隨著全球產業競爭加劇,如何快速地反應顧客的實際需求與有效地降低成本為管理階層所重視之焦點。在傳統的存貨管理之中,上、下游廠商各自追求本身成本的最小化而並非整體供應鏈之利益最佳化,進而無法達成上述之目標。因此,本研究提出一運送前置時間不確定及品質不可靠環境下之多階整理即時化存貨模型,在供應鏈中各階層成員可以交換相關資訊的條件下同時進行生產、採購和運送等決策,以達成整體供應鏈總成本最小化之目標。 求解運送前置時間不確定及品質不可靠環境下之多階整理即時化存貨模型等同於求解一個非線性混整數(Mixed integer)問題,此類NP-hard問題通常無法以傳統方法解決。啟發式演算法近年來常被用於複雜度高且難以解決的NP-hard問題,而粒子群最佳化(Particle Swarm Optimization; PSO)即是一種具有高求解效率與品質的啟發式方法。 本研究透過三種不同的PSO權重遞減策略針對序列式多階存貨模型進行求解並分析不同參數下的結果,再與最佳化軟體LINGO比較。經研究結果顯示,在適當的參數設定下PSO能相當有迅速且有效地解決多階整合即時化存貨模型。
How to respond the real require of customer and reduce the total cost effectively are the major focuses in this highly competitive global supply chain environment. In traditional inventory management system, the vendor and the buyer find their own optimal economic-lot-size respectively but not result in an optimal policy for whole supply chain. In addition, uncertain delivery lead time and the quality unreliability are two common factors that could make a great influence the optima inventory policy. Therefore, a serial multi-echelon integrated JIT inventory model with uncertain delivery lead time and quality unreliability (SMEIJI model) was proposed. Particle Swarm Optimization (PSO) is a burgeoning heuristic approach that has proven to have good performance and quality in solving NP-hard problems. In this research, PSO is applied to search the optimal solution of SMEIJI problem. Experiment results show that PSO is rapid and efficient in solving the multi-echelon inventory problem under the proper parameter setting.