CN117993110A - Novel vehicle chassis truss lightweight optimization design system - Google Patents
Novel vehicle chassis truss lightweight optimization design system Download PDFInfo
- Publication number
- CN117993110A CN117993110A CN202410286669.5A CN202410286669A CN117993110A CN 117993110 A CN117993110 A CN 117993110A CN 202410286669 A CN202410286669 A CN 202410286669A CN 117993110 A CN117993110 A CN 117993110A
- Authority
- CN
- China
- Prior art keywords
- design
- model
- shamode
- algorithm
- iwoa
- 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
- 238000013461 design Methods 0.000 title claims abstract description 88
- 238000005457 optimization Methods 0.000 title claims abstract description 60
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 40
- 230000003044 adaptive effect Effects 0.000 claims abstract description 14
- 238000000034 method Methods 0.000 claims description 39
- 230000006870 function Effects 0.000 claims description 35
- 230000008569 process Effects 0.000 claims description 31
- 238000012549 training Methods 0.000 claims description 23
- 238000012360 testing method Methods 0.000 claims description 20
- 238000004364 calculation method Methods 0.000 claims description 12
- 241000283153 Cetacea Species 0.000 claims description 6
- 238000009826 distribution Methods 0.000 claims description 6
- 230000033001 locomotion Effects 0.000 claims description 6
- 238000005452 bending Methods 0.000 claims description 5
- 238000004088 simulation Methods 0.000 claims description 5
- 230000000694 effects Effects 0.000 claims description 4
- 238000012986 modification Methods 0.000 claims description 4
- 230000004048 modification Effects 0.000 claims description 4
- 108010014172 Factor V Proteins 0.000 claims description 3
- 230000002159 abnormal effect Effects 0.000 claims description 3
- UJRRDDHEMZLWFI-UHFFFAOYSA-N aminitrozole Chemical compound CC(=O)NC1=NC=C([N+]([O-])=O)S1 UJRRDDHEMZLWFI-UHFFFAOYSA-N 0.000 claims description 3
- 230000008859 change Effects 0.000 claims description 3
- 239000003795 chemical substances by application Substances 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 230000035772 mutation Effects 0.000 claims description 3
- 238000012847 principal component analysis method Methods 0.000 claims description 3
- 230000008929 regeneration Effects 0.000 claims description 3
- 238000011069 regeneration method Methods 0.000 claims description 3
- 238000011160 research Methods 0.000 claims description 3
- 230000000717 retained effect Effects 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 230000004913 activation Effects 0.000 claims 1
- 238000004458 analytical method Methods 0.000 claims 1
- 238000011161 development Methods 0.000 abstract description 3
- 238000004519 manufacturing process Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000012938 design process Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000002860 competitive effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
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
- G06F30/23—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]
- G06F30/20—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]
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
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)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Software Systems (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Automation & Control Theory (AREA)
- Aviation & Aerospace Engineering (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a novel vehicle chassis truss lightweight optimization design system, and relates to the technical field of vehicle chassis truss optimization design; the design steps of the optimal design system are as follows: firstly, introducing an adaptive spiral search strategy on the basis of a multi-objective mixed element heuristic algorithm based on adaptive differential evolution of success history, and further providing a novel SHAMODE-IWOA algorithm; then to estimate the reliability level of the chassis truss structure under different combinations of design parameters, a new model of ANFIS-SHAMODE-IWOA was constructed by learning the ANFIS model using the proposed SHAMODE-IWOA algorithm. The ANFIS-SHAMODE-IWOA multi-target optimization design system provided by the invention is used as a novel intelligent model, can be used for evaluating the reliability of a chassis truss structure, improves the development and design efficiency, and is convenient for obtaining the optimal design parameter combination.
Description
Technical Field
The invention relates to the technical field of vehicle chassis truss optimization design, in particular to a novel vehicle chassis truss lightweight optimization design system.
Background
The chassis of the vehicle is an important part for ensuring the power output and stable and reliable running of the vehicle, and the optimal design of the chassis is a key problem in the light weight optimization of the vehicle. In the light-weight optimization design process, taking quality and reliability indexes into consideration while topologically optimizing the shape and the overall dimension of the chassis truss structure is a complicated and laborious task for a designer. He often needs to perform multiple work and cooperation across departments through a large number of finite element simulation calculations to achieve the smooth completion and better effect. Currently, intelligent design, efficient high quality design has become a new trend in the manufacturing industry. Meanwhile, a design system capable of developing discipline crossing, improving working efficiency and reducing cross-department coordination has become a target for competitive pursuit of design departments. In the chassis design process, how to reduce the quality and simultaneously maintain the reliable durability of the chassis truss structure is a key element for improving the design quality of the chassis truss structure, reducing the energy consumption of vehicles and improving the subsequent processing efficiency. Therefore, it is especially critical to establish a lightweight auxiliary optimization design system with accurate calculation, reliable design and efficient operation.
