本研究發展一個混整數模擬最佳化方法來求解兩種不同情境的兩階層
存貨問題，兩個情境中存貨系統皆包含一個中心倉庫以及多個零售商，主要目的在於求解最佳的存貨水準與盤點間隔時間，並且在達到一定的服務水準下使總成本最小化。而本研究提出的 cgR-DSSNE 演算法(converge global
Retrospective with Divide Subspace Search and Neighborhood Enumeration) 包含 cgR-
SPLINE(converge global Retrospective Search with Piecewise-Linear Interpolation and
Neighborhood Enumeration) 中內外圈迭代設計與隨機限制式判斷方法，並透過 DSS(Divide Subspace Search) 將決策變數分割為多個二維度空間，進行梯度搜尋，最後在固定整數型變數，利用二分法方法來處理唯一一個連續型變數。 本研究將利用兩個情境來比較 cgR-DSSNE 與 MISO 兩個演算法的求解效率，而實驗情境例子中，情境一包含高維度與低維度例子，而情境二只有低維度例子，每個例子皆透過兩個演算法進行求解。實驗結果發現，當樣本預算較低時， cgR-DSSNE 收斂到的最佳解品質皆比 MISO 來的好，當樣本預算逐漸增加時，MISO 找到的解會慢慢接近 cgR-DSSNE所求得的解，甚至在情境一例子二與情境二例子一中，MISO 在樣本預算較大的情況下，迭代解的品質較 cgR-DSSNE來的好。

We develop a mixed-integer simulation optimization algorithm to solve a two-echelon
inventory problem in two different scenarios. Both scenarios consist of an external sup-
plier, a central warehouse, and some retailers. The objective in the optimization problem
is to determine the order-up-to-levels at the warehouse and retailers, as well as the re-
plenishment interval at the warehouse and retailer that minimizes a cost function.
The cgR-DSSNE algorithm proposed in this study comprises an inner and outer iteration
and feasibility check method for stochastic constraint in cgR-SPLINE, and DSS is used
to divide the solution space into multiple two-dimensional spaces to perform a gradient
search, and finally fix the integer variable using the bisection method to deal with a
single continuous variable.
This study uses two scenarios to compare the performance of the two algorithms, cgR-
DSSNE and MISO, while Scenario 1 contains high-dimensional and low-dimensional
cases, and Scenario 2 contains only a low-dimensional case, each of which is solved
with the two algorithms. The experimental results indicate that with a lower simulation
budget, cgR-DSSNE finds a better solution than MISO, but when the simulation
budget gradually increases, the solution found by MISO will slowly approach that of
cgR-DSSNE. Even in Scenario 1 for case 2 and Scenario 2, the quality of the iterative
solution in the case of MISO with a large simulation budget is better than that of
cgR-DSSNE.

[1] 李維德(2017)。以回溯最佳化求解服務水準限制式之兩階層存貨問題。國立成功大學工業與資訊管理系碩士論文。
[2] Andersson, J., & Melchiors, P. (2001). A two-echelon inventory model with lost sales. International Journal of Production Economics, 69(3), 307-315.
[3] Andradottir, S., & Kim, S.H. (2010). Fully sequential procedures for comparing con-strained systems via simulation. Naval Research Logistics, 57(5), 403-421.
[4] Bashyam, S., & Fu, M.C. (1998). Optimization of (s, S) inventory systems with ran-dom lead times and a service level constraint. Management Science, 44(12), 243-256.
[5] Batur, D., & Kim, S.H. (2010). Finding feasible systems in the presence of constraints on multiple performance measures. ACM Transactions on Modeling and Computer Simulation, 20(3), 13:1-13:26.
[6] Bayliss, C., De Maere, G., Atkin, J.A.D., & Paelinck, M. (2017). A simulation sce-nario based mixed integer programming approach to airline reserve crew scheduling under uncertainty. Annals of Operations Research, 252(2), 335-363.
[7] Bijvank, M., & Vis, I.F. (2012). Lost-sales inventory systems with a service level criterion. European Journal of Operational Research, 220(3), 610-618.
[8] Caglar, D., Li, C.L., & Simchi-Levi, D. (2004). Two-echelon spare parts inventorysystem subject to a service constraint. IIE Transactions, 36(7), 655-666.
[9] Chen, F., & Zheng, Y.S. (1997). One-warehouse multiretailer systems with central-ized stock information. Operations Research, 45(2), 275-287.
[10] Chen, F.Y., & Krass, D. (2001). Inventory models with minimal service level con-straints. European Journal of Operational Research, 134(1), 120-140.
[11] Chen, H., & Schmeiser, B.W. (2001). Stochastic root finding via retrospective ap-proximation. IIE Transactions, 33(3), 259-275.
[12] Cheung, K.L., & Lee, H.L. (2002). The inventory benefit of shipment coordination and stock rebalancing in a supply chain. Management Science, 48(2), 300-306.
