Table of Contents
Fetching ...

Design and Central Pattern Generator Control of a New Transformable Wheel-Legged Robot

Tyler Bishop, Keran Ye, Konstantinos Karydis

TL;DR

The work tackles robust locomotion of transformable wheel‑leg robots on semi‑natural terrains by integrating a novel leg‑wheel mechanism with central pattern generator (CPG) control. It develops a generalized four‑bar kinematic model driven by coaxial shafts, enabling inverse kinematics and torque analysis that inform the CPG design.Three oscillator models—Kuramoto, Modified Hopf, and Van der Pol—are implemented and compared within a ROS/BeagleBone‑Blue framework controlling four wheels with differential steering, validated through a real prototype and simulations over flat, uneven, and obstacle environments. Results demonstrate effective CPG control for stable transitions between wheel and leg modes, with Kuramoto showing the most consistent performance across tests, and highlight areas for future improvements in drive rigidity and controller complexity to enhance obstacle negotiation and terrain adaptability.

Abstract

This paper introduces a new wheel-legged robot and develops motion controllers based on central pattern generators (CPGs) for the robot to navigate over a range of terrains. A transformable leg-wheel design is considered and characterized in terms of key locomotion characteristics as a function of the design. Kinematic analysis is conducted based on a generalized four-bar mechanism driven by a coaxial hub arrangement. The analysis is used to inform the design of a central pattern generator to control the robot by mapping oscillator states to wheel-leg trajectories and implementing differential steering within the oscillator network. Three oscillator models are used as the basis of the CPGs, and their performance is compared over a range of inputs. The CPG-based controller is used to drive the developed robot prototype on level ground and over obstacles. Additional simulated tests are performed for uneven terrain negotiation and obstacle climbing. Results demonstrate the effectiveness of CPG control in transformable wheel-legged robots.

Design and Central Pattern Generator Control of a New Transformable Wheel-Legged Robot

TL;DR

The work tackles robust locomotion of transformable wheel‑leg robots on semi‑natural terrains by integrating a novel leg‑wheel mechanism with central pattern generator (CPG) control. It develops a generalized four‑bar kinematic model driven by coaxial shafts, enabling inverse kinematics and torque analysis that inform the CPG design.Three oscillator models—Kuramoto, Modified Hopf, and Van der Pol—are implemented and compared within a ROS/BeagleBone‑Blue framework controlling four wheels with differential steering, validated through a real prototype and simulations over flat, uneven, and obstacle environments. Results demonstrate effective CPG control for stable transitions between wheel and leg modes, with Kuramoto showing the most consistent performance across tests, and highlight areas for future improvements in drive rigidity and controller complexity to enhance obstacle negotiation and terrain adaptability.

Abstract

This paper introduces a new wheel-legged robot and develops motion controllers based on central pattern generators (CPGs) for the robot to navigate over a range of terrains. A transformable leg-wheel design is considered and characterized in terms of key locomotion characteristics as a function of the design. Kinematic analysis is conducted based on a generalized four-bar mechanism driven by a coaxial hub arrangement. The analysis is used to inform the design of a central pattern generator to control the robot by mapping oscillator states to wheel-leg trajectories and implementing differential steering within the oscillator network. Three oscillator models are used as the basis of the CPGs, and their performance is compared over a range of inputs. The CPG-based controller is used to drive the developed robot prototype on level ground and over obstacles. Additional simulated tests are performed for uneven terrain negotiation and obstacle climbing. Results demonstrate the effectiveness of CPG control in transformable wheel-legged robots.
Paper Structure (14 sections, 11 equations, 7 figures, 5 tables)

This paper contains 14 sections, 11 equations, 7 figures, 5 tables.

Figures (7)

  • Figure 1: (a) The robot prototype has four transforming leg-wheels driven by eight BLDC motors and arranged to use differential steering. (b) The robot is modeled in Webots to perform simulated experiments on various obstacles and terrain
  • Figure 2: CAD design of the drive section and full robot assembly. The drive uses two two-stage belt and pulley systems for power transmission. Different sized pulleys can be 3D-printed and swapped in without re-fabricating the structural elements. The inner wheel hub is driven through a 3D-printed spline; the outer wheel hub is attached rigidly to the outer shaft.
  • Figure 3: (a) Wheel geometry relative to the number of legs (arcs). (Here, a wheel with $n=4$ legs is shown.) (b) The links of the four bar mechanism are made up of the inner hub $\vec{DC}$, outer hub $\vec{DA}$, wheel arc $\vec{AB}$, and a bar link $\vec{CB}$. (c) Planetary gear mechanism added in later iterations which uses the same four bar mechanism but with the inner hub replaced by the planet gear carrier. (d) U-shaped trajectory of the phase offset between wheel hubs relative to a tip position with a constant Y component.
  • Figure 4: The controller runs on a BeagleBone Blue using ROS with separate nodes (grey boxes). Motor control is provided by four Odrive BLDC boards which employ cascaded PID controllers when running in position control mode OdriveController. Position commands and estimates are sent between the Beagle Bone and Odrive controllers over CAN Bus. Other control parameters are sent over USB.
  • Figure 5: Robot trajectory over flat ground tests. (a-d) Horizontal deviation from the commanded straight line. (e-f) Orientation change over time. The tests were performed using (a) direct drive, and (b, e) Hopf, (c, f) Kuramoto and (d) Van der Pol oscillators.
  • ...and 2 more figures