CN113591230B - Multi-objective optimization method for commercial vehicle cab based on beam section - Google Patents
Multi-objective optimization method for commercial vehicle cab based on beam section Download PDFInfo
- Publication number
- CN113591230B CN113591230B CN202111069225.9A CN202111069225A CN113591230B CN 113591230 B CN113591230 B CN 113591230B CN 202111069225 A CN202111069225 A CN 202111069225A CN 113591230 B CN113591230 B CN 113591230B
- Authority
- CN
- China
- Prior art keywords
- section
- sensitivity
- cross
- cab
- model
- 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]
- G06F30/10—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]
- G06F30/20—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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
-
- 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
- G06F2119/02—Reliability analysis or reliability optimisation; Failure analysis, e.g. worst case scenario performance, failure mode and effects analysis [FMEA]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Geometry (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Automation & Control Theory (AREA)
- Aviation & Aerospace Engineering (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Body Structure For Vehicles (AREA)
Abstract
The invention discloses a multi-objective optimization method for a commercial vehicle cab based on a beam section, which utilizes SFE-Concept software to establish an implicit parameterization model of the commercial vehicle cab, and compared with a traditional finite element model, the model can quickly complete modification of any scheme and still maintain good topological relation after modification, and can also quickly generate high-quality grids meeting the topological relation after adjustment. Aiming at the problem of overlong running time of the test design, the method optimizes the whole flow, proposes to combine hyperMorph to perform sensitivity analysis and relative sensitivity analysis, screens out key variables before the test design, greatly reduces the running time of the test design and improves the optimization efficiency. Aiming at the problem of lower fitting precision of a common approximate model, a response surface-radial basis mixed approximate model is constructed, the fitting precision is improved, the correlation coefficient values of five indexes are all above 0.9, and the precision requirement is met.
Description
Technical Field
The invention relates to the technical field of commercial vehicle cab performance optimization, in particular to a multi-objective commercial vehicle cab optimization method based on a beam section.
Background
Various performance indexes of the automobile body not only influence riding experience of a user, but also directly relate to life safety. Before mass production, each automobile is subjected to performance tests such as NVH performance, collision safety, steering stability, structural durability and the like, and the next stage can be performed after the requirements are met. The rigidity of the automobile body greatly influences various performances of the automobile, and the insufficient rigidity of the automobile body can cause the lower mode of the whole automobile, resonance and the like. Proved by research, the method comprises the following steps: the beam section shape is one of main factors influencing the rigidity of the vehicle body, and the main effect relation between the beam section shape and the rigidity of the vehicle body is researched, so that engineers can be helped to design the vehicle which meets the market demands more.
The traditional beam section optimization method mainly has three defects: related optimization is carried out based on a finite element model, and the optimization efficiency is low; the test design has long running time; the fitting precision of the approximate model is low. (1) performing correlation optimization based on a finite element model: this approach has two distinct disadvantages: 1. the upper and lower limits of the variables vary over a smaller range. When the variable of the cross section shape of the beam is recorded, only the deformation body in the finite element model can be finely adjusted, if the adjustment amplitude is too large, the problems of grid deformity, poor grid quality and the like can occur, and the situation of calculation error reporting can often occur; 2. the optimal solution may be a better solution than an optimal solution. The variation range of the upper and lower limits of the variables is too small, which means that the optimizing space is small, the possibility of finding the optimal solution is reduced, the solved optimal solution is probably a local optimal solution rather than a global optimal solution, and the traditional key section optimization method has the defects of more human factors and large error. (2) the test design run time is too long: the beam section optimization design based on the implicit parameterization model is generally performed by sensitivity analysis and relative sensitivity analysis in a test design mode, and then key optimization variables are selected according to the contribution rate. If the optimization variables are not screened in advance, the key optimization variables are selected directly by using a test design mode, all possible design variables need to be recorded through SFE-accept software, the number of recorded variables is large, the more sample points are, and the longer the running time is. And (3) the fitting precision of the approximate model is low: in the fitting process of the approximate model, the mode of the same vibration mode may occur in different orders when the variable changes, and the mode frequency of a certain order can be restricted when the variable is restricted, so that the problem of poor fitting precision of the approximate model is caused.
In addition, researches based on beam sections are mostly carried out for passenger cars, and related researches aim at light weight, so that the optimization effect and the optimization efficiency are required to be further improved.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provides a multi-objective optimization method for a commercial vehicle cab based on a beam section.
