MATHEMATICAL MODEL AND MEMETIC ALGORITHM FOR A VARIANT OF THE CLERK PROBLEM

The Traveling Car Renter Problem (CaRS) is a generalization of the Traveling Merchant Problem (PCV) where the route can be carried out using different rental cars, also characterized as NP-Hard. All the studies reported in the literature addressed the problem from the same perspective, where the same car cannot be rented more than once, the rented car can be returned in any city and there is a pre-defined city for the beginning and end of the route, called the base city. This article analyzes the problem by modifying these characteristics. The mathematical model of the problem is tuned to account for this variance and is implemented in a solver, producing results for fifty-eight instances of the problem. A memetic algorithm is proposed, with local searches appropriate to the variation of the problem. The computational performance of the proposed memetic algorithm is compared with the performance of the solver.