考量多通路需求公平性與彈性原則之主規劃排程演算法

隨著資訊科技的演進，市場也逐漸轉變為多通路的銷售模式，相較於一般的銷售模式，多通路的銷售模式具有更高的效率以及便利性。然而多通路銷售模式在供應鏈下游也擁有更大量且複雜的分支，使得供應鏈上游的製造商及原料供應商更難準確且全面的預估接下來的需求量。尤其對於製造商而言，必須同時承擔來自上游供應商的供給不確定性以及下游批發商的需求不確定性，因此製造商勢必需要更有效率的規劃排程方式來降低製造成本，以因應多通路銷售模式帶來的衝擊。製造商為了降低來自供應鏈上下游不確定性對生產流程造成的影響，在規劃排程時在每個生產周期預留一些空閒產能以應付緊急狀況，且對於每項需求及自家生產的每樣產品同等重視，以避免任何可能的損失。不過目前市面上大部分的排程規劃工具皆以最小化製造成本為唯一目標，所產出的排程雖然製造成本較低但非常難於執行，且必須承擔任何意外狀況所導致的風險。為此本研究提出一考量多通路需求公平性與彈性原則之主規劃排程，期望解決製造商在規劃排程上的難題。本研究所提出的考量多通路需求公平性與彈性原則之主規劃排程將公平性與彈性量化為指標，使用啟發式演算法的方式實作並以整數規劃求解的方式驗證。實際應用層面已使用臺灣一家專業代工生產公司進行情境設計及實測，就規模而言已將整數規劃求解一小時內無法找到任何可行解的案例成功壓縮在0.02秒內完成；就結果而言已將啟發式演算法規劃結果之總成本與整數規劃求解所能求得的最小總成本差距縮小至0.25% 以內。本研究提出之演算法可提供重視排程公平性及排程彈性之製造商迅速且有效的產出公平且具彈性的生產排程。使用一般成本導向規劃排程工具之製造商亦可參考演算法中公平性及彈性相關之設計，並進一步針對使用中的規劃排程工具進行調整。

關鍵字

供應鏈；多通路銷售；排程公平性；排程彈性；主規劃排程

並列摘要

The business environment has dramatically changed to multi-channel marketing en-vironment because of the progress of Information Technology. This phenomenon makes market more efficient and convenient, but it also makes the supply chain more compli-cated. The role of manufacturers in a supply chain network is responsive. Manufacturers have to coordinate with component suppliers, distributors and even retailers directly, but they can control neither the demand from retailers nor the supply from supplier but only its production capacity and on-hand inventory. To deal with the rapid transformation of the business environment, manufacturers reserve certain capacities to handle the unexpected situations. They also treat all their demands and items equally to prevent any possible loss. Almost all master planning tools have the only target of minimizing costs. The planning results are usually unrealistic in spite of minimized cost. This study proposes a master planning algorithm to resolve the difficulty on producing fair master plans. This study proposes a fairness and flexibility master planning algorithm (FFMPA) to produce a master planning result that satisfies the requirement of flexibility fairly. In terms of scale, FFMPA can produce a master planning result within 0.02 second that the MIP model cannot find any feasible solution within an hour for an identical case. From the aspect of quality, FFMPA successfully decreases the differences of overall costs to less than 0.25%, compared with results derived from the MIP model. The result of this study can help the manufacturers that put emphases on fairness and flexibility of a mas-ter plan. For those manufacturers applying cost-oriented planning tools, they can still take fairness and flexibility into account to improve their current planning tools.