CN101567020A - Optimization method for mixture ratio of components of working layer of metal plastic composite material - Google Patents
Optimization method for mixture ratio of components of working layer of metal plastic composite material Download PDFInfo
- Publication number
- CN101567020A CN101567020A CNA2009100313344A CN200910031334A CN101567020A CN 101567020 A CN101567020 A CN 101567020A CN A2009100313344 A CNA2009100313344 A CN A2009100313344A CN 200910031334 A CN200910031334 A CN 200910031334A CN 101567020 A CN101567020 A CN 101567020A
- Authority
- CN
- China
- Prior art keywords
- parameter
- mixture ratio
- proportioning
- working lining
- parameters
- 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.)
- Granted
Links
Images
Landscapes
- Reinforced Plastic Materials (AREA)
Abstract
The invention discloses an optimization method for a mixture ratio of components of a working layer of a metal plastic composite material, which comprises the following steps: obtaining mixture ratio parameters of working layers of polyphenyl thioether and nylon 66 and carbon fibers, and material performance parameters of the vibration reduction and the firmness of composite materials, which correspond to the mixture ratio parameters by a plurality of groups of orthogonal tests; training self-adapting nerve fuzzy reasoning systems and respectively establishing the global map relation; establishing and correcting a system model by three-dimensional analysis software to obtain various parameters; taking optimal performances with 50 percent of the vibration reduction and 50 percent of the firmness as a target; copying, crossing and varying the mixture ratio parameters by a genetic algorithm to obtain mixture ratio parameter values corresponding to the optimal performances; and comparing the mixture ratio parameter values obtained through the optimization by the genetic algorithm and the performance parameters corresponding to the mixture ratio parameter values with the mixture ratio parameters corresponding to the same performance parameters and corrected by the three-dimensional analysis software. The optimization method not only ensures the precision, but also improves the efficiency, and substantially improves the vibration reduction and the firmness of the metal plastic composite material.
Description
Technical field
The present invention relates to the method that draws of compound substance work composition of layer optimal proportion, specifically is a kind of method that draws that is aided with the working layer of metal plastic composite material optimal proportion of finite element analysis based on adaptive neural network fuzzy reasoning genetic algorithm.
Background technology
Metallic composite is a kind of of multifunctional composite, developed into a new branch of compound substance, be that a class is matrix with the metal or alloy, with metal or non-metal wire, silk, fiber, whisker or particulate component is the heterogeneous body potpourri of wild phase, and its common ground is to have continuous metallic matrix.And plastimets is made up of metallic matrix and plastic layer, metallic matrix is mainly played a supporting role, plastic layer is made up of middle layer and working lining, the connection effect is played in the middle layer, working lining plays vibration damping, effect such as wear-resistant and corrosion-resistant, and the proportioning of components of working layer has directly determined the quality of compound substance overall performance.At present, for the optimization of components of working layer of metal plastic composite material proportioning, because the principal element that relates to has some, study prioritization scheme with the method that enumeration method is manually tested, workload is too big, and inaccurate.
Nerual network technique development in recent years is used widely in fields such as economic, military affairs, commercial production and biomedicines, and has been produced profound influence rapidly.Neural network has the ability of very strong self-adaptation, self-organization, self study and the ability of large-scale parallel computing.But in actual applications, neural network has also exposed some self inherent shortcoming: the initialization of weights is at random, and local minimum, convergence time is long and poor robustness etc.
Genetic algorithm provides a kind of general framework of finding the solution the complication system optimization problem, and it does not rely on the specific field of problem, the kind of problem is had very strong robustness, so be widely used in a lot of subjects.The main application fields of genetic algorithm has: function optimization, Combinatorial Optimization, production scheduling problems, control automatically, robotics, Flame Image Process, artificial life, genetic programming, machine learning and data mining etc.
