Um iterated greedy e uma mateurística para o problema de roteamento de veículos elétricos com dois níveis e janela de tempo

Abstract

The Two-Echelon Electric Vehicle Routing Problem with Time Windows comprises a complex variant of the Vehicle Routing Problem that has to attend customs from a central deposit using intermediary depots (satellites). The satellites attend to customers using Electric Vehicles, and the satellites are attended by Combustion Vehicles that leave the central depot. Recharging stations can be used to extend the limited range of Electric Vehicles. The present work proposes two approaches based on Iterated Greedy (IG) to solve the optimization problem. The first one consists of an IG that’s utilizes operators of destruction and rebuilding. The local search Random Variable Neighborhood Descent is also utilized to get a local minimal solution after the rebuilding phase. The second approach consists in a matheuristic that uses the proposed IG method with a set covering of routes. The matheuristic uses the proposed IG to create a pool of routes for the Electric Vehicles. This pool is used to formulate a Mixed Integer Linear Programming of the set covering problem. The solution to this problem consists in a selection of a sub-set of Electric Vehicles that attend all customs while minimizing the sum of the select routes distances. Both proposed methods were compared with the state of the art method and obtained significantly better results, concerning distance (objective function) and computational time.