成功大學博碩士論文系統

簡易檢索 / 詳目顯示

回結果列表

研究生：	王邑璽 Wang, Yi-Xi
論文名稱：	結合基因演算法與模擬最佳化改善車燈零組件排程-以 T 公司為例 A Study on Combination of Genetic Algorithm and Simulation Optimization to Improve the Schedule of Vehicle Headlight Components - Take T Company as an Example
指導教授：	洪郁修 Hung, Yu-Hsiu
共同指導教授:	楊大和 Yang, Taho
學位類別：	碩士 Master
系所名稱：	規劃與設計學院 - 工業設計學系 Department of Industrial Design
論文出版年：	2019
畢業學年度：	107
語文別：	英文
論文頁數：	111
中文關鍵詞：	混合流程式生產排程、專用機台、整備時間、模擬最佳化、基因演算法
外文關鍵詞：	Hybrid flow shop problem, Machine eligibility, Setup time, Simulation optimization, Genetic algorithm
相關次數：	點閱：121 下載：0
分享至:	分享至facebook 分享至twitter

查詢本校圖書館目錄查詢臺灣博碩士論文知識加值系統勘誤回報

隨著車燈設計越來越複雜，車燈的零組件也越來越繁複。在車燈製程中，車燈零組件製程以成型工程為最上游接著經過塗裝、烘乾工程與蒸鍍工程。關係到品質問題，成型工程生產之半成品必須在2小時內進入塗裝烘乾及蒸鍍工程進行加工，所以在進行排程規劃時必須將成型、塗裝、烘乾和蒸鍍工程一起納入考量。由於機台數量多且採購的時間點不一致，造成機台能力的差異，因此成型工程與蒸鍍工程為混合式流程型工廠問題。改善車燈零組件的排程規劃，將有利於提升車燈製程的生產力。
在典型的最佳化排程規劃中，對實際系統做了過多的假設或簡化，造成模式與真實系統間的落差。若採用以模擬為基礎的智慧搜尋法，也常因為模擬時間過長無法滿足即時性的需求。因此本研究將發展一套減少運算求解時間之結合基因演算法與模擬最佳化排程方法。本研究排程方法以最小化平均在製品水準為目標，考慮車燈零組件製程之四大特性：平行機台加工、專用機台、獨立整備時間與相依整備時間，並探討不同需求水準下混合流程型生產系統排程的問題，期望符合真實系統之需求且可快速得到排程建議。本研究方法不論在高、中、低需求水準的問題中，其求解的績效－庫存水準均有良好的表現。
最後，本研究針提出的排程解將於附件F中呈現，可提供業界針對不同需求水準下機台排程時的決策參考。

As the design of the automotive lights becomes more and more complex, the components of the automotive lights are becoming more and more complicated. In the automotive light component manufacturing process, the injection molding is the most upstream of the automotive light component process, followed by painting, drying, and deposition. Relating to the quality problem, the semi-finished products produced by the injection molding must be processed by the painting, drying and deposition process within 2 hours, so the molding, painting, drying, and deposition must be all taken into account in the scheduling planning. Due to a large number of machines and the inconsistent time of purchase, the difference in machine capacity is caused, so the automotive light component manufacturing process is hybrid flow shop problems(HFSP). Improving the scheduling of automotive light component manufacturing process will help improve the productivity of the automotive light manufacturing process.
In a typical optimization schedule, too many assumptions or simplifications are made to the actual system, resulting in a gap between the model and the real system. If you use the simulation-based wisdom search method, it doesn’t often meet the immediate needs because the simulation time is too long. Therefore, this study will develop a scheduling method combined genetic algorithms(GA) and simulation optimization to reduce computation time. The scheduling method aims to minimize the average work-in-process(WIP) level and considers the four characteristics of the automotive light component manufacture: parallel machine processing, machine eligibility, independent setup time and sequence-dependent setup time(SDST). This study discusses HFSP under different demand levels, expects to meet the needs of real systems, and quickly gives scheduling recommendations. Regardless of the WIP level in the high, medium and low demand levels, the solution method has a good performance. Finally, the solutions proposed in this study for the industrial characteristics of automotive light components are shown in the Appendix, and they can provide decision-making reference for automotive light components scheduling under different levels of demand.