Further, topological optimization of the shape and the overall dimensions of the truss structure of the vehicle chassis while considering quality and reliability indexes is a difficult and arduous task for designers, but the traditional lightweight design model is not fully studied about the joint model of truss multi-objective optimization and reliability estimation.
In the prior art, the optimization algorithm can search the front edges of a plurality of Pareto targets at the same time in one solving process, so that the multi-target problem can be solved conveniently, but the optimization algorithm has the defects of low efficiency and easy sinking into local optimum under the condition of processing a high-dimensional problem or non-micro constraint and the like.
The traditional ANFIS model can only be optimized by back-propagation algorithms or hybrid learning algorithms, which limit further development of the ANFIS model. Furthermore, the ANFIS model, while advantageous in many respects, has several drawbacks, including:
1. The complexity is high: the structure of the ANFIS model is relatively complex, including multiple parts such as fuzzification, rule base, reasoning mechanism, etc., requiring significant computational resources and time to train and optimize the model.
2. Overfitting: because the ANFIS model has larger flexibility and parameter quantity, when the training data quantity is less or the noise is more, the problem of fitting is easy to occur, so that the generalization capability of the model on new data is poorer.
3. Rule setting is difficult: the fuzzy rules in the ANFIS model require manual setup, selecting the appropriate number and form of rules is a challenge, and the setup of rules is typically dependent on the experience of the domain expert.
4. Training convergence is difficult: training of the ANFIS model requires updating the model parameters by iterative optimization algorithms, but sometimes may fall into locally optimal solutions, requiring careful adjustment of learning rates and other super-parameters.
5. Not applicable to large-scale data sets: due to the complexity of the ANFIS model and the high computational requirements of the training process, the model is not well suited for processing large-scale data sets, and the training time is long and resources are consumed.
Disclosure of Invention
The invention aims to solve the defects in the prior art, and provides a novel lightweight optimization design system for a vehicle chassis truss.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
a novel vehicle chassis truss lightweight optimization design system comprises the following design steps:
S1: firstly, introducing an adaptive spiral search strategy on the basis of a multi-objective mixed element heuristic algorithm based on adaptive differential evolution of success history, and further providing a novel SHAMODE-IWOA algorithm;
S2: then, in order to estimate the reliability level of the chassis truss structure under different design parameter combinations, a novel ANFIS-SHAMODE-IWOA model is constructed by learning the ANFIS model by using the proposed SHAMODE-IWOA algorithm;
S3: finally, to obtain the optimal design parameter combination, a SHAMODE-IWOA algorithm is used to study multi-objective optimization based on minimum design quality and optimal reliability metric functions.
Preferably: the specific steps of the S1 are as follows:
S11: initializing parameters; randomly generating NP initialization solution sets;
xi,G＝[x1,G,x2,G,x3,G,…,xNP,G]
wherein i is index, NP is index extremum, equal to the number of design variables, G is iteration number;
Filling non-dominant solutions in the initial Pareto archive Pareto 1; creating an empty external archive (a 1) for the regeneration process; in the process, initializing and setting initial values of all adaptive parameters;
S12: variation, in the variation process, a random strategy is adopted to generate a variation;
Wherein V i,G ε [0,1] is the scaling factor that controls the influence of differential variation;
x i,G is a viable solution for the G generation;
x pbest is a solution randomly selected from the external Pareto profile;
x r1,G is a randomly selected solution from the current population (xG), Then it is a randomly selected solution from the union (x G∪AG) of the current population and the external archive;
S13: crossing;
S14: selecting; extracting a current individual x i,G and a test individual h i,G to rank the combined population (x G∪uG) using a non-dominant ranking scheme of NSGA-II; then, the NP solutions with the highest non-dominant levels will be retained in the next iteration; if the number of solutions with the highest non-dominant level exceeds NP, some of them will be randomly removed to maintain a constant population size; finally, the NP survivors of the current iteration will be stored in x G+1; after the selection process, all non-dominant solutions ordered from h G∪ParetoG will be saved in Pareto G+1, some of which will be randomly removed from the archive if the number of non-dominant solutions exceeds the maximum Pareto archive size;
S15: parameter self-adaptive regulation strategies; all adaptive parameters, including external profile (a), history of scaling factors (MF) and cross-ratio (MCR), are updated at the end of each iteration; at the end of each iteration, the index of the offspring that survived the selection process that was successfully updated is stored in a vector named sind; then, the parent vector that generated the successful child (x sind, G) is stored in external archive a G+1; if the number of solutions stored in the external archive exceeds a specified value, some of the solutions are randomly deleted to maintain a constant archive size;
S16: improving whale self-adaptive spiral strategy, and integrating spiral motion of WOA into SHAMODE algorithm; modifications are made during the mutation process, where each mutated vector has the opportunity to update further with the helical movement of the WOA, and then activate the crossover process.
