a total of five partitions.

[0115] When the obtained support reaction force data curve is a curve with a negative stiffness interval, because the positive stiffness interval has fewer characteristics and is easy to predict, the negative stiffness interval has complex characteristics and the curve changes rapidly, which is difficult to predict, so the positive and negative stiffness regions and the different regions of the negative stiffness are separated by curve reconstruction, and similar characteristics are concentrated in a specific region, which not only improves the prediction accuracy, but also reduces the number of training sets and facilitates training. As shown in Figure 4 , the specific method is as follows:

[0116] For the positive stiffness curve, machine learning classification training can be directly performed due to the simple curve characteristics, and for the non-monotonic curve of negative stiffness, there are mainly two types, smooth negative stiffness curve and non-smooth jump negative stiffness curve. For the smooth negative stiffness curve, the curve is divided into three regions, i.e., two positive stiffness regions and one negative stiffness region, during curve reconstruction.

[0117] For the curve with rapid change of negative stiffness, the curve is reconstructed into five regions. The negative stiffness region has rapid change of slope, and the local characteristics at both ends of the negative stiffness region are obvious. The negative stiffness part is further subdivided according to the continuity characteristics, including two smooth segments and one non-smooth segment with rapid decline, plus two positive stiffness regions, a total of five regions.

[0118] Then, an interpolation calculation method is used to make the same partition of the same type of curve contain the same number of feature points, and a mechanical response vector describing the curve characteristics is obtained by the feature points on each partition and the respective points between different partitions. After curve reconstruction, the sampling points in the negative stiffness region are increased, which improves the above curve preprocessing techniques based on curve classification and curve reconstruction, and realizes the method of fast and accurate prediction of strongly nonlinear curves, which is called feature-driven method.

[0119] S5, establishing a curve feature prediction model: based on the response vector describing the curve characteristics, multiple shallow lightweight neural networks are trained for different types of curves to construct a curve feature prediction model; and an error evaluation mechanism with minimum Euclidean distance is established for the curve feature prediction model.

[0120] With the size parameter as the input, the mechanical response vector as the output, a plurality of shallow lightweight LSTM neural networks are trained for a plurality of classified and reconstructed curves, then the test set is used to test the training result, and finally the curve feature prediction model is established. When testing the training result, the minimum Euclidean distance is used as the error evaluation method, the force-displacement curve is discretized into 10000 evaluation points, the Euclidean distance of each prediction point to the 10000 evaluation points is calculated, the minimum distance is extracted to evaluate the deviation of the prediction point from the reference curve, and the formula of MED (minimum Euclidean distance) is as follows:

[0121]

[0122] Wherein, i represents the i th sampling point, j represents the j th evaluation point, ( , ) is the prediction point coordinate, ( , ) is the reference point coordinate, m is the number of evaluation points, and n is the number of sampling points.

[0123] Those skilled in the art can understand that various aspects of the present application can be implemented as a system, a method or a program product. Therefore, various aspects of the present application can be specifically implemented in the following forms, namely: a complete hardware implementation, a complete software implementation (including firmware, microcode, etc.), or a combination of hardware and software aspects, which can be collectively referred to as "circuit", "module" or "platform" here.

[0124] In another embodiment of the present application, a complex superstructure multi-class strong nonlinear curve prediction system is provided, which can be used to implement the complex superstructure multi-class strong nonlinear curve prediction method described above. Specifically, the complex superstructure multi-class strong nonlinear curve prediction system comprises a data module, a classification module, a construction module, a reconstruction module and a prediction module.

