Combining Genetic Algorithms with Traffic Simulations for restructuring traffic networks subject to increases in population

ENRIGUE GABRIEL BAQUELA; ANA CAROLINA OLIVERA

Marrakech

Conferencia; 5th International Conference on Metaheuristics and Nature Inspired Computing; 2014

Universidad de Lille

Nowadays, population growth seems to have a direct impact on daily life due mainly to society progress. Traffic jams, pollution and parking problems are some of the most common problems in relation to traffic networks. In this paper, we present a methodology based on Optimization via Simulation (OvS) to select, from a group of nodes of a traffic network, the best alternative to increase the capacity of that network in the Origin-Destiny Traffic Assignment Problem (ODTAP), a subtype both of the Facility Location Problem (FLP) and the Network Traffic Design Problem (NTDP) [1-2]. Given a specific traffic network, which will suffer an increase in population, the main goal is to decide where to foment the installation of urban, industrial and/or commercial complexes so that it has a minimum effect over the total traffic system: considering the network already has a certain structure and a dynamic flow restricting possible configurations. To deal with the resolution of ODTAP, a Genetic Algorithm (GA) was used in conjunction with Traffic Simulations (TS), in an OvS based formalism.