Preferably: in S13, the crossover phase is performed as shown in the following formula:
Wherein R i,G is the cross ratio, the range R i,G E [0,1]; rand ([ 0, 1)) represents a simple random number; j rand is the x index randomly generated from [1,2, …, n ], n is the number of design variables.
Preferably: in S15, the update of VC and RG is performed according to the formula as follows:
Vi,G＝randci(μv，0.1)
Ri,G＝randni(μR，0.1)
Wherein μ v and μ R are mean values whose initial values are 0.5, ramdct (μ, 0.1) and randnt (μR, 0.1) are random numbers generated based on the Cauchy distribution and the normal distribution, where the variance Is (0.1);
In addition, let M V and M R be the history of the scaling factor V i,G and the crossover ratio R i,G, from which μ v and μ R for each individual can be randomly selected; at the end of each iteration, one element of memory (MF and MCR) will be updated with the Lehmer mean of the parameters that generated the offspring of the successful update, as follows:
Wherein, L mean(Vsind,G) and L mean(Rsind,G) are the Lehmer mean of V i,G and R i,G, respectively, and subscript sind indicates individuals of successful offspring, namely: if there is at least one successful offspring, the kth memory location will be updated; otherwise, such elements will remain unchanged; index k is a memory element index that is initially equal to 1 and increases linearly as the process continues; if k > H, k is reset to 1.
Preferably: in S16, the specific formula of the active crossover process is as follows:
Di＝|xpbest2-ui,G|
Wherein, Is a new variant with a spiral update process, and x pbest2 is another feasible solution (different from x pbest1) that is randomly selected from the current Pareto profile; the variables 1 and rand are random numbers in the intervals [ -1,1] and [0,1], respectively.
Preferably: in S16, the spiral shape parameter l is set to a dynamic value that varies with the number of iterations, so that the whale individual can dynamically adjust the spiral shape during searching after entering the spiral searching stage, and the global searching capability of the algorithm is enhanced to improve the convergence accuracy of the algorithm, specifically as follows:
Di＝|xpbest2-ui,G|
Wherein, gamma is the spiral shape adjustment coefficient, G max is the maximum iteration number, and G is the current iteration number.
Preferably: the specific mode of constructing the model in the step S2 is as follows:
Layer 1: the self-adaptive node is a membership function layer of input variables, converts the input into a fuzzy set, and the output functions are as follows:
Wherein x 1 and x 2 are inputs; a i and B i-2 represent fuzzy sets, i.e. linguistic variables calculated; And/> As the membership function, a gaussian function is selected:
Wherein parameters c ij and σ ij are initial parameters that need to be adjusted by a learning algorithm;
layer 2: the effect of this layer is a regular intensity release, the node function is multiplied by the input, represented by the fuzzy rule:
layer 3: the number of nodes of the layer 3 is the same as that of the layer 2, and the result of the layer 2 is normalized:
layer 4: layer 4 is a back-piece network, obtaining fuzzy if-then rules, and calculating the output of the fuzzy rules:
layer 5: layer 5 is the output layer, calculates the sum of the input signals as the whole output:
in the step S2, after the structure of the ANFIS model is established, training the ANFIS model by using SHAMODE-iWOA algorithm; the root mean square error is used as a fitness function to train the ANFIS as follows:
Wherein n is the data quantity of the training set; r is the r data, dr is the actual value of the r training set data; pr is the r-th predictor.
Preferably: the optimal design system and the acquisition method of the data set comprise the following steps:
S31: constructing a database required by chassis truss structure research;
s32: constructing a chassis truss structure SFE model;
s33: after solving, extracting design variables in the SFE model according to the contribution by using a principal component analysis method;
S34: selecting a design variable with a contribution quantity in front as a design parameter;
s35: generating a DOE matrix for the design variables, and then solving, wherein Bending, torion and Mass solving relies on a finite element solver to carry out calculation and solving;
S36: removing abnormal data after solving to obtain a data set;
S37: the experimental data set obtained after model solving is divided into two parts: a training data set and a test data set; the training dataset was used to train the ANFIS-SHAMODE-IWOA model and the test dataset was used to test the performance of the ANFIS-SHAMODE-IWOA model.
