Yapay Zeka Hızlı Bakış & Çıkarımlar
AI InsightHowever, it should be noted that the MILP model outperforms the CP model in terms of solution quality and CPU time for the instances with a long planning horizon.
📌 3 Maddede Bu Makale
Abstract
This study considers the milk dispatching problem for a small-sized distribution company. The milk dispatching problem can be seen in many real-life applications. Under a social responsibility project, several organizations, including companies and municipalities, distribute bottled milk for children to primary schools and impoverished families without any charge. These companies generally use capacitated vehicles for distribution and should consider available hours of the schools as well as the families. The planners generally want to minimize their expenses, such as fuel oil and storage costs. Under those restrictions, the problem turns out to be a capacitated vehicle routing problem with time windows (CVRPTW). One of the main objectives is to minimize total traveled distance considering the vehicle type to reduce the fuel consumption of the vehicles. Another objective is to minimize serving the customers late to reduce the storage cost of undelivered milk. To achieve those objectives, we formulated mixed-integer linear programming (MILP) and constraint programming (CP) models for the problem. To verify and compare our mathematical models, we modified well-known instances from the literature, including problem-specific parameters. The comprehensive computational results show that both models are very competitive for the problem. However, it should be noted that the MILP model outperforms the CP model in terms of solution quality and CPU time for the instances with a long planning horizon.