隨著資訊技術的蓬勃發展,溝通與資訊交流的距離已經走向全球化。若是企業經營模式可以擁有良好的配銷管理讓供應鏈流程更加流暢、成本的降低,那最終的利潤以及績效都會有相當大的提升。基於以上的原因,本研究透過二階規劃技術模擬了供應鏈中也具有分層式以及資訊分享性的特性,再將供應鏈中配銷模式各階的目標以數學模式套入目標函數中進行規劃,試圖找出最佳的配銷系統。 本研究提出了一種改良式演算法,它整合了遺傳演算法以及向量式粒子群最佳化演算法的特性,提出了Immune Genetic Vector-Controlled Particle Swarm Optimization Algorithm (IGVPSO) 演算法。並透過不同配銷模式的供應鏈模型來進行驗證的動作。求解的結果證明了所提出的改良式演算法的精確性即穩定性皆優於傳統的遺傳演算法、粒子群最佳化演算法,以及韓永祥 (2008) 所提出的GAPSO演算法。
With the rapid development of information technology, communication and information is moving toward globalization. If the business model could have a good management of supply chain management to make the flow more smooth and reduce the costs, the profit and performance will have a considerable improvement. Based on the above reasons, this research utilized bi-level linear programming technique to simulate the hierarchical relationship and information sharing between upper and lower levels of supply chain by transforming each level's objective of the supply chain into the mathematical model's objective function in order to determine the best distribution model. This research also proposes a hybrid method which integrates the characteristic of immune genetic algorithm (IGA) and the vector-controlled particle swarm optimization algorithm. The proposed method is called Immune Genetic Vector-Controlled Particle Swarm Optimization Algorithm (IGVPSO). It is validated by using different distribution models. The results demonstrated that the modified algorithm is more accurate and stable than genetic algorithm, particle swarm optimization algorithm and GAPSO designed by Han (2008).