Preferably: in the optimal design system, the multi-objective optimization of the chassis truss structure comprises the following steps:
s41: establishing a multi-objective optimization equation;
S42: solving and analyzing multi-objective optimization;
the multi-objective optimization equation is specifically established as follows:
Truss reliability and mass multi-objective equations are built as follows:
Wherein Pr is the fault probability; x is a design variable vector; y is a physical vector comprising yield strength, torsional stiffness and applied load; f1 is the structural mass, which is the sum of the products of the mass and the density of the topological unit; f2 is a reliability metric function: f2 =1/β, β being a reliability coefficient;
the reliability coefficient (beta) refers to the shortest distance between the limit state line and the origin of the transformation space; the larger β represents higher reliability (higher safety), and is represented by the following formula:
Wherein, mu, m can be characterized as the average value of the stress in the state of the limit failure function, and sigma m is characterized as the variance of the mechanical change rate in the state of the limit failure function.
Preferably: in the solving and analyzing of the multi-objective optimization, SHAMODE-IWOA algorithm is used for carrying out optimization solving on the constructed multi-objective function; based on the iteration solving result, obtaining a design index optimization value by using a compromise method; and further performing simulation calculation on the finite element model.
The beneficial effects of the invention are as follows:
1. the ANFIS-SHAMODE-IWOA multi-target optimization design system provided by the invention is used as a novel intelligent model, can be used for evaluating the reliability of a chassis truss structure, improves the development and design efficiency, is convenient for obtaining the optimal design parameter combination, and is beneficial to improving the green intelligent manufacturing design level.
Drawings
FIG. 1 is a general structure diagram of the overall operation and model of a novel vehicle chassis truss lightweight optimization design system provided by the invention;
FIG. 2 is a schematic diagram of database construction in a novel vehicle chassis truss lightweight optimization design system;
FIG. 3 is a comparison chart of design calculation optimization indexes of the novel vehicle chassis truss lightweight optimization design system;
FIG. 4 is a graph of the results of stress acquisition testing according to the present invention.
Detailed Description
The technical scheme of the invention is further described in detail below with reference to the specific embodiments.
Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are illustrative only and are not to be construed as limiting the invention.
Example 1:
a novel vehicle chassis truss lightweight optimization design system comprises the following design steps:
S1: firstly, introducing an adaptive spiral search strategy on the basis of a multi-objective mixed element heuristic algorithm (SHAMODE) based on adaptive differential evolution of success history, and further providing a novel SHAMODE-IWOA algorithm;
s2: then, in order to estimate the reliability level of the chassis truss structure under different design parameter combinations, a novel ANFIS-SHAMODE-IWOA model is constructed by learning the ANFIS model by using the proposed SHAMODE-IWOA algorithm;
s3: finally, to obtain the optimal design parameter combination, a SHAMODE-IWOA algorithm is used to study multi-objective optimization based on minimum design quality and optimal reliability metric functions.
The specific steps of the S1 are as follows:
S11: initializing parameters; randomly generating NP initialization solution sets;
xi,G＝[x1,G,x2,G,x3,G,…,xNP,G]
wherein i is index, NP is index extremum, equal to the number of design variables, G is iteration number;
Filling non-dominant solutions in the initial Pareto archive Pareto 1; creating an empty external archive (a 1) for the regeneration process; in the process, initializing and setting initial values of all adaptive parameters;
S12: variation, in the variation process, a random strategy is adopted to generate a variation;
Wherein V i,G ε [0,1] is the scaling factor that controls the influence of differential variation;
x i,G is a viable solution for the G generation;
x pbest is a solution randomly selected from the external Pareto profile;
x r1,G is a randomly selected solution from the current population (xG), Then it is a randomly selected solution from the union (x G∪AG) of the current population and the external archive;
S13: the crossover phase is performed by the following formula:
Wherein R i,G is the cross ratio, the range R i,G E [0,1]; rand ([ 0, 1)) represents a simple random number; j rand is the x index randomly generated from [1,2, …, n ], n is the number of design variables;
S14: selecting; extracting a current individual x i,G and a test individual h i,G to rank the combined population (x G∪uG) using a non-dominant ranking scheme of NSGA-II; then, the NP solutions with the highest non-dominant levels will be retained in the next iteration; if the number of solutions with the highest non-dominant level exceeds NP, some of them will be randomly removed to maintain a constant population size; finally, the NP survivors of the current iteration will be stored in x G+1; after the selection process, all non-dominant solutions ordered from h G∪ParetoG will be saved in Pareto G+1, some of which will be randomly removed from the archive if the number of non-dominant solutions exceeds the maximum Pareto archive size;
