Optimization of a humanoid robot gait: multilocal optimization approach

• Autor(es): Lima, José ; Pereira, Ana I. ; Costa, Paulo ; Gonçalves, José
• Data: 2013
• Identificador Persistente: http://hdl.handle.net/10198/9870
• Origem: Biblioteca Digital do IPB
• Assunto(s): Robotics; Optimization
• Tipo de Documento: Artigo
• Idioma: Inglês

Descrição

The humanoid robot gait planning presents a large number of unknown parameters that should be found to make the humanoid robot to walk. There are several approaches to achieve the gait but an accurate simulation can be used to compute it. A stable joint model of a humanoid robot is used in simulation to optimize the gait parameters. The optimization is based on the stretched simulated annealing with the multilocal algorithm approach. Final results prove the benefits of the presented optimization algorithm.