Metaheuristic solution for vehicle transportations to improve Moroccan logistics

Abstract

The objective of this study is to minimize transportation costs of the Moroccan supply chain and enhance the overall logistics performance of its different sectors. To achieve that we studied the Vehicle Routing Problem (VRP), which is a well-known problem, where multiple vehicles has to deliver goods to different customers efficiently, meanwhile the Pickup and Delivery Problem (PDP) is one of its variants where the task of each vehicle is to pickup goods and take them to their delivery location, which could be a representation of the supply chain transporting goods, especially since the founding of a new logistic zone in Morocco that offers transportation services. To calculate an optimized solution that would reduce the costs of the transportations, we used the cluster first, route second approach, we start by assigning a set of orders to each vehicle, which we consider a cluster, the assignment is done by using a hybrid of the genetic algorithm and a centroid based local optimization process, to further enhance the effectiveness of this stage, we used an immigration operator to delay the convergence of the solutions, we called the algorithm the Centroid Optimized Genetic Immigration Clustering (COGIC) after finishing this stage, we are left with multiple Travelling Salesman Problem (TSP) instances, we route each solution by using an insertion algorithm, this algorithm is very optimal computational time wise, we exploit its speed further by not only routing the solutions passed from the first stage, but also other solutions derived from them.