Finite element method is exactly a kind of computer modeling technique, makes people need not a thing with the generating process of an engineering problem of software simulation on computers and really works it out.The benefit that this technology is brought is exactly, the product that just can allow people observe on computers to design in the drawing design phase in use any problem may occur in the future, need not work it out what problem can appear in check in experiment to model machine, can effectively reduce the cost of product development, shorten the cycle of product design, thereby become effective engineering analysis means.
Summary of the invention
The objective of the invention is to cause inaccuracy and inefficient deficiency, provide a kind of efficient height, accuracy height, computing velocity is fast and antijamming capability is strong is aided with the optimization method of the components of working layer of metal plastic composite material proportioning of finite element analysis based on orthogonal test and adaptive neural network fuzzy reasoning genetic algorithm for overcoming in the prior art in the proportioning of only carrying out by rule of thumb that exists aspect the material composition proportioning.
The technical solution used in the present invention is: the material of this working lining is made up of polyphenylene sulfide, nylon 66 and carbon fiber, and the working lining proportioning parameter that draws polyphenylene sulfide, nylon 66 and these three kinds of materials of carbon fiber by some groups of orthogonal tests reaches the vibration damping of corresponding with it compound substance and the material property parameter of fastness; Adaptive Neuro-fuzzy Inference is trained and sets up respectively global map relation between each related work layer proportioning parameter input and the material property parameter output, the Adaptive Neuro-fuzzy Inference model of acquisition composite property with the working lining proportioning parameter of gained and material property parameter; Utilize three dimensional analysis software to set up this system model simultaneously, and each parameter in this system model is revised, obtain working lining proportioning parameter and the material property parameter corresponding with it according to gained working lining proportioning parameter and material property parameter; Respectively account for 50% optimal performance as target with composite material vibration damping and fastness, utilize genetic algorithm to the working lining proportioning parameter of described three kinds of materials duplicate, intersection and mutation operation, draw and the pairing working lining proportioning of compound substance optimal performance parameter value, simultaneously constantly to the optimal performance of seeking material through the system model input service layer proportioning parameter of revising; Learn from else's experience the working lining proportioning parameter value of genetic algorithm optimization gained and corresponding material property parameter and are compared with the corresponding working lining proportioning parameter of the identical materials performance parameter of three dimensional analysis software correction, if numerical value is close, then optimize end; If differ bigger, then optimize again.
The present invention is aided with finite element analysis by adaptive neural network fuzzy reasoning genetic algorithm and is optimized, and has both guaranteed precision, has improved efficient again, and the vibration damping of plastimets and fastness are significantly improved.This optimization method of the present invention has very important reference to the structure proportioning of optimizing various compound substances.
Description of drawings
Below in conjunction with the drawings and specific embodiments the present invention is described in further detail.
Fig. 1 is that the carbon fiber content that orthogonal test of the present invention draws is 5% o'clock, and content and the corresponding material structure thereof of nylon 66 in remaining material bears maximum impact and try hard to;
Fig. 2 is that the carbon fiber content that orthogonal test of the present invention draws is 5% o'clock, content and corresponding material the amplitude image under identical vibration source effect thereof of nylon 66 in remaining material;
Fig. 3 is that the carbon fiber content that orthogonal test of the present invention draws is 10% o'clock, and content and the corresponding material structure thereof of nylon 66 in remaining material bears maximum impact and try hard to;
Fig. 4 is that the carbon fiber content that orthogonal test of the present invention draws is 10% o'clock, content and corresponding material the amplitude image under identical vibration source effect thereof of nylon 66 in remaining material;
Fig. 5 is that the carbon fiber content that orthogonal test of the present invention draws is 15% o'clock, and content and the corresponding material structure thereof of nylon 66 in remaining material bears maximum impact and try hard to;
Fig. 6 is that the carbon fiber content that orthogonal test of the present invention draws is 15% o'clock, content and corresponding material the amplitude image under identical vibration source effect thereof of nylon 66 in remaining material;
Fig. 7 is the neural network structure with the fuzzy neural system equivalence;
Fig. 8 is the web materials performance map in the makeover process;
Fig. 9 is the hereditary optimizing figure of coupling adaptive neural fuzzy inference system model.
