CN109408941A - Flight vehicle aerodynamic optimization method based on data mining and genetic algorithm - Google Patents
Flight vehicle aerodynamic optimization method based on data mining and genetic algorithm Download PDFInfo
- Publication number
- CN109408941A CN109408941A CN201811216160.4A CN201811216160A CN109408941A CN 109408941 A CN109408941 A CN 109408941A CN 201811216160 A CN201811216160 A CN 201811216160A CN 109408941 A CN109408941 A CN 109408941A
- Authority
- CN
- China
- Prior art keywords
- genetic algorithm
- optimization
- flight vehicle
- data mining
- optimality
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- 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
-
- 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
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Abstract
The invention discloses a kind of flight vehicle aerodynamic optimization method based on data mining and genetic algorithm, convergence rate is slow when for solving the problems, such as to carry out aerodynamic optimization using traditional genetic algorithm, and this method carries out the stage without manual intervention in final optimization pass, it is able to achieve automatic Iterative calculating, improves optimization efficiency.The method that semiempirical estimation is first passed through headed by technical solution obtains design sample library, then data mining is carried out to sample database, the principle of optimality of high confidence level is obtained using clustering, variance analysis and decision tree analysis, these rules are fused in genetic algorithm as priori knowledge again, specific manifestation is that crossover rule, variation rule are set dynamically according to priori rules, improved genetic algorithm will finally be merged and be used for the aerodynamic optimization based on high-precision fluid emulation, obtain excellent design parameter.Compared to traditional optimization method based on genetic algorithm, the present invention substantially increases convergence rate, has very big engineering value to flight vehicle aerodynamic optimization.
Description
Technical field
The invention belongs to aircraft field of engineering technology, in particular to a kind of flight based on data mining and genetic algorithm
Device Aerodynamic optimization method.
Background technique
Aerodynamic optimization refers to the shape of each main component of aircraft and the design of relative position, needs meeting given pact
Under the conditions of beam, the optimal design scheme of aeroperformance is obtained.Genetic algorithm is a kind of common Aerodynamic optimization method, is had preferable
Global optimization ability, suitable for complicated multiextremal optimization problem, but convergence rate is more slow, especially with it is time-consuming
When high-precision fluid emulation combines progress aerodynamic optimization, which is more highlighted.
There are many derived type of genetic algorithm, compared to original genetic algorithm in terms of optimizing ability and convergence rate
It has a certain upgrade, but convergence rate is still not satisfactory enough, the mainly not addition and benefit of priori knowledge of tracing it to its cause
With since the performance of initial population has a great impact to the performance of genetic algorithm, so commonly accelerating convergence speed in engineering
The method of degree is that population artificially is added as priori knowledge in performance preferably individual, although this method is effectively, to higher-dimension
For the optimization problem of degree, it is also more difficult to obtain performance preferably individual itself, and the operation for artificially modifying population can be big
The efficiency of entire optimization process is reduced greatly.To sum up, existing genetic algorithm sufficiently can not extract and utilize aerodynamic optimization to lead
Priori knowledge in domain, so that the convergence rate of optimization process is not ideal enough.
Summary of the invention
Priori knowledge cannot be made full use of in order to solve existing genetic algorithm in the utilization of aerodynamic optimization, lead to convergence speed
Slow problem is spent, the invention proposes a kind of flight vehicle aerodynamic optimization method based on data mining and genetic algorithm, and the party
Method carries out the stage without manual intervention in final optimization pass, is able to achieve automatic Iterative calculating, improves optimization efficiency.This method is directed to
Given aerodynamic optimization problem extracts the principle of optimality as first by the method for data mining first from semiempirical evaluation method
Knowledge is tested, then these principles of optimality are dissolved into genetic algorithm and are used, is embodied in crossover rule, heredity rule
Dynamic setting, with the help of priori knowledge, the convergence rate of optimization algorithm is greatly speeded up, and due to semiempirical evaluation method
Calculating speed be significantly faster than high-precision fluid emulation, so data mining link spend time can ignore substantially, to base
There is very big engineering value in the aerodynamic optimization of high-precision fluid emulation.
The present invention solves technical solution used by flight vehicle aerodynamic optimization problem: one kind is calculated based on data mining and heredity
The flight vehicle aerodynamic optimization method of method, feature the following steps are included:
Step 1 establishes the parametric method of flight vehicle aerodynamic shape, and the parameter after Selecting All Parameters is as design variable.
