Published April 2010 | Version public
Journal Article

CPG-based control of a turtle-like underwater vehicle

Abstract

This paper presents biologically inspired control strategies for an autonomous underwater vehicle (AUV) propelled by flapping fins that resemble the paddle-like forelimbs of a sea turtle. Our proposed framework exploits limit cycle oscillators and diffusive couplings, thereby constructing coupled nonlinear oscillators, similar to the central pattern generators (CPGs) in animal spinal cords. This paper first presents rigorous stability analyses and experimental results of CPG-based control methods with and without actuator feedback to the CPG. In these methods, the CPG module generates synchronized oscillation patterns, which are sent to position-servoed flapping fin actuators as a reference input. In order to overcome the limitation of the open-loop CPG that the synchronization is occurring only between the reference signals, this paper introduces a new single-layered CPG method, where the CPG and the physical layers are combined as a single layer, to ensure the synchronization of the physical actuators in the presence of external disturbances. The key idea is to replace nonlinear oscillators in the conventional CPG models with physical actuators that oscillate due to nonlinear state feedback of the actuator states. Using contraction theory, a relatively new nonlinear stability tool, we show that coupled nonlinear oscillators globally synchronize to a specific pattern that can be stereotyped by an outer-loop controller. Results of experimentation with a turtle-like AUV show the feasibility of the proposed control laws.

Additional Information

© 2010 Springer Science+Business Media, LLC. Received: 1 April 2009, Accepted: 17 December 2009, Published online: 22 January 2010. The authors thank Prof. Michael Triantafyllou for allowing them to borrow the robotic turtle developed in his laboratory, as well as Stephen Licht for his extensive help with the implementation and inspiring discussions on the experiments at the MIT Tow Tank. The first author was supported by the Institute for Information Technology Advancement of South Korea. The first and second authors were partially supported by the Air Force Office of Scientific Research (AFOSR), and thank Prof. James Oliver at the Virtual Reality Application Center, Iowa State University.

Additional details

Identifiers

Eprint ID
72462
DOI
10.1007/s10514-009-9169-0
Resolver ID
CaltechAUTHORS:20161130-142353121

Related works

Funding

Institute for Information Technology Advancement of South Korea
Air Force Office of Scientific Research (AFOSR)

Dates

Created
2016-11-30
Created from EPrint's datestamp field
Updated
2023-10-23
Created from EPrint's last_modified field

Caltech Custom Metadata

Caltech groups
GALCIT