Optimization of vehicle routing problem using guided local search and simulated annealing

Abstract

Transportation concerns in the supply chain are important because they have the potential to greatly raise logistical requirements. The proposed solution to the issue is the vehicle routing problem. The Heterogeneous Fleet Vehicle Routing Problem (HFVRP) model will be implemented in this study using the metaheuristic methodology together with the Guided Local Search (GLS) and Simulated Annealing (SA) approaches for case studies. Python 3.10 software was used to analyze the SA and GLS, and the Google OR-Tools module was used. Implementing the SA and GLS algorithm based on data from the case study in a logistics company involved in the distribution of chicken feed and gas is the goal of this project. By doing so, it will be possible to determine whether the current situation is ideal or if it could be improved to reduce the number of vehicles used, find the best route, lower shipping costs, and increase efficiency. The findings of this study show that when compared to the GLS algorithm, research employing SA produces better outcomes in terms of cost and route management. According to actual data, SA increases income by an average amount between 3% and 10%.