Advanced studies of selection schemes for dual virtual-metrology outputs

Wei Ming Wu, Fan-Tien Cheng, Tung Ho Lin, Deng Lin Zeng, Jyun Fang Chen, Min Hsiung Hung

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Advanced Studies of selection schemes between neural-network (NN) and multiple-regression (MR) outputs of a virtual metrology system (VMS) are presented in this paper. Both NN and MR are applicable algorithms for implementing VM conjecture models. But a MR algorithm may achieve better accuracy only with a stable process, whereas a NN algorithm may has superior accuracy when equipment property drift or shift occurs. To take advantage of the merits of both MR and NN algorithms, the simple-selection scheme (SS-scheme) was proposed in CASE 2008 to enhance virtual-metrology (VM) conjecture accuracy. This SS-scheme simply selects either NN or MR output. Recently, with advanced studies, a weighted-selection scheme (WS-scheme), which computes the VM output with a weighted sum of NN and MR results, has been developed. Besides the example with the CVD process of fifth generation TFT-LCD used in the CASE 2008 paper, a new example with the photo process is also adopted in this paper to test and compare the conjecture accuracy among solo NN, solo MR, SS-scheme, and WS-scheme. One-hidden-layered back-propagation neural network (BPNN-I) is adopted for establishing the NN conjecture model. Test results show that the conjecture accuracy of the WS-scheme is the best among those of solo NN, solo MR, SS-scheme, and WS-scheme algorithms.

Original languageEnglish
Title of host publication2009 IEEE International Conference on Automation Science and Engineering, CASE 2009
Pages421-426
Number of pages6
DOIs
Publication statusPublished - 2009 Nov 12
Event2009 IEEE International Conference on Automation Science and Engineering, CASE 2009 - Bangalore, India
Duration: 2009 Aug 222009 Aug 25

Publication series

Name2009 IEEE International Conference on Automation Science and Engineering, CASE 2009

Other

Other2009 IEEE International Conference on Automation Science and Engineering, CASE 2009
CountryIndia
CityBangalore
Period09-08-2209-08-25

Fingerprint

Neural networks
Liquid crystal displays
Backpropagation
Chemical vapor deposition

All Science Journal Classification (ASJC) codes

  • Computer Networks and Communications
  • Computer Science Applications
  • Software

Cite this

Wu, W. M., Cheng, F-T., Lin, T. H., Zeng, D. L., Chen, J. F., & Hung, M. H. (2009). Advanced studies of selection schemes for dual virtual-metrology outputs. In 2009 IEEE International Conference on Automation Science and Engineering, CASE 2009 (pp. 421-426). [5234137] (2009 IEEE International Conference on Automation Science and Engineering, CASE 2009). https://doi.org/10.1109/COASE.2009.5234137
Wu, Wei Ming ; Cheng, Fan-Tien ; Lin, Tung Ho ; Zeng, Deng Lin ; Chen, Jyun Fang ; Hung, Min Hsiung. / Advanced studies of selection schemes for dual virtual-metrology outputs. 2009 IEEE International Conference on Automation Science and Engineering, CASE 2009. 2009. pp. 421-426 (2009 IEEE International Conference on Automation Science and Engineering, CASE 2009).
@inproceedings{651537c6136a4b5c90cf2a447e2125c7,
title = "Advanced studies of selection schemes for dual virtual-metrology outputs",
abstract = "Advanced Studies of selection schemes between neural-network (NN) and multiple-regression (MR) outputs of a virtual metrology system (VMS) are presented in this paper. Both NN and MR are applicable algorithms for implementing VM conjecture models. But a MR algorithm may achieve better accuracy only with a stable process, whereas a NN algorithm may has superior accuracy when equipment property drift or shift occurs. To take advantage of the merits of both MR and NN algorithms, the simple-selection scheme (SS-scheme) was proposed in CASE 2008 to enhance virtual-metrology (VM) conjecture accuracy. This SS-scheme simply selects either NN or MR output. Recently, with advanced studies, a weighted-selection scheme (WS-scheme), which computes the VM output with a weighted sum of NN and MR results, has been developed. Besides the example with the CVD process of fifth generation TFT-LCD used in the CASE 2008 paper, a new example with the photo process is also adopted in this paper to test and compare the conjecture accuracy among solo NN, solo MR, SS-scheme, and WS-scheme. One-hidden-layered back-propagation neural network (BPNN-I) is adopted for establishing the NN conjecture model. Test results show that the conjecture accuracy of the WS-scheme is the best among those of solo NN, solo MR, SS-scheme, and WS-scheme algorithms.",
author = "Wu, {Wei Ming} and Fan-Tien Cheng and Lin, {Tung Ho} and Zeng, {Deng Lin} and Chen, {Jyun Fang} and Hung, {Min Hsiung}",
year = "2009",
month = "11",
day = "12",
doi = "10.1109/COASE.2009.5234137",
language = "English",
isbn = "9781424445783",
series = "2009 IEEE International Conference on Automation Science and Engineering, CASE 2009",
pages = "421--426",
booktitle = "2009 IEEE International Conference on Automation Science and Engineering, CASE 2009",

