Development and Implementation of Walking Algorithms for Humanoid Robots

Abstract

The objective of this proposal is to develop walking algorithms for humanoid robots. This area of research is at the state-of-the-art in the context of robotics, because of the importance of integrating the robots to the environments in which humans live and operate. Currently, model-based approaches have been the most successful being able to realize the walking movement of several distinct humanoid robotic platforms. The present proposal aims to improve the performance of the walk (in terms of speed and energy expenditure) and the stability (with respect to avoiding falls) in relation to the state-of-the-art. Therefore, it is intended to develop walking algorithms that overcome some of the difficulties encountered in the more recent approaches such as the use of a single mathematical model for the robot throughout the walking motion, the imposition of a constant height for the Center of Mass of the robot and pre-setting of the period for execution of a walking cycle. In order to validate the proposed algorithms with experiments, it is planned to acquire a software for simulation with a high-fidelity model and a humanoid robot suitable for research. (AU)