CN115270284A - Shock absorber dynamic characteristic identification model establishing method and simulation method - Google Patents
Shock absorber dynamic characteristic identification model establishing method and simulation method Download PDFInfo
- Publication number
- CN115270284A CN115270284A CN202210561440.9A CN202210561440A CN115270284A CN 115270284 A CN115270284 A CN 115270284A CN 202210561440 A CN202210561440 A CN 202210561440A CN 115270284 A CN115270284 A CN 115270284A
- Authority
- CN
- China
- Prior art keywords
- shock absorber
- simulation
- parameters
- network
- input
- 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.)
- Pending
Links
- 239000006096 absorbing agent Substances 0.000 title claims abstract description 62
- 238000000034 method Methods 0.000 title claims abstract description 62
- 230000035939 shock Effects 0.000 title claims abstract description 62
- 238000004088 simulation Methods 0.000 title claims abstract description 33
- 238000012360 testing method Methods 0.000 claims abstract description 18
- 230000008569 process Effects 0.000 claims abstract description 17
- 238000013528 artificial neural network Methods 0.000 claims abstract description 16
- 238000003062 neural network model Methods 0.000 claims abstract description 14
- 238000012549 training Methods 0.000 claims abstract description 14
- 238000012545 processing Methods 0.000 claims abstract description 12
- 238000004364 calculation method Methods 0.000 claims abstract description 9
- 239000000725 suspension Substances 0.000 claims abstract description 9
- 238000007781 pre-processing Methods 0.000 claims abstract description 8
- 238000013461 design Methods 0.000 claims abstract description 7
- 238000004458 analytical method Methods 0.000 claims abstract description 4
- 238000013016 damping Methods 0.000 claims description 33
- 230000035945 sensitivity Effects 0.000 claims description 9
- 230000000694 effects Effects 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 6
- 230000008859 change Effects 0.000 claims description 4
- 238000004891 communication Methods 0.000 claims description 3
- 238000005538 encapsulation Methods 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000010206 sensitivity analysis Methods 0.000 claims description 3
- 239000003381 stabilizer Substances 0.000 claims description 3
- 230000002159 abnormal effect Effects 0.000 claims description 2
- 238000005094 computer simulation Methods 0.000 abstract description 4
- 238000004422 calculation algorithm Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 6
- 230000006399 behavior Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 239000003795 chemical substances by application Substances 0.000 description 2
- 238000007476 Maximum Likelihood Methods 0.000 description 1
- 238000000418 atomic force spectrum Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000012938 design process Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000011664 nicotinic acid Substances 0.000 description 1
Images
Classifications
-
- 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
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Geometry (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Pure & Applied Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Aviation & Aerospace Engineering (AREA)
- Automation & Control Theory (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Vehicle Body Suspensions (AREA)
Abstract
A shock absorber dynamic characteristic identification model establishing method and a shock absorber dynamic characteristic identification model simulation method relate to the technical field of dynamic simulation, solve the problem that dynamic characteristics of a shock absorber are difficult to describe through a traditional function mode, and can be applied to the process of developing a whole vehicle system to obtain loads of parts of a suspension system. Identification model building the method comprises the following steps: acquiring sample data; preprocessing test data; defining input and output parameters related to dynamic characteristics of the shock absorber; carrying out standardization processing on the parameters; setting an input layer, an output layer and a hidden layer structure of a neural network; distributing training and predicting samples according to a proportion, setting a training step length and a convergence error, and executing network operation; determining whether the network meets the precision requirement or not according to the precision requirement of a user; and (5) modularizing the neural network model. The simulation method also comprises the steps of establishing a complete vehicle dynamic model in dynamic simulation software according to the complete vehicle design parameters; and (3) applying load according to the simulation requirement of a user, defining analysis types and simulation time, and executing simulation calculation.
Description
Technical Field
The invention relates to the technical field of dynamics simulation, in particular to a dynamic characteristic simulation technology of a shock absorber.
Background
The shock absorber is used as a main damping element in a chassis structure, forms a basic element of an automobile suspension system together with an elastic element and a guide mechanism, and plays an important role in quickly attenuating vibration from an uneven road surface, improving the stability and smoothness of the whole automobile and improving the strength and durability of the whole automobile.
