Transshipment demand in vehicle routing problem using genetic algorithm

The Transshipment Demand in Vehicle Routing Problem (VRPTD) is a problem that involves transferring items between retailers due to shortages at retail stores and availability at another retail store. Strategically transferring items between retailers minimizes lost sales by ensuring item availability which leads to increasing customer satisfaction. In this study, the VRPTD considered real world scenarios involving valuable bulk items. The objective is to minimize the transportation cost while efficiently meeting customer demand by finding the best delivery routes that fulfill both regular and transshipment demand. A metaheuristic method, Genetic Algorithm (GA) was proposed for the problem. Three main genetic operators employed are Stochastic Universal Sampling (SUS) is used for selection, a modified Edge Recombination Operator (ERO) as the crossover operator and two different swap strategies in mutation operator. The control parameters: population size and maximum generation were determined through a small-scale experiment. GA was run for 10 independent runs, and the results obtained were compared with previous literature. The best sequence found has slightly higher distance with a running time of less than 3 minutes. In conclusion, this study successfully developed GA for solving VRPTD with near-optimal solutions with difference of not more than 10% when compared to the previous literature.