[13] Fu, M.C., Glover, F.W., & April, J. (2005, December). Simulation optimization: A review, new developments, and applications. Paper presented at 2005 Winter Simula-tion Conference, Orlando, FL.
[14] Gruen, T.W., Corsten, D.S., & Bharadwaj, S. (2002). Retail out-of-stocks: A world-wide examination of extent causes and consumer responses. Washington, DC: Gro-cery Manufacturers of America.
[15] Guijarro, E., Cardös, M., & Babiloni, E. (2012). On the exact calculation of the fill rate in a periodic review inventory policy under discrete demand patterns. European Journal of Operational Research, 218(2), 442-447.
[16] Hill, R.M., Seifbarghy, M., & Smith, D.K. (2007). A two-echelon inventory model with lost sales. European Journal of Operational Research. 181(2), 753-766.
[17] Hong, L.J., & Nelson, B.L. (2006). Discrete optimization via simulation using COMPASS. Operations Research, 54(1), 115-129.
[18] Huh, W.T., Janakiraman, G., Muckstadt, J.A., & Rusmevichientong, P., 2009.
Asymptotic optimality of order-up-to policies in lost sales inventory systems. Man-agement Science, 55(3), 404-420.
[19] Lee, L.H., Pujowidianto, N.A., Li, L.W., Chen, C., & Yap, C.M., (2012). Approx-imate simulation budget allocation for selecting the best design in the presence of stochastic constraints. IEEE Transactions on Automatic Control, 57(11), 2940-2945.
[20] Li, J.C., Gong, B., & Wang, H. (2016). Mixed integer simulation optimization for optimal hydraulic fracturing and production of shale gas fields. Engineering Opti-mization, 48(8), 1378–1400.
[21] Nagaraj, K. (2014). Stochastically constrained simulation optimization on mixed-integer spaces (Doctoral dissertation, Virginia Polytechnic Institute and State Univer-sity).
[22] Nagaraj, K., & Pasupathy, R. (2017). Stochastically constrained simulation opti-mization on integer-ordered spaces: The cgR-SPLINE algorithm. Operations Re-search, forthcoming.
[23] Ozdemir, D., Yücesan, E., & Herer, Y.T., (2013). Multi-location transshipment prob-lem with capacitated production. European Journal of Operational Research, 226(3), 425-435.
[24] Park, C., & Kim, S.H., (2015). Penalty function with memory for discrete optimiza- tion via simulation with stochastic constraints. Operations Research, 63(5), 1195- 1212.
[25] Schwarz, L.B., (1989). A model for assessing the value of warehouse risk-pooling: risk-pooling over outside-supplier leadtimes. Management Science, 35(7), 828-842.
[26] Sezen, B., (2006). Changes in performance under various lengths of review periods in a periodic review inventory control system with lost sales: A simulation study.
International Journal of Physical Distribution Logistics Management, 36(5), 360-373.
[27] Sherbrooke, C.C. (1968). METRIC: A multi-echelon technique for recoverable item control. Operations Research, 16(1), 122-141.
[28] Truong, T. H., & Azadivar, F. (2003, December). Simulation based optimization for supply chain configuration design. Proceedings of the Winter Simulation Conference 2003, New Orleans, Louisiana.
[29] Tsai, S.C., & Zheng, Y.X. (2013). A simulation optimization approach for a two-echelon inventory system with service level constraints. European Journal of Opera-tional Research, 229(2), 364-374.
[30] Wang, H. (2012). Retrospective optimization of mixed-integer stochastic systems using dynamic simplex linear interpolation. European Journal of Operational Re- search, 217(1), 141-148.
[31] Wang, H. (2013, December). Mixed integer simulation optimization for petroleum fielddevelopmentundergeologicaluncertainty.ProceedingsoftheWinterSimulation Conference 2013, Washington, DC.
[32] Wang, H. (2017). Subspace dynamic-simplex linear interpolation search for mixed-integer black-box optimization problems. Naval Research Logistics, 64(4), 305-322.
[33] Wang, H., Pasupathy, R., & Schmeiser, B.W. (2013). Integer-ordered simulation op-timization using R-SPLINE: Retrospective search with piecewise-linear interpolation and neighborhood enumeration. ACM Transactions on Modeling and Computer Sim- ulation, 23(3), 17:1-17:24.
[34] Wang, Q. (2013). A periodic-review inventory control policy for a two-level supply chainwithmultipleretailersandstochasticdemand.EuropeanJournalofOperational Research, 230(1), 53-62.
[35] Wang, Q., & Axsäter, S. (2010). Fixed-interval joint-replenishment policies for dis-tribution systems with multiple retailers and stochastic demand. Naval Research Lo- gistics, 60(8), 637-651.
[36] Zhang, J., Bai, L., & He, Y. (2010). Fill rate of general periodic review two-stage inventory systems. International Journal of Operational Research, 8(1), 62-84.