S15: parameter self-adaptive regulation strategies; all adaptive parameters, including external profile (a), history of scaling factors (MF) and cross-ratio (MCR), are updated at the end of each iteration; the maximum number of solutions in the external archive (a) is 1.4 xnp; at the end of each iteration, the index of the offspring that survived the selection process that was successfully updated is stored in a vector named sind; then, the parent vector that generated the successful child (x sind, G) is stored in external archive a G+1; if the number of solutions stored in the external archive exceeds a specified value, some of the solutions are randomly deleted to maintain a constant archive size;
the updating of VC and RG is performed according to the formula as follows:
Vi,G＝randci(μv，0.1)
Ri,G＝randni(μR，0.1)
Wherein μ v and μ R are mean values whose initial values are 0.5, ramdct (μ, 0.1) and randnt (μR, 0.1) are random numbers generated based on the Cauchy distribution and the normal distribution, where the variance Is (0.1);
In addition, let M V and M R be the history of the scaling factor V i,G and the crossover ratio R i,G, from which μ v and μ R for each individual can be randomly selected; at the end of each iteration, one element of memory (MF and MCR) will be updated with the Lehmer mean of the parameters that generated the offspring of the successful update, as follows:
Wherein, L mean(Vsind,G) and L mean(Rsind,G) are the Lehmer mean of V i,G and R i,G, respectively, and subscript sind indicates individuals of successful offspring, namely: if there is at least one successful offspring, the kth memory location will be updated; otherwise, such elements will remain unchanged; index k is a memory element index that is initially equal to 1 and increases linearly as the process continues; if k > H, k is reset to 1;
S16: improving whale self-adaptive spiral strategy, and integrating spiral motion of WOA into SHAMODE algorithm; modification is performed during the mutation process, wherein each mutated vector has the opportunity to update further with the helical movement of WOA, and then the crossover process is activated as follows:
Di＝|xpbest2-ui,G|
Wherein, Is a new variant with a spiral update process, and x pbest2 is another feasible solution (different from x pbest1) that is randomly selected from the current Pareto profile; the variables 1 and rand are random numbers in the intervals [ -1,1] and [0,1], respectively;
The spiral shape parameter l is set to be a dynamic value which changes along with the iteration times, so that a whale individual can dynamically adjust the spiral shape during searching after entering a spiral searching stage, the global searching capability of an algorithm is enhanced to improve the convergence accuracy of the algorithm, and the method is specifically as follows:
Di＝|xpbest2-ui,G|
Wherein, gamma is the spiral shape adjustment coefficient, G max is the maximum iteration number, and G is the current iteration number.
The specific way of constructing the model in S2 is as follows:
Layer 1: the self-adaptive node is a membership function layer of input variables, converts the input into a fuzzy set, and the output functions are as follows:
Wherein x 1 and x 2 are inputs; a i and B i-2 represent fuzzy sets, i.e. linguistic variables calculated; And/> As the membership function, a gaussian function is selected:
Wherein parameters c ij and σ ij are initial parameters that need to be adjusted by a learning algorithm;
layer 2: the effect of this layer is a regular intensity release, the node function is multiplied by the input, represented by the fuzzy rule:
layer 3: the number of nodes of the layer 3 is the same as that of the layer 2, and the result of the layer 2 is normalized:
layer 4: layer 4 is a back-piece network, obtaining fuzzy if-then rules, and calculating the output of the fuzzy rules:
layer 5: layer 5 is the output layer, calculates the sum of the input signals as the whole output:
in the step S2, after the structure of the ANFIS model is established, training the ANFIS model by using SHAMODE-iWOA algorithm; the root mean square error is used as a fitness function to train the ANFIS as follows:
Wherein n is the data quantity of the training set; r is the r data, dr is the actual value of the r training set data; pr is the r-th predictor.
Constructing an ANFIS model for obtaining the design parameter combination with the best reliability and the lowest quality; the parameters input by the model are Torsion, mass, bending, DESIGN PARAMETER, relability Coefficient and the like, and the reliability prediction result is output. In addition, an ANFIS model result is used as one of objective functions to construct a multi-objective optimization model, and finally, the optimal design parameters are obtained.
The optimal design system and the acquisition method of the data set comprise the following steps:
S31: constructing a database required by chassis truss structure research;
s32: constructing a chassis truss structure SFE model;
s33: after solving, extracting design variables in the SFE model according to the contribution by using a principal component analysis method;
S34: selecting a design variable with a contribution quantity in front as a design parameter;
s35: generating a DOE matrix for the design variables, and then solving, wherein Bending, torion and Mass solving relies on a finite element solver to carry out calculation and solving;
S36: removing abnormal data after solving to obtain a data set;
S37: the experimental data set obtained after model solving is divided into two parts: a training data set and a test data set; the training data set is used for training the ANFIS-SHAMODE-IWOA model, and the test data set is used for testing the performance of the ANFIS-SHAMODE-IWOA model;
Wherein, to exclude sample selection from affecting model training, a random extraction method is adopted to extract 70% of the data set as training data set and 30% as test data set.