[0125] Wherein, the data module selects the geometric parameters of the complex superstructure as the size parameters according to the influencing factors of the mechanical properties of the complex superstructure, changes the size parameters to obtain different structures, performs quasi-static compression simulation to obtain the correlation database of the structure size parameters and the mechanical response, and normalizes the curve and the size parameters to obtain the curve data set;

[0126] The classification module classifies the curve data set according to the curve characteristics, and obtains the monotone increasing positive stiffness curve, the smooth negative stiffness curve and the non-smooth snap negative stiffness curve;

[0127] The construction module trains the neural network according to the monotone increasing positive stiffness curve, the smooth negative stiffness curve and the non-smooth snap negative stiffness curve, and constructs the curve type prediction model;

[0128] The reconstruction module reconstructs the curve with strong nonlinearity by interpolation calculation according to the obtained initial data set, and concentrates the similar features of the curve in a specific area, i.e. certain specific sections of the curve, so as to facilitate feature prediction.

[0129] The prediction module trains a neural network by combining a machine learning algorithm for the reconstructed curve, constructs a curve feature prediction model, and the curve feature prediction model takes the size parameter as the input parameter and outputs a predicted quasi-static compression mechanical property curve.

[0130] In another embodiment of the present application, a terminal device is provided, which includes a processor and a memory, the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, and is specifically suitable for loading and executing one or more instructions to realize the corresponding method flow or corresponding function. The processor in the embodiment of the present application can be used for the operation of the complex superstructure multi-class strong nonlinear curve prediction method, including:

[0131] According to the influence factors of the mechanical properties of the complex superstructure, the geometric parameters of the complex superstructure are selected as the size parameters, different structures are obtained by changing the size parameters, a correlation database of the structure size parameters and the mechanical response is obtained by carrying out quasi-static compression simulation, and the curve and the size parameter are normalized to obtain a curve data set; the curve data set obtained is classified according to the monotonicity and smoothness of the curve to obtain a monotonic increasing positive stiffness curve, a smooth negative stiffness curve and a non-smooth snap negative stiffness curve; according to the monotonic increasing positive stiffness curve, the smooth negative stiffness curve and the non-smooth snap negative stiffness curve obtained, a neural network is trained by combining a machine learning algorithm to construct a curve type prediction model; according to the initial data set obtained, the curves with strong nonlinearity, that is, the non-smooth snap negative stiffness curves obtained by the curve prediction model are reconstructed by interpolation calculation, the similar features of the curves are concentrated in a specific area, that is, some specific sections of the curves, so that feature prediction is facilitated; for the reconstructed curves, a neural network is trained by combining a machine learning algorithm to construct a curve feature prediction model, and the curve feature prediction model takes the size parameter as an input parameter and outputs a predicted quasi-static compression mechanical property curve.

[0132] In another embodiment of the present application, the present application also provides a storage medium, specifically a computer readable storage medium (Memory), which is a memory device in a terminal device, used for storing programs and data. It can be understood that the computer readable storage medium herein can include a built-in storage medium in the terminal device, and of course can also include an expansion storage medium supported by the terminal device, and can be any tangible medium containing or storing a program, which can be used by or in combination with an instruction execution system, device or apparatus. The computer readable storage medium provides a storage space, which stores an operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space, and the instructions can be one or more computer programs (including program codes). It should be noted that more specific examples (non-exhaustive list) of the computer readable storage medium include an electrical connection with one or more conductive wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0133] Computer readable storage media further includes data signals transported through a carrier wave and a communications medium, including or excluding wired storage media.

[0134] The program code can be executed by using any combination of one or more programming languages including an object oriented programming language such as Java, C++ or the like, and a conventional procedural programming language such as a "C" language or similar programming languages. The program code can execute entirely on the user's computing device, partly on the user's computing device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device or entirely on the remote computing device or server. In the latter scenario, the remote computing device can be connected to the user's computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computing device, such as through the Internet using an Internet Service Provider (ISP).