摘要	i
SUMMARY	iii
ACKNOWLEDGEMENTS	iv
TABLE OF CONTENTS	v
LIST OF TABLES	ix
LIST OF FIGURES	x
CHAPTER 1 INTRODUCTION	1
1.1 Research Background	1
1.2 Research Motivation	3
1.3 Research Purposes	4
1.4 Research Contributions	5
1.5 Research Framework	6
CHAPTER 2 LITERATURE REVIEW	7
2.1 Automotive Headlight Process Introduction	7
2.2 Scheduling Theory	8
2.3 Hybrid Flow Shop Production Scheduling Related Research	10
2.4 Genetic Algorithm	15
2.5 Simulation Mode Construction	16
2.6 Simulation Optimization	18
CHAPTER 3 RESEARCH METHODS	20
3.1 Case Profile And Current Status	20
3.2 Research Process	24
3.3 Observation And Analysis Of The Current Status Of The Production System	25
3.3.1 Drawing A Value Stream Map	25
3.4 Current Status System  Simulation Mode Construction	27
3.4.1 Drawing A Simulation Mode Blueprint Based On The Current System	29
3.4.2 Collecting The Information Required By The Model	30
3.4.3 Reasonable Assumptions	30
3.4.4 Current System Simulation Mode Construction	31
3.4.5 Verification And Validation	31
3.5 Propose Scheduling Improvement Strategies	32
3.5.1 Strategy 1: Split Customer Demand Into Orders, Order Production Sequence Is Random	32
3.5.2 Strategy 2: Split Customer Demand Into Orders, Order Production Sequence Is Determined By Genetic Algorithm	33
3.6 Comparison Of Experimental Result	34
CHAPTER 4 CASE STUDY	35
4.1 Drawing A Value Stream Map	35
4.2 Current System Analysis And Problem Description	39
4.3 Current Simulation Model Construction	41
4.3.1 Collecting Information And Model Assumptions	41
4.3.2 Constructing Current Simulation Model	41
4.4 Propose Scheduling Improvement Strategies	55
4.4.1 Strategy 1: Split Customer Demand Into Orders, Order Production Sequence Is Random	56
4.4.2 Strategy 2: Split Customer Demand Into Orders, Order Production Sequence Is Determined By Genetic Algorithm	56
4.5 Comparison of Experimental Data and solution suggestion	61
CHAPTER 5 CONCLUSION AND RECOMMENDATIONS	64
5.1 Conclusion	64
5.2 Recommendations	64
REFERENCES	66
Appendix A MOLDING MACHINE PRODUCTION INFORMATION	68
A.1 Low-demand Molding Machine 1 Production Information	68
A.2 Low-demand Molding Machine 2 Production Information	69
A.3 Low-demand Molding Machine 3 Production Information	69
A.4 Low-demand Molding Machine 4 Production Information	70
A.5 Medium-demand Molding Machine 1 Production Information	71
A.6 Medium-demand Molding Machine 2 Production Information	72
A.7 Medium-demand Molding Machine 3 Production Information	73
A.8 Medium-demand Molding Machine 4 Production Information	73
A.9 High-demand Molding Machine 1 Production Information	74
A.10 High-demand Molding Machine 2 Production Information	75
A.11 High-demand Molding Machine 3 Production Information	76
A.12 High-demand Molding Machine 4 Production Information	77
Appendix B PARALLEL PROCESSING BATCH IN DEPOSITION	78
B.1 Parallel Processing Batch Of Product Family 1 In Deposition	78
B.2 Parallel Processing Batch Of Product Family 2 & 3 In Deposition	78
B.3 Parallel Processing Batch Of Product Family 3 In Deposition	79
B.4 Parallel Processing Batch Of Product Family 4 In Deposition	79
Appendix C OPTQUEST SETTING	80
C.1 Objective Function:	80
C.2 Constraints: Inventory	80
C.3 Controls: Production Size, Safety Stock	81
Appendix D GENETIC ALGORITHM CODE	84
Appendix E SUGGESTED PRODUCTION BATCH AND SEQUENCE	89
E.1 High-demand Molding Machine 1 Suggested Production Batch	89
E.2 High-demand Molding Machine 1 Suggested Production Sequence	89
E.3 High-demand Molding Machine 2 Suggested Production Batch	90
E.4 High-demand Molding Machine 2 Suggested Production Sequence	90
E.5 High-demand Molding Machine 3 Suggested Production Batch	90
E.6 High-demand Molding Machine 3 Suggested Production Sequence	91
E.7 High-demand Molding Machine 4 Suggested Production Batch	91
E.8 High-demand Molding Machine 4 Suggested Production Sequence	91
E.9 Medium-demand Molding Machine 1 Suggested Production Batch	92
E.10 Medium-demand Molding Machine 1 Suggested Production Sequence	92
E.11 Medium-demand Molding Machine 2 Suggested Production Batch	93
E.12 Medium-demand Molding Machine 2 Suggested Production Sequence	93
E.13 Medium-demand Molding Machine 3 Suggested Production Batch	93
E.14 Medium-demand Molding Machine 3 Suggested Production Sequence	94
E.15 Medium-demand Molding Machine 4 Suggested Production Batch	94
E.16 Medium-demand Molding Machine 4 Suggested Production Sequence	95
E.17 Low-demand Molding Machine 1 Suggested Production Batch	95
E.18 Low-demand Molding Machine 1 Suggested Production Sequence	96
E.19 Low-demand Molding Machine 2 Suggested Production Batch	97
E.20 Low-demand Molding Machine 2 Suggested Production Sequence	97
E.21 Low-demand Molding Machine 3 Suggested Production Batch	97
E.22 Low-demand Molding Machine 3 Suggested Production Sequence	98
E.23 Low-demand Molding Machine 4 Suggested Production Batch	98
E.24 Low-demand Molding Machine 4 Suggested Production Sequence	99
Appendix F SCHEDULE SOLUTION	100
F.1 High-demand Molding Machine 1 Schedule	100
F.2 High-demand Molding Machine 2 Schedule	101
F.3 High-demand Molding Machine 3 Schedule	102
F.4 High-demand Molding Machine 4 Schedule	103
F.5 Medium-demand Molding Machine 1 Schedule	104
F.6 Medium-demand Molding Machine 2 Schedule	105
F.7 Medium-demand Molding Machine 3 Schedule	106
F.8 Medium-demand Molding Machine 4 Schedule	107
F.9 Low-demand Molding Machine 1 Schedule	108
F.10 Low-demand Molding Machine 2 Schedule	109
F.11 Low-demand Molding Machine 3 Schedule	110
F.12 Low-demand Molding Machine 4 Schedule	111
                                    

