Implementação de um mecanismo de controle de preços em um sistema multiagente para gerenciamento de vagas em um smart parking

Abstract

Long periods of time searching for a parking space causes wasted fuel and more congestion. In large urban centers, where population growth has been faster than infrastructure investments, a mobility problem occurs that needs innovative solutions. The Smart City paradigm consists of the development and integration of new technologies in order to solve the traffic problems of a city. The lack of available parking spaces and the mobility problem are covered in the Smart Parking area and the MAPS (Multi-Agent Parking System) project aims to create solutions for parking lots with the use of Multi-agent Systems. This work is an extension of this project and has the objective to develop a dynamic pricing module that alters the price of parking spaces considering the characteristics of its environment. This pricing mechanism was implemented in a Multiagent System and a managing agent decides the moment of its use. The module has the function of observing the demand for parking spaces, the time of day and the weather condition and decides from rules contained in a configuration file how the price should be changed. Simulations of parking lots with different aspects were run using six different configuration files and control scenarios in which the module was not used. The results found showed that the pricing module has significant influence on the operation of the parking facility, but has oscillating values due to pseudorandom characteristics. In general, the use of any configuration provided better results in at least one aspect, be it higher total credits at the end of the month or higher average occupancy of the spaces.