Open AccessDesign of Fin-Curvature-Based Feedback Controller for Efficient Swimming
Author(s) -
Fumiaki Nose,
Yuichiro Sueoka,
Daisuke Nakanishi,
Yasuhiro Sugimoto,
Koichi Osuka
Publication year - 2021
Publication title -
journal of robotics and mechatronics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.257
H-Index - 19
eISSN - 1883-8049
pISSN - 0915-3942
DOI - 10.20965/jrm.2021.p0955
Subject(s) - controller (irrigation) , robot , fish fin , control theory (sociology) , fin , curvature , feed forward , propulsion , motion control , trajectory , computer science , mechanism (biology) , simulation , open loop controller , control engineering , fish <actinopterygii> , engineering , artificial intelligence , control (management) , mathematics , physics , mechanical engineering , aerospace engineering , closed loop , biology , geometry , quantum mechanics , fishery , astronomy , agronomy
Over the past few decades, biologists and engineers have attempted to elucidate the swimming mechanism of fish and developed a fish-like robot to perform fast swimming in water. Such a robot will have wide applicability in investigations and exploration in the sea. There have been many studies on fish-type robots; however, the propulsion efficiency of the introduced robots is far from that of the actual fish. The main reason is that the robot controller for generating motions is conventionally designed by trial and error, and little attention has been placed on designing a motion controller that matches the body structure of a real fish. In this paper, we present an approach that uses fin-curvature-based feedback to design a motion controller. A swimming robot composed of a body with two actuated joints and a flexible tail fin is developed. After examining the relationship between the swimming speed and tail fin curvature in feedforward (open-loop) system experiments, we propose to reflect the tail fin curvature to the actuation inputs (phase difference between the two cyclic oscillations), which will perform the efficient swimming motion. Further, the results show that implementing the proposed feedback controller in a fish-type robot makes it swim similar to a real fish. In addition, the proposed controller functions to find inappropriate actuation according to the body structure.