[0135] The one or more instructions stored in the computer readable storage medium can be loaded and executed by a processor to implement the steps of the complex superstructure multi-class strongly nonlinear curve prediction method described above. The one or more instructions stored in the computer readable storage medium are loaded and executed by the processor to perform the following steps:

[0136] According to the influence factors of the mechanical properties of the complex superstructure, the geometric parameters of the complex superstructure are selected as the size parameters, different structures are obtained by changing the size parameters, a correlation database of the size parameters and the mechanical responses is obtained by performing quasi-static compression simulation, and the curve and the size parameters are normalized to obtain a curve dataset; the monotonicity and smoothness of the curve are used to classify the curve dataset to obtain a monotonic increasing positive stiffness curve, a smooth negative stiffness curve and a non-smooth snap negative stiffness curve; a curve type prediction model is constructed by training a neural network based on the monotonic increasing positive stiffness curve, the smooth negative stiffness curve and the non-smooth snap negative stiffness curve obtained, and a machine learning algorithm; according to the initial dataset obtained, the curves with strong nonlinearity, that is, the non-smooth snap negative stiffness curve obtained by the curve prediction model, are reconstructed by interpolation calculation, the similar features of the curve are concentrated in a specific area, that is, a certain specific section of the curve, so as to facilitate feature prediction; for the reconstructed curve, a neural network is trained based on a machine learning algorithm to construct a curve feature prediction model, and the curve feature prediction model takes the size parameters as input parameters and outputs a predicted quasi-static compression mechanical property curve.

[0137] Please refer to Figure 7 , the terminal device is a computer device, the computer device 60 of the embodiment includes a processor 61, a memory 62, and a computer program 63 stored in the memory 62 and executable on the processor 61, and the computer program 63 implements the complex superstructure multi-class strong nonlinear curve prediction method in the embodiment when executed by the processor 61, to avoid repetition, which will not be described here. Alternatively, the computer program 63 is executed by the processor 61 to implement the functions of each model / unit in the complex superstructure multi-class strong nonlinear curve prediction system of the embodiment, to avoid repetition, which will not be described here.

[0138] The computer device 60 can be a desktop computer, a notebook computer, a palm computer and a cloud server, etc. The computer device 60 can include, but is not limited to, a processor 61 and a memory 62. Those skilled in the art can understand that Figure 7 is only an example of the computer device 60 and does not constitute a limitation on the computer device 60, and can include more or fewer components than shown, or combine certain components, or different components, for example, the computer device can also include an input / output device, a network access device, a bus, etc.

[0139] The processor 61 can be a central processing unit (CPU), and can also be other general-purpose processors, central processing units, graphics processing units, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, discrete gates or transistor logic components, quantum computing-based data processing logic components, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0140] The memory 62 can be an internal storage unit of the computer device 60, such as a hard disk or a memory of the computer device 60. The memory 62 can also be an external storage device of the computer device 60, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc.

[0141] Further, the memory 62 can include both an internal storage unit and an external storage device of the computer device 60. The memory 62 is used to store computer programs and other programs and data required by the computer device. The memory 62 can also be used to temporarily store data that has been output or will be output.

[0142] Any reference to storage, databases or other media used to store data in the embodiments provided herein is intended to include at least one of volatile and non-volatile storage. Non-volatile storage can include, for example, optical, floppy disks, hard disks, or solid state drives. Volatile storage can include, for example, random access memory (RAM). A basic input / output system (BIOS), containing the basic routines that help to transfer information between elements within the electronic device, such as during startup, can typically be stored in non-volatile memory. By way of illustration, and not limitation, a basic input / output system based on the BIOS, can include a BIOS, a unified extensible firmware interface (UEFI), or the like, including without limitation basic input / output system software stored in nonvolatile memory that

[0143] The database referred to in the embodiments provided herein can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a blockchain, and the like, without being limited thereto. The processor referred to in the embodiments provided herein can be a general processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, and the like, without being limited thereto.

[0144] Referring to FIG. 6, Figure 8 The electronic device 600 is in the form of a general computing device. Components of the electronic device can include, but are not limited to, at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different platform components including the storage unit 620 and the processing unit 610, a display unit 640, and the like.

