Quadrupedal Robots Whole-Body Motion Control Based on Centroidal Momentum Dynamics

Liu, Mingmin; Qu, Daokui; Xu, Fang; Zou, Fengshan; Di, Pei; Tang, Chong

doi:10.3390/app9071335

Cited by 15 publications

(10 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Center-of-mass tracking is an effective policy for keeping quadrupedal robots stable [21]. The function of the target RMP is to maintain the COM of the robot's trunk at the desired position by outputting the leg's locomotion.…”

Section: Target Rmp For Center-of-mass Trackingmentioning

confidence: 99%

Learning Advanced Locomotion for Quadrupedal Robots: A Distributed Multi-Agent Reinforcement Learning Framework with Riemannian Motion Policies

Wang,

Sagawa,

Yoshiyasu

2024

Robotics

View full text Add to dashboard Cite

Recent advancements in quadrupedal robotics have explored the motor potential of these machines beyond simple walking, enabling highly dynamic skills such as jumping, backflips, and even bipedal locomotion. While reinforcement learning has demonstrated excellent performance in this domain, it often relies on complex reward function tuning and prolonged training times, and the interpretability is not satisfactory. Riemannian motion policies, a reactive control method, excel in handling highly dynamic systems but are generally limited to fully actuated systems, making their application to underactuated quadrupedal robots challenging. To address these limitations, we propose a novel framework that treats each leg of a quadrupedal robot as an intelligent agent and employs multi-agent reinforcement learning to coordinate the motion of all four legs. This decomposition satisfies the conditions for utilizing Riemannian motion policies and eliminates the need for complex reward functions, simplifying the learning process for high-level motion modalities. Our simulation experiments demonstrate that the proposed method enables quadrupedal robots to learn stable locomotion using three, two, or even a single leg, offering advantages in training speed, success rate, and stability compared to traditional approaches, and better interpretability. This research explores the possibility of developing more efficient and adaptable control policies for quadrupedal robots.

show abstract

Section: Target Rmp For Center-of-mass Trackingmentioning

confidence: 99%

Learning Advanced Locomotion for Quadrupedal Robots: A Distributed Multi-Agent Reinforcement Learning Framework with Riemannian Motion Policies

Wang,

Sagawa,

Yoshiyasu

2024

Robotics

View full text Add to dashboard Cite

show abstract

“…Since the 1980s [1], the locomotion of quadruped robots has been widely investigated to ascertain a stable and more effective walking motion. Studies have typically involved either the moment of inertia for the entire robot system [2][3][4] or the zero-moment point (ZMP) [5,6] to optimize the trajectory for the robot's center of mass in order to allow the robot to walk in a stable fashion. To this end, the locomotion control inputs the optimized center-of-mass (COM) trajectory result into the inverse kinematics calculations to obtain the joint solution.…”

Section: Introductionmentioning

confidence: 99%

Time Efficiency Improvement in Quadruped Walking with Supervised Training Joint Model

et al. 2023

View full text Add to dashboard Cite

To generate stable walking of a quadruped, the complexity of the configuration of the robot involves a significant amount of optimization that decreases to its time efficiency. To address this issue, a machine learning method was used to build a simplified control policy using joint models for the supervised training of quadruped robots. This study considered 12 joints for a four-legged robot, and each joint value was determined based on the conventional method of walking simulation and prepossessed, equaling 2508 sets of data. For data training, the multilayer perceptron model was used, and the optimized number of epochs used to train the model was 5000. The trained models were implemented in robot walking simulations, and they improved performance with an average distance error of 0.0719 m and a computational time as low as 91.98 s.

show abstract

“…The advantages of legged locomotion depend on the postures, the number of legs and the functionality of leg. The static motion has been implemented in multi-legged robots system to make their motion stable, but which decreases the motion of the robot (Raibert, 1986; Liu et al , 2019). Most of the robot has the characteristic of dynamic and static motion directly by keeping at least one or two legs in ground surfaces.…”

Section: Introductionmentioning

confidence: 99%

Design, analysis and method of foot trajectory generation for quadruped robots in swing-stance phase

Biswal

Mohanty

2022

WJE

View full text Add to dashboard Cite

Purpose Legged walking robots have numerous advantages over the wheel or tracked robots due to their strong operational ability and exposure to the complex environment. This paper aims to present details about the mechanical formation and a new conceptual elliptical trajectory generation discussed throughout the paper of the quadruped robot. Design/methodology/approach Initially, a realistic CAD model of the four-legged robot is developed in Solidwork-2019. The proposed model’s forward and inverse kinematics equations are deduced using Denavit–Hartenberg parameters. Based on geometry and kinematics, manipulability and obstacle avoidance are investigated. A method of galloping trajectory is proposed for aiming the increase of upright direction impulse, which is produced by ground reaction force at each step frequency. Furthermore, the locomotion equation of the ellipse trajectory is derived by setting transition angle polynomial of free-fall phase, stance phase and swing phase and the constraints. Findings Finally, a successive simulation on a 2D sagittal plane is performed to check and verify the usefulness of the proposed trajectory. Before the development of the full quadruped, a single prototype leg is generated for experimental verification of the dynamic simulations. Originality/value The proposed trajectory is novel in that it uses force tracking control, which is intended to improve the quadruped robot’s robustness and stability.

show abstract

Quadrupedal Robots Whole-Body Motion Control Based on Centroidal Momentum Dynamics

Cited by 15 publications

References 37 publications

Learning Advanced Locomotion for Quadrupedal Robots: A Distributed Multi-Agent Reinforcement Learning Framework with Riemannian Motion Policies

Learning Advanced Locomotion for Quadrupedal Robots: A Distributed Multi-Agent Reinforcement Learning Framework with Riemannian Motion Policies

Time Efficiency Improvement in Quadruped Walking with Supervised Training Joint Model

Design, analysis and method of foot trajectory generation for quadruped robots in swing-stance phase

Contact Info

Product

Resources

About