Proposta de algoritmo otimizador de caminhos para manipuladores robóticos em ambientes com presença de obstáculos

• Ano de defesa: 2020
• Autor(a) principal: Magalhães, Aline de Cássia
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; Brasil; Programa de Pós-graduação em Engenharia Elétrica
• Palavras-chave em Português: Inteligência artificial; Manipuladores robóticos; Otimização; Planejamento de caminho; Artificial intelligence; Otimization; Pathfinding; Robotic manipulators; CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA; Robótica
• Link de acesso: https://repositorio.ufu.br/handle/123456789/29815; http://doi.org/10.14393/ufu.di.2020.515

Based on the increasing use of robotic manipulators mainly in industries, the need arose to insert in them the capacity of self-optimization and reaction to the environment. Thus, the present work proposes the creation of a generalized algorithm for different anatomies of manipulators, which is able to map the environment, calculating a minimum path between two desired points, using smooth joint movements and avoiding collisions with possible static or dynamic obstacles. This algorithm, called Path Optimizer for Manipulators (POM), uses the fusion of the Denavit-Hartenberg method with the D* Lite and CCD algorithms, with some modifications made to them in order to adapt to the manipulators’ operation. The results obtained from various experiments reported in this document prove the effectiveness of the method, although it still has some vulnerabilities in relation to the mapping and definition of neighboring nodes.