國立虎尾科技大學 |

Efficient Sequential Sampling for Neural Network-Based Surrogate Modeling.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	Efficient Sequential Sampling for Neural Network-Based Surrogate Modeling./
作者:	Koratikere, Pavankumar.
面頁冊數:	1 online resource (97 pages)
附註:	Source: Masters Abstracts International, Volume: 85-01.
Contained By:	Masters Abstracts International85-01.
標題:	Mechanics. -
電子資源:	click for full text (PQDT)
ISBN:	9798379875169

Efficient Sequential Sampling for Neural Network-Based Surrogate Modeling.
Koratikere, Pavankumar.

Efficient Sequential Sampling for Neural Network-Based Surrogate Modeling. - 1 online resource (97 pages)

Source: Masters Abstracts International, Volume: 85-01.

Thesis (M.Sc.)--Purdue University, 2023.

Includes bibliographical references

Gaussian Process Regression (GPR) is a widely used surrogate model in efficient global optimization (EGO) due to its capability to provide uncertainty estimates in the prediction. The cost of creating a GPR model for large data sets is high. On the other hand, neural network (NN) models scale better compared to GPR as the number of samples increase. Unfortunately, the uncertainty estimates for NN prediction are not readily available. In this work, a scalable algorithm is developed for EGO using NN-based prediction and uncertainty (EGONN). Initially, two different NNs are created using two different data sets. The first NN models the output based on the input values in the first data set while the second NN models the prediction error of the first NN using the second data set. The next infill point is added to the first data set based on criteria like expected improvement or prediction uncertainty. EGONN is demonstrated on the optimization of the Forrester function and a constrained Branin function and is compared with EGO. The convergence criteria is based on the maximum number of infill points in both cases. The algorithm is able to reach the optimum point within the given budget. The EGONN is extended to handle constraints explicitly and is utilized for aerodynamic shape optimization of the RAE 2822 airfoil in transonic viscous flow at a free-stream Mach number of 0.734 and a Reynolds number of 6.5 x 106. The results obtained from EGONN are compared with the results from gradient-based optimization (GBO) using adjoints. The optimum shape obtained from EGONN is comparable to the shape obtained from GBO and is able to eliminate the shock. The drag coefficient is reduced from 200 drag counts to 114 and is close to 110 drag counts obtained from GBO. The EGONN is also extended to handle uncertainty quantification (uqEGONN) using prediction uncertainty as an infill method. The convergence criteria is based on the relative change of summary statistics such as mean and standard deviation of an uncertain quantity. The uqEGONN is tested on Ishigami function with an initial sample size of 100 samples and the algorithm terminates after 70 infill points. The statistics obtained from uqEGONN (using only 170 function evaluations) are close to the values obtained from directly evaluating the function one million times. uqEGONN is demonstrated on to quantifying the uncertainty in the airfoil performance due to geometric variations. The algorithm terminates within 100 computational fluid dynamics (CFD) analyses and the statistics obtained from the algorithm are close to the one obtained from 1000 direct CFD based evaluations.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2024

Mode of access: World Wide Web

ISBN: 9798379875169Subjects--Topical Terms:

527684
Mechanics.
Index Terms--Genre/Form:

554714
Electronic books.

Efficient Sequential Sampling for Neural Network-Based Surrogate Modeling.
LDR:03914ntm a22003737 4500 001 1145131
005 20240617111351.5
006 m o d
007 cr mn ---uuuuu
008 250605s2023 xx obm 000 0 eng d
020 $a 9798379875169
035 $a (MiAaPQ)AAI30540054
035 $a (MiAaPQ)Purdue22695622
035 $a AAI30540054
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Koratikere, Pavankumar. $3 1470350
245 1 0 $a Efficient Sequential Sampling for Neural Network-Based Surrogate Modeling.
264 0 $c 2023
300 $a 1 online resource (97 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Masters Abstracts International, Volume: 85-01.
500 $a Advisor: Leifur, Leifur.
502 $a Thesis (M.Sc.)--Purdue University, 2023.
504 $a Includes bibliographical references
520 $a Gaussian Process Regression (GPR) is a widely used surrogate model in efficient global optimization (EGO) due to its capability to provide uncertainty estimates in the prediction. The cost of creating a GPR model for large data sets is high. On the other hand, neural network (NN) models scale better compared to GPR as the number of samples increase. Unfortunately, the uncertainty estimates for NN prediction are not readily available. In this work, a scalable algorithm is developed for EGO using NN-based prediction and uncertainty (EGONN). Initially, two different NNs are created using two different data sets. The first NN models the output based on the input values in the first data set while the second NN models the prediction error of the first NN using the second data set. The next infill point is added to the first data set based on criteria like expected improvement or prediction uncertainty. EGONN is demonstrated on the optimization of the Forrester function and a constrained Branin function and is compared with EGO. The convergence criteria is based on the maximum number of infill points in both cases. The algorithm is able to reach the optimum point within the given budget. The EGONN is extended to handle constraints explicitly and is utilized for aerodynamic shape optimization of the RAE 2822 airfoil in transonic viscous flow at a free-stream Mach number of 0.734 and a Reynolds number of 6.5 x 106. The results obtained from EGONN are compared with the results from gradient-based optimization (GBO) using adjoints. The optimum shape obtained from EGONN is comparable to the shape obtained from GBO and is able to eliminate the shock. The drag coefficient is reduced from 200 drag counts to 114 and is close to 110 drag counts obtained from GBO. The EGONN is also extended to handle uncertainty quantification (uqEGONN) using prediction uncertainty as an infill method. The convergence criteria is based on the relative change of summary statistics such as mean and standard deviation of an uncertain quantity. The uqEGONN is tested on Ishigami function with an initial sample size of 100 samples and the algorithm terminates after 70 infill points. The statistics obtained from uqEGONN (using only 170 function evaluations) are close to the values obtained from directly evaluating the function one million times. uqEGONN is demonstrated on to quantifying the uncertainty in the airfoil performance due to geometric variations. The algorithm terminates within 100 computational fluid dynamics (CFD) analyses and the statistics obtained from the algorithm are close to the one obtained from 1000 direct CFD based evaluations.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Mechanics. $3 527684
650 4 $a Fluid mechanics. $3 555551
650 4 $a Reynolds number. $3 839681
650 4 $a Aerospace engineering. $3 686400
650 4 $a Pressure distribution. $3 1413590
650 4 $a Neural networks. $3 1011215
650 4 $a Fluid dynamics. $3 528074
650 4 $a Space exploration. $3 1467124
650 4 $a Monte Carlo simulation. $3 1468022
650 4 $a Sample size. $3 1464660
650 4 $a Aircraft. $3 580976
655 7 $a Electronic books. $2 local $3 554714
690 $a 0538
690 $a 0800
690 $a 0204
690 $a 0338
690 $a 0346
710 2 $a Purdue University. $3 1184550
710 2 $a ProQuest Information and Learning Co. $3 1178819
773 0 $t Masters Abstracts International $g 85-01.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30540054 $z click for full text (PQDT)