Logo image
System Identification of a Novel Amphibious Cycloidal Propeller Unmanned Underwater Vehicle
Journal article   Peer reviewed

System Identification of a Novel Amphibious Cycloidal Propeller Unmanned Underwater Vehicle

Jonathan Lephuoc, Chenliang Zhang, David Coleman, Casey Harwood and Moble Benedict
IEEE journal of oceanic engineering
02/03/2026
DOI: 10.1109/JOE.2026.3651186

View Online

Abstract

The littoral zone presents several operational challenges for most unmanned underwater vehicles (UUVs) due to strong breaking waves, obstacles in the surf zone, and close proximity to land. A novel amphibious vehicle, the cycloidal propeller unmanned underwater vehicle (Cyclo-UUV), was designed and developed to help overcome these challenges. The vehicle features two parallel tracks for ground locomotion and four retractable cycloidal propellers (cyclo-propellers) for better agility, disturbance rejection, and controllability underwater than vehicles that utilize conventional screw propellers. Due to the uniqueness of the design, little is known about its natural dynamics. As such, this article presents an experimentally derived linear state-space dynamics model of the Cyclo-UUV in forward level motion. Manual inputs were provided to perturb the vehicle from steady-state conditions to excite the system modes. System identification programs for aircraft were used to extract the bare airframe dynamics of the Cyclo-UUV using least squares parameter estimation between the inputs and outputs. The longitudinal and lateral dynamics were found to be coupled and featured two oscillatory modes, one stable and one unstable. These results further our understanding of cyclo-propeller UUV dynamics, and can help with the development of robust model-based controllers for autonomous undersea operations.
Aerodynamics Hydrodynamics Propulsion Vehicle Design Vehicle Dynamics Amphibious unmanned vehicles Autonomous underwater vehicles Blades Complexity theory cycloidal propellers Land vehicles Propellers remotely operated vehicles Robots

Details

Metrics

1 Record Views
Logo image