Parametric method is the geometric parameter that can determine that flight vehicle aerodynamic shape to be chosen, with aircraft wings according to given optimization problem
For the optimization of face, including the span, wing tip chord length, wing root chord length, aerofoil leading-edge sweep angle etc.；
Step 2 is sampled in design space by the method for experimental design for given aerodynamic optimization problem, makes sample
This is sufficiently distributed in design space, and the aeroperformance of these samples is then calculated using the evaluation method with high confidence level,
To establish estimation sample database；
Step 3 is carried out data mining to estimation sample database, is obtained different designs parameter pair using variance analysis first
Sample data discretization is then obtained the cluster data of estimation sample using the method for clustering by the influence of design result
Then library is handled cluster data library using traditional decision-tree, the principle of optimality with high confidence level is extracted, and forms optimization
Rule base；
Step 4 merges the principle of optimality that previous step is extracted with genetic algorithm, for instructing the friendship of genetic process
Fork and variation link, be specifically divided into two aspect, first is that according to individual state and matching rule dynamic setting intersect, variation it is general
Rate, second is that being dynamically selected the mode of individual variation according to the matched principle of optimality；
Above-mentioned fused genetic algorithm is applied to the aerodynamic optimization based on high-precision fluid emulation, passed through by step 5
Iterative calculation obtains optimal design parameter.
The features of the present invention and beneficial effect are:
1, for given aerodynamic optimization problem, elder generation is extracted from semiempirical evaluation method by the means of data mining
Test knowledge；
2, by the way that intersection, the variation rule of genetic algorithm is set dynamically, the priori knowledge of extraction is fused to genetic algorithm
In；
3, aerodynamic optimization is applied to fused genetic algorithm, it can be big while guaranteeing original global optimizing ability
It is big to improve convergence rate, and manual intervention is eliminated in the optimizing phase, the efficiency of Optimizing Flow is substantially increased, is realized quick
Obtain the purpose of high-precision optimal design parameter.
Detailed description of the invention
Fig. 1 is the Aerodynamic optimization method schematic diagram based on data mining and genetic algorithm
Fig. 2 is to obtain sample database and carry out the flow chart that the principle of optimality is extracted in data mining to it
Fig. 3 is the genetic algorithm flow chart incorporated after the principle of optimality
Fig. 4 is a kind of hereditary variation method flow diagram for merging the principle of optimality
Specific embodiment
Using embodiment, the present invention will be further described below, and software described herein, file format and platform are used
To provide a further understanding of the present invention, but therefore by protection scope of the present invention the range that embodiment describes is not limited in it
Among.
Choosing guided missile first is that object optimizes the aerodynamic configuration of its missile wing, and optimization aim is to keep aerodynamic center
Lift resistance ratio is improved under the premise of being basically unchanged, with distance, spanwise length, wing tip chord length, the wing of the wing root leading edge at the top of bullet
Root chord length and aerofoil front and rear edge sweepback angle optimize for design parameter.
For above-mentioned missile aerodynamic optimization problem, sampled in design space using Latin Hypercube Sampling method,
So that the design parameter of sample is evenly distributed in as far as possible in entire design space, software then is estimated using semiempirical
Missile DATCOM calculates the aeroperformance of these samples, obtains sample database.
Subsequent decision Tree algorithms carry out calculation processing for convenience, first to use K-means clustering method by database
In sample discretization, obtain each design parameter to the influence degree of aeroperformance result, so using ANOVA method of analysis of variance
Afterwards according to rough set theory, advised using the optimization that can promote aeroperformance that decision tree C4.5 method extracts 90% or more confidence level
Then, formation rule library.
Then the principle of optimality of extraction is applied in genetic algorithm, specific method is will be at individual clusters all in population
Then reason compares one by one with the rule in rule base, find out the highest rule of matching degree as alternation rule, the calculating of matching degree
It is shown below:
I is individual in formula, and R is rule, wjFor j-th of design parameter weight that ANOVA is analyzed, ijAnd rjRespectively
J-th of design parameter of individual and rule,WithThe respectively mean value of individual and regular design parameter.
After obtaining matching rule, applied to the intersection and variation of individual, a kind of relatively simple and effective mode is
Traditional interleaved mode is still used in crossover process, only the mutation process of individual is set dynamically, when individual performance
When more excellent, multinomial variation is carried out using free variation mode without restriction, such as by probability, when individual performance is bad,
Then design parameter is allowed targetedly to be made a variation according to matching rule by probability, the improvement variation method such as Fig. 4 used herein
It is shown.
