Uma análise comparativa sobre as estratégias de inicialização da população em algoritmos genéticos para o problema de eficiência energética em trens urbanos

Abstract

Over time mankind has been causing great impacts to the environment. In the last few years there has been a great world population growth, which has aggravated these impacts. From this situation, a worldwide concern about the environment was generated and the need to provide Sustainable Development arose. With rapid urbanization it is necessary to make cities sustainable, and an important factor for this demand is urban mobility. The use of electric public transport is an alternative to reduce the use of fossil fuels and reduce polluting gases. However, electrical energy also has negative impacts. In this context, several studies were carried out in order to nd ways to enable energy eficiency of electric trains, providing better solutions for the driving pro les through the use of Genetic Algorithms. This work aims to implement and make a comparative analysis of the population initiation strategies for the studied problem, and then was integrated with the implementations with the latest version of GeneticBee software in order to show the behavior of each type of initialization and the impact generated in the Genetic Algorithm. With the present work it was possible to identify which of the initializations of the population obtained better performance, and when being integrated to the existing software it is possible to make con guration changes and verify the impact in the fi nal solution.