Luiz Carvalho Ottoni, André and Nepomuceno, Erivelton and Santos de Oliveira, Marcos (2019) Estimação de Parâmetros do Aprendizado por Reforço para o Problema de Planejamento de Rotas com Reabastecimento. In: SBAI 2019, 2019.

Preview

Download (466kB) | Preview

Abstract

Path planning is a important problem in mobile robotics. One of the aspects of this type of autonomous vehicles planning refers to observe the fuel-constraints. In this sense, the objective of this work is to estimate the Reinforcement Learning parameters for the path planning problem with refueling. The results indicate that the parameters estimated with the Response Surface Methodology reached the best solutions in most of the experiments. Resumo: O planejamento de rotas ´e um importante problema na rob´otica m´ovel. Uma das vertentes desse tipo de planejamento para ve´ıculos autˆonomos, refere-se a observar as restri¸c˜oes operacionais com combust´ıvel. Nesse sentido, o objetivo deste trabalho ´e estimar os parˆametros do Aprendizado por Refor¸co para o problema planejamento de rotas com reabastecimento. Os resultados apontam que os parˆametros estimados com a Metodologia de Superf´ıcie de Resposta alcan¸caram as melhores solu¸c˜oes na maioria dos experimentos.

Item Type:	Conference or Workshop Item (Paper)
Keywords:	Reinforcement learning; Path planning; Refueling; Response surface methodology; Parameter estimation; Aprendizado por refor¸co; Planejamento de rotas; Reabastecimento; Metodologia de superf´ıcie de resposta; Estima¸c˜ao de parˆametros;
Academic Unit:	Faculty of Science and Engineering > Electronic Engineering Faculty of Science and Engineering > Research Institutes > Hamilton Institute
Item ID:	16862
Identification Number:	https://doi.org/10.17648/sbai-2019-111113
Depositing User:	Erivelton Nepomuceno
Date Deposited:	17 Jan 2023 14:32
Refereed:	Yes
URI:
Use Licence:	This item is available under a Creative Commons Attribution Non Commercial Share Alike Licence (CC BY-NC-SA). Details of this licence are available here

Repository Staff Only(login required)

Item control page

Downloads

Origin of downloads