Multi-objective Optimization of Vehicle Routing with Environmental Penalty
Luis Bernardo Pulido-Gaytan,Andrei Tchernykh,Sergio Nesmachnow,Alfredo Cristóbal-Salas,Arutyun Avetisyan,Harold Enrique Castro Barrera,Carlos Jaime Barrios Hernandez
DOI: https://doi.org/10.1007/978-3-030-38043-4_13
2019-01-01
Abstract:AbstractCities host more than half of the population in only 2% of the earth’s surface and consume 75% of the resources extracted from the planet, this abrupt demographic growth in urban areas has worsened the level of pollution in the city, as well as the problems of road congestion. Therefore, smart cities propose the incorporation of technologies to optimize the use of existing infrastructure and thus, achieve a sustainable and inclusive city. In this context, we propose an optimization method based on a multi-objective cellular genetic algorithm to determine an adequate routing of vehicles that allows to diminish the environmental impact in highly concurred areas, while providing alternative routes that minimize the associated costs of moving from one place to another, this under the premise that not always the shortest path represents the best solution.The presented algorithm simultaneously minimizes three important objectives: the time, the amount of pollutants emitted and an environmental penalty, where the latter represents the implicit cost of moving through a certain segment of the map. That is, an approximate set of Pareto is obtained, which contains alternative routes that avoid traveling through areas with a high degree of pollution and that, in turn, minimize the time and quantity of emissions when traveling through a certain network of routes. Our experimental analysis based on several quality indicators, like Hypervolume and Spread, shows a competitive performance of the proposed approach in terms of convergence and diversity, this with respect to NSGA II and SPEA2, well-known multi-objective algorithms in the literature.