CN114626278B - Method for evaluating structural performance and monitoring health of power battery bracket - Google Patents
Method for evaluating structural performance and monitoring health of power battery bracket Download PDFInfo
- Publication number
- CN114626278B CN114626278B CN202210527233.1A CN202210527233A CN114626278B CN 114626278 B CN114626278 B CN 114626278B CN 202210527233 A CN202210527233 A CN 202210527233A CN 114626278 B CN114626278 B CN 114626278B
- Authority
- CN
- China
- Prior art keywords
- sampling
- model
- data
- battery bracket
- space
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/23—Design optimisation, verification or simulation using finite element methods [FEM] or finite difference methods [FDM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/15—Vehicle, aircraft or watercraft design
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/17—Mechanical parametric or variational design
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/20—Ensemble learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/01—Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C5/00—Registering or indicating the working of vehicles
- G07C5/08—Registering or indicating performance data other than driving, working, idle, or waiting time, with or without registering driving, working, idle or waiting time
- G07C5/0808—Diagnosing performance data
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C5/00—Registering or indicating the working of vehicles
- G07C5/08—Registering or indicating performance data other than driving, working, idle, or waiting time, with or without registering driving, working, idle or waiting time
- G07C5/0816—Indicating performance data, e.g. occurrence of a malfunction
- G07C5/0825—Indicating performance data, e.g. occurrence of a malfunction using optical means
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/08—Probabilistic or stochastic CAD
Abstract
The invention discloses a method for evaluating the structural performance and monitoring the health of a power battery bracket, which comprises the following steps: collecting design data of the battery bracket; establishing a finite element model according to the design data of the battery bracket; expanding data calculated by the finite element model and establishing a sampling space; sampling the sampling space to obtain a sample, and calculating the sample through a finite element model to obtain calculated data; establishing an artificial intelligence model according to the calculated data; carrying out structural performance evaluation on the power battery bracket in an automobile research and development stage according to the artificial intelligence model to obtain an evaluation result, and carrying out structural performance health monitoring on the power battery bracket in an automobile service stage according to the artificial intelligence model to obtain a health monitoring result; and constructing a digital twin model according to the artificial intelligence model, and carrying out visual display on the evaluation result and the health monitoring result by the digital twin model. The method can save labor cost and time cost, and provides guarantee for the working and driving safety of the automobile.
Description
Technical Field
The invention relates to the technical field of power battery bracket performance evaluation and health monitoring, in particular to a method for evaluating the structural performance and monitoring the health of a power battery bracket.
Background
In 2022 in the major meeting of the commercial vehicle industry in China, experts predict that more than half of commercial vehicles in 2035 can be used for realizing new energy, the situation that the quantity of new-energy commercial vehicles in China is increased by policy is finished, and the market share is improved by quality at present, so that more technical and institutional innovations are needed. The power battery bracket of the new energy commercial vehicle is used as a bearing structure of the battery module, bears severe working conditions of variable load, impact vibration, high-low temperature circulation and other factors, and the structural performance of the power battery bracket directly determines the safety problem of vehicle running, so that the power battery bracket has great significance in real-time evaluation and health monitoring of the structural performance in research, development and service stages.
In the research and development stage of new energy commercial vehicles, the main method for evaluating the performance of the power battery bracket at present is as follows: through finite element preliminary analysis, then based on the road test of big mileage, whether the structure has the problems of weld joint cracking, section bar deformation, fracture and the like is checked through manual work, and the problems are returned to the designer for key optimization, and finally the steps are repeated until the whole structure has no cracking risk. But the process is slow, time consuming and labor intensive. And because the road test mileage is large, the time is long, the time that the structural performance problem of battery bracket appears can not be accurately judged by manual work, when a plurality of composite performance problems appear, the inducement of the performance problem and the sequence or coupling effect of the structural performance problems are more difficult to judge, thereby increasing the research and development difficulty of designers and prolonging the design cycle.
In the service stage of the new-energy commercial vehicle, the new-energy commercial vehicle is severe in working condition and complex in working condition, the battery bracket is easy to generate structural deformation and accumulate fatigue, and finally, an automobile accident can be caused, so that the life risk and the economic loss of a driver are caused, and therefore, the structural performance of the battery bracket is evaluated and the real-time health monitoring is carried out, so that the important significance is realized for guaranteeing the driving safety of the automobile and promoting the development of the new-energy commercial vehicle. In other fields, for example, a measuring method of structural deformation such as a visual measuring method and a deformation monitoring method based on a displacement sensor are widely applied, but for a new energy commercial vehicle, due to the defects that the use condition is limited, the global deformation, cracking, failure and the like of the structure cannot be obtained, mature methods in other fields cannot be applied to the online monitoring of the power battery bracket of the new energy commercial vehicle.
