Analysis of the Stability and Maneuverability of Bottlenose Dolphins (Tursiops Truncatus) with Applications to Biomimetic Robotics

Publication Date

8-2002

Type of Culminating Activity

Thesis

Degree Title

Master of Science in Engineering, Mechanical Engineering

Department

Mechanical and Biomechanical Engineering

Supervisory Committee Chair

John F. Gardner

Abstract

Animals have inspired various technological advances including flight and robotics. The biomimetic approach of copying animals seeks solutions from engineering and biology for increased efficiency and specialization. Biological designs have resulted from the evolutionary Darwinian process of natural selection; therefore, it is assumed that Mother Nature has already subjected these animals to a cost-benefit-analysis, optimizing particular designs for specific functions. Engineers can target the morphological parameters responsible for this optimum design for technology transfer; effectively reducing the time required to develop innovative designs. This particular application examines aquatic animals, specifically bottlenose dolphins (Tursiops truncatus), for design principles which may be responsible for their highly maneuverable, yet stable motions. The equations of motion for a dolphin are developed using Newton-Euler mechanics. The forces and moments referred to the center of mass are linked to the linear and angular velocity vectors and the inertia tensor about the center of mass. The variables within these equations are resolved using the physical and hydrodynamic parameters of Robo-Dolphin, a robotic version of a bottlenose dolphin developed at Boise State University. A transformation routine is used to convert these results from the body-fixed coordinate system to the inertial coordinate system. An attitude control routine is also developed to allow for closed-loop control of the pitch angle and depth by modifying the pectoral flipper deflection, thus changing the lift and drag forces on the body of the dolphin. This initial SIMULINK block diagram is passed to a linearization and model order reduction program to construct the root locus and find the locations of the dominant poles. Three of the four dominant poles are located on the right side of the real axis, indicating the initial system is inherently unstable. The unstable poles are moved using pole placement and the resulting gains are implemented into the simulation using state feedback. The modified SIMULINK block diagram is passed back to the linearization and model order reduction program to construct the new root locus and Bode plot for the system. The model is then analyzed during a depth control task. The results of the depth response for a pitch gain of K pitch = 0.10 can be compared to an underdamped second order system. The graph shows the dolphin reaching its desired depth of Zd = 1.00 meter with a settling time of TS = 16.0 seconds and a percent overshoot of %OS = 5.88%. The depth control experimentation with Robo-Dolphin qualitatively confirms the results found using the SIMULINK models. The settling time and percent overshoot found using the SIMULINK models, however, is much larger than what we would expect from real bottlenose dolphins, indicating that maneuverability is not solely achieved by the hydrodynamic characteristics of the pectoral flippers. The inherent instability in the initial simulation demonstrates the pectoral flippers do not produce the forces and moments necessary to complete large pitch and depth control tasks. The caudal peduncle apparently plays a larger role in the quick and accurate maneuvering ability of bottlenose dolphins.

This document is currently not available here.

Files over 30MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS