An Improved Imperialist Competitive Algorithm to Solve the Dynamic Vehicle Routing Problem in Logistics Management

• Main Authors: Jen-Chun Song, 宋仁鈞
• Other Authors: Shih-Che Lo
• Format: Others
• Language: en_US
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/w98232

碩士 === 國立臺灣科技大學 === 工業管理系 === 106 === In today's fierce competitive environment, any enterprise considers increasing its profitability as main purpose. Apart from reducing production costs and increasing sales, the most important thing is to increase the efficiency of logistics. After the fourth industrial revolution, the construction and integration of the intelligent logistics system will be even more important. Therefore, companies must use efficient logistics networks to rapid respond customer’s requests. The concept of intelligent logistics is a new way of thinking to effectively dispatch vehicles to reduce the company's fixed and various costs. This thesis proposed a method based on the Imperialist Competition Algorithm combined with the sweep method, called the sICA, to generate a near optimal solution quickly to the dynamic vehicle routing problem (DVRP) at different time segments. The main goal is to satisfy all constraints while minimizing the total cost (transportation and operation). In order to verify the effectiveness of the proposed sICA algorithm, we compared the sICA to the genetic algorithm (GA). Experimental results from 60 DVRP problems show that the sICA algorithm has better performance compared with GA in combinatory optimization problems.