Disclosure of Invention
Therefore, the technical problem to be solved by the invention is to overcome the defects that manual inspection is required to be carried out after the battery bracket is disassembled, time and labor are wasted, the occurrence sequence of structural performance problems is difficult to judge, and online health monitoring cannot be carried out on the power battery bracket of a new energy commercial vehicle in the prior art, so that the method for carrying out structural inspection without disassembly, saving labor cost and time cost, and realizing real-time health monitoring on the structural performance of the battery bracket, in particular to the method for evaluating and monitoring the structural performance of the power battery bracket.
The invention provides a method for performance evaluation and health monitoring of a power battery bracket structure, which comprises the following steps:
s1: collecting design data of the battery bracket;
s2: establishing a finite element model according to the design data of the battery bracket; expanding data calculated by the finite element model and establishing a sampling space;
s3: sampling the sampling space to obtain a sample, and calculating the sample through a finite element model to obtain calculated data; establishing an artificial intelligence model according to the calculated data;
s4: carrying out structural performance evaluation on the power battery bracket in an automobile research and development stage according to an artificial intelligence model to obtain an evaluation result; carrying out structural performance health monitoring on the power battery bracket in the automobile service stage according to the artificial intelligence model to obtain a health monitoring result;
s5: and constructing a digital twin model according to the artificial intelligence model, carrying out visual display on the evaluation result and the health monitoring result by the digital twin model, feeding back the evaluation result to a designer, and feeding back the health monitoring result to a driver.
Preferably, the design data of the battery carrier includes operating condition parameter data, shape information data, size information data, connection information data, material information data, and battery weight data.
Preferably, in S2, the data calculated by the finite element model is expanded by using a test design method, a sampling factor and an initial sample point are set, and a sampling space is established according to the sampling factor, the initial sample point and the expanded data, wherein the sampling space includes multiple layers; the sampling space is used for obtaining data of all working conditions, and therefore the precision of the digital twin model is improved; the expression of the sampling space is:
wherein the content of the first and second substances,Mto representmA sampling space of layers;Ban upsampling boundary representing a first layer of sampling space;A 1 representing a sampling boundary of a first layer of sampling space and a second layer of sampling space;A 3 representing a sampling boundary of the second layer of sampling space and the third layer of sampling space;A 5 representing a sampling boundary of a third layer of sampling space and a fourth layer of sampling space;A n-1 is shown asm-1 layer of sampling space andma sampling boundary of a layer sampling space;A n+1 is shown asmThe downsampled boundary of the layer sample space,nis an even number; and isB=A 0 +10%A 0 ,A 0 Representing an initial sample point;
wherein the content of the first and second substances,A n is shown asmA central sampling point of the layer sampling space;A 1 =A 0 -10%A 0 。
preferably, the process of obtaining the first layer sampling space is as follows: taking the initial sample point as the center, and obtaining the sampling space of the initial sample point according to the sampling factor, which is marked as [ 2 ]A 0 -10%A 0 ,A 0 +10%A 0 ];A 0 -10%A 0 A downsampled boundary of the sample space representing the initial sample point,A 0 +10%A 0 an upsampled boundary of the sampling space representing an initial sample point; the sampling space of the initial sample points is taken as the first layer sampling space,the upsampling boundary of the first layer sampling space isA 0 +10%A 0 I.e. byB=A 0 +10%A 0 (ii) a The sampling boundary between the first layer of sampling space and the second layer of sampling space isA 0 -10%A 0 I.e. byA 1 =A 0 -10%A 0 。
Preferably, S2 further includes: and setting the loading condition of the finite element model of the battery bracket under the limit working condition, and calculating the strength performance of the battery bracket under the limit working condition by adopting a finite element.
Preferably, in S3, the samples include a first sample and a random sample, and the calculated data includes training data and verification data;
sampling a sampling space by adopting a Latin hypercube sampling method to obtain a first sample, and calculating the first sample through a finite element model to obtain training data; training a hybrid machine learning model according to training data, and establishing an artificial intelligence model;
sampling the sampling space by adopting a random sampling method to obtain a random sample, calculating the random sample through a finite element model to obtain verification data, and verifying the generalization of the artificial intelligence model according to the verification data.
Preferably, the hybrid machine learning model includes an extreme gradient boosting model, a k-fold cross validation model, and a random forest model.