Embodiment
Working layer of metal plastic composite material is made up of three kinds of materials: polyphenylene sulfide, nylon 66 and carbon fiber.Under the constant situation of the shape that guarantees metallic matrix, middle layer and working lining and thickness, draw the vibration damping of the working lining proportioning parameter of polyphenylene sulfide, nylon 66 and these three kinds of materials of carbon fiber and corresponding with it compound substance and the material property parameter of fastness by some groups of orthogonal tests, by the arrange work orthogonal test of composition of layer proportioning of high efficiency principle.The present invention is an example with compound substance fastness and vibration damping sexual intercourse value, does a small amount of test, draws some groups of components of working layer proportionings and reaches the compound substance fastness corresponding with it and the relation value of vibration damping; The present invention arranges six big group tests altogether, and in these six groups tests, the number percent of carbon fiber is respectively 5%, 5%, 10%, 10%, 15%, 15%.In first group of test, the content of carbon fiber is 5%, the number percent that nylon 66 accounts for all the other two kinds of materials improves constantly from 5%, draws the structure corresponding with it at the maximum impact force of not throwing off between the genetic horizon or plastics split etc. and can bear under the situation of phenomenon, as shown in Figure 1.In second group of test, the content of carbon fiber is 5%, the number percent that nylon 66 accounts for all the other two kinds of materials improves constantly from 5%, draws the structure corresponding with it and is not throwing off between the genetic horizon or the plastics amplitude of material under the situation of phenomenon such as split, as shown in Figure 2.In the 3rd group of test, the content of carbon fiber is 10%, the number percent that nylon 66 accounts for all the other two kinds of materials improves constantly from 5%, draws the structure corresponding with it at the maximum impact force of not throwing off between the genetic horizon or plastics split etc. and can bear under the situation of phenomenon, as shown in Figure 3.In the 4th group of test, the content of carbon fiber is 10%, the number percent that nylon 66 accounts for all the other two kinds of materials improves constantly from 5%, draws the structure corresponding with it and is not throwing off between the genetic horizon or the plastics amplitude of material under the situation of phenomenon such as split, as shown in Figure 4.In the 5th group of test, the content of carbon fiber is 15%, the number percent that nylon 66 accounts for all the other two kinds of materials improves constantly from 5%, draws the structure corresponding with it at the maximum impact force of not throwing off between the genetic horizon or plastics split etc. and can bear under the situation of phenomenon, as shown in Figure 5.In the 6th group of test, the content of carbon fiber is 15%, the number percent that nylon 66 accounts for all the other two kinds of materials improves constantly from 5%, draws the structure corresponding with it and is not throwing off between the genetic horizon or the plastics amplitude of material under the situation of phenomenon such as split, as shown in Figure 6.
Use the step and with the data that draw the adaptive fuzzy nervous system is trained and set up the proportioning parameter input of each relevant aforementioned three kinds of material and the global map relation between the composite property output, obtain the Adaptive Neuro-fuzzy Inference model of composite property, utilize three dimensional analysis software (ABAQUS software) to set up composite material model simultaneously, and according to gained data of last step each parameter in the model is revised, obtain working lining proportioning parameter and the material property parameter corresponding with it; The Adaptive Neuro-fuzzy Inference model is imported genetic algorithm,, draw optimum composition proportion parameter, constantly to model input service layer proportioning parameter, seek the optimum performance of compound substance simultaneously through revising by evolving; The learn from else's experience vibration damping of the working lining proportioning parameter value of genetic algorithm optimization gained and corresponding compound substance, firm performance parameter, with compare with the corresponding working lining proportioning parameter of identical vibration damping, the fastness parameter of three dimensional analysis software optimization, if numerical value is close, then optimizes and finish; If differ bigger, then analyze reason, optimize again.
