Adaptive control traffic system based on genetic algorithms

Identifier:  TFM:995

Handle: https://hdl.handle.net/20.500.11797/TFM995

Authors:  Barrero Díaz, Miguel Ángel

Abstract:
From the beginning of the road traffic, the infrastructures and regulations have evolved with the objective of being adapted to the increasingly volume of vehicles and its interaction with pedestrians. Traffic lights were developed aimed to automate the traffic regulation in road intersections and the police officers were replaced by traffic lights geared by automated regulators. This solutions have evolved towards complex systems where many parameters are taking into account and are ,in general, managed from a control center. Usually, these kind of infrastructures has large implementation costs and maintenance expenses and are not feasible in many cases. This paper proposes a solution based on genetic algorithms were each traffic light regulator has it own intelligent agent, which learns from traffic data and takes time distribution decisions in a decentralized way, based only on its own parameters taken from the environment.