scholarly journals Preference-Based Learning for User-Guided HZD Gait Generation on Bipedal Walking Robots

2021 ◽  
Author(s):  
Maegan Tucker ◽  
Noel Csomay-Shanklin ◽  
Wen-Loong Ma ◽  
Aaron D. Ames
Keyword(s):  
Bipedal Walking ◽  
Walking Robots ◽  
Gait Generation

Enhanced Footsteps Generation Method for Walking Robots Based on Convolutional Neural Networks

2019 ◽  
pp. 16-39 ◽  
Author(s):  
Sergei Savin ◽  
Aleksei Ivakhnenko
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Convolutional Neural Network ◽  
Network Design ◽  
Bipedal Walking ◽  
Training Dataset ◽  
Walking Robots ◽  
Foot Placement ◽  
Gait Generation ◽  
Free Region

In this chapter, the problem of finding a suitable foothold for a bipedal walking robot is studied. There are a number of gait generation algorithms that rely on having a set of obstacle-free regions where the robot can step to and there are a number of algorithms for generating these regions. This study breaches the gap between these algorithms, providing a way to quickly check if a given obstacle free region is accessible for foot placement. The proposed approach is based on the use of a classifier, constructed as a convolutional neural network. The study discusses the training dataset generation, including datasets with uncertainty related to the shapes of the obstacle-free regions. Training results for a number of different datasets and different hyperparameter choices are presented and showed robustness of the proposed network design both to different hyperparameter choices as well as to the changes in the training dataset.

ZMP-Based Trajectory Generation for Bipedal Robots Using Quadratic Programming

2020 ◽  
pp. 266-285
Author(s):  
Sergei Savin
Keyword(s):  
Numerical Simulation ◽  
Quadratic Programming ◽  
Center Of Mass ◽  
Trajectory Generation ◽  
Bipedal Walking ◽  
Walking Robots ◽  
Bipedal Robots ◽  
Change Of Variables ◽  
Viable Solution ◽  
Modern Optimization

In this chapter, the problem of trajectory generation for bipedal walking robots is considered. A number of modern techniques are discussed, and their limitations are shown. The chapter focuses on zero-moment point methods for trajectory generation, where the desired trajectory of that point can be used to allow the robot to keep vertical stability if followed, and presents an instrument to calculate the desired trajectory for the center of mass for the robot. The chapter presents an algorithm based on quadratic programming, with an introduction of a slack variable to make the problem feasible and a change of variables to improve the numeric properties of the resulting optimization problem. Modern optimization tools allow one to solve such problems in real time, making it a viable solution for trajectory planning for the walking robots. The chapter shows a few results from the numerical simulation made for the algorithm, demonstrating its properties.

Simplified Triped Robot for Analysis of Three-Dimensional Gait Generation

2017 ◽  
Vol 29 (3) ◽  
pp. 528-535
Author(s):  
Yoichi Masuda ◽  
◽  
Masato Ishikawa
Keyword(s):  
High Speed ◽  
Force Feedback ◽  
Three Dimensional ◽  
Radial Symmetry ◽  
The Body ◽  
Walking Robots ◽  
Body Structure ◽  
Gait Patterns ◽  
Gait Generation ◽  
Proposed Model

[abstFig src='/00290003/08.jpg' width='230' text='The tripedal robot “Martian petit”' ] Significant efforts to simplify the body structure of multi-legged walking robots have been made over the years. Of these, the Spring-Loaded-Inverted-Pendulum (SLIP) model has been very popular, therefore widely employed in the design of walking robots. In this paper, we develop a SLIP-based tripedal walking robot with a focus on the geometric symmetry of the body structure. The proposed robot possesses a compact, light-weight, and compliant leg modules. These modules are controlled by a distributed control law that consists of decoupled oscillators with only local force feedback. As demonstrated through experiments, the simplified design of the robot makes possible the generation of high-speed dynamic locomotion. Despite the structural simplicity of the proposed model, the generation of several gait-patterns is demonstrated. The proposed minimalistic design approach with radial symmetry simplifies the function of each limb in the three-dimensional gait generation of the robot.

Sign in / Sign up

Export Citation Format

Share Document