Method and device for evaluating structural strength of rail transit converter and medium

Abstract

The invention discloses a method and device for evaluating the structural strength of a rail transit converter and a medium, and relates to the technical field of rail transit, and the method comprises the steps: determining the sample input of basic parameters and a plurality of key design variables according to the structural characteristics of the converter, and determining the sample output of stress as a response; according to the response stress, generating a sampling position through a Latin hypercube sampling method; respectively constructing finite element models for the sampling positions and carrying out simulation calculation to form a simulation database; performing statistical analysis and machine learning modeling based on the simulation database, and constructing a function relationship between sample input and sample output to obtain an evaluation model; and evaluating the structural strength of the rail transit converter according to the reconstructed evaluation model. Rapid evaluation of the structural strength is achieved, and then the design period is shortened.

Description

Technical Field

[0001] The present invention relates to the field of rail transit technology, and in particular to a method, device and medium for evaluating the structural strength of a rail transit converter. Background Art

[0002] The rapid development of the current rail transit industry has put forward increasingly stringent requirements on the structural safety of rail transit equipment. Rail transit inverters and other electrical equipment, as the core traction power supply equipment of rail transit vehicles, are subject to static, impact, vibration and other loads during service. The rationality and reliability of their structure are crucial to ensuring driving safety.

[0003] However, the current evaluation of the structural strength of rail transit inverters can no longer meet the needs of rapid iterative optimization of the design process. How to accurately and effectively evaluate the structural strength of rail transit inverters is an important issue that needs to be urgently resolved in the industry. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides a method, device and medium for evaluating the structural strength of a rail transit converter, so as to solve the problem that it is difficult to accurately and effectively evaluate the structural strength of a rail transit converter in the prior art.

[0005] According to a first aspect, an embodiment of the present invention provides a method for evaluating the structural strength of a rail transit converter, comprising:

[0006] Determine sample inputs of basic parameters and several key design variables based on the structural characteristics of the converter, and determine sample outputs of stress as a response;

[0007] According to the stress of the response, the sampling positions are generated by the Latin hypercube sampling method;

[0008] Construct finite element models for the sampling positions respectively and perform simulation calculations to form a simulation database;

[0009] Perform statistical analysis and machine learning modeling based on the simulation database, build a functional relationship between sample input and sample output, and obtain an evaluation model;

[0010] Rebuild the evaluation model and evaluate the model accuracy, and adjust the model accuracy until the model accuracy meets the requirements;

[0011] Based on the reconstructed evaluation model, the structural strength of the rail transit converter is evaluated.

[0012] In combination with the first aspect, in the first embodiment of the first aspect, the basic parameters and several key design variable parameters used as sample inputs include at least: total weight of the inverter, center of gravity position, material grade, installation dimensions, inverter cabinet dimensions, sheet metal thickness and shape and dimensions of key components.

[0013] In combination with the first aspect, in a second implementation of the first aspect, the stress as a response output by the sample includes: maximum stress, static stress and random vibration stress.

[0014] In combination with the first aspect, in a third implementation of the first aspect, the evaluation model is a Kriging model.

[0015] In combination with the third implementation of the first aspect, in a fourth implementation of the first aspect, the method further includes:

[0016] For evaluation models that do not meet the model accuracy requirements, adaptive sampling is performed based on the criterion of maximizing the mean square prediction error.

[0017] In combination with the first aspect, in a fifth implementation of the first aspect, the statistical analysis and machine learning modeling are performed based on the simulation database to construct a functional relationship between the sample input and the sample output to obtain the evaluation model, which specifically includes:

[0018] Normalize the sampling positions:

[0019] The method optimizes the model's hyperparameters by maximum likelihood estimation:

[0020] The accuracy of the constructed evaluation model was evaluated using the fitting quality coefficient and the maximum absolute error.

[0021] In combination with the fifth implementation of the first aspect, in the sixth implementation of the first aspect, the sum of the number of basic parameters and the number of design variables at the sampling position.

[0022] According to a second aspect, an embodiment of the present invention provides a method and device for evaluating the structural strength of a rail transit converter, the device comprising:

[0023] An output module, for determining sample inputs of basic parameters and several key design variables according to structural characteristics of the converter, and determining sample outputs of stress as a response;

[0024] A generation module, used for generating sampling positions by a Latin hypercube sampling method according to the stress of the response;

[0025] The construction module also constructs finite element models for the sampling positions and performs simulation calculations to form a simulation database;

[0026] The optimization module is used to perform statistical analysis and machine learning modeling based on the simulation database, build a functional relationship between sample input and sample output, and obtain an estimation model;

[0027] The adjustment module is used to rebuild the estimation model and evaluate the model accuracy, and adjust the model accuracy until the model accuracy meets the requirements;

[0028] The evaluation module is used to estimate the structural strength of the rail transit converter according to the reconstructed estimation model.

