Metaheuristic applications on electric vehicle traveling salesman problem

Abstract

The energy consumption behaviors of the vehicles with electric motors are different compared to traditional internal combustion engines. As a result of regenerative braking systems, electric vehicles also have the possibility to recover energy during the journey. This feature causes a considerable variation in consumption functions, especially on negative slopes. This study focuses on finding the optimal or near-optimal TSP tours for electric vehicles on real-time fed data with the consideration of the road grades, transported loads, the speed of the vehicle, and acceleration-deceleration. These conditions mean that much more complexity in a traveling salesman problem, whose exact methods are already requiring a significant amount of computation time. The ultimate aim was obtaining high-quality solutions using efficient steepest ascent and simulated annealing metaheuristics while reducing the computation times.