The current development process of the whole vehicle system has the characteristics of short period requirement, high reliability requirement and the like, and the whole vehicle road test cannot envelop all road working conditions. Therefore, the whole vehicle subsystem component load is obtained by simulating the whole vehicle behavior through dynamics simulation. The subsystems are often connected by elastic elements, wherein the damping characteristics of the shock absorber have a particularly significant effect on the simulation accuracy. The dynamic model construction of the shock absorber and the accurate identification of the parameters are the basis for obtaining the loads of the suspension system components. The dynamic characteristic of the shock absorber is identified after the sample car is produced, and the conditions that the risk is high after the design problem occurs, the improvement difficulty is high and the like exist. In addition, the design scheme can be evaluated through a bench test in the development process, but the matching problem with the whole vehicle cannot be considered.
The dynamic damping of the shock absorber has strong nonlinearity, the real speed-damping dynamic characteristic of the shock absorber is shown in figure 1, and the dynamic characteristic of the shock absorber is mostly described by adopting a speed-damping force curve when a traditional shock absorber dynamic model is built, so that the influence of complex and severe road conditions on the dynamic behavior of the shock absorber is difficult to consider. For nonlinear parameter identification, a learner provides a plurality of nonlinear agent models, and provides a plurality of parameter optimization algorithms, such as a least square algorithm, a gradient optimization method, a maximum likelihood estimation algorithm, a genetic algorithm and the like, for the corresponding agent models, so as to identify characteristic parameters of a nonlinear system, but when the conditions of more parameters and stronger nonlinearity are met, the algorithm is difficult to converge, and a local optimal solution can also occur, so that the design condition is misled.
In summary, the dynamic characteristics of the shock absorber are difficult to describe through a traditional function mode at present, and therefore load calculation accuracy of the whole vehicle dynamic simulation is low.
Disclosure of Invention
In order to solve the problems, the invention provides a method for identifying and simulating dynamic characteristics of a shock absorber.
The technical scheme of the invention is as follows:
a shock absorber dynamic characteristic identification model building method, the method comprising the steps of:
s1, obtaining sample data through a whole vehicle road test;
s2, preprocessing the test data;
s3, defining input and output parameters related to the dynamic characteristics of the shock absorber, wherein the output parameters are defined as damping force of the shock absorber, and the input parameters are defined as related parameters influencing damping of the shock absorber;
s4, carrying out standardization processing on the parameters obtained by screening;
s5, setting an input layer, an output layer and a hidden layer structure of the neural network;
s6, distributing training and prediction samples according to a proportion, taking 80% of a training set and 20% of a testing set, setting a training step length and a convergence error, and executing network operation;
s7, determining whether the network meets the precision requirement or not according to the precision requirement of the user, and ending the network training process if the network meets the requirement; if the network precision does not meet the requirement, modifying the network structure parameters and repeating S5 and S6 until the network precision reaches the standard;
and S8, performing modular processing on the neural network model with the optimal final performance.
Preferably, the sample data in step S1 includes wheel center load, vehicle running speed, vehicle body attitude, and shock absorber state data.
Preferably, the preprocessing in step S2 includes filling missing values of the data, removing abnormal values, and filtering frequency band data affected by noise.
Preferably, the specific definition method of the input parameters in step S3 is as follows:
to influence the damping force FdamperIs subjected to sensitivity analysis for all test parameters including the wheel center load FwhellVehicle running attitude DvehicleAnd damper state data DdamperThe effect on the damping force when the parameter is changed is expressed as:
wherein,
the sensitivity of the damping force to the influence parameters is calculated through the formula, and the parameters with high sensitivity are reserved as the input parameters of the neural network.
Preferably, the normalization process in step S4 is performed by the Z-score method, and the formula of the normalization process is as follows:
where X represents the pre-processed data, Y represents the post-processed data, μ is the mean of X, and σ is the variance of X.
Preferably, the input layer, the output layer and the hidden layer structure of the neural network in step S5 are set as follows:
inputting X (X) for parameter1,x2,…,xn) Representing, for output parameters of said output layerY(y1,y2,…,yn) The relationship between the input parameter and the output parameter is represented by Y = f (WX + b), W (W)i1,wi2,…,win) Representing the weight from the input of the upper layer to the output of the adjacent layer, b representing the threshold value for controlling the network precision, setting the number of units of the input layer and the output layer according to the parameters determined by the sensitivity, and setting the number of layers and the number of units of the hidden layer according to the empirical formula
Determining that m and n respectively represent the number of input and output layer units, and a takes an integer of 1-10 according to experience.
