DC 欄位 值 語言 DC.contributor 土木系營建管理博士班 zh_TW DC.creator 陳致霖 zh_TW DC.creator Chih-Lin Chen en_US dc.date.accessioned 2020-7-24T07:39:07Z dc.date.available 2020-7-24T07:39:07Z dc.date.issued 2020 dc.identifier.uri http://ir.lib.ncu.edu.tw:444/thesis/view_etd.asp?URN=105385001 dc.contributor.department 土木系營建管理博士班 zh_TW DC.description 國立中央大學 zh_TW DC.description National Central University en_US dc.description.abstract 國內預鑄產業最近面臨著三大挑戰，缺乏資源分配活動之間的邏輯關係、難以確定複雜構件的優先次序及因勞工密集加班而導致人力成本不斷增加。本研究之目的(1)利用約略集合理論增強K-近鄰演算法(Rough-set enhanced KNN)對預鑄結構體大樑、小樑及柱等14個生產工項的關鍵項目進行探索；(2)透過基於自我組織特徵映射圖之動作軌跡相似度測量法(SOMMTS)確定專案資源項目複雜度；(3)利用Arena模擬建立工人生產人力成本預測模型。 在綜合文獻回顧的基礎上，專家訪談提出了五個主要假設，包含按工作訂單成本法計算成本、合理的人力成本範圍、加班小時費率的範圍、加班生產進度庫存及生產力評估方式。本研究依構件類型彙集出臺灣近10年來共有55,157根預鑄構件生產數量，並蒐集預鑄生產數據的總體情況，包含專案資源項目、每月工人加班成本及工項生產時間等數據。 本研究之結論共分為三項:(1)透過約略集合理論增強K-近鄰演算法(Rough-set enhanced KNN)分別識別了大樑、小樑及柱構件生產的5個、7個及3個關鍵生產項目，以作為影響預鑄構件生產效率之決策因子；(2)利用基於自我組織特徵映射圖之動作軌跡相似度測量法(SOMMTS)為所有專案類型的預鑄構件生產提供了3個複雜度級別；(3)透過Arena系統模擬降低6.9%總生產人力成本。本研究之成果使決策者能針對預鑄廠各項資源規畫提供最佳化配置之參考依據。 zh_TW dc.description.abstract Construction precast factories have recently faced three major challenges of lack of logical relationships among activities to allocate resources, difficulties to identify priorities for complex components, and cost saving on worker fee due to intensive overtime. The research objectives are (1) to explore the key activities among 14 activities using rough set enhanced k-mean nearest neighbor (KNN), (2) to determine component complexity using SOM-based motion trajectory similarity measure (SOMMTS), and (3) to establish a prediction model for worker costs using Arena simulation. Based on a comprehensive literature review, the expert interviews suggest five major assumptions involving cost calculation by job order costing method, reasonable manpower cost range, range for overtime hourly rates, inventory availability by work overtime, and ignorance for individual difference. Expertise also gives us an overall picture to collect production data including activity sequences, activity duration in minutes, and worker wage in hourly rate for construction precast components. This results in a total of 55,157 precast components or 772,198 activities in details, making up 90% of total data for precast component productivity in recent 10 years in Taiwan. The findings show that (1) set enhanced KNN identifies 5, 7, and 3 key activities for main beam, beam, and column component production, respectively; (2) SOMMTS yields 3 complexity levels for all types of precast component production; (3) Arena simulation brings about 6.9% saving for the total cost. en_US DC.subject 預鑄工法 zh_TW DC.subject 生產力 zh_TW DC.subject 人力彈性運用 zh_TW DC.subject K-近鄰演算法 zh_TW DC.subject 約略集合理論 zh_TW DC.subject 自我組織特徵映射圖網路 zh_TW DC.subject 複雜度 zh_TW DC.subject 節省人力成本 zh_TW DC.subject Construction precast component en_US DC.subject Productivity en_US DC.subject KNN en_US DC.subject Rough set en_US DC.subject SOM en_US DC.subject Simulation en_US DC.subject Complexity level en_US DC.subject Manpower cost saving en_US DC.title 運用組織特徵映射圖動作軌跡相似度測量法探索預鑄工項生產效率與資源規劃之研究 zh_TW dc.language.iso zh-TW zh-TW DC.title Exploring productivity and resource allocation for construction precast compomements using SOMMTS en_US DC.type 博碩士論文 zh_TW DC.type thesis en_US DC.publisher National Central University en_US