Targetedly variation is carried out according to matching rule to be shown below:
C is a random number being evenly distributed in section [0.9,1.1] in formula,For j-th of design in matching rule
The cluster centre value of parameter.Finally the above-mentioned genetic algorithm for having merged optimization design rule is mutually tied with high-precision fluid emulation
It closes, is iterated calculating until convergence, obtains optimal design parameter value.
Above-described specific embodiment has carried out further the purpose of the present invention, technical scheme and beneficial effects
Be described in detail, but be not intended to limit the scope of protection of the present invention, all within the spirits and principles of the present invention, done based on
Equivalent transformation, equivalent replacement and obvious change of the invention etc., are all included in the scope of protection of the present invention.
Claims (1)
1. a kind of flight vehicle aerodynamic optimization method based on data mining and genetic algorithm, it is characterised in that the following steps are included:
Step 1 establishes the parametric method of flight vehicle aerodynamic shape, and the parameter after Selecting All Parameters is as design variable.Parameter
Change method is to choose the geometric parameter that can determine that flight vehicle aerodynamic shape according to given optimization problem, excellent with aircraft aerofoil
Turn to example, including the span, wing tip chord length, wing root chord length, aerofoil leading-edge sweep angle etc.；
Step 2 is sampled in design space by the method for experimental design for given aerodynamic optimization problem, fills sample
It is distributed in design space, the aeroperformance of these samples is then calculated using the evaluation method with high confidence level, thus
Establish estimation sample database；
Step 3 is carried out data mining to estimation sample database, is obtained different designs parameter using variance analysis first to design
Sample data discretization is then obtained the cluster data library of estimation sample, so using the method for clustering by influence as a result
Cluster data library is handled using traditional decision-tree afterwards, extracts the principle of optimality with high confidence level, forms the principle of optimality
Library；
Step 4 merges the principle of optimality that previous step is extracted with genetic algorithm, for instruct the intersection of genetic process with
Make a variation link, is specifically divided into two aspects, first is that intersected according to individual state and matching rule dynamic setting, the probability of variation,
Second is that being dynamically selected the mode of individual variation according to the matched principle of optimality；
Above-mentioned fused genetic algorithm is applied to the aerodynamic optimization based on high-precision fluid emulation, passes through iteration by step 5
It calculates and obtains optimal design parameter.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811216160.4A CN109408941B (en) | 2018-10-18 | 2018-10-18 | Aircraft pneumatic optimization method based on data mining and genetic algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811216160.4A CN109408941B (en) | 2018-10-18 | 2018-10-18 | Aircraft pneumatic optimization method based on data mining and genetic algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109408941A true CN109408941A (en) | 2019-03-01 |
| CN109408941B CN109408941B (en) | 2022-06-03 |
Family
ID=65468549
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811216160.4A Active CN109408941B (en) | 2018-10-18 | 2018-10-18 | Aircraft pneumatic optimization method based on data mining and genetic algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109408941B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110008639A (en) * | 2019-04-24 | 2019-07-12 | 东莞理工学院 | A kind of micro flapping wing air vehicle wing intelligent parameter design method |
| CN110134007A (en) * | 2019-05-22 | 2019-08-16 | 南昌航空大学 | Multiple no-manned plane cooperates with target assignment method |
| CN110175373A (en) * | 2019-05-10 | 2019-08-27 | 河北工业大学 | Motor optimized design method and system |
| CN111967090A (en) * | 2020-08-04 | 2020-11-20 | 中国空气动力研究与发展中心计算空气动力研究所 | Dynamic improvement method for optimizing design space |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103440377A (en) * | 2013-08-27 | 2013-12-11 | 北京航空航天大学 | Aircraft aerodynamic configuration optimum design method based on improved parallel DE algorithm |
