An efficient evolutionary computation approach for the vending machine inventory control problem

Abstract

This paper considers the policy of economic order quantity for the geographically distributed vending machine inventory control problem. In the problem, the economic order policy for each product and the quantity of each type of product for every vending machine are to be decided simultaneously so as to maximize the total expected profit. Due to each ordering cost follows the corresponding skip function, the proposed problem belongs to the polynomial concave problem. Also, the optimal economic ordering quantity is based on the random demand for each type of product in different vending machine. Since the above reasons, it will be very hard to find the close form solution effectively and efficiently by using the mathematical programming approaches. In this research, a hybrid evolutionary computation (EC) approach has been applied to solve the problem efficiently and effectively. Experimental results are presented to exhibit the superior performance of the proposed methodology for solving the economic order quantity with quantity discount problem.

摘要

在這篇論文裏探討對於分散於不同地點之自動販賣機存貨控制的經濟訂購量政策制定之問題。於此問題，必須同時決定對各類產品的經濟訂購政策以及分配至每部自動販賣機裏各類產品分別的數量必須同時被決定，以使期望利潤最大化。由於訂購成本爲階梯關數(Skip function)，所以這個問題屬於多凸整數計劃問題。另外，由於每一自動販賣機裏不同產品具有不同的隨機需求，因此若以數學規劃的方法來求解將會遭遇到若干困難，諸如無法獲得封閉解(Close form)、無法有效且迅速的求得最佳的訂購量等。爲能迅速且有效快速的獲得訂購政策，於此研究，一個進化演算法被應用來解決此一問題並且能夠迅速且有效的來求解。由實驗數據顯示此方法可有效且快速的對於解決具有折扣多品目多販賣機的經濟訂購量問題的優越性。

Keywords:

• Economic order quantity
• Quantity discounts
• Evolutionary Computation

關鍵詞:

• 經濟訂購量
• 折扣訂購量
• 進化演算法

Additional information

Notes on contributors

Ta-Cheng Chen

Correspondence addressee