[0029] According to the third aspect, an embodiment of the present invention provides an electronic device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the method for evaluating the structural strength of a rail transit inverter as described in any preferred embodiment of the first aspect by executing the computer instructions.

[0030] According to a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the method for evaluating the structural strength of a rail transit inverter as described in any preferred embodiment of the first aspect.

[0031] The method, device and medium for evaluating the structural strength of a rail transit inverter provided by the present invention construct a sample database based on a finite element model and an experimental design method, construct a Kriging proxy model by statistically approximating the sample data in a machine learning manner, and improve the model accuracy through an adaptive sampling method. The stress response of the inverter under different working conditions is predicted based on the information provided by the proxy model, thereby achieving rapid evaluation of the structural strength and shortening the design cycle. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The features and advantages of the present invention will be more clearly understood by referring to the accompanying drawings, which are schematic and should not be construed as limiting the present invention in any way. In the accompanying drawings:

[0033] Figure 1 It is one of the flow charts of the method for evaluating the structural strength of a rail transit converter according to an embodiment of the present application;

[0034] Figure 2 This is a second flow chart of a method for evaluating the structural strength of a rail transit converter according to an embodiment of the present application;

[0035] Figure 3 This is a third flow chart of a method for evaluating the structural strength of a rail transit converter according to an embodiment of the present application;

[0036] Figure 4 is a structural schematic diagram of a method for evaluating the structural strength of a rail transit converter according to an embodiment of the present application;

[0037] Figure 5It is a schematic diagram of the hardware structure of the electronic device provided in the embodiment of the present application. DETAILED DESCRIPTION

[0038] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0039] The rapid development of the rail transit industry has put forward increasingly stringent requirements on the structural safety of rail transit equipment. As the core traction power supply equipment of rail transit vehicles, electrical equipment such as converters are subjected to static, impact, vibration and other loads during service. The reasonable and reliable structure is crucial to ensure driving safety. Nowadays, the structural strength assessment of rail transit converters and other equipment is mainly based on physical tests and computer-aided engineering (CAE) analysis. Since physical tests require prototypes, the single cost is large, and CAE analysis requires the construction of complex finite element models. The modeling cycle is usually as long as several weeks, and the cost and efficiency are severely restricted, which makes it difficult to meet the needs of rapid iteration and optimization of the design process.

[0040] For example, due to the limitation of sample number, the constructed Kriging model may have low model accuracy, and the structural strength of the rail transit converter may have a high degree of nonlinearity, so the evaluation results may be too small when evaluated by the Kriging model.

[0041] Therefore, how to accurately and effectively evaluate the structural strength of rail transit converters is an important issue that needs to be urgently addressed in the industry.

[0042] In view of the above situation, the present invention proposes a method, device and medium for evaluating the structural strength of a rail transit converter. The main solutions of the embodiments of the present invention are:

[0043] Determine sample inputs of basic parameters and several key design variables based on the structural characteristics of the converter, and determine sample outputs of stress as a response;

[0044] According to the stress of the response, the sampling positions are generated by the Latin hypercube sampling method;

[0045] Construct finite element models for the sampling positions respectively and perform simulation calculations to form a simulation database;

[0046] Perform statistical analysis and machine learning modeling based on the simulation database, build a functional relationship between sample input and sample output, and obtain an evaluation model;

[0047] Rebuild the evaluation model and evaluate the model accuracy, and adjust the model accuracy until the model accuracy meets the requirements;

[0048] Based on the reconstructed evaluation model, the structural strength of the rail transit converter is evaluated.

[0049] In an embodiment of the present invention, a sample database is constructed based on a finite element model and an experimental design method, a Kriging proxy model is constructed by statistically approximating the sample data in a machine learning manner, and the model accuracy is improved through an adaptive sampling method. The stress response of the converter under different working conditions is predicted based on the information provided by the proxy model, thereby achieving a rapid evaluation of the structural strength and shortening the design cycle.

[0050] The technical solutions of the embodiments of the present invention and how the technical solutions of the embodiments of the present invention solve the above-mentioned technical problems will be described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present invention will be described in detail below with reference to specific drawings.

