講稿star文章類優(yōu)化fluid_第1頁(yè)
講稿star文章類優(yōu)化fluid_第2頁(yè)
講稿star文章類優(yōu)化fluid_第3頁(yè)
講稿star文章類優(yōu)化fluid_第4頁(yè)
講稿star文章類優(yōu)化fluid_第5頁(yè)
已閱讀5頁(yè),還剩17頁(yè)未讀 繼續(xù)免費(fèi)閱讀

下載本文檔

版權(quán)說(shuō)明:本文檔由用戶提供并上傳,收益歸屬內(nèi)容提供方,若內(nèi)容存在侵權(quán),請(qǐng)進(jìn)行舉報(bào)或認(rèn)領(lǐng)

文檔簡(jiǎn)介

1、The use of race simulation models as an alternative to robust design optimisationLuca Nardin, Ph.D. cand.,EADS Military Air Systems, Munich, GermanyUniversity of Trieste, ItalyThe use of race simulation models as an alternative to robust design optimisationOutlineWhy do we need Robust Design Optimis

2、ation“Classical” ways of addressing Robust Design:Weighted multi-point optimisationRisk functionMean value and variance (MORDO)Race Simulation methodology:IdeaExampleConclusionsThe use of race simulation models as an alternative to robust design optimisationWhy do we need Robust Design Optimisation?

3、Uncertainty (AIAA definition): A potential deficiency in any phase or activity of the modelling process that is due to lack of knowledge: Uncertainties on geometry parameters due to manufacturing tolerance. Uncertainties on operative conditions (design point).The use of race simulation models as an

4、alternative to robust design optimisationWhy do we need Robust Design Optimisation?Over-optimisation: when optimising the objective function with fixed operative conditions, the final solution has usually good performance at the design point but poor off-design characteristics.Usually the determinis

5、tic solution presents good performance only at the design point.The use of race simulation models as an alternative to robust design optimisationOutlineWhy do we need Robust Design Optimisation“Classical” ways of addressing Robust Design:Weighted multi-point optimisationRisk functionMean value and v

6、ariance (MORDO)Race Simulation methodology:IdeaExampleConclusionsThe use of race simulation models as an alternative to robust design optimisationWays of addressing Robust Design OptimisationWeighted multi-point optimisation: the objective function is a weighted sum of different optimisations subjec

7、ted to different operative conditions (to address the uncertainties)Disadvantage: arbitrary definition of weights.Minimize a risk function :Disadvantage: there is still the possibility of obtaining “unstable” solution.The use of race simulation models as an alternative to robust design optimisationW

8、ays of addressing Robust Design OptimisationOptimise mean value of the objective function(s) and minimize its(their) standard deviation (MORDO):2 different directions in the optimisation: minimizing the variance of objective function, will minimize the off-design performance degradation; optimising

9、the objective mean value, the performances will be privileged.The use of race simulation models as an alternative to robust design optimisationWays of addressing Robust Design OptimisationMean and variance values methodology (MORDO):Advantages: No arbitrary definition of weights. Set of solutions (P

10、areto front) from which to choose: high performance or stability of performance.But the Pareto front presents also some drawbacks: Incrementing the number of objectives results in an increase of the number of evaluations. Additional efforts are demanded in order to choose among the different designs

11、. If the uncertainties change slightly the Pareto front can still have the same behaviour.The use of race simulation models as an alternative to robust design optimisationOutlineWhy do we need Robust Design Optimisation“Classical” ways of addressing Robust Design:Weighted multi-point optimisationRis

12、k functionMean value and variance (MORDO)Race Simulation methodology:IdeaExampleConclusionsThe use of race simulation models as an alternative to robust design optimisationRace Simulation modelThe idea comes from the Americas Cup:The most challenging design is the one which wins most of the matches

13、under stochastic perturbative conditions.The design which wins (maybe with a great gap) only one match under the most probable condition usually do not win the tournament.This idea is directly applied to design under perturbative conditions (they are not part of the design itself: AoA, speed, etc.)