In the optimal design system, the multi-objective optimization of the chassis truss structure comprises the following steps:
s41: establishing a multi-objective optimization equation;
s42: and solving and analyzing multi-objective optimization.
The multi-objective optimization equation is specifically established as follows:
Truss reliability and mass multi-objective equations are built as follows:
Wherein Pr is the fault probability; x is a design variable vector; y is a physical vector comprising yield strength, torsional stiffness and applied load; f1 is the structural mass, which is the sum of the products of the mass and the density of the topological unit; f2 is a reliability metric function: f2 =1/β, β being the reliability coefficient.
The reliability coefficient (beta) refers to the shortest distance between the limit state line and the origin of the transformation space; the larger β represents higher reliability (higher safety), and is represented by the following formula:
Wherein, mu, m can be characterized as the average value of the stress in the state of the limit failure function, and sigma m is characterized as the variance of the mechanical change rate in the state of the limit failure function.
In the multi-objective optimization solving and analyzing, the SHAMODE-IWOA algorithm is used for carrying out optimization solving on the constructed multi-objective function; based on the iteration solving result, obtaining a design index optimization value by using a compromise method; and further performing simulation calculation on the finite element model.
And (3) testing:
In order to further comprehensively verify the optimization result, researching and adopting an electrical measurement method to perform stress test on the optimized key position of the truss structure; the vehicle carries the chassis truss after optimization trial production, selects typical test pavement to run, and runs 10 groups of related severe pavement, each group of 20 cycles;
Strain data under each stress level is separated through a rain flow counting method, the separation result is shown in the following graph, and the damage coupling values corresponding to the stress at each level of four positions shown in the upper graph are calculated respectively.
The chassis truss framework of the dumper of a certain brand is optimized by adopting the proposed SHAMODE-IWOA algorithm, related design parameters of the truss are adjusted based on the optimized calculation result, and then trial-manufacturing and loading are carried out, and three tests of mathematical model calculation, finite element simulation calculation and stress actual test are carried out. The test results show that: the optimized chassis truss structure improves torison, bending performance and mechanical performance while reducing total truss quality and improving reliability, and finally shows that the algorithm has good engineering application value.
The foregoing is only a preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art, who is within the scope of the present invention, should make equivalent substitutions or modifications according to the technical scheme of the present invention and the inventive concept thereof, and should be covered by the scope of the present invention.
Claims (10)
1. The utility model provides a novel vehicle chassis truss lightweight optimal design system which characterized in that, optimal design system's design step does:
S1: firstly, introducing an adaptive spiral search strategy on the basis of a multi-objective mixed element heuristic algorithm based on adaptive differential evolution of success history, and further providing a novel SHAMODE-IWOA algorithm;
S2: then, in order to estimate the reliability level of the chassis truss structure under different design parameter combinations, a novel ANFIS-SHAMODE-IWOA model is constructed by learning the ANFIS model by using the proposed SHAMODE-IWOA algorithm;
S3: finally, to obtain the optimal design parameter combination, a SHAMODE-IWOA algorithm is used to study multi-objective optimization based on minimum design quality and optimal reliability metric functions.
2. The novel vehicle chassis truss lightweight optimization design system according to claim 1, wherein the specific steps of S1 are as follows:
S11: initializing parameters; randomly generating NP initialization solution sets;
wherein i is index, NP is index extremum, equal to the number of design variables, G is iteration number;
Filling non-dominant solutions in the initial Pareto archive Pareto 1; creating an empty external archive (a 1) for the regeneration process; in the process, initializing and setting initial values of all adaptive parameters;
S12: variation, in the variation process, a random strategy is adopted to generate a variation;
Wherein V i,G ε [0,1] is the scaling factor that controls the influence of differential variation;
x i,G is a viable solution for the G generation;
x pbest is a solution randomly selected from the external Pareto profile;
x r1,G is a randomly selected solution from the current population (xG), Then it is a randomly selected solution from the union (x G∪AG) of the current population and the external archive;
S13: crossing;
S14: selecting; extracting a current individual x i,G and a test individual h i,G to rank the combined population (x G∪uG) using a non-dominant ranking scheme of NSGA-II; then, the NP solutions with the highest non-dominant levels will be retained in the next iteration; if the number of solutions with the highest non-dominant level exceeds NP, some of them will be randomly removed to maintain a constant population size; finally, the NP survivors of the current iteration will be stored in x G+1; after the selection process, all non-dominant solutions ordered from h G∪ParetoG will be saved in Pareto G+1, some of which will be randomly removed from the archive if the number of non-dominant solutions exceeds the maximum Pareto archive size;
S15: parameter self-adaptive regulation strategies; all adaptive parameters, including external profile (a), history of scaling factors (MF) and cross-ratio (MCR), are updated at the end of each iteration; at the end of each iteration, the index of the offspring that survived the selection process that was successfully updated is stored in a vector named sind; then, the parent vector that generated the successful child (x sind, G) is stored in external archive a G+1; if the number of solutions stored in the external archive exceeds a specified value, some of the solutions are randomly deleted to maintain a constant archive size;
S16: improving whale self-adaptive spiral strategy, and integrating spiral motion of WOA into SHAMODE algorithm; modifications are made during the mutation process, where each mutated vector has the opportunity to update further with the helical movement of the WOA, and then activate the crossover process.
