Título: Rede Neural Modular e Aprendizagem por Reforço para Sistema de Navegação Autônoma
Autores: Calvo, Rodrigo; Figueiredo, Maurício
Resumo: This work describes an autonomous navigation system based on a modular neural network. The environment is unknown and the system is not able for navigation. As the robot collides and takes targets, it improves its navigation strategy and efficiently guides the robot to targets. A reinforcement learning mechanism adjusts parameters of the neural networks at target capture and collision moments. Simulation experiments show performance comparisons. Only the proposed system reaches targets if the environment presents a high risk (dangerous) configuration (targets are very close to obstacles).
Código DOI: 10.21528/CBRN2003-081
Artigo em PDF: 6CBRN_081.PDF
Arquivo BibTex: 6CBRN_081.bib