[0145] The storage unit stores program codes that can be executed by the processing unit 610, so that the processing unit 610 performs steps according to various exemplary embodiments of the present application described in the method part of the present specification. For example, the processing unit 610 can perform steps as shown in Figure 1

[0146] ​The storage unit 620 can include a readable medium in the form of volatile storage such as random access memory (RAM) 6201 and / or cache memory 6202, and also can include a non-volatile storage such as read only memory (ROM) 6203.

[0147] The storage unit 620 also can include a program / utility 6204 having a set of programs / modules 6205, including an operating system, one or more application programs, other program modules, and program data, each of which can implement aspects of a network environment, for example, as each or some combination of these examples.

[0148] The bus 630 can represent one or more of several types of bus structures, including a storage bus or bus for storage controller, peripheral bus, graphics bus, processor or local bus using any of a variety of bus architectures.

[0149] The electronic device 600 also can communicate with one or more external devices 700 such as a keyboard or pointing device, a Bluetooth device, etc.; other devices that enable a user to interact with the electronic device 600; and / or any devices (e.g., a router, a modem, a printer, etc.) that enable the electronic device 600 to communicate with one or more other computing devices. Such communication can occur via an input / output (I / O) interface 650. Still yet, the electronic device 600 can communicate with one or more networks, such as a local area network (LAN), a general wide area network (WAN), and / or a public network such as the Internet, via a network adapter 660. The network adapter 660 can be communicatively coupled to the other components of the electronic device 600 via the bus 630. It should be appreciated that the electronic device 600 can be a part of one or more networks, such as virtual networks, which further can include more than one network.

[0150] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some embodiments of the present application but not all embodiments of the present application. The components of the embodiments of the present application described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application but only represents selected embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments in the present application belong to the scope of protection of the present application.

[0151] Please refer to Figure 5 The figure reflects the advantage of using the minimum Euclidean distance method when dealing with non-smooth curves with large slope changes.

[0152] Horizontal axis: "Strain" represents the degree of deformation of the material when subjected to external forces, usually expressed in percentage or specific value. In this graph, the strain ranges from -0.1 (compressive strain) to 1.0 (tensile strain), with a step of 0.2.

[0153] Vertical axis: "Absolute deviation (MPa)" represents the difference between the actual measured value and a certain reference value or predicted value at a certain strain, with units of MPa (megapascal). This indicator reflects the accuracy and reliability of the measurement results.

[0154] Reference curve: Starting from the lower left corner, as the strain increases (from negative to positive, i.e., from compression to tension), the absolute deviation gradually decreases, and at higher tensile strain, the material's performance is more stable, or the measurement system is more accurate under these conditions.

[0155] Predicted curve: In contrast to the reference curve trend, it starts from the upper left corner, and as the strain increases, the absolute deviation also gradually decreases, but it is overall below the reference curve. This indicates that the predicted value is closer to the actual value than the reference value in most cases, or the prediction model is more accurate in some aspects than the reference standard in reflecting the relationship between strain and absolute deviation.

[0156] Small box on the right side of the graph: labeled "Absolute deviation (MPa)" and "Minimum Euclidean distance", here is a quantitative description of the difference between the curves in the graph or the ideal state, the minimum Euclidean distance is a commonly used method to measure the distance between two points in space, used here to evaluate the closeness or difference between the predicted curve and the reference curve.

[0157] If the prediction accuracy does not meet the set standard, continue to increase the training set for training until the set condition is met. In addition, due to the use of curve classification and curve reconstruction preprocessing methods, BP neural network and RBF neural network can also be used for training in the curve feature set, and the accuracy is significantly improved, indicating that the feature-driven machine learning method based on the invention is suitable for any neural network.

[0158] Please refer to Figure 6 Compare the minimum Euclidean distance of the feature-driven machine learning algorithm with the traditional algorithm, where the X-axis represents the type of neural network, from left to right, BP, LSTM, RBF.