Preferably, the artificial intelligence model is evaluated by accuracy, and the accuracy of the artificial intelligence model is recorded asAnd the expression is:
wherein the content of the first and second substances,αa weight representing the prediction accuracy of the random forest model,βa weight representing the prediction accuracy of the extreme gradient boosting model,γweights representing prediction accuracy of the k-fold cross validation model;the prediction accuracy of the random forest model is shown,representing the prediction accuracy of the extreme gradient boost model,representing the prediction precision of the k-fold cross validation model; and is
Wherein the content of the first and second substances,R,irepresents the first under random forest modeliSampling points;X,irepresenting extreme gradient lifting modeliSampling points;C,irepresents the first under a k-fold cross-validation modeliA number of sample points are sampled at the time of sampling,Nrepresenting the number of sampling points;expressing the second place under the random forest modeliOutput parameter values at the sampling points;represents the first under the extreme gradient lifting modeliOutput parameter values at the sampling points;representing the first under a k-fold cross-validation modeliOutput parameter values at the sampling points;expressing the second place under the random forest modeliRegression model values at each sampling point;representing extreme gradient lifting modeliRegression model values at each sampling point;representation under the model of k-fold cross validationiRegression model values at each sampling point;is composed ofIs calculated as the arithmetic mean of the average of the values,is composed ofIs calculated as the arithmetic mean of the average of the values,is composed ofIs calculated as the arithmetic mean of (1).
Preferably, in S4, the process of obtaining the evaluation result is:
in the automobile research and development stage, experimental acceleration data are uploaded to the artificial intelligence model in real time through the acceleration sensor, and the artificial intelligence model carries out real-time intensity response calculation on the experimental acceleration data to obtain an evaluation result and realize the structural performance evaluation of the power battery bracket;
the process of obtaining the health monitoring result is as follows:
in the service stage of the automobile, service acceleration data are uploaded to the artificial intelligence model in real time through the acceleration sensor, the artificial intelligence model carries out real-time intensity response calculation on the service acceleration data, a health monitoring result is obtained, and the structural performance health monitoring of the power battery bracket is achieved.
Preferably, in S5, a digital twin model is constructed by combining the digital model with the artificial intelligence model, and the digital twin model visually displays the evaluation result and the health monitoring result through the digital model.
The technical scheme of the invention has the following advantages:
1. by the method, the structural performance of the power battery bracket is evaluated and the health of the power battery bracket is monitored by carrying out real-time strength response on the acceleration data through the artificial intelligence model; the method can save labor cost and time cost;
2. under severe service conditions and working conditions of the new energy commercial vehicle, the method provided by the invention can be used for monitoring the structural performance of the power battery bracket on line in real time through the artificial intelligence model, further carrying out real-time visual display on the health monitoring result through the digital twin model, and feeding the health monitoring result back to a driver, so that the life danger of the driver and the economic loss of the driver caused by the occurrence of an automobile accident are avoided, the working and running safety of the vehicle is guaranteed, and the development of the new energy commercial vehicle is further promoted.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a flow chart of a method for power battery carrier structural performance assessment and health monitoring in the practice of the present invention;
FIG. 2 is a schematic flow chart of training data and validation data obtained in the practice of the present invention;
FIG. 3 is a schematic flow chart illustrating the process of creating and validating an artificial intelligence model in the practice of the present invention.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings, and it should be understood that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the description of the present invention, it should be noted that the terms "first", "second", and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
Furthermore, the technical features involved in the different embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
The power battery of the new energy commercial vehicle is a main power source, the battery bracket is a bearing part of the power battery, and the structural performance of the battery bracket directly determines the driving safety of the vehicle, so that the good structural performance of the battery bracket is very important. At the present stage, the new energy commercial vehicle has no good method for evaluating the structural performance of the battery bracket in the research and development stage, and the main method is to disassemble the battery bracket for manual inspection after a long-distance road test. As shown in fig. 1, the present embodiment provides a method for performance evaluation and health monitoring of a power battery bracket structure, the method comprising the steps of:
s1: collecting design data of the battery bracket;
in the present embodiment, the design data of the battery bracket includes operating condition parameter data, shape information data, size information data, connection information data, material information data, and battery weight data; and iteratively updating the design data of the battery bracket in the automobile development stage.
