Multi-objective Shipment Allocation using Extreme Nondominated Sorting Genetic Algorithm-III (E-NSGA-III)
Kittichai Lavangnananda, Peerasak Wangsom · 2019
Dealing with logistical problems are common among companies where delivery of goods are involved. An item to be delivered is commonly known as a shipment and hence efficient shipment allocation is essential for such companies to maintain profits and company competitiveness. This work is concerned with shipment allocation of a large and worldwide delivery company, which has a distribution center in a country where this is still carried manually. Determination of an efficient shipment allocation becomes a multi-objective optimization. Three objectives are identified, these are minimizing number of vehicles used, maximizing vehicle utilization and maximizing operator's route familiarity. Three Multi-Objective Evolution Algorithms (MOEAs) are utilized in this work, these are the commonly known, Nondominated Sorting Genetic Algorithm-III (NSGA-II), Nondominated Sorting Genetic Algorithm-III (NSGA-III) and the recently developed Extreme Nondominated Sorting Genetic Algorithm-III (E-NSGA-III). Hypervolume is used as the metric to measure the quality of solutions. The results from MOEAs reveal superiority over the existing manual solutions.