[0051] This embodiment provides a method for evaluating the structural strength of a rail transit converter. It can be understood that the method for evaluating the structural strength of a rail transit converter according to the embodiment of the present invention can be used in electronic devices, including but not limited to computers, mobile terminals, etc. Figure 1 , Figure 2 and Figure 3 is a schematic diagram of the structural strength of a rail transit converter according to an embodiment of the present invention. Figure 1 , Figure 2 and Figure 3 As shown, the method comprises the following steps:

[0052] S1. According to the structural characteristics of the converter, determine the sample input of basic parameters and several key design variables, and determine the sample output of stress as the response. Among them, the basic parameters and several key design variable parameters used as sample input include: converter total weight, center of gravity position, material grade, installation size, converter cabinet size, sheet metal thickness and key component shape size, and the stress obtained as the response of the sample output includes: maximum von Mises stress, static stress (four different working conditions) and random vibration stress (three different working conditions).

[0053] S2. Based on the stress obtained as the response from the combing results of S1, 10d sampling positions (i.e., a 10d×d sampling matrix) are generated through the Latin hypercube sampling method, where d is the sum of the number of basic parameters and the number of design variables.

[0054] S3. Construct finite element models for the 10d sampling positions generated in step S2 respectively and perform simulation calculations to form a simulation database.

[0055] S4. Perform statistical analysis and machine learning modeling on the samples obtained in step S3 to construct a functional relationship between the sample input (basic parameters of the converter, designed converter) and the sample output.

[0056] S5. For the Kriging model that does not meet the model accuracy requirements in step S3, adaptive sampling is performed based on the criterion of maximizing the mean square prediction error:

[0057]

[0058] Among them, x new Represents the input of the prediction sample; s 2 Represents the mean square prediction error; r represents the correlation function vector; F represents the regression function matrix; u represents the correlation of the regression term; the x here represents the same parameter as the x mentioned before.

[0059] The Kriging model is then rebuilt and the model accuracy is evaluated, and the model accuracy is repeatedly adjusted until the model accuracy meets the requirements.

[0060] S6. Based on the constructed Kriging model, the structural strength of the converter is quickly evaluated:

[0061]

[0062] It should be noted that among these parameters:

[0063] is the predicted mean, is the root mean square value of the prediction error, λ is the model correction coefficient, λ∈[1,3].

[0064] The method for evaluating the structural strength of a rail transit inverter provided by the present invention constructs a sample database based on a finite element model and an experimental design method, constructs a Kriging proxy model by statistically approximating the sample data in a machine learning manner, and improves the model accuracy through an adaptive sampling method. The stress response of the inverter under different working conditions is predicted based on the information provided by the proxy model, thereby achieving rapid evaluation of the structural strength and shortening the design cycle.

[0065] In an embodiment of the present invention, the Kriging proxy model method is used for machine learning modeling:

[0066] The specific modeling process includes:

[0067] S41, normalize the 10d×d sampling matrix:

[0068]

[0069] in, is the input normalized to the interval [0,1]; x is the original input; ub and lb are the upper and lower bounds of the input respectively.

[0070] S42. Use the JADE algorithm to optimize the hyperparameters of the Kriging model through maximum likelihood estimation:

[0071]

[0072] stθ∈[10 -6 ,10 2 ] k

[0073] Among them, the regression term of Kriging adopts f(x)=1, that is, ordinary Kriging; θ is the hyperparameter that characterizes the influence of the input variable in the correlation function of the Kriging model (needs to be obtained through maximum likelihood estimation during the modeling process); n is the number of samples; is the estimated value of the Kriging model process variance; R represents the correlation matrix of the Kriging model; k represents the number of input variables.

[0074] The correlation function takes a Gaussian shape:

[0075] Corr[Y(x (i) ),Y(x (j) )]=exp(-l 2 )

[0076] Matérn (ν=5 / 2) type

[0077]

[0078] in,

[0079] Y represents the random variable form corresponding to the sample output; x represents the sample input; i and j represent the i-th sample input and the j-th sample input respectively; exp represents the exponential operation with e as the base, that is, exp(x) = e x

[0080] More specifically, h is the variable used for loop counting in the summation process from 1 to k, indicating that it is the turn of the hth input variable.

[0081] S43. The following indicators are used to evaluate the accuracy of the constructed Kriging model:

[0082] Fitting quality coefficient

[0083] Maximum absolute error Not more than 5% of the response range.