S2: establishing a finite element model according to the design data of the battery bracket; expanding data calculated by the finite element model and establishing a sampling space; then setting the loading condition of the finite element model of the battery bracket under the limit working condition, and calculating the strength performance of the battery bracket under the limit working condition by adopting a finite element;
specifically, expanding data calculated by a finite element model by using a test design method, setting a sampling factor and an initial sample point, and establishing a sampling space according to the sampling factor, the initial sample point and the expanded data, wherein the sampling space comprises a plurality of layers; the sampling space is used for obtaining data of all working conditions, so that the precision of the digital twin model is improved, and the sampling space is convenient for obtaining loading condition data of the global working conditions only by using loading condition data of the limit working conditions meeting the design working condition requirements.
In this embodiment, because the loading that needs to satisfy all operating modes, so set up a central sampling point for every layer of sampling space, the central sampling point in first layer of sampling space is initial sample point promptly to set up sampling factor as 10%, every layer of sampling space is marked as respectively:A 0 、A 2 、A 4 、…、A n ;A 0 representing the center sample point of the first layer of sample space,A 2 representing the center sample point of the second layer sample space,A 4 representing the center sample point of the third tier of sample space,A n is shown asmA center sampling point of the layer sampling space, annIs determined as an even number; therefore, it ismThe expression for the layer sample space can be written as:
wherein the content of the first and second substances,Mto representmThe space is sampled by the layers of the image,nis an even number and is provided with a plurality of groups,ais even and comprises 0,2,4,6, …n。
Obtaining the space of each layer of sampling space according to the sampling factors and each central sampling point; which are respectively as follows:
the total spacing of the sampling spaces of each layer is expressed as:
wherein the content of the first and second substances,represents the total spacing of m layers of sampling spaces, j represents the j-th layer of sampling spaces, and n is an even number starting from 0;
then the above-mentioned layers are sampled according to the space between the sampling spacesmAnd rewriting an expression of the layer sampling space to obtain:
wherein the content of the first and second substances,Mto representmA sampling space of layers;Ban upsampling boundary representing a first layer of sampling space;A 1 representing a sampling boundary of a first layer of sampling space and a second layer of sampling space;A 3 representing a sampling boundary of the second layer of sampling space and the third layer of sampling space;A 5 representing a sampling boundary of a third layer of sampling space and a fourth layer of sampling space;A n-1 is shown asm-1 layer of sampling space andma sampling boundary of a layer sampling space;A n+1 to representFirst, themThe downsampled boundary of the layer sample space,nis an even number; and isB=A 0 +10%A 0 ,A 0 Representing an initial sample point;
wherein the content of the first and second substances,A n is shown asmA center sampling point of a layer sampling space;A 1 =A 0 -10%A 0 。
for example: the process of obtaining the first layer of sampling space is as follows: taking the initial sample point as the center, and obtaining the sampling space of the initial sample point according to the sampling factor, which is marked as [ 2 ]A 0 -10%A 0 ,A 0 +10%A 0 ];A 0 -10%A 0 A downsampled boundary of the sample space representing the initial sample point,A 0 +10%A 0 an upsampled boundary of the sampling space representing an initial sample point; taking the sampling space of the initial sample point as the first layer sampling space, the up-sampling boundary of the first layer sampling space isA 0 +10%A 0 I.e. byB=A 0 +10%A 0 (ii) a The sampling boundary between the first layer of sampling space and the second layer of sampling space isA 0 -10%A 0 I.e. byA 1 =A 0 -10%A 0 。
Constructed in this waymThe layer sample space ignores the value of 0,A n+1 ]can choose whether to discard according to the working requirement of a specific digital twin body, andmthe layer sampling space should also select a suitable number of layers according to specific working requirements, and not the greater the number of layers, the better.
S3: sampling the sampling space to obtain a sample, and calculating the sample through a finite element model to obtain calculated data; establishing an artificial intelligence model according to the calculated data;
specifically, the samples include a first sample and a random sample, and the calculated data includes training data and verification data;
as shown in fig. 2 and 3, a first sample is obtained by sampling a sampling space by a latin hypercube sampling method (LHS), and the first sample is calculated by a finite element model to obtain training data; mixing an extreme gradient lifting model (XGboost model), a k-fold cross validation model and a random forest model by adopting a hybrid machine learning method to obtain a hybrid machine learning model, training the hybrid machine learning model according to part of training data, testing the hybrid machine learning model according to part of the training data, and establishing an artificial intelligence model; in the present embodiment, 70% of the training data is used for training the hybrid machine learning model, and 30% of the training data is used for testing the hybrid machine learning model;
sampling the sampling space by adopting a random sampling method to obtain a random sample, calculating the random sample through a finite element model to obtain verification data, and verifying the generalization of the artificial intelligence model according to the verification data; in this embodiment, 100 sets of verification data are selected to verify the artificial intelligence model.