As Fig. 7, have a kind of fuzzy inference system (being that the T2S pattern is stuck with paste inference system) of m the single output of input to be stated as:
Ri：If?X?is?A
i?Then?y
i＝θ
i T.X，i＝1，2，…，n (1)
In the formula: Ri represents the i bar rule in the fuzzy inference system rule base;
X=(x
1, x
2..., x
m)
T∈ R
mInput for system;
Ai=(A
1i, A
2i..., A
Mi) be the semantic variant vector;
y
iBe part output;
X=(1, x
1..., x
m)
T∈ R
M+1Be the augmentation input;
θ
i=(θ
0i, θ
1i..., θ
Mi)
T∈ R
M+1Be the conclusion parameter vector;
N is a fuzzy inference rule collection number.
For the input X of system
0, have:
In the formula, W
i(X)=T[A
1i(x
1), A
2i(x
2) ..., A
Mi(x
m)], being the excitation intensity or the matching degree of corresponding i bar rule, T gets product calculation for the T-modular operator (being the fuzzy inference system operator) corresponding to " AND " (be logic and); A
Ki(x
k) be i linguistic variable subordinate function of k input component.In the said method, each prerequisite parameter (prerequisite proportioning and prerequisite parameter) determine to have very big subjectivity, and lack adaptive ability.For this reason, come this fuzzy system of equal value, so just can improve with neural algorithm with feedforward neural network structure shown in Figure 7.To one group of input and output mode (sample) (X
i, Y
i), i=1 ..., p takes error to return algorithm for inversion or comprehensively adopts error to return algorithm for inversion and least square method, seeks optimum set of parameters Ψ
OptSatisfy:
In the formula: λ
i〉=0 is weighting coefficient, y (X
i) be corresponding input X
iOutput.
Number percent (difference X with the polyphenylene sulfide in the working lining, nylon 66, carbon fiber
1, X
2..., X
n) be input, be output (Y with the maximum impact force that can bear and the amplitude of material
i), the adaptive fuzzy nervous system is trained, set up the Adaptive Neuro-fuzzy Inference model of composite property.
Set up the compound substance system model with three dimensional analysis software, carry out finite element analysis, and each parameter in the system model is revised according to gained data of last step, obtain working lining proportioning parameter and the material property parameter corresponding with it, Fig. 8 is the web materials performance map in the makeover process, compound substance is apart from the performance parameter of the each point of center of circle different distance as seen from Figure 8, the input of a kind of parameter just corresponding a width of cloth and Fig. 8 similarly scheme, like this, continuous correction along with parameter, just can from these figure, see the variation tendency of material property, and can effectively must seek the material optimal performance by the scope of dwindling parameter with parameter.
With the compound substance optimal performance as target, be that vibration damping and fastness respectively account for 50% optimal performance as target, utilize genetic algorithm the working lining proportioning parameter of three kinds of materials to be duplicated, intersected and operation such as variation, finally draw each working lining proportioning parameter value corresponding with the compound substance optimal performance, simultaneously, constantly to the optimal performance of seeking material through the system model input service layer proportioning parameter of revising.
The mode of optimization aim with the appropriateness value is embedded in the genetic algorithm, directly the expressed relation of above-mentioned Adaptive Neuro-fuzzy Inference model carried out optimizing.As shown in Figure 9, after receiving initiation command, the large quantities of parameters of material mixture ratio that computing machine will indicate with gene expression are included into initial population, introduce adaptive neural network structural model (being called for short " ANFIS model ") then, draw the material property corresponding, material property is estimated, if result's convergence then is " Y " with these parameters, finish immediately, the parameter of this moment is qualified parameter; Otherwise be " N ", the step that enters selection, duplicates, intersects and make a variation, and then enter " adaptive neural network structural model ", carry out successively, up to end.It is 30 that population number is set like this, hybrid rate 0.1, and aberration rate 0.01, each input variable is encoded with 8 binary strings, and evolution generation number n is 80, the qualified parameter after finally being optimized.
Learn from else's experience the working lining proportioning parameter value of genetic algorithm optimization gained and the vibration damping of corresponding compound substance, the performance parameter of fastness, the working lining proportioning parameter corresponding with the performance parameter of the vibration damping of aforesaid identical materials with the correction of three dimensional analysis software, fastness compared, if numerical value is close, then optimizes and finish; If differ bigger, then analyze reason, optimize again.
