Motion planning of a fish-like piezoelectric actuated robot using model-based predictive control

Autor(es)

Barbosa, Arthur S. ^[1] ; Tahara, Lucas Z. ^[1] ; da Silva, Maira M. ^[1]

Afiliação do(s) autor(es)

^[1] Univ Sao Paulo, Sao Carlos Sch Engn, Av Trab Sao Carlense, 400 Parque Arnold Schimidt, BR-13566590 Sao Carlos - Brazil

• Tipo de documento: Artigo Científico
• Fonte: JOURNAL OF VIBRATION AND CONTROL; OCT 2021.
• Citações Web of Science: 0

Resumo

This work proposes a novel methodology for planning the motion of fish-like soft robots actuated by macro-fiber composite (MFC) pairs. These structures should mimic oscillatory and undulation movements, which can be accomplished if the amplitude of the tail motion is larger than that of the head motion. Design strategies, such as the use of concentrated and distributed masses, are addressed to mimic fish-like motion since they guarantee suitable mode shapes for the structure. The motion planning proposal explores a model-based predictive control (MPC) strategy for deriving the input signals for the MFC actuators. This model-based control strategy requires the use of reasonably small-sized models. This is accomplished by extracting modal state-space models based on the free-free Euler-Bernoulli beam theory considering the electro-mechanical coupling of the MFC actuator pairs. Numerical results demonstrate the capability of the proposal for deriving bounded input signals that generate oscillatory and undulation movements even in the presence of disturbances. This general approach can be further extended for other applications. (AU)

• Processo FAPESP: 18/15894-0 - Projeto e otimização de estruturas periódicas para aprimoramento de desempenho vibroacústico
• Beneficiário: Carlos de Marqui Junior
• Modalidade de apoio: Auxílio à Pesquisa - Temático
• Processo FAPESP: 18/21336-0 - Alcançando altas acelerações com manipuladores paralelos, Fase II: instrumentação, modelagem e controle de um manipulador flexível
• Beneficiário: Maira Martins da Silva
• Modalidade de apoio: Auxílio à Pesquisa - Regular