| Abstract: | 本研究探討網路通設備製造業之物料規劃，其物料規劃具有以下特殊考量：(1)原物料多具有共用性；(2)生產排程必須考量齊料，使製程能連續加工及避免產線平衡異動。因此，為了達到『物料最大化』，使物料規劃的整體收益最大，齊料邏輯與共用料配置具有關鍵性的影響，成了當前網路通訊設備製造業在物料規劃重要的議題。現今企業不論是透過ERP或先進規劃排程 (Advanced Planning and Scheduling)系統等企業應用軟體進行物料規劃，多採用以優先權為基礎(priority-based)啟發式的規劃方法。然而，這樣的方法雖然求得可行解，卻無法達到最佳解或近似最佳解。過去物料規劃的研究中，甚少針對網路通訊設備業之齊料邏輯與共用料配置進行探討，大多數的研究多採用由上而下(top-down)的規劃邏輯，關注在未來應如何規劃，考量批量、產能限制、需求變動等種種因素，希望達到最小化的成本；卻鮮少以由下而上(bottom-up)的規劃邏輯，在現階段的需求無法被完全滿足，如何規劃及配置資源，使整體規劃收益能更佳。因此，本研究考量網路通訊設備製造業之物料規劃特性與齊料邏輯，在多階產品、多品項、多共用料的環境下，提出一基因演算法求解如何配置共用料以提升物料規劃之收益。最後，進行演算法之績效評估並與優先權為基礎(priority-based)的啟發式規劃方法比較及驗證。
Enterprises today are facing ever increasing challenges because of evolving globalization and the increasingly severe competitive environment, how to efficiently using and planning resource has become an important issue to enhance a business's competitiveness. This study uses a case of network communication equipment manufacturer. A network communication equipment manufacturer, not only produces switch, router and wireless modem, but also develops the network communication software. Because the product specifications are very complicated and vary dramatically, many materials are common parts for different products, moreover, must consider the process continuous and avoid changing the line balance. To achieve “maximizing the material”, that means maximizing the profit after material planning, become an important issue of material planning in network communication equipment manufacturer.In past, most of researches for material planning focus on establish the lot size of production and purchase, few consideration product structure, using the Top-down planning logic, or not allow backorders. Actually few in view of in existing materials and limited resources, using Bottom-up planning logic to efficiently allocate the current resource to fulfill the demand.Hence, this study proposed an algorithm, consider the material planning characteristic in network communication equipment manufacturer, for environment of multi-level product, multi items, multi common parts, and also consider the full-scale material planning(齊料) logic, then using genetic algorithm to solve the issue of “maximizing the material” in network communication equipment manufacturer. |