The technical scheme for realizing the aim of the invention is as follows:
a multi-objective optimization method for a commercial vehicle cab based on a beam section is shown in fig. 1, and comprises the following steps:
1) Establishing an implicit parameterization model of a commercial vehicle cab by using SFE-accept software: firstly, importing the existing finite element model into SFE-accept software, and dividing the whole cab into five assemblies, namely a top cover, a floor, a side wall, a rear wall and a front wall; secondly, different base points are established according to the shapes and positions of the parts, base lines with different curvatures are generated by the base points, different base planes are established, and the shapes and positions of any part are controlled through three basic elements; for a cab with a bilateral symmetry structure, firstly, establishing a left-side cab hidden parameterization model, and generating a right-side cab hidden parameterization model by using mirror images, namely, completing the establishment of the whole cab hidden parameterization model;
2) The hidden parameterized model of the commercial vehicle cab is exported from SFE-accept software and stored as a commercial cab finite element model, and the file format is bdf format, so that the hidden parameterized model is conveniently imported into Hypermesh to establish working conditions for basic performance analysis;
3) The method comprises the steps of utilizing an OptiStruct solver in Hypermesh to carry out mass, torsional rigidity, bending rigidity, first-order torsional mode and first-order bending mode basic performance analysis on a derived commercial cab finite element model to obtain a basic performance analysis value, wherein the minimum mass, the maximum torsional rigidity and the maximum bending rigidity are used as optimization targets, and the first-order torsional mode frequency value and the first-order bending mode frequency value are used as constraint conditions;
4) Carrying out mechanical characteristic analysis on a beam section of a cab of the commercial vehicle, determining the relation between mechanical characteristics and the beam section shape, taking the beam section shape as a variable, and controlling the shape of the beam section by adopting a proportional vector method;
5) Beam section sensitivity analysis was performed using HyperMorph: setting a plurality of deformation bodies on the beam section of the cab by utilizing a shape function in hyperMorph, changing the shape of the beam section by adjusting the deformation bodies, finishing the recording of the beam section variable, and calculating the sensitivity value of the target to the beam section variable by sensitivity analysis; when sensitivity analysis is carried out on the beam section variables of different targets, the variables of the different targets are balanced by adopting relative sensitivity analysis;
6) Screening out the variable of the cross section shape of the beam conforming to the problem according to the results of sensitivity analysis and relative sensitivity analysis, and carrying out DOE (Design of Experiment test design) analysis by adopting an optimal Latin hypercube method by using the screened variable to obtain a group of sample points;
7) And (3) properly selecting the sample points obtained in the step (6), and constructing a response surface-radial basis mixed approximation model, wherein a fitting formula of the model is as follows:
in the formula (1), y (x) is a response actual value;
for response approximations; epsilon is the random error between the actual value of the response and the approximation of the response, where: the response surface-radial basis mixed approximation model achieves the accuracyAfter solving, replacing the original model, and adopting a correlation coefficient R for precision requirement 2 Evaluation, R 2 The calculation formula is as follows:
in the above formula, the number of n sample points, y i For the simulation value, y si In order to be able to predict the value,
is y i Is the average value of (2); as shown in fig. 2, the flow of the response surface-radial basis hybrid approximation model is:
7-1) analyzing and collecting required sample points through experimental design;
7-2) selecting an approximate model fitting method conforming to the characteristics of the optimization problem, and constructing an approximate model by using the collected sample points;
7-3) verifying the fitting precision of the approximate model, and substituting the original model to participate in optimization after meeting the requirement;
7-4) if the accuracy of the approximate model is too low, the accuracy requirement cannot be met, and further increasing the number of sample points or replacing a more applicable approximate model fitting method to improve the fitting accuracy until the engineering requirement is met.
Response surface method: fitting a design space by using a polynomial function, wherein a fitting formula is as follows:
the response surface method fitting approximation model can reduce sample points as much as possible, and approximates the objective function to the maximum extent; the mathematical model formed by fitting is simple, has stronger robustness and stronger practicability. However, because the number of the sample points is small, all indexes cannot be considered in the process of constructing an approximate model, and the defects are overcome by adopting a radial basis method, so that a radial basis-response surface mixed approximate model is constructed.
Radial basis method: by the point to be measured and the sampleThe Euclidean distance between the points is an argument, i.e. it is assumed that
Representing a set of input vectors, ">
Is a basis function, wherein x-x j， I is the euclidean distance: (x-x) j ) T (x-x j ) C is more than or equal to 0.2 and less than or equal to 3;
8) Taking the minimum mass, the maximum torsional rigidity and the maximum bending rigidity as optimization targets, taking a first-order torsional mode frequency value and a first-order bending mode frequency value not lower than an initial calculated value as constraint conditions, and adopting an NSGA-II (Nondominated Sorting Genetic Algorithm II) algorithm to build a response surface-radial basis mixed approximate model in the step 7) to perform multi-target optimization so as to obtain an optimal solution.