成沢俊子（2010）.大家來學TPS-豐田改善直達車.台北市：財團法人中衛發展中心。
麥克．魯斯、約翰．舒克（2010）.學習觀察 : 增加價值、消除浪費的價值流圖.台北市：財團法人中衛發展中心。
產業價值鏈資訊平台（2010）. 汽車產業鏈簡介. 取自http://ic.tpex.org.tw/introduce.php?ic=3000
經濟部統計處（2016）. 取自
http://www.moea.gov.tw/MNS/dos/home/Home.aspx
工研院經濟產業與趨勢研究中心（2015）.台灣汽車零組件產業概況與廠商機會. 取自http://ieknet.iek.org.tw
Arbel, A. , & Vargas, L.G.(1993). Preference simulation and preference programming: robustness issues in priority derivation, European Journal of Operational Research, 69, 200-209.
Brucker, P.(1998).Scheduling Algorithm 2e. Springer- Germany :Verlag.
Chankonk, V., & Haimes, Y. Y. (1983). Multiobjective decision making –Theory and methodology. New York: Elsevier Science Publishing Co.
Gupta, J.N.D., Ruiz-Torres, & A. J.(2005). Generating efficient schedules for identical parallel machines involving flow-time and tardy jobs. European Journal of Operation Research, 167, pp. 679-695.
Hopp, W., & Spearman, M. L.(2000). Factory Physics 2e. New York: McGraw-Hill.
Horowitz, E., & Sahni, S.(1976). An exact and approximate algorithm for scheduling nonidentical processors. Journal of the Association for Computing Machinery, 23, pp. 317-327
Hwang, C.L., & Yoon, K.P.(1981). Multiple attribute decision making: Methods and Applications. New York: Springer-Verlag.
Jungwattanakit, J., Reodecha, M., Chaovalitwongse, P.,& Werner, F. (2008). Algorithms for flexible flow shop problems with unrelated parallel machines, setup times, and dual criteria. The International Journal of Advanced Manufacturing Technology, 37(3), 354-370.
Kelton, W. D., Sadowski, R. P., & Sadowaski, D. A.(2007). Simulation with ARENA 4e. New York: McGraw-Hill.
Liu, M., & Wu, C.(2003). Scheduling algorithm based on evolutionary computing in an identical machine production line. Robotics and Computer Integrated Manufacturing, 19, pp. 401-407.
Low, C. (2005). Simulated annealing heuristic for flow shop scheduling problems with unrelated parallel machines. Computers & Operations Research, 32(8), 2013-2025.
Mosheiov, G. & Oron, D. A note on the SPT heuristic for solving scheduling problems with generalized due dates. Computers & Operations Research, 31, pp. 645-655.
Pinedo, M.(1995). Scheduling-Theory, algorithms, and systems. New Jersey: Prentice-Hall.
Rockwell Technical support(2003). Arena user’s guide. Milwaukee: Rockwell Software.
Satty, T. L.(1980). The Analytic hierarchy process. New York: McGrew-Hill.
Van der Velde, S.L.(1993). Duality based algorithm for scheduling unrelated parallel machines. ORSA Journal on Computing, 5, pp. 192-205.
Weng, M. X., Lu, J., & Ren, H.(2001). Unrelated parallel machine scheduling with setup consideration and a total weighted completion time objective. International Journal of Production Economics, 70, pp. 215-226.
Yu, L., Helosia, M.S., Pefund, M., Automotivelyle, W. M., & Fowler, J. W.(2002). Scheduling of unrelated parallel machines: an application to PWB manufacturing. IIE Transactions, 34, pp. 921-931

校內：2024-08-30公開
校外：不公開電子論文尚未授權公開，紙本請查館藏目錄

簡易檢索 / 詳目顯示

相關論文