[Paper Review] Online Dynamic Motion Planning and Control for Wheeled Biped Robots

Songyan Xin, Sethu Vijayakumar|arXiv (Cornell University)|Mar 7, 2020

Robotic Locomotion and Control24 references4 citations

TL;DR

This paper presents a novel online dynamic motion planning and control framework for wheeled biped robots using a hybrid Cart-Linear Inverted Pendulum Model (Cart-LIPM) for rolling and under-actuated LIPM for gait transitions. The approach enables real-time whole-body motion generation via model predictive control and inverse dynamics, successfully demonstrating dynamic hybrid locomotion over obstacles in simulation for the first time.

ABSTRACT

Wheeled-legged robots combine the efficiency of wheeled robots when driving on suitably flat surfaces and versatility of legged robots when stepping over or around obstacles. This paper introduces a planning and control framework to realise dynamic locomotion for wheeled biped robots. We propose the Cart-Linear Inverted Pendulum Model (Cart-LIPM) as a template model for the rolling motion and the under-actuated LIPM for contact changes while walking. The generated motion is then tracked by an inverse dynamic whole-body controller which coordinates all joints, including the wheels. The framework has a hierarchical structure and is implemented in a model predictive control (MPC) fashion. To validate the proposed approach for hybrid motion generation, two scenarios involving different types of obstacles are designed in simulation. To the best of our knowledge, this is the first time that such online dynamic hybrid locomotion has been demonstrated on wheeled biped robots.

Motivation & Objective

• To enable dynamic, online motion planning for wheeled biped robots navigating complex environments.
• To address the challenge of coordinating wheeled locomotion with legged gait transitions in real time.
• To develop a hierarchical control framework that integrates rolling and stepping motions seamlessly.
• To validate the framework on diverse obstacle scenarios through simulation.

Proposed method

• Proposes the Cart-Linear Inverted Pendulum Model (Cart-LIPM) as a template for stable rolling motion.
• Introduces an under-actuated LIPM variant to model and control contact transitions during walking.
• Employs a hierarchical model predictive control (MPC) framework for online trajectory generation.
• Uses an inverse dynamic whole-body controller to coordinate all joints, including wheels, for precise motion tracking.
• Designs a hybrid motion generation strategy that switches between rolling and stepping modes based on terrain and obstacle conditions.
• Implements the framework in simulation to test robustness across diverse obstacle types.

Experimental results

Research questions

• RQ1 How can wheeled biped robots achieve dynamic locomotion that combines efficient rolling with agile stepping over obstacles?
• RQ2 What control model enables stable and adaptive motion planning during transitions between rolling and legged gaits?
• RQ3 Can online, real-time motion planning be effectively achieved using MPC and inverse dynamics for a fully actuated wheeled biped?
• RQ4 How does the hybrid motion generation framework perform under varying obstacle configurations in simulation?
• RQ5 What is the feasibility of integrating rolling and stepping behaviors in a unified control architecture for wheeled biped robots?

Key findings

• The proposed framework successfully enables online dynamic motion planning and control for wheeled biped robots in simulation.
• The Cart-LIPM and under-actuated LIPM models effectively capture rolling and contact transition dynamics.
• The hierarchical MPC-based controller achieves stable and accurate motion tracking across both rolling and stepping phases.
• The system demonstrates robust performance in navigating diverse obstacle types, including gaps and raised obstacles.
• To the best of the authors' knowledge, this is the first demonstration of online dynamic hybrid locomotion on wheeled biped robots.
• The integration of wheel and leg control via inverse dynamics ensures coordinated whole-body motion.