東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Threshold production control policie...

Yan, Houmin.

FindBook

Google Book

Amazon

博客來

Threshold production control policies in dynamic stochastic manufacturing systems.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Threshold production control policies in dynamic stochastic manufacturing systems./
作者:	Yan, Houmin.
面頁冊數:	181 p.
附註:	Source: Dissertation Abstracts International, Volume: 56-07, Section: A, page: 2770.
Contained By:	Dissertation Abstracts International56-07A.
標題:	Business Administration, Management. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=NN97331
ISBN:	9780315973312

Threshold production control policies in dynamic stochastic manufacturing systems.
Yan, Houmin.

Threshold production control policies in dynamic stochastic manufacturing systems. - 181 p.

Source: Dissertation Abstracts International, Volume: 56-07, Section: A, page: 2770.

Thesis (Ph.D.)--University of Toronto (Canada), 1994.

Most manufacturing systems are large systems characterized by the presence of stochastic interferences such as machine failures, demand fluctuations, stochastic set up times, and random yields. These uncertainties cause difficulties for manufacturers in planning production, regulating inventories, and meeting customer demands. On one hand, it is generally agreed that it is difficult to devise optimal policies for such manufacturing systems. On the other hand, an easy-to-implement, well performing control policy is in great demand by practitioners. Therefore, we investigate the application of suboptimal policies, called threshold control policies, to manufacturing systems under uncertainty, and particularly, semiconductor wafer fabrication. We shall specially consider threshold policies known as WIP Control, Surplus Control and Two Boundary Control (the combination of WIP and Surplus Control). Algorithms attempting to optimize threshold values required for implementing these policies are developed. First, perturbation analysis techniques are employed to obtain the gradient estimates based on a single simulation run. Iterative algorithms of the stochastic optimization type are constructed. The issue of consistency is explored. We show that gradient estimates are consistent. It is also illustrated, both analytically and empirically, that the iterative algorithms converge to the optimal threshold values. We test these threshold control policies on a CMOS4 wafer fabrication model consisting of about 250 sequential operations and 60 machines. A simulation study is conducted to compare the performance of these threshold control policies and other policies commonly used in practice. The simulation results demonstrate that Two Boundary Control policy outperforms others in terms of reducing cycle time and work-in-process. Moreover, it is the most robust policy of all against random disturbances such as machine failures, demand fluctuations, and system parameter changes.

ISBN: 9780315973312Subjects--Topical Terms:

626628
Business Administration, Management.

Threshold production control policies in dynamic stochastic manufacturing systems.
LDR:02865nmm a2200265 4500 001 2055757
005 20150213104838.5
008 170521s1994 ||||||||||||||||| ||eng d
020 $a 9780315973312
035 $a (MiAaPQ)AAINN97331
035 $a AAINN97331
040 $a MiAaPQ $c MiAaPQ
100 1 $a Yan, Houmin. $3 784017
245 1 0 $a Threshold production control policies in dynamic stochastic manufacturing systems.
300 $a 181 p.
500 $a Source: Dissertation Abstracts International, Volume: 56-07, Section: A, page: 2770.
500 $a Supervisor: Suresh Sethi.
502 $a Thesis (Ph.D.)--University of Toronto (Canada), 1994.
520 $a Most manufacturing systems are large systems characterized by the presence of stochastic interferences such as machine failures, demand fluctuations, stochastic set up times, and random yields. These uncertainties cause difficulties for manufacturers in planning production, regulating inventories, and meeting customer demands. On one hand, it is generally agreed that it is difficult to devise optimal policies for such manufacturing systems. On the other hand, an easy-to-implement, well performing control policy is in great demand by practitioners. Therefore, we investigate the application of suboptimal policies, called threshold control policies, to manufacturing systems under uncertainty, and particularly, semiconductor wafer fabrication. We shall specially consider threshold policies known as WIP Control, Surplus Control and Two Boundary Control (the combination of WIP and Surplus Control). Algorithms attempting to optimize threshold values required for implementing these policies are developed. First, perturbation analysis techniques are employed to obtain the gradient estimates based on a single simulation run. Iterative algorithms of the stochastic optimization type are constructed. The issue of consistency is explored. We show that gradient estimates are consistent. It is also illustrated, both analytically and empirically, that the iterative algorithms converge to the optimal threshold values. We test these threshold control policies on a CMOS4 wafer fabrication model consisting of about 250 sequential operations and 60 machines. A simulation study is conducted to compare the performance of these threshold control policies and other policies commonly used in practice. The simulation results demonstrate that Two Boundary Control policy outperforms others in terms of reducing cycle time and work-in-process. Moreover, it is the most robust policy of all against random disturbances such as machine failures, demand fluctuations, and system parameter changes.
590 $a School code: 0779.
650 4 $a Business Administration, Management. $3 626628
690 $a 0454
710 2 $a University of Toronto (Canada). $3 1017674
773 0 $t Dissertation Abstracts International $g 56-07A.
790 $a 0779
791 $a Ph.D.
792 $a 1994
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=NN97331