3. The novel vehicle chassis truss lightweight optimization design system according to claim 2, wherein in S13, the crossing stage is performed by the following formula:
Wherein R i,G is the cross ratio, the range R i,G E [0,1]; rand ([ 0, 1)) represents a simple random number; j rand is the x index randomly generated from [1,2, …, n ], n is the number of design variables.
4. The novel vehicle chassis truss lightweight optimization design system according to claim 3, wherein in S15, the updating of VC and RG is performed according to the formula as follows:
Vi,G＝randci(μv,0.1)
Ri,G＝randni(μR,0.1)
wherein μ v and μ R are mean values whose initial values are 0.5, ramdct (μ, 0.1) and randnt (μR, 0.1) are random numbers generated based on the Cauchy distribution and the normal distribution, where the variance Is (0.1);
In addition, let M V and M R be the history of the scaling factor V i,G and the crossover ratio R i,G, from which μ v and μ R for each individual can be randomly selected; at the end of each iteration, one element of memory (MF and MCR) will be updated with the Lehmer mean of the parameters that generated the offspring of the successful update, as follows:
Wherein, L mean(Vsind,G) and L mean(Rsind,G) are the Lehmer mean of V i,G and R i,G, respectively, and subscript sind indicates individuals of successful offspring, namely: if there is at least one successful offspring, the kth memory location will be updated; otherwise, such elements will remain unchanged; index k is a memory element index that is initially equal to 1 and increases linearly as the process continues; if k > H, k is reset to 1.
5. The novel vehicle chassis truss lightweight optimization design system according to claim 4, wherein in S16, the specific formula of the activation crossover process is as follows:
Di＝|xpbest2-ui,G|
Wherein, Is a new variant with a spiral update process, and x pbest2 is another feasible solution (different from x pbest1) that is randomly selected from the current Pareto profile; the variables l and rand are random numbers in the intervals [ -1,1] and [0,1], respectively.
6. The system according to claim 5, wherein in S16, the spiral shape parameter i is set to a dynamic value that varies with the number of iterations, so that the whale individual can dynamically adjust the spiral shape during searching after entering the spiral searching stage, and the global searching capability of the algorithm is enhanced to improve the convergence accuracy of the algorithm, specifically as follows:
Di＝|xpbest2-ui,G|
Wherein, gamma is the spiral shape adjustment coefficient, G max is the maximum iteration number, and G is the current iteration number.
7. The novel vehicle chassis truss lightweight optimization design system according to claim 1, wherein the specific mode of constructing the model in S2 is as follows:
Layer 1: the self-adaptive node is a membership function layer of input variables, converts the input into a fuzzy set, and the output functions are as follows:
Wherein x 1 and x 2 are inputs; a i and B i-2 represent fuzzy sets, i.e. linguistic variables calculated; And/> As the membership function, a gaussian function is selected:
Wherein parameters c ij and σ ij are initial parameters that need to be adjusted by a learning algorithm;
layer 2: the effect of this layer is a regular intensity release, the node function is multiplied by the input, represented by the fuzzy rule:
layer 3: the number of nodes of the layer 3 is the same as that of the layer 2, and the result of the layer 2 is normalized:
layer 4: layer 4 is a back-piece network, obtaining fuzzy if-then rules, and calculating the output of the fuzzy rules:
layer 5: layer 5 is the output layer, calculates the sum of the input signals as the whole output:
in the step S2, after the structure of the ANFIS model is established, training the ANFIS model by using SHAMODE-iWOA algorithm; the root mean square error is used as a fitness function to train the ANFIS as follows:
Wherein n is the data quantity of the training set; r is the r data, dr is the actual value of the r training set data; pr is the r-th predictor.
8. The novel vehicle chassis truss lightweight optimization design system of claim 7, wherein the optimization design system, the data set acquisition method comprises the following steps:
S31: constructing a database required by chassis truss structure research;
s32: constructing a chassis truss structure SFE model;
s33: after solving, extracting design variables in the SFE model according to the contribution by using a principal component analysis method;
S34: selecting a design variable with a contribution quantity in front as a design parameter;
s35: generating a DOE matrix for the design variables, and then solving, wherein Bending, torion and Mass solving relies on a finite element solver to carry out calculation and solving;
S36: removing abnormal data after solving to obtain a data set;
S37: the experimental data set obtained after model solving is divided into two parts: a training data set and a test data set; the training dataset was used to train the ANFIS-SHAMODE-IWOA model and the test dataset was used to test the performance of the ANFIS-SHAMODE-IWOA model.
