運用蟻群演算法於電子商務配送之研究

Ant Colony System on E-Commerce Delivery Model

鄭旭軒(Hsu-Hsuan Cheng)

指導教授 ： 郭瑞祥

共同指導教授 ： 蔣明晃(Ming-Huang Chiang)

國立臺灣大學/工學院/工業工程學研究所/碩士(2016年)

https://doi.org/10.6342/NTU201600505

摘要

隨著科技日新月異，網路越來越發達，現代人的生活與消費模式也跟著改變，從以往必須到實體店面選購商品，到現在藉由電腦和手機在家中也可以輕鬆完成購物，連帶著也促進了B2C電子商務的發展，許多企業爭相提出不同的策略搶奪市場這塊大餅，其中在配送策略上，時效性便扮演了成本以外不可或缺的腳色，如何在規定時間內將商品準確的送達顧客手中便是致勝的關鍵。
有鑑於此，本研究之目的為建置一套決策系統，幫助B2C業者在滿足其提出之時間限制前提下，能夠快速決定設施設立位置、數量、車輛數及配送繞行路線，卻又不會花費大量成本。
過去有關B2C設施途程問題的文獻，通常使用較為傳統的演算法作為求解手段，或在演算過程中耗費太多時間，造成求解效率不彰，因此本研究採用兩階段模型並修改時窗限制下多重蟻群演算法，提出新的啟發式演算法，用以分別求解最小車輛數與最短配送路線，減少B2C業者車輛配送所需成本。
最後以國內知名B2C業者之訂單資料為研究對象，導入本研究所提出之啟發式演算法，探討不同參數組合下對設施數量、位置、車輛數與繞行路線之影響，彙整結論後提供給B2C業者作為發展自有決策系統或訂定策略之參考依據。

關鍵字

B2C電子商務物流 ； 區位途程問題 ； 蟻群演算法 ； 配送模型

並列摘要

Flowing by the advance of technology and the innovation, the internet has become more and more developed. Nowadays, people have changed their life and consumption patterns. We don’t have to go and purchase in the physical stores but shopping on the internet by our computer or cell phone easily.
These changes promoted the development of B2C e-commerce. Many enterprises competing to propose different strategies in order to get B2C e-commerce market. Wherein the delivery strategy, timeliness will play an indispensable role other than cost. How to deliver goods to the hands of customers accurately in the specified time is a sticking point.
The purpose of this study is to establish a decision-making system which can help B2C e-commerce enterprise to decide the facility location, number of facilities, the number of vehicle required and delivery route rapidly under the premise of the time limit enterprise proposed without spending a lot of cost.
In the earlier literature, the location routing problem associated with B2C e-commerce commonly used traditional algorithm as the method of solving or spend too much time during the process.
Therefore, this study uses a two-stage model and modify the multiple ant colony system for vehicle routing problem with time windows(MACS-VRPTW)and propose a new heuristic algorithm to solve facility location problem and vehicle routing problem respectively and also minimize the distribution cost.
At last, the study draws data from the well-known B2C e-commerce enterprise and import in the heuristic algorithm which proposed in this study to explore the impact on the facility location, number of facilities, the number of vehicle required and delivery route by different combinations of parameters.
The findings in this study can provide the managers a reference to develop its own decision-making system or setting strategies.

並列關鍵字

B2C E-commerce Logistic ； Distribution Model ； Location-Routing Problem ； Ant Algorithm

參考文獻

1. 施朝翔 (2015). "基於總成本考量下 B2C 電子商務配送模式之研究." 臺灣大學工業工程學研究所學位論文: 1-48.

3. 張易晟 (2014). "B2C 電子商務下都會區配送模型之研究." 臺灣大學工業工程學研究所學位論文.

21. Nemoto, T., J. Visser and R. Yoshimoto (2001). Impacts of information and communication technology on urban logistics system. Joint OECD/ECMT Seminar on the impacts of E-commerce on Transport, Citeseer.

5. Bullnheimer, B., R. F. Hartl and C. Strauss (1999). "An improved ant System algorithm for thevehicle Routing Problem." Annals of operations research 89: 319-328.

6. Clarke, G. and J. W. Wright (1964). "Scheduling of vehicles from a central depot to a number of delivery points." Operations Research 12(4): 568-581.