Best-Match in a Set of Single-Vehicle Dynamic Pickup and Delivery Problem Using Ant Colony Optimization

The Pickup and Delivery problem (PDP) is an optimization problem used in logistics, where goods must be transported in a certain order. This paper proposes the Set of Single-Vehicle Dynamic Pickup and Delivery Problem (SVPDP), which extends PDP to a set of vehicles that are already on the road when a new pickup and delivery order arrives. A ranking is created to determine, which vehicles are best suited to fulfill this order with the smallest detour. A modified Ant Colony Optimization algorithm (ACO) dynamically defines for each vehicle new routes while minimizing the overall costs when integrating the additional locations. The problem is modeled, different sizes are examined and suitable parameter sets for the ACO are proposed to solve this particular SVPDP.