Applying genetic algorithm for solving multi-depot vehicle routing problem with time windows

Abstract

The optimization of the multi-depot Vehicle Routing Problem (VRP) with time windows is a complex issue within the logistics and transportation field that traditional optimization methods struggle to solve due to its complexity. In contrast, Genetic Algorithm (GA) is a metaheuristic optimization method that has shown promising results in addressing various optimization issues, including the VRP. This research aims to implement GA for solving the multi-depot VRP with time windows and assess its performance regarding solution quality and computation time. The problem is transformed into a set of chromosomes, and genetic operators are employed to generate novel solutions. The fitness of each solution is assessed using multiple objective functions, including the total distance traveled, the number of vehicles utilized, and the total delivery time. The findings indicate that GA can identify nearly optimal solutions in a reasonable amount of time and handle dynamic and uncertain environments. Future studies may explore the potential of integrating machine learning algorithms and other metaheuristic optimization techniques to enhance the performance of GA in solving the multi-depot VRP with time windows.