Faculty Advisor

Beck, Joseph E.

Faculty Advisor

Gennert, Michael A.

Abstract

Existing methods to allow humanoid robots to walk suffer from a lack of adaptability to new and unexpected environments, due to their reliance on using only higher-level motion control with relatively fixed sub-motions, such as taking an individual step. These conventional methods require significant knowledge of controls and assumptions about the expected surroundings. Humans, however, manage to walk very efficiently and adapt to new environments well due to the learned behaviors. Our approach is to create a reinforcement learning framework that continuously chooses an action to perform, by utilizing a neural network to rate a set of joint values based on the current state of the robot. We successfully train the Boston Dynamics Atlas robot to learn how to walk with this framework.

Publisher

Worcester Polytechnic Institute

Date Accepted

April 2016

Major

Computer Science

Major

Robotics Engineering

Project Type

Major Qualifying Project

Accessibility

Unrestricted

Advisor Department

Computer Science

Share

COinS