As shown in fig. 3, the optimization steps of the NSGAII algorithm are:
8-1) setting basic parameters of an algorithm, including population scale, cross variation probability and iteration times;
8-2) generating an initialization population P 1 ；
8-3) selecting parent population individuals, generating a child population through crossover and mutation operations, and calculating the fitness value of the child population individuals;
8-4) combining the parent population and the offspring population to form a new population, and carrying out rapid non-dominant sorting on individuals of the new population;
8-5) calculating the individual crowding degree distance of the new population, screening out individuals with high fitness, and entering the next generation P t+1 ；
8-6) judging a termination condition, if the termination condition is met, terminating the conditional algorithm, otherwise adding 1 to the iteration times, and turning to the step 8-3).
In the step 3), the mass, torsional rigidity, bending rigidity, first-order torsional mode and first-order bending mode basic performance analysis is specifically as follows:
when analyzing quality, the one-key analysis is realized by utilizing a built-in quality calculation function in the Hypermesh software; when analyzing the low-order modal frequency, adding an Eigra card into the Hypermesh, setting a modal analysis frequency value range and a loading step, and directly obtaining a first-order torsional modal frequency value and a first-order bending modal frequency value by using an Optigstruct solver; when the torsional rigidity is analyzed, the left and right rear suspensions of the cab are restrained, and two forces with the same magnitude and opposite directions are loaded on the left and right front suspensions, wherein the calculation formula of the torsional rigidity is as follows:
in the above formula, T is the applied torque; d (D) 1 、D 2 The displacement of the left loading point and the right loading point in the gravity direction respectively; theta is D 2 And D 1 The displacement difference value is the arc tangent value in the horizontal direction, and L is the distance between the left loading point and the right loading point;
when the bending rigidity is analyzed, the cab is restrained from left, right, front and back suspension, the loading force of the left and right seats and the sleeper is restrained, and the bending rigidity calculation formula is as follows:
in the above formula, F Front seat Total loading force for the front seat; f (F) Sleeping berth The sleeper is provided with a total loading force; d (D) 1 、D 2 Is the maximum displacement of the bottom left and right longitudinal beams in the gravity direction.
In the step 4), the mechanical characteristics comprise a cross section area, a cross section moment of inertia, a cavity sealing area and a torsion constant, wherein the cross section is formed by cutting a certain part of a white car body in a vertical direction, and the cut surface is a cross section; the section area refers to the amount of materials used for the section; the section moment of inertia reflects the bending resistance of the section and has positive correlation with the bending resistance of the section; the area of the sealing cavity is the size of the space dimension of the section; the torsion constant reflects the torsion resistance of the reaction section and has positive correlation with the torsion resistance; the calculation formula is as follows:
wherein I is y 、I z Representing moment of inertia in mm 4 The method comprises the steps of carrying out a first treatment on the surface of the A is the cross-sectional area of the part in mm 2 ；
Wherein I is t Is a torsion constant; c is the area surrounded by the outline; t is the thickness of the thin-wall rod; s is the circumference of the center line of the cross section, and the cross section shape is closely related to the mechanical property of the cross section according to the calculation principle and the calculation result, so that the cross section shape is selected as a design variable;
the proportional vector method is to control the section change through an angle value and a deformation amount, determine the change direction of a control point firstly, and then realize the section change through the deformation measurement value; the proportional vector method includes two variables of a rotation angle θ and a deformation measurement value SV, assuming that the coordinate of a certain point B in a yoz coordinate system is (y, z), the coordinate system is rotated by a certain angle θ, and in a new coordinate system y 'oz', the new coordinate of the point B is (y ', z'), and the calculation formula is as follows:
after the rotation angle is determined, the value of the deformation measurement value SV is determined, after deformation occurs, the coordinate (y ', z') of the deformed B point is further obtained, and the calculation formula is as follows:
bringing equation (9) into equation (10) yields the following equation (11):
according to the calculation formula, when θ=0°, the base point coordinates change along with the Y direction; when θ=90°, the base point coordinates change with the Z direction; the commercial vehicle cab beam section optimization is studied when θ=0°.
