Evolutionary algorithms for the Traveling Car Renter with Passengers

The Traveling Car Renter with Passengers (CaRSP) is a generalization of the Traveling Salesman Problem (TSP) where the tour can be decomposed into contiguous paths that are travelled by different rented cars and allows the vehicles to be shared with other passengers to reduce expenses. The definition of the proposed problem involves the combination of two important concepts that are currently being widely used in the field of transportation: car rental and ride-sharing. This paper defines the CaRSP and presents four evolutionary algorithms to solve it: a genetic algorithm, a memetic algorithm and hybridizations of both using transgenetic vectors (transposons and recombinant plasmids). The metaheuristics are tested in a set of 30 Euclidean, non-Euclidean, symmetric and asymmetric instances. We use a framework to define the best parameter settings for each algorithm. Computational experiments are performed in two stages to ensure the algorithms are compared properly. The obtained results are submitted to statistical analysis. An algorithmic study is reported presenting the first heuristic results for CaRSP.