Abstract
dc:descriptionHumanoid robot research and development have been an ongoing effort since the 1900sand can be broken down to two problems. A mechanical problem, getting a humanoid robot to move human-like or a software problem, getting a humanoid robot to behave human-like. These problems of moving and behaving human-like can be often solved using control theory as research advances. For the premise of this research, we explore how to balance and walk on non-flat terrain for the humanoid robot Darwin-Op. Since the focus was on the control theory, the vision control to detect the non-flat terrain was a side objective. The vision control was implemented by detecting if a non-flat terrain such as a ramp or stairs had a specific color. Once the object had been identify, some computer vision techniques can be applied to get the distance and orientation from the object. While the vision control isn’t as robust, it was enough to guide the robot to the obstacle. For the control theory aspect of the research, the robot was able to walk up stairs and ramps by using zero moment point trajectory planning with a cart-table model for the center of mass. After running some experiment, it was concluded that smooth surfaces such as ramp work the best compared to stairs. With the edges on a stairs, there were problems getting a proper step landing for step heights higher than the robots ankle. But it shows that autonomous movement on non-flat surfaces is viable.
Degree
thesis:*- Name thesis:degree_name
- MS in Computer Engineering (M.S.C.P.)
- Level thesis:degree_level
- Thesis
- Discipline thesis:degree_discipline
- Computer Science and Engineering
- Grantor
- University of South Florida
- Year dc:date
- 2022
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Truong, Tommy
- Contributors dc:contributor
-
- Alfredo Weitzenfeld, Ph.D.
- Les Piegl, Ph.D.
- Yu Sun, Ph.D.
Subjects
dc:subject × 6Identifiers
dc:identifier.*- Repository record dc:identifier
- https://digitalcommons.usf.edu/etd/9485