[0159] The Y-axis represents the scale of the minimum Euclidean distance, divided into two main parts, the bottom part with a scale of (x10 -2 ) representing smaller distance values, and the top part with a scale of (x10²) representing larger distance values. This dual-scale design allows the chart to display data changes at different orders of magnitude simultaneously.

[0160] BP Neural Network: In the smaller scale (x10 -2 ) region of the Y-axis, the data points of the BP network are relatively low, indicating that the output performance of the BP network is better at smaller scales (i.e., when the minimum Euclidean distance is small). However, in the larger scale (x10²) region of the Y-axis, the data points of the BP network rise significantly, indicating that the performance of the BP network is poorer at larger scales (i.e., when the minimum Euclidean distance is large).

[0161] LSTM Neural Network: The LSTM network maintains a relatively low and stable data point position at each scale of the Y-axis, indicating that the LSTM network performs well at different scales, with good stability and robustness.

[0162] RBF Neural Network: Similar to the BP network, the RBF network also performs well at smaller scales (x10 -2 ), with lower data point positions. However, at larger scales (x10²), the performance of the RBF network declines, but the decline is smaller compared to the BP network, and the RBF network still outperforms the BP network.

[0163] By comparing the minimum Euclidean distance values of BP, LSTM, and RBF neural networks at different scales, the performance differences of these networks in handling specific problems are demonstrated. Among them, the LSTM network stands out with its excellent stability and robustness, while the BP and RBF networks perform well at smaller scales but have limitations when dealing with larger scale data.

[0164] In summary, the complex superstructure multi-class strong non-linear curve prediction method and system greatly improves the prediction accuracy of multi-class strong non-linear curves by accurately describing the curve features. It effectively avoids the problem of mixing multiple curve features, and only requires a small amount of data to complete the construction of the prediction model, reducing the cost of data generation and neural network training. It provides a new approach for predicting complex superstructure multi-class strong non-linear mechanical performance curves and has good application prospects in the field of complex superstructure research.

[0165] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above-mentioned division of each functional unit and module is exemplified, and in actual application, the above-mentioned functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be realized in the form of hardware or software. In addition, the specific names of each functional unit and module are only for the convenience of mutual distinction, and do not limit the protection scope of the present application. The specific working process of the units and modules in the above system can refer to the corresponding process in the foregoing method embodiments, which will not be described here.

[0166] In the above embodiments, the description of each embodiment has its own emphasis, and the parts not described or recorded in detail in a certain embodiment can be referred to the related description of other embodiments.

[0167] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in the present application can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0168] In the embodiments provided by the present application, it should be understood that the disclosed apparatus / terminal and method can be implemented by other ways. For example, the above-mentioned apparatus / terminal embodiments are only schematic, and the division of the modules or units is only a logical function division, and there can be another division way in actual implementation, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual coupling or direct coupling or communication connection can be indirect coupling or communication connection through some interface, device or unit, and can be electrical, mechanical or other forms.

[0169] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0170] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0171] The integrated module / unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on such understanding, the present application realizes all or part of the processes in the above-mentioned embodiment methods, and can also be completed by instructing related hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can realize the steps of each method embodiment when executed by a processor. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate forms, etc. The computer-readable medium can include any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0172] The present application is described with reference to flowcharts and / or block diagrams according to the methods, devices (systems), and computer program products of the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to produce a machine, so that the instructions executed by the computer or other programmable data processing devices produce a device that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The device that implements the functions specified in one flow or multiple flows and / or blocks Figure 1 The device that implements the functions specified in one flow or multiple flows and / or blocks

[0173] These computer program instructions can also be stored in a computer-readable storage medium that can guide the computer or other programmable data processing devices to work in a specific way, so that the instructions stored in the computer-readable storage medium produce a manufactured product that includes instruction devices that implement the functions specified in the flowcharts and / or block diagrams. Figure 1one or more processes and / or blocks Figure 1 the function specified in the one or more blocks.