9. The novel vehicle chassis truss lightweight optimization design system of claim 1, wherein the multi-objective optimization of the chassis truss structure in the optimization design system comprises the steps of:
s41: establishing a multi-objective optimization equation;
S42: solving and analyzing multi-objective optimization;
the multi-objective optimization equation is specifically established as follows:
Truss reliability and mass multi-objective equations are built as follows:
Wherein Pr is the fault probability; x is a design variable vector; y is a physical vector comprising yield strength, torsional stiffness and applied load; f1 is the structural mass, which is the sum of the products of the mass and the density of the topological unit; f2 is a reliability metric function: f2 =1/β, β being a reliability coefficient;
the reliability coefficient (beta) refers to the shortest distance between the limit state line and the origin of the transformation space; the larger β represents higher reliability (higher safety), and is represented by the following formula:
Wherein, mu, m can be characterized as the average value of the stress in the state of the limit failure function, and sigma m is characterized as the variance of the mechanical change rate in the state of the limit failure function.
10. The novel vehicle chassis truss lightweight optimization design system according to claim 9, wherein in the multi-objective optimization solution and analysis, the SHAMODE-IWOA algorithm is used to perform optimization solution on the constructed multi-objective function; based on the iteration solving result, obtaining a design index optimization value by using a compromise method; and further performing simulation calculation on the finite element model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410286669.5A CN117993110A (en) | 2024-03-13 | 2024-03-13 | Novel vehicle chassis truss lightweight optimization design system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410286669.5A CN117993110A (en) | 2024-03-13 | 2024-03-13 | Novel vehicle chassis truss lightweight optimization design system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117993110A true CN117993110A (en) | 2024-05-07 |
Family
ID=90888898
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410286669.5A Pending CN117993110A (en) | 2024-03-13 | 2024-03-13 | Novel vehicle chassis truss lightweight optimization design system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117993110A (en) |
-
2024
- 2024-03-13 CN CN202410286669.5A patent/CN117993110A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106600059B (en) | Intelligent power grid short-term load prediction method based on improved RBF neural network | |
| CN105930901B (en) | A kind of Diagnosis Method of Transformer Faults based on RBPNN | |
| CN110428175B (en) | Hot rolled strip steel plate convexity prediction method based on deep learning | |
| CN107609671A (en) | A kind of Short-Term Load Forecasting Method based on composite factor evaluation model | |
| CN116362398A (en) | Power load prediction method based on modal decomposition and reconstruction and LSTM-MLR hybrid model | |
| CN111859733B (en) | Method for optimizing reliability of automobile exhaust system based on ant colony algorithm | |
| CN113836823A (en) | Load combination prediction method based on load decomposition and optimized bidirectional long-short term memory network | |
| CN113065268A (en) | Rapid multi-objective space reflector optimization design method based on proxy model | |
| CN114036850A (en) | Runoff prediction method based on VECGM | |
| CN116774088A (en) | Lithium ion battery health state estimation method based on multi-objective optimization | |
| CN115856633A (en) | Lithium ion battery capacity estimation method based on graph neural network | |
| CN114169251A (en) | Ultra-short-term wind power prediction method | |
| CN111325308B (en) | Nonlinear system identification method | |
| CN112036651A (en) | Electricity price prediction method based on quantum immune optimization BP neural network algorithm | |
| CN111105005B (en) | Wind power prediction method | |
| CN117407802A (en) | Runoff prediction method based on improved depth forest model | |
| CN115963407A (en) | ICGWO (intensive care unit) optimization ELM (element-based robust model) based lithium battery SOC estimation method | |
| CN113239547B (en) | Core backbone network frame construction method based on TLBO algorithm | |
| CN114564787A (en) | Bayesian optimization method, device and storage medium for target-related airfoil design | |
| CN117993110A (en) | Novel vehicle chassis truss lightweight optimization design system | |
| CN113449912B (en) | Space load situation sensing method based on artificial intelligence technology | |
| CN115796327A (en) | Wind power interval prediction method based on VMD (vertical vector decomposition) and IWOA-F-GRU (empirical mode decomposition) -based models | |
| CN112464541B (en) | Multi-scale uncertainty considered mixed composite material layering method | |
| CN114202063A (en) | Fuzzy neural network greenhouse temperature prediction method based on genetic algorithm optimization | |
| CN114548400A (en) | Rapid flexible full-pure embedded neural network wide area optimization training method |
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 |