In the step 5), the sensitivity S of the performance parameter of the vehicle body structure to the section shape variable of the part is:
in the formula (12), S is sensitivity;
for the variable x as an objective function f (x) i Inverse, x i Is the cross-sectional shape of the ith component; mass sensitivity S of mass to be analyzed to cross-sectional shape for optimizing cross-section of cab beam of commercial vehicle M Torsional stiffness sensitivity S of torsional stiffness to cross-sectional shape T First order torsional mode sensitivity S of first order torsional mode to cross-sectional shape TM Bending stiffness sensitivity S of bending stiffness to cross-sectional shape B First-order bending mode sensitivity S of first-order bending mode to cross-sectional shape BM A total of 5 different objective functions; />
The relative sensitivity analysis is the ratio of other performance sensitivity values to mass sensitivity values, wherein:
the mass relative sensitivity formula is:
m is mass, x i Is the cross-sectional shape of the ith component;
the relative sensitivity of the first order torsional mode is:
r tm ＝S TM /S M (14)
the relative sensitivity of the first order bending mode is:
r bm ＝S BM /S M (15)
the relative sensitivity of torsional stiffness is:
r t ＝S T /S M (16)
the relative sensitivity of bending stiffness is:
r b ＝S B /S M (17)
in step 6), the screening is to obtain the quality sensitivity S M The first 20 cross-section shape variables are taken as design variables and are arranged in descending order according to the sensitivity values, and the first-order torsion mode sensitivity S is respectively used TM First-order bending mode sensitivity S BM Sensitivity to torsional stiffness S T Bending stiffness sensitivity S B And mass sensitivity S M In comparison, the first 20 section shape variables are respectively taken as design variables according to the ascending order of the relative sensitivity values, and a proper amount of section shape variables are finally selected as optimization variables through relative sensitivity analysis.
The optimal Latin super-standing method is to make each dimension coordinate interval in n dimension space
Dividing into m sections uniformly, each section being marked as +.>
And randomly selecting m points, and ensuring that each level of a factor is studied only once, namely forming an n-dimensional space, and recording an Latin hypercube design with the sample number of m as m x n LHD.
The invention provides a multi-objective optimization method for a commercial vehicle cab based on a beam section, which mainly has the following three advantages:
(1) And the hidden parameterization model of the cab of the commercial vehicle is established by using SFE-accept software, and compared with the traditional finite element model, the hidden parameterization model of the cab of the commercial vehicle can quickly complete the modification of any scheme, still maintains good topological relation after the modification, and can also quickly generate high-quality grids meeting the topological relation after adjustment.
(2) Aiming at the problem of overlong running time of the test design, the method optimizes the whole flow, proposes to combine hyperMorph to perform sensitivity analysis and relative sensitivity analysis, screens out key variables before the test design, greatly reduces the running time of the test design and improves the optimization efficiency.
(3) Aiming at the problem of lower fitting precision of a common approximate model, a response surface-radial basis mixed approximate model is constructed, the fitting precision is improved, the correlation coefficient values of five indexes are all above 0.9, and the precision requirement is met.
Drawings
FIG. 1 is a cross-sectional optimized block diagram of a commercial vehicle cab beam;
FIG. 2 is a block diagram of an approximation model fit;
FIG. 3 is an NSGAII algorithm optimization block diagram;
FIG. 4 is a schematic diagram of an implicit parameterization model of a commercial vehicle cab;
FIG. 5 is a schematic diagram of a finite element model of a commercial vehicle cab;
FIG. 6 is a schematic diagram of a proportional vector method;
FIG. 7 is a graph of partial sensitivity analysis results.
Detailed Description
The present invention will now be further illustrated with reference to the drawings and examples, but is not limited thereto.
Examples:
by adopting the technical scheme, the cab of a certain commercial vehicle is optimized, an implicit parameterized model is built by using SFE-accept software as shown in fig. 4, a corresponding finite element model is derived as shown in fig. 5, and the basis performance analysis of mass, torsional rigidity, bending rigidity, first-order torsional mode, first-order bending mode and the like is completed by using an Optics solving device, and the basis performance values are shown in table 1.
Table 1 basic properties
The mechanical properties of the cross section are analyzed, the cross section shape which is a key control factor of the cross section is found, the cross section shape is controlled by using a proportional vector method, and a flow chart of the proportional vector method is shown in figure 6. And the sensitivity analysis and the relative sensitivity analysis are carried out by combining hyperMorph, key optimization variables are screened out before experimental design, the running time of the experimental design is shortened, and the partial sensitivity analysis results are shown in figure 7. Shortening the test design running time: advanced sensitivity analysis in combination with HyperMorph can shorten the run time of the test design. The total of the section variables of the cab of a certain commercial vehicle is 63, the optimal Latin hypercube test design method is used for selecting 64 sample points at least, and the total of the test design operation time is 64 x 15/60 h=16 h by taking a computer with an operation memory of 16G as an example. And 33 cross-sectional shape variables are selected by combining hyperMorph, a minimum of 34 sample points are needed, the running time of the test design is 34 times 15/60=8.5 hours, and the time is shortened by 44% by adding 9 hours of HyperMesh sensitivity analysis time of 0.5 hours. And (3) performing test design by using the screened variables in an optimal Latin hypercube mode, and collecting a group of sample points. And selecting a proper amount of sample points to construct a response surface-radial basis mixed approximation model so as to improve fitting accuracy, wherein the fitting accuracy is shown in table 2.