In this embodiment, the artificial intelligence model is evaluated by the accuracy, and the accuracy of the artificial intelligence model is recorded asAnd the expression is:
wherein the content of the first and second substances,αa weight representing the prediction accuracy of the random forest model,βa weight representing the prediction accuracy of the extreme gradient boosting model,γweights representing prediction accuracy of the k-fold cross validation model;the prediction accuracy of the random forest model is shown,representing the prediction accuracy of the extreme gradient boost model,representing the prediction precision of the k-fold cross validation model; and is
Wherein the content of the first and second substances,R,iexpressing the second place under the random forest modeliSampling points;X,irepresenting extreme gradient lifting modeliSampling points;C,irepresenting the first under a k-fold cross-validation modeliA number of sample points are sampled at the time of sampling,Nrepresenting the number of sampling points;expressing the second place under the random forest modeliOutput parameter values at the sampling points;representing extreme gradient lifting modeliOutput parameter values at the sampling points;representing the first under a k-fold cross-validation modeliOutput parameter values at the sampling points;expressing the second place under the random forest modeliRegression model values at each sampling point;representing extreme gradient lifting modeliRegression model values at each sampling point;representation under the model of k-fold cross validationiRegression model values at each sampling point;is composed ofIs calculated as the arithmetic mean of the average of the values,is composed ofIs calculated as the arithmetic mean of the average of the values,is composed ofIs calculated as the arithmetic mean of (1).
S4: carrying out structural performance evaluation on the power battery bracket in an automobile research and development stage according to an artificial intelligence model to obtain an evaluation result; carrying out structural performance health monitoring on the power battery bracket in the automobile service stage according to the artificial intelligence model to obtain a health monitoring result;
specifically, the process of obtaining the evaluation result is as follows:
in an automobile research and development stage, experimental acceleration data are uploaded to an artificial intelligence model in real time through an acceleration sensor for a power battery bracket, the artificial intelligence model carries out real-time intensity response calculation on the experimental acceleration data to obtain an evaluation result, and the structural performance evaluation of the power battery bracket is realized;
the process of obtaining the health monitoring result is as follows:
in the service stage of the automobile, service acceleration data are uploaded to the artificial intelligence model in real time through the acceleration sensor, the artificial intelligence model carries out real-time intensity response calculation on the service acceleration data, a health monitoring result is obtained, and the structural performance health monitoring of the power battery bracket is achieved.
S5: constructing a digital twin model according to the artificial intelligence model, and carrying out visual display on an evaluation result and a health monitoring result by the digital twin model; and the digital twin model feeds the evaluation result back to the designer and feeds the health monitoring result back to the driver.
In the embodiment, an artificial intelligence model and a digital model are used as carriers of a digital twin model to construct the digital twin model; and the digital twin model visually displays the evaluation result and the health monitoring result through the digital model, so that the real-time structural performance prediction is realized, and the aim of visually monitoring the structural health is fulfilled.
Further, integrating the finite element model, the artificial intelligence model and the digital twin model, wherein the integration process is the above steps, and deploying the integrated finite element model, the artificial intelligence model and the digital twin model on the digital twin cloud platform, wherein the deployment mode includes: the digital twin cloud platform can be deployed by utilizing the existing big data platform (such as the Alibara cloud); the digital twin cloud platform is used for realizing the structural performance evaluation and health monitoring of the power battery bracket and carrying out visual display on the power battery bracket.
Furthermore, in the automobile development stage, the changed power battery bracket structure after performance evaluation is fed back to the digital twin cloud platform, and each model is updated to adapt to a new round of evaluation work.
The experimental acceleration data and the experimental acceleration data are uploaded to the artificial intelligent model through the acceleration sensor in real time, so that the digital twin model can predict and display the sequence of the structural performance problems in real time through the artificial intelligent model, and whether the structural performance problem appearing first produces coupling influence on the structural performance problem appearing later can be quickly judged through the load transfer path principle due to the determination of the sequence of the structural performance problems.
In this embodiment, the digital twin cloud platform may be replaced by an edge server or an individual machine.
The method provided by the embodiment can be used for evaluating the structural performance of the power battery bracket of the new-energy commercial vehicle in real time and monitoring the health, not only can the structure be optimized and the time be saved, but also the health performance early warning can be carried out to remind a user whether the vehicle needs to be overhauled or not, so that the safety of the new-energy commercial vehicle is improved, and the rapid and high-quality development of the new-energy commercial vehicle is promoted; the specific principle is as follows: extracting physical parameters such as size, shape, quality and material of a designed vehicle to establish a finite element model, establishing loading conditions of extreme working conditions for the finite element model and calculating the strength performance of the finite element model, acquiring more calculation data by using a test design method to establish an artificial intelligence model, and finally establishing a digital twin model according to the artificial intelligence model and the digital model. The structural performance of the power battery bracket is monitored on line in real time through the artificial intelligence model, and then the health monitoring result is visually displayed in real time through the digital twin model, so that the structural performance of the power battery bracket is evaluated and monitored.
