Desenvolvimento de softwares e algoritmo baseado em redes neurais artificiais para suporte à gestão da mobilidade urbana em smart campus com característica multimodal

Abstract

This work presents the development of two software solutions and an algorithm based on artificial neural networks to support the management of urban mobility in a smart campus. The first software, called Norte Rotas, is a web solution whose objective is to support the planning of pedestrian routes, providing relevant information about the physical conditions of the routes of a smart campus. The second solution is an Android mobile software that aims to manage transport modes present in a smart campus. Tests in simulated and real environments were carried out and the results indicate that the proposed tools are good solutions for the planning and management of modal routes in an intelligent university campus. In addition to the software, a computational intelligence algorithm is proposed to determine the best travel route considering the options on foot, by bus and by boat in an IoT system of a smart campus. Data were collected from UFPA's Circular bus routes, and tests with different parameters of an ANN were performed. The results show that the solution based on ANN is promising to be implanted in urban mobility aid systems in a smart campus.