Preferably, the method for determining whether the network meets the accuracy requirement in step S7 specifically includes:
evaluated by the root mean square error RMSE and the correlation coefficient R2,
wherein m represents the number of samples, y represents the value of the sample,
representing the predicted value of the network,
represents the sample mean; the closer the RMSE is to 0,R2The closer to 1, the higher the network accuracy.
Preferably, the modularization processing in step S8 is specifically to adopt a neural network encapsulation module of MATLAB and Python to convert the network model into a function model capable of updating the damping force in real time according to the input parameters.
The invention also provides a dynamic characteristic simulation method of the shock absorber, which comprises the model establishing method and further comprises the following steps:
s9, establishing a complete vehicle dynamics model in dynamics simulation software according to complete vehicle design parameters;
s10, applying a load according to the simulation requirement of a user, defining an analysis type and simulation time, and executing simulation calculation;
and S11, calling input parameters in real time by the neural network model following the iteration step change in the calculation process, and updating the damping force of the shock absorber in real time to realize accurate simulation of the real driving state.
Preferably, step S9 further comprises:
s91, establishing a template of each component of the whole vehicle, wherein the template comprises a suspension, a stabilizer bar, a steering wheel, a wheel and a vehicle body;
s92, defining connection and communication relations between each component and a subsystem, and expressing variable parameters related to simulation by using variables, wherein the variables comprise wheel center load, vehicle body attitude and shock absorber state;
s93, introducing the modularized neural network model into the suspension template in a functional form to realize the deployment of the neural network model;
and S94, assembling the subsystem files generated by the templates into a whole vehicle model.
Compared with the prior art, the invention solves the problem that the dynamic characteristic of the shock absorber is difficult to describe through the traditional function mode, and has the following specific beneficial effects:
1. the method provided by the invention can accurately and quickly identify the dynamic damping characteristic of the shock absorber, considers the dynamic characteristic of the shock absorber closest to the reality in the simulation process, improves the reliability of the whole vehicle dynamic simulation calculation, shortens the development period of the shock absorber, and assists in simulating the whole vehicle road test through the whole vehicle dynamic simulation to check the durability of the parts of the whole vehicle.
2. The neural network algorithm used by the invention is an intelligent bionic algorithm, has high optimization speed and strong adaptability, has strong nonlinear processing capability, is widely applied to the fields of image recognition, stock prediction, damage diagnosis and the like, has good applicability to the nonlinear parameter recognition problem of the shock absorber, and can solve the problems that the traditional algorithm has low convergence speed and is easy to fall into a local optimal solution. Meanwhile, the algorithm has strong mobility, similar parameter identification problems are encountered in the design process, and the neural network model developed at the early stage can be used for expanding application.
Drawings
FIG. 1 is a graph of the relationship between the true speed and the dynamic damping characteristic of the shock absorber in the background art;
FIG. 2 is a schematic flow chart of a method for establishing a dynamic characteristic identification model of a shock absorber provided by the invention;
FIG. 3 is a schematic diagram of the neural network according to embodiment 6;
fig. 4 is a schematic flow chart of a method for simulating dynamic characteristics of a shock absorber provided by the invention.
Detailed Description
In order to make the technical solution of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings of the specification of the present invention, it should be noted that the following examples are only for better understanding of the technical solutions of the present invention, and should not be construed as limiting the present invention.
Example 1.
The embodiment provides a method for establishing a dynamic characteristic identification model of a shock absorber, as shown in fig. 2, the method comprises the following steps:
s1, obtaining sample data through a whole vehicle road test;
s2, preprocessing the test data;
s3, defining input and output parameters related to the dynamic characteristics of the shock absorber, wherein the output parameters are defined as damping force of the shock absorber, and the input parameters are defined as related parameters influencing damping of the shock absorber;
s4, carrying out standardization processing on the parameters obtained by screening;
s5, setting an input layer, an output layer and a hidden layer structure of the neural network;
s6, distributing training and prediction samples according to a proportion, taking 80% of a training set and 20% of a testing set, setting a training step length and a convergence error, and executing network operation;
s7, determining whether the network meets the precision requirement or not according to the precision requirement of the user, and ending the network training process if the network meets the requirement; if the network precision does not meet the requirement, modifying the network structure parameters and repeating S5 and S6 until the network precision reaches the standard;
and S8, performing modular processing on the neural network model with the optimal final performance.
Example 2.