By adopting the method provided by the embodiment, real-time evaluation can be carried out at any mileage stage of a sample vehicle road test in a research and development stage, and a battery module does not need to be disassembled, so that the complex operation of manual retrieval is saved; in the service stage of the vehicle of the user, the real-time health monitoring of the structural performance can be carried out, so that the driving safety of a driver is improved; a digital twin model of the new energy commercial vehicle battery bracket is established based on a digital twin technology and by combining a finite element method and a hybrid machine learning method. The hybrid machine learning technology can integrate the prediction advantages of each sub machine learning on data characteristics, so that the prediction precision and the generalization of the artificial intelligent model are improved; a sampling space construction method is also provided, and the sampling space provides a data base for training an artificial intelligence model and improves the precision.
It should be understood that the above examples are only for clarity of illustration and are not intended to limit the embodiments. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. And are neither required nor exhaustive of all embodiments. And obvious variations or modifications therefrom are within the scope of the invention.
Claims (9)
1. A method for power cell carrier structural performance assessment and health monitoring, comprising:
s1: collecting design data of the battery bracket;
s2: establishing a finite element model according to the design data of the battery bracket; expanding data calculated by the finite element model and establishing a sampling space;
expanding data calculated by a finite element model by using a test design method, setting a sampling factor and an initial sample point, and establishing a sampling space according to the sampling factor, the initial sample point and the expanded data, wherein the sampling space comprises a plurality of layers; the sampling space is used for obtaining data of all working conditions, and therefore the precision of the digital twin model is improved; the expression of the sampling space is:
wherein the content of the first and second substances,Mto representmA sampling space of layers;Ban upsampling boundary representing a first layer of sampling space;A 1 representing a sampling boundary of a first layer of sampling space and a second layer of sampling space;A 3 representing a sampling boundary of the second layer of sampling space and the third layer of sampling space;A 5 representing a sampling boundary of a third layer of sampling space and a fourth layer of sampling space;A n-1 is shown asm-1 layer of sampling space andma sampling boundary of a layer sampling space;A n+1 is shown asmThe downsampled boundary of the layer sample space,nis an even number; and isB=A 0 +10%A 0 ,A 0 Representing an initial sample point;
wherein the content of the first and second substances,A n is shown asmA central sampling point of the layer sampling space;A 1 =A 0 -10%A 0 ;
s3: sampling the sampling space to obtain a sample, and calculating the sample through a finite element model to obtain calculated data; establishing an artificial intelligence model according to the calculated data;
s4: according to the artificial intelligence model, structural performance evaluation is carried out on the power battery bracket in an automobile research and development stage to obtain an evaluation result; carrying out structural performance health monitoring on the power battery bracket in the automobile service stage according to the artificial intelligence model to obtain a health monitoring result;
s5: and constructing a digital twin model according to the artificial intelligence model, carrying out visual display on the evaluation result and the health monitoring result by the digital twin model, feeding back the evaluation result to a designer, and feeding back the health monitoring result to a driver.
2. The method for power battery carrier structural performance assessment and health monitoring of claim 1, wherein the design data of the battery carrier comprises operating condition parameter data, shape information data, size information data, connection information data, material information data, and battery weight data.
3. The method for power battery bracket structure performance assessment and health monitoring according to claim 1, wherein the process of obtaining the first layer of sampling space is: taking the initial sample point as a center, and obtaining a sampling space of the initial sample point according to the sampling factor, which is marked as [ 2 ]A 0 -10%A 0 ,A 0 +10%A 0 ];A 0 -10%A 0 A downsampled boundary of the sample space representing the initial sample point,A 0 +10%A 0 an upsampled boundary of the sampling space representing an initial sample point; taking the sampling space of the initial sample point as a first layer sampling space, and then the up-sampling boundary of the first layer sampling space isA 0 +10%A 0 I.e. byB=A 0 +10%A 0 (ii) a The sampling boundary between the first layer of sampling space and the second layer of sampling space isA 0 -10%A 0 I.e. byA 1 =A 0 -10%A 0 。
4. The method for power battery bracket structure performance assessment and health monitoring as claimed in claim 1, wherein in S2, further comprising: and setting the loading condition of the finite element model of the battery bracket under the limit working condition, and calculating the strength performance of the battery bracket under the limit working condition by adopting a finite element.