[0174] These computer program instructions can also be loaded into computer or other programmable data processing devices, so that a series of operational steps are performed on the computer or other programmable data processing devices to generate a computer-implemented process, so that the instructions executed on the computer or other programmable data processing devices provide a process for implementing the flowchart Figure 1 one or more processes and / or blocks Figure 1 the function specified in the one or more blocks.

[0175] The above is only to illustrate the technical idea of the present application, and cannot limit the protection scope of the present application. Any modification made according to the technical idea of the present application on the basis of the technical scheme falls within the protection scope of the claims of the present application.

Claims

1. A method for predicting multi-class strongly nonlinear curves of complex superstructure, characterized in that, The method comprises the following steps: S1, according to the influence factors of the mechanical properties of the complex superstructure, selecting the geometric parameters of the complex superstructure as the size parameters, changing the size parameters to obtain different structures, obtaining the correlation database of the structure size parameters and the mechanical response by performing quasi-static compression simulation, and normalizing the curve and the size parameter to obtain the curve data set, according to the influence factors of the mechanical properties of the complex superstructure, selecting nine geometric parameters of the complex superstructure as the size parameters, and the value ranges of the nine parameters are as follows: wherein, is the AB rod thickness and C rod thickness, is the AB rod boundary length and C rod boundary length, is the AB rod thickness and C rod thickness, is the C rod length L2, is the AB rod angle, is the C rod angle; S2, the curve data set obtained in step S1 is classified according to the monotonicity and smoothness of the curve, and the monotonically increasing positive stiffness curve, the smooth negative stiffness curve and the non-smooth snap negative stiffness curve are obtained; S3, according to the monotonically increasing positive stiffness curve, the smooth negative stiffness curve and the non-smooth snap negative stiffness curve obtained in step S2, a neural network is trained by combining a machine learning algorithm to construct a curve type prediction model; S4, according to the initial data set obtained in step S1, the curve is reconstructed by interpolation calculation, and the similar features of the curve are concentrated in a specific area, and the monotonically increasing positive stiffness curve, the smooth negative stiffness curve and the non-smooth snap negative stiffness curve are as follows: wherein, F and u are the normalized force and displacement, represents a threshold value set to determine whether the curve is smooth. The reconstruction of the partition is as follows: For the positive stiffness curve, directly perform machine learning classification training; For the smooth negative stiffness curve, when the curve is reconstructed, the curve is divided into three regions, including two positive stiffness regions and one negative stiffness region; For the negative stiffness curve with sharp change, the curve is reconstructed into five regions, including two smooth segments and one non-smooth segment with rapid decline, and two positive stiffness regions, a total of five regions; An interpolation calculation method is used to make the same category curve have the same number of feature points in the same partition, and a mechanical response vector describing the features of the curve is obtained by the feature points in each partition and the respective points between different partitions; S5, for the curve reconstructed in step S4, a neural network is trained by combining a machine learning algorithm to construct a curve feature prediction model, and the curve feature prediction model takes the size parameter as the input parameter and outputs a predicted quasi-static compression mechanical property curve.

2. The method of claim 1, wherein the complex superstructure multiclass strong nonlinear curve prediction method is characterized by, The curve feature classification of the curve data set according to the monotonicity and smoothness of the curve is as follows: When all the first-order derivatives are greater than 0, it is classified as the first type of curve; When there is a derivative less than 0, if it does not exceed the set threshold, it is considered that the curve is smooth, and it is classified as the second type of curve; Otherwise, it is a non-smooth snap behavior with force mutation, and it is classified as the third type of curve.