This embodiment is a further illustration of embodiment 1, wherein the sample data in step S1 includes wheel center load, vehicle running speed, body attitude, and shock absorber status data.
In order to consider the dynamic characteristic performance of the shock absorber of the whole vehicle in the running process as much as possible, the road test needs to include smooth and complex road conditions. When the whole vehicle is simulated, the wheel center is loaded and excited to simulate real power behaviors, the dynamic characteristic of the shock absorber is closely related to the road surface characteristic, and the wheel center load data can reflect the change of the road surface characteristic, so that the collected road test data is divided into wheel center load, vehicle running speed, vehicle body posture and shock absorber state data.
Example 3.
This embodiment is a further illustration of embodiment 1, wherein the preprocessing in step S2 includes missing value padding, outlier culling, and filtering of frequency band data affected by noise.
The preprocessing can improve the data quality, so that the neural network has better feature acquisition capability.
Example 4.
This embodiment is a further illustration of embodiment 1, wherein the specific defining method of the input parameters in step S3 includes:
to influence the damping force FdamperIs subjected to sensitivity analysis for all test parameters including the wheel center load FwhellVehicle running attitude DvehicleAnd damper state data DdamperThe effect on the damping force when the parameter is changed is expressed as:
wherein,
the sensitivity of the damping force to the damping force is calculated through the formula, and the parameter with high sensitivity is reserved as the input parameter of the neural network.
Example 5.
This example is a further illustration of example 1, wherein the normalization process in step S4 is performed by using the Z-score method, and the formula of the normalization process is as follows:
where X represents the pre-processed data, Y represents the post-processed data, μ is the mean of X, and σ is the variance of X.
According to the embodiment, data with discrete distribution intervals can be converted into data which obey standard normal distribution through standardization processing, and the learning capacity of the network on data characteristics is enhanced.
Example 6.
This embodiment is a further illustration of embodiment 1, wherein the method for setting the structures of the input layer, the output layer, and the hidden layer of the neural network in step S5 includes:
as shown in FIG. 3, X (X) is used as an input parameter1,x2,…,xn) The output parameter of the output layer is represented by Y (Y)1,y2,…,yn) The relationship between the input parameter and the output parameter is represented by Y = f (WX + b), W (W)i1,wi2,…,win) Representing the weight from the input of the upper layer to the output of the adjacent layer, b representing the threshold value for controlling the network precision, setting the number of units of the input layer and the output layer according to the parameters determined by the sensitivity, and setting the number of layers and the number of units of the hidden layer according to the empirical formula
Determining that m and n respectively represent the number of input and output layer units, and a takes an integer of 1-10 according to experience.
Example 7.
This embodiment is a further example of embodiment 1, wherein the method for determining whether the network meets the accuracy requirement in step S7 specifically includes:
evaluated by the root mean square error RMSE and the correlation coefficient R2,
wherein m represents the number of samples, y represents the value of the sample,
representing the predicted value of the network,
represents the sample mean; the closer the RMSE is to 0,R2The closer to 1, the higher the network accuracy.
Example 8.
This embodiment is a further example of embodiment 1, wherein the modularization processing in step S8 is specifically to use a neural network encapsulation module of MATLAB and Python to convert a network model into a function model capable of updating the damping force in real time according to an input parameter.
In the embodiment, the network model is converted into the function model capable of updating the damping force in real time according to the input parameters, so that the effects of standardization and real-time calling are achieved.
Example 9.
The present embodiment provides a shock absorber dynamic characteristics simulation method, as shown in fig. 4, where the simulation method includes the model building method described in any one of embodiments 1 to 8, and further includes:
s9, establishing a complete vehicle dynamics model in dynamics simulation software according to complete vehicle design parameters;
s10, applying a load according to the simulation requirement of a user, defining an analysis type and simulation time, and executing simulation calculation;
and S11, calling input parameters in real time by the neural network model following the iteration step change in the calculation process, and updating the damping force of the shock absorber in real time to realize accurate simulation of the real driving state.
The method can realize the identification of the damping dynamic characteristics of the shock absorber based on the identification model established in the embodiment, and the embodiment calls the model through the vehicle dynamics simulation stage, and can update the function model of the damping force in real time according to the input parameters, so as to achieve the effects of standardization and real-time calling and realize the accurate simulation of the real driving state.
Example 10.
