Solving multi-objective vehicle routing problem with time windows by FAGA

Abstract

The vehicle routing problem with time windows (VRPTW) is a complex transportation issue. In this paper, multi objective VRPTW is considered in which the total distance travelled, total number of vehicles used and route balance are minimized. Genetic algorithm with fitness aggregation approach and specialized operators like selection based on aggregate fitness value, best cost route crossover called Fitness Aggregated Genetic Algorithm (FAGA) is introduced for solving the multi objective problem. The algorithm was tested on large number of Solomon's benchmarks for bi-objective model that is minimization of total distance travelled and total number of vehicles used. The results produced by FAGA are highly competitive to best known results reported in the literature. After validation the third objective that is route balance is incorporated into bi-objective model and it is observed that FAGA produces better balanced routes without affecting the total distance travelled and total number of vehicles used.