| US9141756B1 (en) * | 2010-07-20 | 2015-09-22 | University Of Southern California | Multi-scale complex systems transdisciplinary analysis of response to therapy |
| CN107480335A (en) * | 2017-07-12 | 2017-12-15 | 南京航空航天大学 | A kind of hypersonic vehicle Iterative Design method |
-
2018
- 2018-10-18 CN CN201811216160.4A patent/CN109408941B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9141756B1 (en) * | 2010-07-20 | 2015-09-22 | University Of Southern California | Multi-scale complex systems transdisciplinary analysis of response to therapy |
| CN103440377A (en) * | 2013-08-27 | 2013-12-11 | 北京航空航天大学 | Aircraft aerodynamic configuration optimum design method based on improved parallel DE algorithm |
| CN107480335A (en) * | 2017-07-12 | 2017-12-15 | 南京航空航天大学 | A kind of hypersonic vehicle Iterative Design method |
Non-Patent Citations (4)
| Title |
|---|
| 彭鑫 等: "遗传算法在机翼气动外形优化中的应用研究", 《中国自动化学会系统仿真专业委员会 会议论文集》 * |
| 徐波: "遗传算法及其在数据挖掘中的应用", 《电脑编程技巧与维护》 * |
| 李迅等: "基于分段进化的遗传算法在机翼气动外形设计上的应用", 《力学季刊》 * |
| 许平等: "基于遗传算法及Hicks-Henne型函数的层流翼型优化设计", 《空军工程大学学报(自然科学版)》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110008639A (en) * | 2019-04-24 | 2019-07-12 | 东莞理工学院 | A kind of micro flapping wing air vehicle wing intelligent parameter design method |
| CN110008639B (en) * | 2019-04-24 | 2020-10-27 | 东莞理工学院 | Intelligent parametric design method for wings of miniature flapping-wing aircraft |
| CN110175373A (en) * | 2019-05-10 | 2019-08-27 | 河北工业大学 | Motor optimized design method and system |
| CN110134007A (en) * | 2019-05-22 | 2019-08-16 | 南昌航空大学 | Multiple no-manned plane cooperates with target assignment method |
| CN111967090A (en) * | 2020-08-04 | 2020-11-20 | 中国空气动力研究与发展中心计算空气动力研究所 | Dynamic improvement method for optimizing design space |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109408941B (en) | 2022-06-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109408941A (en) | Flight vehicle aerodynamic optimization method based on data mining and genetic algorithm | |
| CN109460566B (en) | Aerodynamic robust optimization design method for thick airfoil section on inner side of wind turbine blade | |
| CN102682172B (en) | Numerous-parameter optimization design method based on parameter classification for supercritical aerofoil | |
| CN104834772B (en) | Aircraft wing based on artificial neural network/wing inverse design method | |
| CN110334449A (en) | A kind of aerofoil profile Fast design method based on online agent model algorithm | |
| CN107480335A (en) | A kind of hypersonic vehicle Iterative Design method | |
| CN108459505B (en) | Unconventional layout aircraft rapid modeling method suitable for control iterative design | |
| CN105718634A (en) | Airfoil robust optimization design method based on non-probability interval analysis model | |
| CN106772387A (en) | A kind of wind shear recognition methods | |
| CN112464366B (en) | Multi-fidelity shape optimization method of autonomous underwater vehicle based on data mining | |
| CN106126860A (en) | A kind of hypersonic wing Robust Optimal Design considering mismachining tolerance | |
| CN110516318A (en) | Airfoil Design method based on radial basis function neural network agent model | |
| Levesque et al. | An overset grid 2D/infinite swept wing URANS solver using recursive cartesian bucket method | |
| CN109977526A (en) | A method of the adjustment wing finite element model based on three-dimensional CST technology | |
| Asouti et al. | PCA-enhanced metamodel-assisted evolutionary algorithms for aerodynamic optimization | |
| Lengyel-Kampmann et al. | Generalized optimization of counter-rotating and single-rotating fans | |
| CN111310282A (en) | Helicopter rotor wing profile generation method and system suitable for plateau environment | |
| CN110348575A (en) | A kind of impeller reverse engineering approach based on genetic algorithm | |
| CN105260498A (en) | Variable camber design method of large civil aircraft wing | |
| CN112926132B (en) | Fixed wing airfoil aerodynamic shape design method considering influence of three-dimensional effect | |
| Tong et al. | Multi-objective aerodynamic optimization of supercritical wing with substantial pressure constraints | |
| Campbell et al. | A Knowledge-Based Optimization Method for Aerodynamic Design | |
| CN114266202A (en) | Modified actuating line model method for simulating wake flow of wind turbine | |
| Wang et al. | Co-kriging based multi-fidelity aerodynamic optimization for flying wing UAV with multi-shape wingtip design | |
| CN113626935B (en) | Design method of transonic moon-shaped wing with high cruising efficiency |
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 |