Claims (1)
1. the optimization method of a components of working layer of metal plastic composite material proportioning, the material of this working lining is made up of polyphenylene sulfide, nylon 66 and carbon fiber, it is characterized in that may further comprise the steps:
A. draw the vibration damping of the working lining proportioning parameter of polyphenylene sulfide, nylon 66 and these three kinds of materials of carbon fiber and corresponding with it compound substance and the material property parameter of fastness by some groups of orthogonal tests;
B. with the working lining proportioning parameter of step a gained and material property parameter Adaptive Neuro-fuzzy Inference is trained and sets up respectively global map relation between each related work layer proportioning parameter input and the material property parameter output, the Adaptive Neuro-fuzzy Inference model of acquisition composite property; Utilize three dimensional analysis software to set up this system model simultaneously, and each parameter in this system model is revised, obtain working lining proportioning parameter and the material property parameter corresponding with it according to step a gained working lining proportioning parameter and material property parameter;
C. respectively account for 50% optimal performance as target with composite material vibration damping and fastness, utilize genetic algorithm to the working lining proportioning parameter of described three kinds of materials duplicate, intersection and mutation operation, draw and the pairing working lining proportioning of compound substance optimal performance parameter value, simultaneously constantly to the optimal performance of seeking material through the system model input service layer proportioning parameter of revising;
The working lining proportioning parameter value of the step of d. learning from else's experience c genetic algorithm optimization gained and corresponding material property parameter, compare with the corresponding working lining proportioning parameter of the identical materials performance parameter of three dimensional analysis software correction with step b, if numerical value is close, then optimizes and finish; If differ bigger, then optimize again.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009100313344A CN101567020B (en) | 2009-05-08 | 2009-05-08 | Optimization method for mixture ratio of components of working layer of metal plastic composite material |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009100313344A CN101567020B (en) | 2009-05-08 | 2009-05-08 | Optimization method for mixture ratio of components of working layer of metal plastic composite material |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101567020A true CN101567020A (en) | 2009-10-28 |
| CN101567020B CN101567020B (en) | 2011-06-01 |
Family
ID=41283169
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2009100313344A Active CN101567020B (en) | 2009-05-08 | 2009-05-08 | Optimization method for mixture ratio of components of working layer of metal plastic composite material |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101567020B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663151A (en) * | 2012-03-05 | 2012-09-12 | 西安交通大学 | Nuclear radiation shielding material optimization design method |
| CN106624166A (en) * | 2016-12-27 | 2017-05-10 | 沈阳航空航天大学 | Optimization method for CFRP (carbon fiber reinforced plastics) and titanium alloy laminated structure reaming process |
| CN109558664A (en) * | 2018-11-22 | 2019-04-02 | 广东工业大学 | A kind of compound material formula formulating method of injection molding manufacture |
| CN110765546A (en) * | 2018-07-09 | 2020-02-07 | 上海汽车集团股份有限公司 | Method and device for calculating performance of composite material and electronic equipment |
| CN114496125A (en) * | 2022-02-10 | 2022-05-13 | 西南交通大学 | Preparation method, device and equipment of similar material and readable storage medium |
-
2009
- 2009-05-08 CN CN2009100313344A patent/CN101567020B/en active Active
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663151A (en) * | 2012-03-05 | 2012-09-12 | 西安交通大学 | Nuclear radiation shielding material optimization design method |
| CN106624166A (en) * | 2016-12-27 | 2017-05-10 | 沈阳航空航天大学 | Optimization method for CFRP (carbon fiber reinforced plastics) and titanium alloy laminated structure reaming process |
