Application of particle swarm optimization algorithm to dynamic vehicle routing problem

Abstract

In this paper we apply Particle Swarm Optimization (PSO) algorithm to Dynamic Vehicle Routing Problem (DVRP) for solving the client assignment problem and optimizing the routes. Our approach to solving the DVRP consists of two stages. First we apply an instance of PSO to finding clients' requests that should be satisfied by the same vehicle (thus solving the Bin Packing Problem (BPP) part of the DVRP task) and then we apply several other instances of PSO, one for each of the vehicles assigned to particular requests (thus solving the Travelling Salesman Problem (TSP)). Proposed algorithm found better solutions then other tested PSO approaches in 6 out of 21 benchmarks and better average solutions in 5 of them. © 2013 Springer-Verlag.