Decision Support Models for Multi-plant Production Planning

本研究針對供應鏈中多廠生產規劃的問題進行研究，為了彌補現行企業採用的經驗法則 之不足，首先運用混合型整數線性規劃，建構「單階決策模式」，而為了降低問題的複雜 度及運算時間，乃進而建構「二階決策模式」。單階決策模式以成本極小化為目標式，考 量產能限制與物料限制，可求出企業最佳的生產規劃與資源分配。二階決策模式則是將 單階決策模式中所考量的成本與限制分成兩個階段考量：第一階段為「訂單分配階段」； 第二階段為「工廠規劃階段」。透過系統化之情境設計與模擬分析，可得以下結論：1. 單階決策模式能得到最佳的生產規劃計畫，但運算耗時，不適合實務應用。2.二階決策 模式能求得接近最佳的生產規劃計畫，運算時間合理，若配合企業其他供應鏈管理方案，可提升企業整體營運效能。The production planning among many sites is one of the key supply chain management issues. In this research, two decision support models for multi-plant production planning based on mixed-integer linear programming techniques are developed. The first one is a “monolithic model” that considers both the capacity and material constraints and provides an optimal production plan to minimize the overall production cost. The second one is a “two-stage model”, which first allocates the customers’ orders to different plants under the capacity constraint and the minimum cost objective. With the orders being determined, then in the second stage a different model is used to provide the optimal production plans for individual plants under the material and capacity constraints. Results of simulation study show that the two-stage decision support model is superior and provides the near optimal solutions and requires a very efficient computation effort