This embodiment is a further illustration of embodiment 9, wherein step S9 further includes:
s91, building templates of all components of the whole vehicle, including a suspension, a stabilizer bar, a steering wheel, wheels and a vehicle body;
s92, defining connection and communication relations between each component and a subsystem, and expressing variable parameters related to simulation by using variables, wherein the variables comprise wheel center load, vehicle body attitude and shock absorber state;
s93, introducing the modularized neural network model into the suspension template in a functional form to realize the deployment of the neural network model;
and S94, assembling the subsystem files generated by the templates into a whole vehicle model.
Claims (10)
1. A method for establishing a dynamic characteristic identification model of a shock absorber, comprising the steps of:
s1, obtaining sample data through a whole vehicle road test;
s2, preprocessing the test data;
s3, defining input and output parameters related to the dynamic characteristics of the shock absorber, wherein the output parameters are defined as damping force of the shock absorber, and the input parameters are defined as related parameters influencing damping of the shock absorber;
s4, carrying out standardization processing on the parameters obtained by screening;
s5, setting an input layer, an output layer and a hidden layer structure of the neural network;
s6, distributing training and prediction samples according to a proportion, taking 80% of a training set and 20% of a testing set, setting a training step length and a convergence error, and executing network operation;
s7, determining whether the network meets the precision requirement or not according to the precision requirement of the user, and ending the network training process if the network meets the requirement; if the network precision does not meet the requirement, modifying the network structure parameters and repeating S5 and S6 until the network precision reaches the standard;
and S8, performing modular processing on the neural network model with the optimal final performance.
2. The shock absorber dynamic characteristic identification model building method according to claim 1, wherein the sample data in step S1 includes wheel center load, vehicle running speed, body attitude, and shock absorber state data.
3. The method for establishing a dynamic characteristics identification model of a shock absorber according to claim 1, wherein the preprocessing in step S2 comprises filling missing values of data, removing abnormal values, and filtering frequency band data affected by noise.
4. The method for establishing the dynamic characteristic identification model of the shock absorber according to claim 1, wherein the specific definition method of the input parameters in the step S3 is as follows:
to influence the damping force FdamperIs subjected to sensitivity analysis for all test parameters including the wheel center load FwhellVehicle running attitude DvehicleAnd damper state data DdamperThe effect on the damping force when the parameter is changed is expressed as:
wherein,
the sensitivity of the damping force to the damping force is calculated through the formula, and the parameter with high sensitivity is reserved as the input parameter of the neural network.
5. The dynamic characteristics identification model building method of shock absorber according to claim 1, characterized in that said normalization process in step S4 is a Z-score method, and the formula of said normalization process is as follows:
where X represents the pre-processed data, Y represents the post-processed data, μ is the mean of X, and σ is the variance of X.
6. The method for establishing a dynamic characteristic recognition model of a shock absorber according to claim 1, wherein the input layer, the output layer and the hidden layer structure of the neural network in the step S5 are set by:
inputting X (X) for parameter1,x2,…,xn) The output parameter of the output layer is represented by Y (Y)1,y2,…,yn) The relationship between the input parameter and the output parameter is represented by Y = f (WX + b), W (W)i1,wi2,…,win) Representing the weight from the input of the upper layer to the output of the adjacent layer, b representing the threshold value for controlling the network precision, setting the number of units of the input layer and the output layer according to the parameters determined by the sensitivity, and setting the number of layers and the number of units of the hidden layer according to the empirical formula
Determining that m and n represent input and output layers, respectivelyThe number of units, a, is empirically an integer from 1 to 10.
7. The method for establishing the dynamic characteristic identification model of the shock absorber according to claim 1, wherein the method for determining whether the network meets the accuracy requirement in step S7 specifically comprises:
evaluated by the root mean square error RMSE and the correlation coefficient R2,
wherein m represents the number of samples, y represents the value of the sample,
representing the predicted value of the network,
represents the sample mean; the closer the RMSE is to 0,R2The closer to 1, the higher the network accuracy.
8. The method for establishing the dynamic characteristic identification model of the shock absorber according to claim 1, wherein the modularization processing in the step S8 is specifically to adopt a neural network encapsulation module of MATLAB and Python to convert the network model into a function model capable of updating the damping force in real time according to the input parameters.
9. A method for simulating dynamic characteristics of a shock absorber, the method comprising the method for modeling according to any one of claims 1 to 8, and further comprising:
s9, establishing a complete vehicle dynamics model in dynamics simulation software according to complete vehicle design parameters;
s10, applying a load according to the simulation requirement of a user, defining an analysis type and simulation time, and executing simulation calculation;
and S11, calling input parameters in real time by the neural network model following the iteration step change in the calculation process, and updating the damping force of the shock absorber in real time to realize accurate simulation of the real driving state.