[0084] Among them, m represents the number used to calculate Q 2 The number of validation samples used; y i Represents the true response value of the validation sample.

[0085] This embodiment provides a device for evaluating the structural strength of a rail transit converter. It can be understood that the device for evaluating the structural strength of a rail transit converter according to the embodiment of the present invention can be used in electronic devices, including but not limited to computers, mobile terminals, etc. Figure 5 is a schematic diagram of the structural strength of a rail transit converter according to an embodiment of the present invention. Figure 5 As shown, the device specifically includes:

[0086] The output module 10 is used to determine the sample input of basic parameters and several key design variables according to the structural characteristics of the converter, and determine the sample output of stress as a response. The basic parameters and several key design variable parameters as sample inputs include: converter total weight, center of gravity position, material grade, installation size, converter cabinet size, sheet metal thickness and key component shape size, and the stress as a response obtained as sample output includes: maximum von Mises stress, static stress (four different working conditions) and random vibration stress (three different working conditions).

[0087] The generation module 20 is used to generate 10d sampling positions (ie, a 10d×d sampling matrix) by a Latin hypercube sampling method according to the stress obtained as a response from the previous combing results, where d is the sum of the number of basic parameters and the number of design variables.

[0088] The construction module 30 is used to construct finite element models for the 10d sampling positions generated above and perform simulation calculations to form a simulation database.

[0089] The optimization module 40 is used to perform statistical analysis and machine learning modeling on the previously obtained samples, and to construct a functional relationship between the sample input (basic parameters of the converter, designed converter) and the sample output.

[0090] The adjustment module 50 is used to perform adaptive sampling for the Kriging model that does not meet the model accuracy requirement, taking the maximization of the mean square prediction error as the criterion:

[0091]

[0092] Then, the Kriging model is rebuilt and the model accuracy is evaluated, and the model accuracy is repeatedly adjusted.

[0093] Until the model accuracy meets the requirements;

[0094] The evaluation module 60 is used to quickly evaluate the structural strength of the converter according to the constructed Kriging model:

[0095]

[0096] It should be noted that among these parameters:

[0097] is the predicted mean, is the root mean square value of the prediction error, λ is the model correction coefficient, λ∈[1,3].

[0098] The present invention provides a device and system for evaluating the structural strength of a rail transit inverter. A sample database is constructed based on a finite element model and an experimental design method. A Kriging proxy model is constructed by statistically approximating the sample data in a machine learning manner and an adaptive sampling method is used to improve the model accuracy. The stress response of the inverter under different working conditions is predicted based on the information provided by the proxy model, thereby achieving rapid evaluation of the structural strength and shortening the design cycle.

[0099] Figure 5 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 5 As shown, the electronic device may include: a processor 510, a communications interface 520, a memory 530 and a communication bus 540, wherein the processor 510, the communications interface 520 and the memory 530 communicate with each other through the communication bus 540. The processor 510 may call the logic command in the memory 530 to execute the method for detecting the standardization of the measurement of the finished oil tank, and the method includes:

[0100] Determine sample inputs of basic parameters and several key design variables based on the structural characteristics of the converter, and determine sample outputs of stress as a response;

[0101] According to the stress of the response, the sampling positions are generated by the Latin hypercube sampling method;

[0102] Construct finite element models for the sampling positions respectively and perform simulation calculations to form a simulation database;

[0103] Perform statistical analysis and machine learning modeling based on the simulation database, build a functional relationship between sample input and sample output, and obtain an evaluation model;

[0104] Rebuild the evaluation model and evaluate the model accuracy, and adjust the model accuracy until the model accuracy meets the requirements;

[0105] Based on the reconstructed evaluation model, the structural strength of the rail transit converter is evaluated.

[0106] In addition, the logic commands in the above-mentioned memory 530 can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software medium, and the computer software medium is stored in a storage medium, including several commands to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0107] On the other hand, the present invention further provides a computer program medium, the computer program medium includes a computer program, the computer program can be stored in a non-transitory computer-readable storage medium, when the computer program is executed by a processor, the computer can execute the method for detecting the standardization of the measurement of finished oil tanks provided by the above methods, the method comprising:

[0108] Determine sample inputs of basic parameters and several key design variables based on the structural characteristics of the converter, and determine sample outputs of stress as a response;

[0109] According to the stress of the response, the sampling positions are generated by the Latin hypercube sampling method;

[0110] Construct finite element models for the sampling positions respectively and perform simulation calculations to form a simulation database;

[0111] Perform statistical analysis and machine learning modeling based on the simulation database, build a functional relationship between sample input and sample output, and obtain an evaluation model;

[0112] Rebuild the evaluation model and evaluate the model accuracy, and adjust the model accuracy until the model accuracy meets the requirements;

[0113] Based on the reconstructed evaluation model, the structural strength of the rail transit converter is evaluated.