| CN110765546A (en) * | 2018-07-09 | 2020-02-07 | 上海汽车集团股份有限公司 | Method and device for calculating performance of composite material and electronic equipment |
| CN109558664A (en) * | 2018-11-22 | 2019-04-02 | 广东工业大学 | A kind of compound material formula formulating method of injection molding manufacture |
| CN109558664B (en) * | 2018-11-22 | 2023-04-18 | 广东工业大学 | Method for formulating composite material manufactured by injection molding |
| CN114496125A (en) * | 2022-02-10 | 2022-05-13 | 西南交通大学 | Preparation method, device and equipment of similar material and readable storage medium |
| CN114496125B (en) * | 2022-02-10 | 2023-05-02 | 西南交通大学 | Preparation method, device and equipment of similar materials and readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101567020B (en) | 2011-06-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101567020B (en) | Optimization method for mixture ratio of components of working layer of metal plastic composite material | |
| Ursem et al. | Analysis and modeling of control tasks in dynamic systems | |
| CN101533423A (en) | Method for optimizing structure of metallic-plastic composite material | |
| CN103729694B (en) | The method that improvement GA based on polychromatic sets hierarchical structure solves Flexible workshop scheduling | |
| Kubalík et al. | Symbolic regression methods for reinforcement learning | |
| Liu et al. | Modeling and simulation of robot inverse dynamics using LSTM-based deep learning algorithm for smart cities and factories | |
| CN114186791A (en) | Multi-model small-batch-oriented dynamic scheduling method for debugging production of complex equipment products | |
| Protalinskiy et al. | Strategic decision support in the process of manufacturing systems management | |
| Farazmand et al. | Multi-objective optimization of time-cost-quality-carbon dioxide emission-plan robustness in construction projects | |
| Zhou et al. | Advances in teaching-learning-based optimization algorithm: A comprehensive survey | |
| Sun et al. | Evaluation on Symbiotic Performance of Regional Technological Entrepreneurship Ecosystem | |
| Nelles et al. | Entrepreneurial by design: theorizing the entrepreneurial transformation of contemporary universities | |
| Shafiq et al. | Decisional DNA based framework for representing Virtual Engineering Objects | |
| Xie et al. | Estimating a civil aircraft’s development cost with a GM (1, N) model and an MLP neural network | |
| Liu et al. | Mixed-flow assembly line balancing with uncertain assembly times in remanufacturing | |
| Hanafy et al. | Recent trends in soft computing techniques for solving real time engineering problems | |
| Zhao et al. | GA-based resource leveling optimization for construction project | |
| Kang | Role and Policies of STP in the Era of 4 th Industrial Revolution from Triple Helix Viewpoint | |
| Dawal et al. | Multi-attribute decision-making for CNC machine tool selection in FMC based on the integration of the improved consistent fuzzy AHP and TOPSIS | |
| Growiec et al. | Hardware and Software over the Course of Long-Run Growth: Theory and Evidence | |
| Cai et al. | Optimization method of machining parameters based on intelligent algorithm | |
| Kundapur et al. | Simulating knowledge worker adoption rate of KMS: An organizational perspective | |
| Cheng et al. | Ceramic Tile Production Intelligent Decision Research Based on Reinforcement Learning Algorithm | |
| Wang et al. | Research on the Integration Strategy of Traditional Advantage Engineering Specialty and Artificial Intelligence Based on TRIZ | |
| Mosalam | Transdisciplinary, expediency, and rigor of CACAIE. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20170307 Address after: 212300 Zhenjiang city of Jiangsu province Danyang City Yanling Town East Village Patentee after: Jiangsu Tong Tong Machinery Manufacturing Co., Ltd. Address before: 212013 Zhenjiang City, Jiangsu Province University Road, No. 301 Patentee before: Jiangsu University |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20180205 Address after: 212300 Zhenjiang city of Jiangsu province Danyang city Beizhen Hou Xiang Dan Xi Huan Lu Qian Xiang Wu Xing Patentee after: Danyang Yongfeng hardware and Electronic Co., Ltd. Address before: 212300 Zhenjiang city of Jiangsu province Danyang City Yanling Town East Village Patentee before: Jiangsu Tong Tong Machinery Manufacturing Co., Ltd. |
|
| TR01 | Transfer of patent right |