10. The dynamic characteristics simulation method of shock absorber according to claim 9, wherein step S9 further comprises:
s91, establishing a template of each component of the whole vehicle, wherein the template comprises a suspension, a stabilizer bar, a steering wheel, a wheel and a vehicle body;
s92, defining connection and communication relations between each component and a subsystem, and expressing variable parameters related to simulation by using variables, wherein the variables comprise wheel center load, vehicle body attitude and shock absorber state;
s93, introducing the modularized neural network model into the suspension template in a functional form to realize the deployment of the neural network model;
and S94, assembling the subsystem files generated by the templates into a whole vehicle model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210561440.9A CN115270284A (en) | 2022-05-23 | 2022-05-23 | Shock absorber dynamic characteristic identification model establishing method and simulation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210561440.9A CN115270284A (en) | 2022-05-23 | 2022-05-23 | Shock absorber dynamic characteristic identification model establishing method and simulation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115270284A true CN115270284A (en) | 2022-11-01 |
Family
ID=83759911
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210561440.9A Pending CN115270284A (en) | 2022-05-23 | 2022-05-23 | Shock absorber dynamic characteristic identification model establishing method and simulation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115270284A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116839783A (en) * | 2023-09-01 | 2023-10-03 | 华东交通大学 | Method for measuring stress value and deformation of automobile leaf spring based on machine learning |
-
2022
- 2022-05-23 CN CN202210561440.9A patent/CN115270284A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116839783A (en) * | 2023-09-01 | 2023-10-03 | 华东交通大学 | Method for measuring stress value and deformation of automobile leaf spring based on machine learning |
| CN116839783B (en) * | 2023-09-01 | 2023-12-08 | 华东交通大学 | Method for measuring stress value and deformation of automobile leaf spring based on machine learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Rauh | Virtual development of ride and handling characteristics for advanced passenger cars | |
| CN112287551B (en) | Driving performance system level index decomposition method based on whole vehicle conceptual model | |
| CN107480395B (en) | Method and system for constructing vehicle steering knuckle load spectrum prediction model | |
| CN108168577A (en) | MEMS gyro random error compensation method based on BP neural network | |
| CN116090111B (en) | Automobile leaf spring fatigue life prediction method based on deep learning model | |
| CN115270284A (en) | Shock absorber dynamic characteristic identification model establishing method and simulation method | |
| CN116049989A (en) | Bogie digital twin model construction method, system, electronic equipment and medium | |
| CN108228994B (en) | Stress calculation method for vehicle and working device under random excitation of off-road | |
| CN115169167A (en) | Method and system for optimizing and matching motion stroke parameters of automobile plate spring | |
| CN113704872A (en) | Method for improving fatigue simulation analysis efficiency of automobile plate spring | |
| CN116415494A (en) | Road noise optimization method, system and storage medium | |
| Albeaik et al. | Deep Truck: A deep neural network model for longitudinal dynamics of heavy duty trucks | |
| KR100305835B1 (en) | Method for durability evaluation analysis of body | |
| CN115577615A (en) | Automobile multi-body dynamics modeling method and system | |
| CN114676648A (en) | Vehicle load spectrum prediction method and device based on machine learning | |
| CN116702628A (en) | Automobile wheel center vertical displacement solving method based on convolution depth neural network | |
| Shintani et al. | Surrogate modeling of waveform response using singular value decomposition and Bayesian optimization | |
| CN113785333A (en) | Method for determining a load prediction of a component of a motor vehicle | |
| Khandavalli et al. | Potentials of modern active suspension control strategies–from model predictive control to deep learning approaches | |
| Seeger | Neural Network-Based Wheel Suspension Modeling | |
| CN114861323A (en) | Vehicle running deviation analysis method, device and equipment and readable storage medium | |
| Rezayat et al. | Influence of internal loads on the accuracy of durability tests of a vehicle on a test rig [C] | |
| CN117725668A (en) | Data-driven vehicle suspension K & C system and construction method | |
| Van der Auweraer et al. | Hybrid testing in aerospace and ground vehicle development | |
| Chen et al. | Ride Comfort Optimization Method for Commercial Vehicle Based on Nonlinear Damping and PSO |
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 |