[0114] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is implemented when the computer program is executed by a processor to perform the method for detecting the standardization of the measurement of finished oil tanks provided by the above methods, the method comprising:

[0115] Determine sample inputs of basic parameters and several key design variables based on the structural characteristics of the converter, and determine sample outputs of stress as a response;

[0116] According to the stress of the response, the sampling positions are generated by the Latin hypercube sampling method;

[0117] Construct finite element models for the sampling positions respectively and perform simulation calculations to form a simulation database;

[0118] Perform statistical analysis and machine learning modeling based on the simulation database, build a functional relationship between sample input and sample output, and obtain an evaluation model;

[0119] Rebuild the evaluation model and evaluate the model accuracy, and adjust the model accuracy until the model accuracy meets the requirements;

[0120] Based on the reconstructed evaluation model, the structural strength of the rail transit converter is evaluated.

[0121] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, i.e., they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative effort.

[0122] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software medium, and the computer software medium can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of commands to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of each embodiment or some parts of the embodiment.

[0123] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for evaluating the structural strength of a rail transit converter, characterized in that: include: Determine sample inputs of basic parameters and several key design variables based on the structural characteristics of the converter, and determine sample outputs of stress as a response; According to the stress of the response, the sampling positions are generated by the Latin hypercube sampling method; Construct finite element models for the sampling positions respectively and perform simulation calculations to form a simulation database; Perform statistical analysis and machine learning modeling based on the simulation database, build a functional relationship between sample input and sample output, and obtain an evaluation model; Rebuild the evaluation model and evaluate the model accuracy, and adjust the model accuracy until the model accuracy meets the requirements; Based on the reconstructed evaluation model, the structural strength of the rail transit converter is evaluated.

2. The method for evaluating the structural strength of a rail transit converter according to claim 1, characterized in that: The basic parameters and several key design variable parameters used as sample inputs include at least: total weight of the converter, center of gravity position, material grade, installation dimensions, converter cabinet dimensions, sheet metal thickness, and shape and dimensions of key components.

3. The method for evaluating the structural strength of a rail transit converter according to claim 1, characterized in that: The stresses that are the responses of the sample output include: maximum stress, static stress and random vibration stress.

4. The method for evaluating the structural strength of a rail transit converter according to claim 1, characterized in that: The evaluation model is a Kriging model.

5. The method for evaluating the structural strength of a rail transit converter according to claim 4, characterized in that: The method further comprises: For evaluation models that do not meet the model accuracy requirements, adaptive sampling is performed based on the criterion of maximizing the mean square prediction error.

6. The method for evaluating the structural strength of a rail transit converter according to claim 1, characterized in that: The statistical analysis and machine learning modeling based on the simulation database are performed to construct a functional relationship between sample input and sample output to obtain an evaluation model, which specifically includes: Normalize the sampling positions: The method optimizes the model's hyperparameters by maximum likelihood estimation: The accuracy of the constructed evaluation model was evaluated using the fitting quality coefficient and the maximum absolute error.

7. The method for evaluating the structural strength of a rail transit converter according to claim 6, characterized in that: The sum of the number of basic parameters and the number of design variables at the sampling locations.

8. A method and device for evaluating the structural strength of a rail transit converter, characterized in that: The device comprises: An output module, for determining sample inputs of basic parameters and several key design variables according to structural characteristics of the converter, and determining sample outputs of stress as a response; A generation module, used for generating sampling positions by a Latin hypercube sampling method according to the stress of the response; A construction module also constructs finite element models for the sampling positions and performs simulation calculations to form a simulation database; The optimization module is used to perform statistical analysis and machine learning modeling based on the simulation database, build a functional relationship between sample input and sample output, and obtain an estimation model; The adjustment module is used to rebuild the estimation model and evaluate the model accuracy, and adjust the model accuracy until the model accuracy meets the requirements; The evaluation module is used to estimate the structural strength of the rail transit converter according to the reconstructed estimation model.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the method for evaluating the structural strength of a rail transit converter as described in any one of claims 1 to 7 are implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for evaluating the structural strength of a rail transit converter as claimed in any one of claims 1 to 7 are implemented.

Images