Table 2 fitting accuracy
And finally, optimizing the mixed approximation model by using NSGAII, and comparing the performance after optimizing. The optimization selects 33 cross-sectional shapes as design variables T n (n=1, 2, …, 33) with the constraint that the first order torsional mode is 21Hz or more and the first order bending mode is 39.5Hz or more, with minimum mass, maximum torsional stiffness, maximum bending stiffnessIs the object of optimization. Optimization was performed using NSGAII. Mathematical description of optimization problem:
Variable：T＝(T1，T2，…，T33)
Objective：{M min (T)，F Tmax (T)，F Bmax (T)}
s.t.
f TMmin ≥21；f BMmin ≥39.5 (18)
wherein: m is M min The mass of the cab of the commercial vehicle is minimum, and the unit is kg; f (F) Tmax Is the maximum torsional stiffness in Nm/°; f (F) Bmax For the maximum bending stiffness, N/mm. f (f) TMmin The unit Hz is the minimum value of the first-order torsional mode; f (f) BMmin Is the minimum value of the first-order bending mode, and is expressed in Hz.
Through multi-objective optimization, the optimization result is shown in table 3, the quality of the cab is reduced by 16.5kg under the condition that the first-order torsional mode is not lower than 21Hz and the first-order bending mode is not lower than 39.5Hz, the bending stiffness is improved by 15.3%, the torsional stiffness is improved by 7.7%, the first-order torsional mode is improved by 0.5%, the first-order bending mode is reduced by 0.5%, the frequency value change of the low-order bending mode is small and can be ignored, the multi-objective optimization effect is good, and the multi-objective optimization method has a certain guiding significance for engineering practice.
Table 3 comparison of performance values before and after optimization
Claims (6)
1. The multi-objective optimization method for the commercial vehicle cab based on the beam section is characterized by comprising the following steps of:
1) Establishing an implicit parameterization model of a commercial vehicle cab by using SFE-accept software: firstly, importing the existing finite element model into SFE-accept software, and dividing the whole cab into five assemblies, namely a top cover, a floor, a side wall, a rear wall and a front wall; secondly, different base points are established according to the shapes and positions of the parts, base lines with different curvatures are generated by the base points, different base planes are established, and the shapes and positions of any part are controlled through three basic elements; for a cab with a bilateral symmetry structure, firstly, establishing a left-side cab hidden parameterization model, and generating a right-side cab hidden parameterization model by using mirror images, namely, completing the establishment of the whole cab hidden parameterization model;
2) The hidden parameterized model of the commercial vehicle cab is exported from SFE-accept software and stored as a commercial cab finite element model, and the file format is bdf format, so that the hidden parameterized model is conveniently imported into Hypermesh to establish working conditions for basic performance analysis;
3) The method comprises the steps of utilizing an OptiStruct solver in Hypermesh to carry out mass, torsional rigidity, bending rigidity, first-order torsional mode and first-order bending mode basic performance analysis on a derived commercial cab finite element model to obtain a basic performance analysis value, wherein the minimum mass, the maximum torsional rigidity and the maximum bending rigidity are used as optimization targets, and the first-order torsional mode frequency value and the first-order bending mode frequency value are used as constraint conditions;
4) Carrying out mechanical characteristic analysis on a beam section of a cab of the commercial vehicle, determining the relation between mechanical characteristics and the beam section shape, taking the beam section shape as a variable, and controlling the shape of the beam section by adopting a proportional vector method;
5) Beam section sensitivity analysis was performed using HyperMorph: setting a plurality of deformation bodies on the beam section of the cab by utilizing a shape function in hyperMorph, changing the shape of the beam section by adjusting the deformation bodies, finishing the recording of the beam section variable, and calculating the sensitivity value of the target to the beam section variable by sensitivity analysis; when sensitivity analysis is carried out on the beam section variables of different targets, the variables of the different targets are balanced by adopting relative sensitivity analysis;
6) Screening out the variable of the cross section shape of the beam conforming to the problem according to the results of sensitivity analysis and relative sensitivity analysis, and carrying out DOE analysis by adopting an optimal Latin hypercube method by using the screened variable to obtain a group of sample points;
7) And (3) properly selecting the sample points obtained in the step (6), and constructing a response surface-radial basis mixed approximation model, wherein a fitting formula of the model is as follows:
in the formula (1), y (x) is a response actual value;
for response approximations; epsilon is the random error between the actual value of the response and the approximation of the response, where: the response surface-radial basis mixed approximation model replaces an original model after reaching the precision requirement, and the precision requirement adopts a correlation coefficient R 2 Evaluation, R 2 The calculation formula is as follows:
in the above formula, the number of n sample points, y i For the simulation value, y si In order to be able to predict the value,
is y i Is the average value of (2); the procedure of the response surface-radial basis mixed approximation model is as follows:
7-1) analyzing and collecting required sample points through experimental design;
7-2) selecting an approximate model fitting method conforming to the characteristics of the optimization problem, and constructing an approximate model by using the collected sample points;
7-3) verifying the fitting precision of the approximate model, and substituting the original model to participate in optimization after meeting the requirement;
7-4) if the accuracy of the approximate model is too low, the accuracy requirement cannot be met, and further increasing the number of sample points or replacing a more applicable approximate model fitting method to improve the fitting accuracy until the engineering requirement is met;
response surface method: fitting a design space by using a polynomial function, wherein a fitting formula is as follows:
radial basis method: taking Euclidean distance between the point to be measured and the sample point as an independent variable, namely supposing
Representing a set of input vectors, ">
Is a basis function, wherein x-x j And, || is the euclidean distance: (x-x) j ) T (x-x j ) C is more than or equal to 0.2 and less than or equal to 3;
8) Taking the minimum mass, the maximum torsional rigidity and the maximum bending rigidity as optimization targets, taking a first-order torsional mode frequency value and a first-order bending mode frequency value not lower than an initial calculated value as constraint conditions, and adopting an NSGA-II algorithm to perform multi-target optimization on the response surface-radial basis mixed approximate model established in the step 7), so as to obtain an optimal solution;
the optimization steps of the NSGAII algorithm are as follows:
8-1) setting basic parameters of an algorithm, including population scale, cross variation probability and iteration times;
8-2) generating an initialization population P 1 ；
8-3) selecting parent population individuals, generating a child population through crossover and mutation operations, and calculating the fitness value of the child population individuals;
8-4) combining the parent population and the offspring population to form a new population, and carrying out rapid non-dominant sorting on individuals of the new population;
8-5) calculating the individual crowding degree distance of the new population, screening out individuals with high fitness, and entering the next generation P t+1 ；
8-6) judging a termination condition, if the termination condition is met, terminating the conditional algorithm, otherwise adding 1 to the iteration times, and turning to the step 8-3).
2. The multi-objective optimization method for the cab of the commercial vehicle based on the beam section according to claim 1, wherein in the step 3), the mass, torsional rigidity, bending rigidity, first-order torsional mode and first-order bending mode basic performance analysis is specifically as follows:
when analyzing quality, the one-key analysis is realized by utilizing a built-in quality calculation function in the Hypermesh software; when analyzing the low-order modal frequency, adding an Eigra card into the Hypermesh, setting a modal analysis frequency value range and a loading step, and directly obtaining a first-order torsional modal frequency value and a first-order bending modal frequency value by using an Optigstruct solver; when the torsional rigidity is analyzed, the left and right rear suspensions of the cab are restrained, and two forces with the same magnitude and opposite directions are loaded on the left and right front suspensions, wherein the calculation formula of the torsional rigidity is as follows:
in the above formula, T is the applied torque; d (D) 1 、D 2 The displacement of the left loading point and the right loading point in the gravity direction respectively; theta is D 2 And D 1 The displacement difference value is the arc tangent value in the horizontal direction, and L is the distance between the left loading point and the right loading point;
when the bending rigidity is analyzed, the cab is restrained from left, right, front and back suspension, the loading force of the left and right seats and the sleeper is restrained, and the bending rigidity calculation formula is as follows:
in the above formula, F Front seat Total loading force for the front seat; f (F) Sleeping berth The sleeper is provided with a total loading force; d (D) 1 、D 2 Is the maximum displacement of the bottom left and right longitudinal beams in the gravity direction.
