☆ 4.7 Article

VAE-Loco: Versatile Quadruped Locomotion by Learning a Disentangled Gait Representation

IEEE TRANSACTIONS ON ROBOTICS (2023)

Related references

Note: Only part of the references are listed.

Article Computer Science, Artificial Intelligence

A feasibility-driven approach to control-limited DDP

Carlos Mastalli et al.

Summary: Differential dynamic programming (DDP) is a direct method for trajectory optimization, but suffers from numerical instability and limited initialization options. This study proposes a feasibility-driven approach to address these issues by regulating dynamic feasibility and ensuring control limits. Experimental results demonstrate the competitiveness of the new method in terms of numerical convergence, convergence rate, and runtime performance.

AUTONOMOUS ROBOTS (2022)

Add to Collection

Article Robotics

RLOC: Terrain-Aware Legged Locomotion Using Reinforcement Learning and Optimal Control

Siddhant Gangapurwala et al.

Summary: We propose a unified approach for quadrupedal planning and control using both model-based and data-driven methods. By utilizing reinforcement learning, sensory information and velocity commands are mapped into footstep plans, which are then tracked by a model-based motion controller. The results demonstrate the robustness of our method over a variety of complex terrains.

IEEE TRANSACTIONS ON ROBOTICS (2022)

Add to Collection

Proceedings Paper Automation & Control Systems

Vision-aided Dynamic Quadrupedal Locomotion on Discrete Terrain using Motion Libraries

Ayush Agrawal et al.

Summary: This paper presents a framework that allows quadrupedal robots to traverse challenging terrains with discrete footholds using visual feedback. By using an offline library of periodic gaits and a geometric model predictive controller, the robot is able to achieve aperiodic motions and incorporate visual feedback using terrain mapping tools and a depth camera.

2022 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA 2022) (2022)

Add to Collection

Article Robotics

Planning in Learned Latent Action Spaces for Generalizable Legged Locomotion

Tianyu Li et al.

Summary: Hierarchical learning framework presented in this letter jointly learns low-level controller and high-level latent action space, generalizes to multiple robots, outperforming baselines on multiple tasks and two simulations. Results also show robustness and sample efficiency of the approach on hardware, achieving locomotion tasks with only 2000 hardware samples in an unstructured outdoor setting.

IEEE ROBOTICS AND AUTOMATION LETTERS (2021)

Add to Collection

Proceedings Paper Automation & Control Systems

Receding-Horizon Perceptive Trajectory Optimization for Dynamic Legged Locomotion with Learned Initialization

Oliwier Melon et al.

Summary: This study presents a pipeline for flexible parameterized trajectories that optimize perceptive and dynamic quadruped locomotion in an online, receding-horizon manner. The initial guess passed to the optimizer affects the computation needed for convergence and solution quality. Two methods for generating good guesses are considered, one being a heuristic initializer and the other using latent-mode trajectory regression to imitate expert data.

2021 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA 2021) (2021)

Add to Collection

Article Computer Science, Information Systems

Model Predictive Control With Environment Adaptation for Legged Locomotion

Niraj Rathod et al.

Summary: The study presents a real-time Nonlinear Model Predictive Control (NMPC) tailored to legged robots to achieve dynamic locomotion while enhancing leg mobility and terrain adaptation. Experimental results show that the NMPC is successful in traversing various terrains and obstacles.

IEEE ACCESS (2021)

Add to Collection

Article Computer Science, Software Engineering

Character Controllers using Motion VAEs

Hung Yu Ling et al.

ACM TRANSACTIONS ON GRAPHICS (2020)

Add to Collection

Proceedings Paper Automation & Control Systems

Extracting Legged Locomotion Heuristics with Regularized Predictive Control

Gerardo Bledt et al.

2020 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA) (2020)

Add to Collection

Proceedings Paper Automation & Control Systems

First Steps: Latent-Space Control with Semantic Constraints for Quadruped Locomotion

Alexander L. Mitchell et al.

2020 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS) (2020)

Add to Collection

Article Robotics

Guided Constrained Policy Optimization for Dynamic Quadrupedal Robot Locomotion

Siddhant Gangapurwala et al.

IEEE ROBOTICS AND AUTOMATION LETTERS (2020)

Add to Collection

Article Robotics

Learning agile and dynamic motor skills for legged robots

Jemin Hwangbo et al.

SCIENCE ROBOTICS (2019)

Add to Collection

Article Robotics

Per-Contact Iteration Method for Solving Contact Dynamics

Jemin Hwangbo et al.

IEEE ROBOTICS AND AUTOMATION LETTERS (2018)

Add to Collection

Article Robotics

Fast Trajectory Optimization for Legged Robots Using Vertex-Based ZMP Constraints

Alexander W. Winkler et al.

IEEE ROBOTICS AND AUTOMATION LETTERS (2017)

Add to Collection

Article Computer Science, Artificial Intelligence

Centroidal dynamics of a humanoid robot

David E. Orin et al.

AUTONOMOUS ROBOTS (2013)

Add to Collection

Article Engineering, Industrial

Gait parameters as predictors of slip severity in younger and older adults

BE Moyer et al.

ERGONOMICS (2006)

Add to Collection

© Peeref 2019-2024. All rights reserved.