}

Wu, WM, Cheng, F-T, Lin, TH, Zeng, DL, Chen, JF & Hung, MH 2009, Advanced studies of selection schemes for dual virtual-metrology outputs. in 2009 IEEE International Conference on Automation Science and Engineering, CASE 2009., 5234137, 2009 IEEE International Conference on Automation Science and Engineering, CASE 2009, pp. 421-426, 2009 IEEE International Conference on Automation Science and Engineering, CASE 2009, Bangalore, India, 09-08-22. https://doi.org/10.1109/COASE.2009.5234137

Advanced studies of selection schemes for dual virtual-metrology outputs. / Wu, Wei Ming; Cheng, Fan-Tien; Lin, Tung Ho; Zeng, Deng Lin; Chen, Jyun Fang; Hung, Min Hsiung.

2009 IEEE International Conference on Automation Science and Engineering, CASE 2009. 2009. p. 421-426 5234137 (2009 IEEE International Conference on Automation Science and Engineering, CASE 2009).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Advanced studies of selection schemes for dual virtual-metrology outputs

AU - Wu, Wei Ming

AU - Cheng, Fan-Tien

AU - Lin, Tung Ho

AU - Zeng, Deng Lin

AU - Chen, Jyun Fang

AU - Hung, Min Hsiung

PY - 2009/11/12

Y1 - 2009/11/12

N2 - Advanced Studies of selection schemes between neural-network (NN) and multiple-regression (MR) outputs of a virtual metrology system (VMS) are presented in this paper. Both NN and MR are applicable algorithms for implementing VM conjecture models. But a MR algorithm may achieve better accuracy only with a stable process, whereas a NN algorithm may has superior accuracy when equipment property drift or shift occurs. To take advantage of the merits of both MR and NN algorithms, the simple-selection scheme (SS-scheme) was proposed in CASE 2008 to enhance virtual-metrology (VM) conjecture accuracy. This SS-scheme simply selects either NN or MR output. Recently, with advanced studies, a weighted-selection scheme (WS-scheme), which computes the VM output with a weighted sum of NN and MR results, has been developed. Besides the example with the CVD process of fifth generation TFT-LCD used in the CASE 2008 paper, a new example with the photo process is also adopted in this paper to test and compare the conjecture accuracy among solo NN, solo MR, SS-scheme, and WS-scheme. One-hidden-layered back-propagation neural network (BPNN-I) is adopted for establishing the NN conjecture model. Test results show that the conjecture accuracy of the WS-scheme is the best among those of solo NN, solo MR, SS-scheme, and WS-scheme algorithms.

AB - Advanced Studies of selection schemes between neural-network (NN) and multiple-regression (MR) outputs of a virtual metrology system (VMS) are presented in this paper. Both NN and MR are applicable algorithms for implementing VM conjecture models. But a MR algorithm may achieve better accuracy only with a stable process, whereas a NN algorithm may has superior accuracy when equipment property drift or shift occurs. To take advantage of the merits of both MR and NN algorithms, the simple-selection scheme (SS-scheme) was proposed in CASE 2008 to enhance virtual-metrology (VM) conjecture accuracy. This SS-scheme simply selects either NN or MR output. Recently, with advanced studies, a weighted-selection scheme (WS-scheme), which computes the VM output with a weighted sum of NN and MR results, has been developed. Besides the example with the CVD process of fifth generation TFT-LCD used in the CASE 2008 paper, a new example with the photo process is also adopted in this paper to test and compare the conjecture accuracy among solo NN, solo MR, SS-scheme, and WS-scheme. One-hidden-layered back-propagation neural network (BPNN-I) is adopted for establishing the NN conjecture model. Test results show that the conjecture accuracy of the WS-scheme is the best among those of solo NN, solo MR, SS-scheme, and WS-scheme algorithms.

UR - http://www.scopus.com/inward/record.url?scp=70449111622&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70449111622&partnerID=8YFLogxK

U2 - 10.1109/COASE.2009.5234137

DO - 10.1109/COASE.2009.5234137

M3 - Conference contribution

SN - 9781424445783

T3 - 2009 IEEE International Conference on Automation Science and Engineering, CASE 2009

SP - 421

EP - 426

BT - 2009 IEEE International Conference on Automation Science and Engineering, CASE 2009

ER -

Wu WM, Cheng F-T, Lin TH, Zeng DL, Chen JF, Hung MH. Advanced studies of selection schemes for dual virtual-metrology outputs. In 2009 IEEE International Conference on Automation Science and Engineering, CASE 2009. 2009. p. 421-426. 5234137. (2009 IEEE International Conference on Automation Science and Engineering, CASE 2009). https://doi.org/10.1109/COASE.2009.5234137