3. The method for optimizing multiple targets in a commercial vehicle cab based on beam cross section according to claim 1, wherein in step 4), the mechanical properties include cross section area, cross section moment of inertia, cavity sealing area, torsion constant, the cross section is obtained by cutting a certain part of a white vehicle body in a vertical direction, and the cut surface is a cross section; the section area refers to the amount of materials used for the section; the section moment of inertia reflects the bending resistance of the section and has positive correlation with the bending resistance of the section; the area of the sealing cavity is the size of the space dimension of the section; the torsion constant reflects the torsion resistance of the reaction section and has positive correlation with the torsion resistance; the calculation formula is as follows:
wherein I is y 、I z Representing moment of inertia in mm 4 The method comprises the steps of carrying out a first treatment on the surface of the A is the cross-sectional area of the part in mm 2 ；
Wherein I is t Is a torsion constant; c is the area surrounded by the outline; t is the thickness of the thin-wall rod; s is the circumference of the center line of the cross section, and the cross section shape is closely related to the mechanical property of the cross section according to the calculation principle and the calculation result, so that the cross section shape is selected as a design variable;
the proportional vector method is to control the section change through an angle value and a deformation amount, determine the change direction of a control point firstly, and then realize the section change through the deformation measurement value; the proportional vector method includes two variables of a rotation angle θ and a deformation measurement value SV, assuming that the coordinate of a certain point B in a yoz coordinate system is (y, z), the coordinate system is rotated by a certain angle θ, and in a new coordinate system y 'oz', the new coordinate of the point B is (y ', z'), and the calculation formula is as follows:
after the rotation angle is determined, the value of the deformation measurement value SV is determined, after deformation occurs, the coordinate (y ', z') of the deformed B point is further obtained, and the calculation formula is as follows:
bringing equation (9) into equation (10) yields the following equation (11):
according to the calculation formula, when θ=0°, the base point coordinates change along with the Y direction; θ=90°, the base point coordinates vary with the Z direction.
4. The multi-objective optimization method for a commercial vehicle cab based on a beam section according to claim 1, wherein in the step 5), the sensitivity of the sensitivity analysis, the sensitivity S of the vehicle body structural performance parameter to the cross-sectional shape variable of the component is:
in the formula (12), S is sensitivity;
for the variable x as an objective function f (x) i Inverse, x i Is the cross-sectional shape of the ith component; mass sensitivity S of mass to be analyzed to cross-sectional shape for optimizing cross-section of cab beam of commercial vehicle M Torsional stiffness sensitivity S of torsional stiffness to cross-sectional shape T First order torsional mode sensitivity S of first order torsional mode to cross-sectional shape TM Bending stiffness sensitivity S of bending stiffness to cross-sectional shape B First-order bending mode sensitivity S of first-order bending mode to cross-sectional shape BM A total of 5 different objective functions;
the relative sensitivity analysis is the ratio of other performance sensitivity values to mass sensitivity values, wherein:
the mass relative sensitivity formula is:
m is mass, x i Is the cross-sectional shape of the ith component;
the relative sensitivity of the first order torsional mode is:
r tm ＝S TM /S M (14)
the relative sensitivity of the first order bending mode is:
r bm ＝S BM /S M (15)
the relative sensitivity of torsional stiffness is:
r t ＝S T /S M (16)
the relative sensitivity of bending stiffness is:
r b ＝S B /S M (17)。
5. the method for optimizing multiple targets in a commercial vehicle cab based on beam cross-section according to claim 1, wherein in step 6), the screening is to obtain the mass sensitivity S M The first 20 cross-section shape variables are taken as design variables and are arranged in descending order according to the sensitivity values, and the first-order torsion mode sensitivity S is respectively used TM First-order bending mode sensitivity S BM Sensitivity to torsional stiffness S T Bending stiffness sensitivity S B And mass sensitivity S M In comparison, the first 20 section shape variables are respectively taken as design variables according to the ascending order of the relative sensitivity values, and a proper amount of section shape variables are finally selected as optimization variables through relative sensitivity analysis.
