Modeling the Multihop Ridematching Problem with Time Windows and Solving It Using Genetic Algorithms

Abstract

In ridesharing systems, drivers and riders decide to share their trips with each other for cost sharing, fun, reducing congenstion, etc. The ride matching problem with time windows consists of matching a set of drivers' offers and a set of riders' requests based on their sources, destinations and timing with detour willingness. If a request can be matched with only one offer, then the problem is called single hop ride matching. It is called multihop ride matching, if a request can be matched with two offers at different times. In this work, we model the multihop ride matching problem with time windows and provide a genetic algorithm to solve it. Experimentation results on a realistic dataset indicate that the multihop ride matching could increase the number of matched requests as compared with single hop ride matching.