On the Study of Two-stage Artificial Immune System to Solve the Vehicle Routing Problem with Cross-Docking in Logistics Management

• Main Authors: Chuang, Ying-Lin, 莊英林
• Other Authors: Shih-Che Lo
• Format: Others
• Language: en_US
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/23481958070329509529

碩士 === 國立臺灣科技大學 === 工業管理系 === 100 === All enterprises want to make profits in the extremely competitive environment. In addition to expanding sales and reducing manufacturing cost, the efficiency of logistics management is also considered as the additional source of profit. Increasing efficiency of logistics becomes critical in the supply chain due to customer’s quick response requirements. Therefore, cross-docking (CD) system in the supply chain is considered a good method to reduce inventory and improve responsiveness to various customer demands. This thesis focuses on the vehicle routing problem with cross-docking (VRPCD) aiming at synchronizing the shipments in both pickup and delivery processes concurrently. The collective effort of VRPCD is to reduce handling cost, inventory cost and transport cost to fulfill the distribution services. A two-stage algorithm based on the artificial immune systems (AIS) combined with the sweep method, called sAIS, is proposed in this thesis to find the combinatorial optimum solution of the vehicle routing problem with cross-docking. Since the complexity of the problem is NP-hard, the proposed meta-heuristic method can quickly generate a near optimum solution. Comparisons are made between the proposed method and the genetic algorithm (GA) over the experiments of various VRP pickup and delivery benchmark problems to validate the performance of the sAIS approach. Experiment results show that the sAIS algorithm was able to discover new optimum solutions than the GA for all 60 benchmarks problems. Also, the computational results show that the proposed algorithm is robust, converge fast to near optimal solution and competitive with overall improvement of 7.26% over the GA method.