14、but can be extended to uncertainties of the input parameters.The use of race simulation models as an alternative to robust design optimisationRace Simulation modelRace Simulation methodology:Binary matches in which each design competes against another designEach design will perform a number of match

15、es in order to compete against each of the other designsFor each match the conditions will be different.Example of a tournament in a race simulation:4 designs will result in 6 matches under 6 different conditionsA-B condition 1B-C condition 4A-C condition 2B-D condition 5A-D condition 3C-D condition

16、 6The use of race simulation models as an alternative to robust design optimisationOutlineWhy do we need Robust Design Optimisation“Classical” ways of addressing Robust Design:Weighted multi-point optimisationRisk functionMean value and variance (MORDO)Race Simulation methodology:IdeaExampleConclusi

17、onsThe use of race simulation models as an alternative to robust design optimisationRace Simulation modelA numeric example:where:Three peaks with different behaviours Uncertainty on the input parameters: X and Y defined inside the intervals x-x , x+x and y-y , y+y What happens with different values

18、of the uncertainty ?The use of race simulation models as an alternative to robust design optimisationThe best design found by the race simulation depends on the perturbative condition.Race Simulation modelThe use of race simulation models as an alternative to robust design optimisationThe Pareto fro

19、nt found with MORDO presents always the same behaviour also if the perturbative conditions change.Race Simulation model- Comparison with MORDO - = 0.02 = 0.1The use of race simulation models as an alternative to robust design optimisationRace Tournament applied to an Airfoil test case:5 different ai

20、rfoil profiles with different performances.Perturbative condition (uniformly distributed): AoA.Example preparation:RAE 2822 plus 4 optimised designs for 4 different AoA.Simulation of all of the different designs for a range of AoA, to be able to understand their behaviour and to compute mean and sta

21、ndard deviation.Tournament among these designs (each tournament comprise the simulation of each design for 4 different angles of attack).Race Simulation modelThe use of race simulation models as an alternative to robust design optimisationRace Tournament applied to an Airfoil test case:5 different a

22、irfoil profiles with different performances.Perturbative condition (uniformly distributed): AoARace Simulation win probability:2.00 48.5%2.50 37.0%3.00 11.5%3.50 0.4%RAE 2.7%48.5%37%11.5%0.4%2.7%Race Simulation modelThe use of race simulation models as an alternative to robust design optimisationMea

23、n value and variance (“classic” robust design values):The winner of the tournament is on the Pareto front.The worst design is on the Pareto front as well.original RAE 2822Pareto front48.5%37%11.5%0.4%2.7%Race Simulation modelRace Simulation win probability:2.00 48.5%2.50 37.0%3.00 11.5%3.50 0.4%RAE

24、2.7%The use of race simulation models as an alternative to robust design optimisationOutlineWhy do we need Robust Design Optimisation“Classical” ways of addressing Robust Design:Weighted multi-point optimisationRisk functionMean value and variance (MORDO)Race Simulation methodology:IdeaExampleConclu

25、sionsThe use of race simulation models as an alternative to robust design optimisationRace Simulation model: conclusionsAdvantages of Race Simulation methodology:No arbitrary weights definition.No increment of the objectives number.No preliminary preference for performance or stability: the result of the tournament is an implicit weighted function of mean and variance of performance.Disadvantages:A tournament can be very t

溫馨提示

  • 1. 本站所有資源如無(wú)特殊說(shuō)明,都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請(qǐng)下載最新的WinRAR軟件解壓。
  • 2. 本站的文檔不包含任何第三方提供的附件圖紙等,如果需要附件,請(qǐng)聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶所有。
  • 3. 本站RAR壓縮包中若帶圖紙,網(wǎng)頁(yè)內(nèi)容里面會(huì)有圖紙預(yù)覽,若沒(méi)有圖紙預(yù)覽就沒(méi)有圖紙。
  • 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
  • 5. 人人文庫(kù)網(wǎng)僅提供信息存儲(chǔ)空間,僅對(duì)用戶上傳內(nèi)容的表現(xiàn)方式做保護(hù)處理,對(duì)用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯,并不能對(duì)任何下載內(nèi)容負(fù)責(zé)。
  • 6. 下載文件中如有侵權(quán)或不適當(dāng)內(nèi)容,請(qǐng)與我們聯(lián)系,我們立即糾正。
  • 7. 本站不保證下載資源的準(zhǔn)確性、安全性和完整性, 同時(shí)也不承擔(dān)用戶因使用這些下載資源對(duì)自己和他人造成任何形式的傷害或損失。

最新文檔

評(píng)論

0/150

提交評(píng)論