Position-Based Lateral Balance Control for Knee-Stretched Biped Robot

Kajita, Shin; Benallegue, Mehdi; Cisneros, Rafael; Sakaguchi, Toshifumi; Morisawa, Masaaki; Kaminaga, Hiroshi; Kumagai, Iori; Kaneko, Kenji; Kanehiro, Fumio

doi:10.1109/humanoids43949.2019.9035077

Cited by 13 publications

(9 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As aforementioned, LIPM tries to keep the COM's vertical position at a predefined position which causes the knee joints to be always bent. Indeed, walking with bent knees consumes more energy and does not resemble human walking [36]. To release this constraint and generate more energy efficient and human-like walking, a sinusoidal motion is assigned to the vertical motion of the COM:…”

Section: Dynamics Modelmentioning

confidence: 99%

Robust Biped Locomotion Using Deep Reinforcement Learning on Top of an Analytical Control Approach

Kasaei,

Abreu,

Lau

et al. 2021

Preprint

View full text Add to dashboard Cite

This paper proposes a modular framework to generate robust biped locomotion using a tight coupling between an analytical walking approach and deep reinforcement learning. This framework is composed of six main modules which are hierarchically connected to reduce the overall complexity and increase its flexibility. The core of this framework is a specific dynamics model which abstracts a humanoid's dynamics model into two masses for modeling upper and lower body. This dynamics model is used to design an adaptive reference trajectories planner and an optimal controller which are fully parametric. Furthermore, a learning framework is developed based on Genetic Algorithm (GA) and Proximal Policy Optimization (PPO) to find the optimum parameters and to learn how to improve the stability of the robot by moving the arms and changing its center of mass (COM) height. A set of simulations are performed to validate the performance of the framework using the official RoboCup 3D League simulation environment. The results validate the performance of the framework, not only in creating a fast and stable gait but also in learning to improve the upper body efficiency.

show abstract

Section: Dynamics Modelmentioning

confidence: 99%

Robust Biped Locomotion Using Deep Reinforcement Learning on Top of an Analytical Control Approach

Kasaei,

Abreu,

Lau

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…This means that the output for the ZMP tracking control must be calculated based on the future input in walking design control [27]. In robot walking, the ZMP is estimated based on the average value of the force sensors, as given in (10) and (11) corresponding to on the x-axis and y-axis, respectively. Here, n is the number of sensors on a sole, and f j means each sensor of numbering j.…”

Section: Robot Design and Controllermentioning

confidence: 99%

“…The mentioned indices are extremely important as individual walking features, but one index alone is not enough to judge personal walking. Therefore, many researchers have studied the use of one or more indices for the analysis of human walking [2][3][4][5][6][7][8][9][10][11][12][13][14].…”

Section: Introductionmentioning

confidence: 99%

An Approach to Human Walking Analysis Based on Balance, Symmetry and Stability Using COG, ZMP and CP

Kim

Murakami

2020

Applied Sciences

View full text Add to dashboard Cite

The parameters of walking have been studied from the viewpoints of joint rotation and translation of body. The balance and symmetry of walking are indispensable features to understand for healthy walking, while also being a personal characteristic. However, quantification has not been easy to carry out in the case of the conventional gait parameters COG (center of gravity) and ZMP (zero moment point). In this approach, the CP (crossing point) is proposed to quantify the concept of symmetry and balance by comparing it to the COG and ZMP. The CP is estimated based on the intersection between the hip line and the ankle line. While the hip line is fixed on the upper body where the COG is, the ankle line is altered depending on the each footfall, where the ZMP is. Therefore, the values of COG, ZMP, and CP have similar or different tendencies in terms of whether balanced walking results in symmetry or not. The validity of this is verified by carrying out a simulation with robot walking, and an experiment using human walking. Through additional experiments, it was noticed that the CP was able to improve the role of COG and ZMP in terms of not only stability, but also its relationship with the movement range of the lower limbs.

show abstract

“…Ensuring this usually results in the well-known, bent-knee walking to avoid kinematic singularities. Improving this issue has been the aim of multiple research works [20,13,15]. The second limitation stems from the assumption of zero angular momentum around the CoM.…”

Section: Introductionmentioning

confidence: 99%

Direct Centroidal Control for Balanced Humanoid Locomotion

Ficht

Behnke

2022

Lecture Notes in Networks and Systems

View full text Add to dashboard Cite

We present an integrated approach to locomotion and balancing of humanoid robots based on direct centroidal control. Our method uses a five-mass description of a humanoid. It generates whole-body motions from desired foot trajectories and centroidal parameters of the robot. A set of simplified models is used to formulate general and intuitive control laws, which are then applied in real-time for estimating and regulating the center of mass position and orientation of the multibody's principal axes of inertia. The combination of proposed algorithms produces a stretched-leg gait with naturally looking upper body motions. As only a 6-axis IMU and joint encoders are necessary for the implementation, the portability between robots is high. Our method has been experimentally verified using an igus Humanoid Open Platform, demonstrating whole-body locomotion and push rejection capabilities.

show abstract

Position-Based Lateral Balance Control for Knee-Stretched Biped Robot

Cited by 13 publications

References 19 publications

Robust Biped Locomotion Using Deep Reinforcement Learning on Top of an Analytical Control Approach

Robust Biped Locomotion Using Deep Reinforcement Learning on Top of an Analytical Control Approach

An Approach to Human Walking Analysis Based on Balance, Symmetry and Stability Using COG, ZMP and CP

Direct Centroidal Control for Balanced Humanoid Locomotion

Contact Info

Product

Resources

About