notesum.ai

Published at November 12

Design optimization of semiconductor manufacturing equipment using a novel multi-fidelity surrogate modeling approach

cs.CE

Released Date: November 12, 2024

Authors: Bingran Wang, Min Sung Kim, Taewoong Yoon, Dasom Lee, Byeong-Sang Kim, Dougyong Sung, John T. Hwang

Arxiv: http://arxiv.org/abs/2411.08149v1

Ref. design

HF optimal designs

MF optimal designs

Case 1

Case 2

Case 3

Case 1

Case 2

Case 3

Design variables:

CR1

Redacted

+1.019

+1.073

-1.104

+1.023

-0.143

-0.996

CR2

Redacted

+1.881

+1.893

+4.004

+1.876

+2.043

+3.876

Redacted

-6.246

-5.841

-8.923

-6.257

-10.00

-8.877

Redacted

-3.169

-3.752

-5.538

-3.164

-3.389

-5.689

Redacted

-2.698

-2.625

-2.685

-2.702

-2.644

-2.678

Redacted

-2.447

-2.367

-2.619

-2.451

-2.696

-2.573

Redacted

+8.571

+8.275

+6.186

+8.580

+5.798

+6.125

Validation results:

3\sigma_{T}

4.263

2.972

3.048

2.807

2.946

2.890

2.813

\mu_{T}

17.66

16.95

17.01

17.05

16.94

16.97

17.04

\text{max}(T)

22.70

21.11

21.16

20.99

21.07

20.96

No. of violated opt. consts.

(

\mu_{T}>17

and/or

\text{max}(T)>21.5

)

Ranking of feasible designs

(No constraints violation

3rd

2nd

1st

with minimum

3\sigma_{T}

)