應用蟻群演算法於不一致部門大小之動態設施規劃

Applying Ant Colony Optimization To Solve Unequal-Area Dynamic Facility Layout Problem

林芃萱(Peng-Hsuan Lin)

指導教授 ： 陳育欣

中原大學/電機資訊學院/工業與系統工程研究所/碩士(2012年)

https://doi.org/10.6840/cycu201200513

摘要

現今全球化的影響之下，各企業之間競爭越來越激烈，企業為了生存，必須不斷更新生產策略與提高生產效率，同時降低成本。一個完善的設施規劃將成為影響各企業生產效率的問題。
在目前文獻中，鮮少將不一致部門面積問題（Unequal-Area Problem, UAP）與動態設施規劃問題（Dynamic Facility Layout Problem, DFLP）同時間考量。然而，在現實環境中，企業常常會隨著生產策略的更新，進行部門順序上的更換，且在企業中的部門面積並非皆為一致性。因此，本研究將以彈性區帶架構（Flexible Bay Structure, FBS）的設施規劃方式同時考量不一致部門大小以及動態設施規劃，進行設施安排。
在企業中，部門面積通常為不一致大小，以安排設施的成本考量，由於以往文獻多半以部門中心點對中心點的搬運方式，計算部門之間交換的費用。然而，在現實環境下，這種方式太過抽象，無法實現無視牆壁以及周遭工作機台的情況，因此，本研究提出設置部門出入口點，同時搬運方式以出入口點沿著走道移動為搬運距離。
設施規劃問題為一個NP-hard問題，本研究應用蟻群最佳化演算法（Ant Colony Optimization, ACO）於設施規劃問題上，並求解實際案例問題，同時與粒子群演算法進行比較，研究結果顯示，蟻群最佳化演算法的解優於粒子群演算法的解，且部門規劃能應用於實際案例上。

關鍵字

蟻群最佳化演算法 ； 彈性區帶架構 ； 不一致部門大小 ； 動態設施規劃

並列摘要

As the world economy moves toward globalization, the increasing competition between enterprises has been becoming fierce. Enterprises, in order to survive, must continually update production strategy and improve production efficiency while reducing costs. One of important factors that impact production efficiency is the design of facilities.
Currently, little attention is given to considering the Unequal-Area Problem (UAP) and the Dynamic Facility Layout Problem (DFLP) at the same time. However, in the real environment, companies often will update production strategies and rearrange departments in sequence. And in the enterprise, department areas typically are not all equal. Therefore, this study takes Flexible Bay Structure (FBS) approach to consider Unequal-Area and Dynamic Facility Layout at the same time.
Taking a layout’s costs into account, most literatures take the centroid-to-centroid approach to compute costs after departmental exchanges; nevertheless, that approach is abstracted from the real-world scenario, and some distances are skipped from consideration. Accordingly, this study sets the departmental exit points and measures the distances for transportation along aisles to the entry points of some other departments.
Facility Layout Problem is a NP-hard problem. In this study, the Ant Colony Optimization (ACO) is used on the facility layout problem and compared with the Particle Swarm Optimization over data from a metal parts factory. The findings from the study have shown that Ant Colony Optimization is better than Particle Swarm Optimization and is promising to solve the unequal, dynamic facility layout problem in a real-world situation.

並列關鍵字

Ant Colony Optimization ； Keywords : Unequal-Area ； Dynamic Facility Layout ； Flexible Bay Structure

參考文獻

林欣怡. 2011. "改良式蟻群演算法應用於不等面積設施佈置問題."中原大學土木工程學系.

羅儒杰. 2011. "應用蟻群演算法求解多目標動態設施規劃問題."中原大學工業與系統工程學系.

鄭茂宏. 2012. "應用粒子群演算法於不一致部門大小之動態設施規劃."中原大學工業與系統工程學系.

Balakrishnan, J., F.R. Jacobs, and M. A. Venkataramana. 1992. "Solutions for the Constrained Dynamic Facility Layout Problem." European Journal of Operational Research 57 (2): 280-286.

Bozer, Y. A. and C. T. Wang. 2011. "A Graph-Pair Representation and MIP-Model-Based Heuristic for the Unequal-Area Facility Layout Problem." European Journal of Operational Research.