6. The multi-objective optimization method for the cab of a commercial vehicle based on a beam section according to claim 1, which is characterized in thatCharacterized in that the optimal Latin super-vertical method is to divide each one-dimensional coordinate interval in n-dimensional space
Dividing into m sections uniformly, each section being marked as +.>
And randomly selecting m points, and ensuring that each level of a factor is studied only once, namely forming an n-dimensional space, and recording an Latin hypercube design with the sample number of m as m x n LHD. />
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111069225.9A CN113591230B (en) | 2021-09-13 | 2021-09-13 | Multi-objective optimization method for commercial vehicle cab based on beam section |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111069225.9A CN113591230B (en) | 2021-09-13 | 2021-09-13 | Multi-objective optimization method for commercial vehicle cab based on beam section |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113591230A CN113591230A (en) | 2021-11-02 |
| CN113591230B true CN113591230B (en) | 2023-05-23 |
Family
ID=78241901
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111069225.9A Active CN113591230B (en) | 2021-09-13 | 2021-09-13 | Multi-objective optimization method for commercial vehicle cab based on beam section |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113591230B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114186335B (en) * | 2021-11-23 | 2024-04-12 | 广西汽车集团有限公司 | Optimal design method for automobile frame |
| CN114647897B (en) * | 2022-03-30 | 2024-03-29 | 桂林电子科技大学 | Cab structure optimization method based on hierarchical entropy-gray correlation method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106202731A (en) * | 2016-07-12 | 2016-12-07 | 南京理工大学 | Bridge crane multi-flexibl e dynamics structural optimization method |
| CN110147600A (en) * | 2019-05-13 | 2019-08-20 | 桂林电子科技大学 | Cargo vehicle modification methodology of dynamics model based on typical impact operating condition |
| CN110210134A (en) * | 2019-06-04 | 2019-09-06 | 合肥工业大学 | A kind of body-in-white structure optimum design method and Optimal Structure Designing device |
| KR102037945B1 (en) * | 2018-12-10 | 2019-10-29 | 엘아이지넥스원 주식회사 | Composite Optical System for Multi-Target Detection and Apparatus Therefor |
-
2021
- 2021-09-13 CN CN202111069225.9A patent/CN113591230B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106202731A (en) * | 2016-07-12 | 2016-12-07 | 南京理工大学 | Bridge crane multi-flexibl e dynamics structural optimization method |
| KR102037945B1 (en) * | 2018-12-10 | 2019-10-29 | 엘아이지넥스원 주식회사 | Composite Optical System for Multi-Target Detection and Apparatus Therefor |
| CN110147600A (en) * | 2019-05-13 | 2019-08-20 | 桂林电子科技大学 | Cargo vehicle modification methodology of dynamics model based on typical impact operating condition |
| CN110210134A (en) * | 2019-06-04 | 2019-09-06 | 合肥工业大学 | A kind of body-in-white structure optimum design method and Optimal Structure Designing device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113591230A (en) | 2021-11-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113591230B (en) | Multi-objective optimization method for commercial vehicle cab based on beam section | |
| WO2020244325A1 (en) | Relative sensitivity-based method for lightweighting of non-load bearing body-in-white | |
| KR102553414B1 (en) | Systems and methods for additive manufacturing of transport structures | |
| CN112287551B (en) | Driving performance system level index decomposition method based on whole vehicle conceptual model | |
| Wang et al. | A hybrid method of modified NSGA-II and TOPSIS for lightweight design of parameterized passenger car sub-frame | |
| CN111444623B (en) | Collaborative optimization method and system for damping nonlinear commercial vehicle suspension dynamics | |
| CN111209697B (en) | Analysis method and system for vehicle body ceiling structure based on topology optimization | |
| CN108629137B (en) | Optimization design method for structural parameters of mechanical structural part | |
| CN106980744A (en) | Control stability and ride comfort cooperative optimization method in vehicle dynamics simulation analysis | |
| JP5440773B2 (en) | Vehicle planning support system | |
| CN109002599A (en) | The automobile ride method for optimization analysis tested based on field cause for gossip | |
| CN113946911B (en) | Cab structure optimization method based on gray correlation analysis method | |
| JP7390731B2 (en) | Road surface profile estimation device, road surface profile estimation system, road surface profile estimation method, and road surface profile estimation program | |
| Lim et al. | Role of multi-response principal component analysis in reliability-based robust design optimization: an application to commercial vehicle design | |
| CN113127978B (en) | Optimization method for light weight of instrument board beam | |
| CN112711813B (en) | Lightweight method of riveting structure | |
| CN114647897B (en) | Cab structure optimization method based on hierarchical entropy-gray correlation method | |
| CN114580139A (en) | Method for automatically evaluating a simulation model | |
| CN113901585A (en) | White body lightweight optimization method for commercial vehicle cab | |
| CN114841037B (en) | Cab rigidity modal optimization method based on joint sensitivity analysis | |
| Niedermeier et al. | Virtual Optimization of the Interaction between Tires and the Vehicle | |
| CN117708997B (en) | Vehicle suspension design parameter determination method, device and medium based on genetic algorithm | |
| Kodippili et al. | Multi-objective optimization of a cast-preform shape for a magnesium alloy forging application | |
| CN104166774B (en) | Multi-working-condition and multi-target design method for automobile seat framework | |
| Jin et al. | Immune particle swarm optimization of SUV rollover crashworthiness and weight |
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 |