A Genetic Algorithm for Minimizing Carbon Footprint for the Time-Dependent Heterogeneous Fleet Vehicle Routing Problem with Alternative Paths

• Main Author: 曹祐菘
• Other Authors: Lin, Chun-Cheng
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/49243805946778252554

碩士 === 國立交通大學 === 工業工程與管理系所 === 101 === This study proposes a vehicle routing problem for minimizing carbon footprint by selecting the appropriate vehicles and routes. It is referred to as the “Minimizing Carbon Footprint for the Time-Dependent Heterogeneous Fleet Vehicle Routing Problem with Alternative Paths, CTHVRPP.” The objective of this problem is to minimize carbon footprint rather than distance or time. Thus, the resulting solution can help reduce greenhouse gas emissions and global warming. Since the vehicle routing problem is itself an NP-Hard problem, it can be inferred that CTHVRPP is also an NP-Hard problem. So we developed an enhanced genetic algorithm to resolve this problem. In the process of devising an appropriate coding and decoding system, this study developed a sequential chromosome coding method and a parallel chromosome decoding method. The sequential chromosome encoding method is best suited for crossover and mutation encoding operations, while the parallel chromosome decoding method is typically used in decoding sequentially encoded chromosomes. It is known that carbon footprint size is dependent on the vehicle energy consumption, thus this study improved the energy consumption model Bektaş et al. (2011) proposed, making this study possible. In our test scenario, we applied a test benchmark developed by Taillard (1999), which required the identification of cost variables that affect carbon footprint, including: vehicle speed, vehicle traveling on different paths during different time periods of the day, and the speed of each vehicle type when carrying an empty load and maximum load. The test results were then analyzed by applying various objectives such as carbon footprint minimization, time minimization, and distance minimization. These results were compared with results from scenarios without alternative routes available for selection. The results show that the alternative path in this experiment factor is has a significant impact on the experimental results. In this paper, the genetic algorithms optimization of vehicle scheduling, according to the experimental results, the proposed vehicle scheduling results with the experimental results better in minimum carbon footprint as the target case.