Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.


Dynamic Analysis and Control of a Robotic Bipedal Locomotion System


S. Plüss

Semester Thesis, FS16 (10514)

Bipedal locomotion, in its simplest form, consists of the generation of target joint angles to produce a stable movement in some direction. Walking pattern generators face a trade-off between accuracy and computational cost. In this semester thesis, multiple model-based walking pattern generators have been investigated to gain insights into how the walking motion is generated. The current walking engine of the humanoid robot Nao, used by ETH's Z-Knipsers in the RoboCup, has been proven to produce an unstable gait that makes the robots slow. Thus, an alternative control algorithm based on the three-dimensional linear inverted pendulum mode is proposed that puts more weight on the stability in the selection of step sizes. The dynamic equations of the model can be converted into a linear discrete-time system whose system matrix is hybrid pertaining to the step duration. Exploiting the hybrid character, the step duration can be switched to avoid fast and small consecutive steps that may cause the servos to overheat or the robot to stray o course. A family of LQR controllers is suggested to compute the step sizes in order to preserve the model's stability. Simulations using MATLAB showed that the proposed algorithm is robust to noise and parameter uncertainties. Therefore it is promising in terms of stability and speed of the gait. It's implementation in the real robot holds great promise for the future.

Supervisors: Georgios Darivianakis, Benjamin Flamm, John Lygeros


Type of Publication:

(13)Semester/Bachelor Thesis

File Download:

Request a copy of this publication.
(Uses JavaScript)
% Autogenerated BibTeX entry
@PhdThesis { Xxx:2016:IFA_5465
Permanent link