Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Ground and Flight Locomotion for Two-Wheeled Drones via Model Predictive Path Integral Control

Gosuke Kojima, Kohei Honda, Satoshi Nakano, Manabu Yamada

TL;DR

The paper addresses motion planning for a two-wheeled drone that must navigate cluttered environments by leveraging both ground driving and aerial flight. It extends Model Predictive Path Integral (MPPI) control to multimodal dynamics by introducing a mode-specific input space and an auxiliary controller to manage abrupt mode changes. Key contributions include a Lagrangian-based dynamics model with ground-contact impulses, a switched-input MPPI formulation, and integration of non-differentiable obstacle costs within the optimization framework. Simulation results demonstrate obstacle avoidance and effective mode switching, highlighting practical potential for versatile locomotion in real-world, obstacle-rich settings.

Abstract

This paper presents a novel approach to motion planning for two-wheeled drones that can drive on the ground and fly in the air. Conventional methods for two-wheeled drone motion planning typically rely on gradient-based optimization and assume that obstacle shapes can be approximated by a differentiable form. To overcome this limitation, we propose a motion planning method based on Model Predictive Path Integral (MPPI) control, enabling navigation through arbitrarily shaped obstacles by switching between driving and flight modes. To handle the instability and rapid solution changes caused by mode switching, our proposed method switches the control space and utilizes the auxiliary controller for MPPI. Our simulation results demonstrate that the proposed method enables navigation in unstructured environments and achieves effective obstacle avoidance through mode switching.

Ground and Flight Locomotion for Two-Wheeled Drones via Model Predictive Path Integral Control

TL;DR

The paper addresses motion planning for a two-wheeled drone that must navigate cluttered environments by leveraging both ground driving and aerial flight. It extends Model Predictive Path Integral (MPPI) control to multimodal dynamics by introducing a mode-specific input space and an auxiliary controller to manage abrupt mode changes. Key contributions include a Lagrangian-based dynamics model with ground-contact impulses, a switched-input MPPI formulation, and integration of non-differentiable obstacle costs within the optimization framework. Simulation results demonstrate obstacle avoidance and effective mode switching, highlighting practical potential for versatile locomotion in real-world, obstacle-rich settings.

Abstract

This paper presents a novel approach to motion planning for two-wheeled drones that can drive on the ground and fly in the air. Conventional methods for two-wheeled drone motion planning typically rely on gradient-based optimization and assume that obstacle shapes can be approximated by a differentiable form. To overcome this limitation, we propose a motion planning method based on Model Predictive Path Integral (MPPI) control, enabling navigation through arbitrarily shaped obstacles by switching between driving and flight modes. To handle the instability and rapid solution changes caused by mode switching, our proposed method switches the control space and utilizes the auxiliary controller for MPPI. Our simulation results demonstrate that the proposed method enables navigation in unstructured environments and achieves effective obstacle avoidance through mode switching.

Paper Structure

This paper contains 19 sections, 22 equations, 7 figures.

Table of Contents

  1. Introduction
  2. Related Work
  3. Motion Planning for Two-Wheeled Drones
  4. MPPI-based Motion Planning for Drones
  5. Target Task and Requirements
  6. Dynamics Modeling of Two-wheeled Drone
  7. Derivation of Euler-Lagrange Equations
  8. Discontinuous Impulse at Ground Contact
  9. Two-wheeled Drone Navigation using MPPI
  10. Motion planning based on MPPI
  11. Design of Mode-specific Input Space
  12. MPPI Formulation for Two-Wheeled Drone
  13. Prior input sequence by auxiliary controller
  14. Attitude Controller
  15. Experiments
  16. ...and 4 more sections

Figures (7)

  • Figure 1: Our target scenario. A two-wheeled drone follows a reference trajectory while avoiding obstacles by switching between ground driving and flight modes.
  • Figure 2: Coordinate frame, geometry, and modes of two-wheeled drone.
  • Figure 3: Overview of the navigation system for the two-wheeled drone.
  • Figure 4: Simulation results of position and attitude
  • Figure 5: Simulation results of thrust
  • ...and 2 more figures