5. The method for power battery bracket structure performance assessment and health monitoring of claim 1, wherein in S3, the samples comprise a first sample and a random sample, and the calculated data comprises training data and verification data;
sampling the sampling space by adopting a Latin hypercube sampling method to obtain the first sample, and calculating the first sample through a finite element model to obtain training data; training a hybrid machine learning model according to the training data, and establishing an artificial intelligence model;
sampling the sampling space by adopting a random sampling method to obtain a random sample, calculating the random sample through a finite element model to obtain verification data, and verifying the generalization of the artificial intelligence model according to the verification data.
6. A method for power battery carriage structural performance assessment and health monitoring according to claim 5, wherein said hybrid machine learning models comprise extreme gradient boosting models, k-fold cross validation models and random forest models.
7. The method for power battery bracket structural performance assessment and health monitoring according to claim 6, wherein the artificial intelligence model is evaluated by accuracy, and the accuracy of the artificial intelligence model is recorded asAnd the expression is:
wherein the content of the first and second substances,αa weight representing the prediction accuracy of the random forest model,βweights representing the prediction accuracy of the extreme gradient boosting model,γweights representing prediction accuracy of the k-fold cross validation model;the prediction accuracy of the random forest model is shown,representing the prediction accuracy of the extreme gradient boost model,representing the prediction precision of the k-fold cross validation model; and is
Wherein the content of the first and second substances,R,iexpressing the second place under the random forest modeliSampling points;X,irepresenting extreme gradient lifting modeliSampling points;C,irepresents the first under a k-fold cross-validation modeliA number of sample points are sampled at the time of sampling,Nrepresenting the number of sampling points;expressing the second place under the random forest modeliOutput parameter values at the sampling points;representing extreme gradient lifting modeliOutput parameter values at the sampling points;representing the first under a k-fold cross-validation modeliOutput parameter values at the sampling points;expressing the second place under the random forest modeliRegression model values at each sampling point;representing extreme gradient lifting modeliRegression model values at each sampling point;representation under the model of k-fold cross validationiRegression model values at each sampling point;is composed ofIs calculated as the arithmetic mean of the average of the values,is composed ofIs calculated as the arithmetic mean of the average of the values,is composed ofIs calculated as the arithmetic mean of (1).
8. The method for structural performance evaluation and health monitoring of a power battery bracket according to claim 1, wherein in S4, the process of obtaining the evaluation result is:
in an automobile research and development stage, experimental acceleration data are uploaded to the artificial intelligence model in real time through an acceleration sensor for the power battery bracket, and the artificial intelligence model carries out real-time intensity response calculation on the experimental acceleration data to obtain an evaluation result so as to realize structural performance evaluation of the power battery bracket;
the process of obtaining the health monitoring result is as follows:
in the service stage of the automobile, service acceleration data are uploaded to the artificial intelligence model in real time through an acceleration sensor for the power battery bracket, the artificial intelligence model carries out real-time intensity response calculation on the service acceleration data to obtain the health monitoring result, and the structural performance health monitoring of the power battery bracket is realized.
9. The method for power battery bracket structure performance evaluation and health monitoring as claimed in claim 1, wherein in S5, the digital twin model is constructed according to the artificial intelligence model in combination with a digital model, and the digital twin model visually displays the evaluation result and the health monitoring result through the digital model.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210527233.1A CN114626278B (en) | 2022-05-16 | 2022-05-16 | Method for evaluating structural performance and monitoring health of power battery bracket |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210527233.1A CN114626278B (en) | 2022-05-16 | 2022-05-16 | Method for evaluating structural performance and monitoring health of power battery bracket |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114626278A CN114626278A (en) | 2022-06-14 |
CN114626278B true CN114626278B (en) | 2022-08-05 |
Family
ID=81907041