3. The method of claim 1, wherein the complex superstructure multiclass strong nonlinear curve prediction method is characterized by, The construction of the curve type prediction model is as follows: The size parameter is taken as the input, the monotonically increasing positive stiffness curve, the smooth negative stiffness curve and the non-smooth snap negative stiffness curve are marked as 1-3 respectively, the class number is taken as the output, an LSTM neural network is used for classification training, and a curve type prediction model is obtained.

4. The method of claim 3, wherein the complex superstructure multiclass strong nonlinear curve prediction method is characterized by, The forget gate of the curve type prediction model: wherein, is a sigmoid function, is a weight matrix of the forget gate, is the hidden state of the previous time step, is the input of the current time step, is a bias term; The input gate: wherein, is an activation value of the input gate, is a candidate memory cell, and are weight matrices of the input gate and the candidate memory cell, and are bias terms; The update memory unit: wherein, is the memory cell of the current time step, is the memory cell of the previous time step; The output gate: wherein, is an activation value of the output gate, is a hidden state of the current time step, is a weight matrix of the output gate, is a bias term.

5. The method of claim 1, wherein the complex superstructure multiclass strong nonlinear curve prediction method is characterized by, The error evaluation mechanism of the minimum Euclidean distance is established to verify the constructed curve feature prediction model, the force-displacement curve is discretized into 10000 evaluation points, the Euclidean distance of each prediction point to the 10000 evaluation points is calculated, the minimum distance is extracted to evaluate the deviation of the prediction point and the reference curve.

6. A complex superstructure multi-class strong nonlinear curve prediction system characterized by, Comprise: The data module selects the geometric parameters of the complex superstructure as the size parameters according to the influencing factors of the mechanical properties of the complex superstructure, changes the size parameters to obtain different structures, obtains the correlation database of the structure size parameters and the mechanical response by carrying out the quasi-static compression simulation, and carries out the normalization processing on the curve and the size parameter to obtain the curve data set, selects 9 geometric parameters of the complex superstructure as the size parameters according to the influencing factors of the mechanical properties of the complex superstructure, and the value ranges of the 9 parameters are as follows: wherein, is the AB rod thickness and C rod thickness, is the AB rod border length and C rod border length, is the AB rod thickness and C rod thickness, is the C rod length L2, is the AB rod angle, is the C rod angle; The classification module classifies the curve data set according to the monotonicity and smoothness of the curve to obtain the monotone increasing positive stiffness curve, the smooth negative stiffness curve and the non-smooth snap negative stiffness curve; The construction module trains the neural network by combining the machine learning algorithm according to the monotone increasing positive stiffness curve, the smooth negative stiffness curve and the non-smooth snap negative stiffness curve, and constructs the curve type prediction model; The reconstruction module carries out partition reconstruction by interpolation calculation according to the obtained initial data set, concentrates the similar features of the curve in a specific area, i.e. some specific sections of the curve, so as to facilitate feature prediction, and the monotone increasing positive stiffness curve, the smooth negative stiffness curve and the non-smooth snap negative stiffness curve are as follows: wherein, F and u are the normalized force and displacement, represents a threshold value set for judging whether the curve is smooth; the partition reconstruction is specifically: For the positive stiffness curve, machine learning classification training is directly carried out, For the smooth negative stiffness curve, when the curve is reconstructed, the curve is divided into 3 regions, including 2 positive stiffness regions and 1 negative stiffness region; For the curve with negative stiffness changing sharply, the curve is reconstructed into 5 regions, including 2 smooth segments and 1 non-smooth segment of rapid decline, and 2 positive stiffness regions, a total of 5 regions; The interpolation calculation method is adopted to make the same partition of the same type curve contain the same number of feature points, and the mechanical response vector describing the curve features is obtained by the feature points in each partition and the respective points between different partitions; The prediction module trains the neural network by combining the machine learning algorithm for the reconstructed curve, constructs the curve feature prediction model, and the curve feature prediction model takes the size parameter as the input parameter and outputs a predicted quasi-static compression mechanical property curve.

Images