負荷平均排班最佳化模式之研究－以模具加工廠為例

Optimal Fair-Load Scheduling-A Case Study of a Molding Factory

10.6652/JoCICHE.201506_27(2).0007

顏上堯(Shang-Yao Yan)；陳柏華(Albert Chen)；林淑梅(Shou-Mei Lin)

模具廠 ； 排班 ； 二元整數規劃 ； 公平指派模式 ； molding factory ； work shift ； binary integer programming ； optimal fair work assignment

中國土木水利工程學刊

27卷2期（2015 / 06 / 01）

163 - 171

繁體中文

當前模具製造業大多屬24小時加工生產模式，且人工經驗排班作業為各企業單位常用但耗時的作法，常令決策者花費時間於溝通協調、安排整合調度，排班結果亦非最佳之公平指派。本研究因應現實需求及公平性，以模具加工廠之製程人員，於固定週期內從事日夜班勤務需求之人員排班為研究範圍，將現實問題之限制資源予以量化，進而轉換並建構成二元整數規劃模式，求解最佳化負荷平均之指派。本研究並調整相關參數及增減限制條件以推演方案分析，藉此求得所有製程人員之負荷平均排班，幫助決策者適度規劃人力安排，防止產生勞工超時及過勞現象致違反勞動檢查機關之規定，更可作為製造業之人力指派參考。本研究模式與傳統人工經驗指派方式相對比較下，在指派次數上節省21.05%，求解時間縮短97.44%，其負荷加權目標值、目標值、日夜班排班之次數及指定勤務出勤次數，也優於傳統人工經驗指派方式。

In view of the growing mold and die industry, and due to the fact that most current manufacturers utilize 24-hour operations, there is a need to automate the scheduling and assignment of mold and die work-shifts. Manual scheduling and assignment are time-consuming and pose problems to corporate departments. Decisions-makers often spend a great amount of time on communications and work-shift arrangements, while the outcome of the final work-shift may still be non-optimal. In this research, the actual demands and standard of fairness are factors considered for the scheduling and assignment of the process-personnel of a molding factory. The demands on work-shifts during daytime and nighttime in a fixed weekly period are considered. A Binary Integer Programming (BIP) approach is taken for the formulation of the problem. The test results using the real data from a molding factory show that the optimal work assignment with fair loading out performs the manual approach for higher efficiency and fairness. By using the BIP approach, there is a 21.04% reduction in numbers of assignment and a 97.44% reduction in deriving a solution.

1. 蔡雅萍(2016)。整數規劃法於跨院區醫師排班之應用。屏東科技大學工業管理系所學位論文。2016。1-62。