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210527233.1A Active CN114626278B (en) | 2022-05-16 | 2022-05-16 | Method for evaluating structural performance and monitoring health of power battery bracket |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114626278B (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110488629A (en) * | 2019-07-02 | 2019-11-22 | 北京航空航天大学 | A kind of management-control method of the hybrid vehicle based on the twin technology of number |
CN111737811A (en) * | 2020-05-09 | 2020-10-02 | 北京航空航天大学 | Helicopter movable part service life management method, device and medium based on digital twin |
CN113569445A (en) * | 2021-07-01 | 2021-10-29 | 洛阳星派数值仿真研究院有限公司 | Steel structure health monitoring system and method based on digital twinning technology |
CN114254457A (en) * | 2021-12-17 | 2022-03-29 | 中国矿业大学(北京) | Mine hydraulic support fatigue life prediction method based on digital twinning |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107290679B (en) * | 2017-07-03 | 2019-11-12 | 南京能瑞电力科技有限公司 | The Intelligentized battery method for detecting health status of charging pile is shared for electric car |
CN109960877B (en) * | 2019-03-25 | 2022-07-22 | 江铃汽车股份有限公司 | Method and system for analyzing strength of automobile battery pack bracket |
CN110492186B (en) * | 2019-07-02 | 2021-10-01 | 北京航空航天大学 | Power battery module management method based on cloud control technology |
US11433977B2 (en) * | 2019-11-25 | 2022-09-06 | Akselos S.A. | Methods and systems for component-based reduced order modeling for industrial-scale structural digital twins |
CN114117840B (en) * | 2021-10-28 | 2022-09-06 | 中国运载火箭技术研究院 | Structural performance prediction method based on simulation and test data hybrid drive |
CN114139414A (en) * | 2021-11-13 | 2022-03-04 | 一汽解放汽车有限公司 | Vehicle battery bracket assembly performance testing method, device, equipment and medium |
-
2022
- 2022-05-16 CN CN202210527233.1A patent/CN114626278B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110488629A (en) * | 2019-07-02 | 2019-11-22 | 北京航空航天大学 | A kind of management-control method of the hybrid vehicle based on the twin technology of number |
CN111737811A (en) * | 2020-05-09 | 2020-10-02 | 北京航空航天大学 | Helicopter movable part service life management method, device and medium based on digital twin |
CN113569445A (en) * | 2021-07-01 | 2021-10-29 | 洛阳星派数值仿真研究院有限公司 | Steel structure health monitoring system and method based on digital twinning technology |
CN114254457A (en) * | 2021-12-17 | 2022-03-29 | 中国矿业大学(北京) | Mine hydraulic support fatigue life prediction method based on digital twinning |
Also Published As
Publication number | Publication date |
---|---|
CN114626278A (en) | 2022-06-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US8862346B2 (en) | System and method for simulating the performance of a virtual vehicle | |
Al-Wreikat et al. | Driving behaviour and trip condition effects on the energy consumption of an electric vehicle under real-world driving | |
CN106662501B (en) | System and method for analyzing the efficiency of vehicle | |
CN110376003B (en) | Intelligent train service life prediction method and system based on BIM | |
CN105976074A (en) | Vehicle health parameter generation and presentation method and device | |
WO2010041382A1 (en) | Generation of reference value for vehicle failure diagnosis | |
CN106660563A (en) | System and method for analyzing the energy efficiency of a motor vehicle, in particular of an apparatus of the motor vehicle | |
DE102016210453A1 (en) | Vehicle, system in communication with a communication module of the vehicle, and system in communication with a group of vehicles | |
Goharimanesh et al. | More efficiency in fuel consumption using gearbox optimization based on Taguchi method | |
CN104182995B (en) | Highway roadside landscape color evaluation method based on driving fatigue | |
Martin et al. | Can UK passenger vehicles be designed to meet 2020 emissions targets? A novel methodology to forecast fuel consumption with uncertainty analysis | |
DE102020207537A1 (en) | State diagnostic apparatus and method of a movable system component | |
CN112579900A (en) | Method, system and equipment for recommending second-hand vehicle replacement information | |
CN104865346B (en) | Method and apparatus for verification of data acquired by vehicle tail gas detection system | |
CN114626278B (en) | Method for evaluating structural performance and monitoring health of power battery bracket | |
Sun et al. | Long short-term memory network-based emission models for conventional and new energy buses | |
Gaol et al. | Adoption of business intelligence to support cost accounting based financial systems—case study of XYZ company | |
Prakisya et al. | Utilization of K-nearest neighbor algorithm for classification of white blood cells in AML M4, M5, and M7 | |
González et al. | Fuel-efficient driving in the context of urban waste-collection: A Spanish case study | |
CN104216397B (en) | Failure recognition and detection method for intelligent drive axle system | |
CN115079663A (en) | Vehicle power system monitoring method and device based on digital twin technology | |
Liu et al. | Naturalistic drive cycle synthesis for pickup trucks | |
CN116703612A (en) | Automobile insurance risk level assessment method | |
CN116109357A (en) | Method, system and medium for calculating comprehensive scores of online comments of automobiles | |
CN116049989A (en) | Bogie digital twin model construction method, system, electronic equipment and medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |