Market based allocation of transportation orders to vehicles in adaptive multi-objective vehicle routing

Abstract

The article describes a study on vehicle routing problems under multiple objectives. In particular, we investigate the effectiveness of different approaches when assigning orders to vehicles. The resulting clustering problem is studied within a general framework for multi-objective vehicle routing problems where different vehicle agents place bids for orders which are offered on a marketplace. This marketplace gathers information about the current situation and provides the basis for the resolution of the allocation problem. By implementing different specialized but interacting software agents, an adaptation of the concept to various configurations of the studied problem is possible. Experimental investigations of different assignment logics on benchmark instances have been carried out and numerical results are reported. In brief, a tendency towards a particular clustering approach can be observed. © 2008 Springer-Verlag Berlin Heidelberg.