應用和弦搜尋演算法求解多橫向通道之批次人工揀貨問題

Applying Harmony Search to Solve the Multi-Cross Aisle with Batch Order Policies in Manual Picking Problem

黃建興(Chien-Hsing Huang)

指導教授 ： 胡黃德

元智大學/工程學院/工業工程與管理學系/碩士(2011年)

https://doi.org/10.6838/YZU.2011.00361

摘要

在物流中心所涵蓋的作業範圍裡，揀貨作業是其中一項重要且繁雜的工作。平均而言，揀貨作業時間約佔整個物流中心作業時間比的30% ~ 40%，因此若能妥善規劃揀貨作業將能有效降低物流中心之營運成本、人力及時間。過去之研究主要以探討單一定單或批次定單之揀貨路徑為主，較少討論到橫向走道數目對於求解目標值之影響。本研究之目的為使用和弦搜尋演算法 (Harmony Search, HS) 來改善定單揀貨作業初始解之訂單批次數，並以達到最小的揀貨距離的目標。本研究並且透過加入多條橫向走道設計 (0、1、3、4) 及使用隨機、最少揀貨點及最小矩形覆蓋面積策略之三類種子訂單選取策略，搭配隨機化、最小增加走道數及最短平均相互最近直角距離策略之三種訂單選取策略將訂單批次化，並配合穿越策略、最大間隙策略與合成策略之三種路徑法產生初始解，最後利用和弦搜尋演算法進行改善。根據搜集的之實驗結果，本研究使用MINITAB統計軟體分析各種策略組合的求解績效。結果發現，不同訂單批次化方法對於HS之求解績效沒有顯著性差異，並且以橫向走道數3之倉儲佈置，其整體揀貨距離為最佳。此外，本研究與文獻所提之粒子群演算法 (Particle Swarm Optimization, PSO) 進行比較，發現HS在大部分策略組合之求解績效都較PSO為佳。

關鍵字

物流中心 ； 揀貨作業 ； 訂單批次化 ； 多條橫向走道 ； 和弦搜尋演算法

並列摘要

The order picking is one of the important and complicated operations in a distribution center.　On the average, the order picking time accounts for 30%~40% of total operation time in the distribution center.　Hence, a well planned order picking strategy will reduce the operating cost, manpower and time of the distribution center effectively.　Former researches in the order picking problem are focused on the single order or batch order picking policies. However, fewer of them are concerned about the number of cross-aisles in the search of solution quality. This study aimed to apply the Harmony Search (HS) to improve the batch order quantity of the initial solutions with the objective of minimum order picking travel distance.　This study also added in multi-cross aisle design (with 0, 1, 3, 4) and used three seed order selection strategies (Random, Smallest Number of Picking Locations, and Smallest Rectangular Covering Area) combined with three accompanying order selection Strategies (Random, Smallest Number of Additional Picking Aisles, and Shortest Average Mutual Nearest Rectangular Distance) to determine the initial solutions with three route planning methods (Traversal, Largest Gap, and Combined). Then, we applied HS to improve the initial solutions for these test problems. The computational performances of each combined method were tested by the statistical software MINITAB. The results showed that order batching methods had no statistical significance on the performances, and the order picking travel distance achieved optimally with the cross aisle quantity equaled to three.　 Besides, this study also compared with Particle Swam Optimization (PSO) method, the computational results showed that the performances of HS outperformed PSO.

並列關鍵字

Harmony Search ； Distribution Center ； Order picking ； Order Batching ； Multiple Cross Aisle

參考文獻

張有恆，「現代物流管理」，華泰文化事業股份有限公司，2005。

吳舲雨，「應用模擬退火法於批次訂單揀取環境下之物流中心人工揀貨系統」，元智大學工業工程與管理研究所，2008。

羅婉珍，「粒子群演算法應用於物流中心批次人工揀貨之研究」，元智大學工業工程與管理研究所，碩士論文，2009。

Bassan, Y., Roll, Y. and Rosenblatt, M. J. (1980). Internal layout design of a warehouse. IIE Transactions, 12(4), 317-322.

Bozer, Y. A. and Kile, J. W. (2008). Order batching in walk-and-pick order picking systems. International Journal of Production Research, 46(7), 1887-1909.