Legged-wheel control and optimized jumping forward for Five-star wheel robot
Loading...
Authors
Nguyen, Luan V
Issue Date
2020
Type
Thesis
Language
Keywords
Five-star wheel robot , Hybrid legged-wheel control , Legged robot jumping forward , Legged-wheel , Legged-wheel control
Alternative Title
Abstract
The most powerful capability of a legged robot is that it acquires not only the excellent flexibility and freedom to walk on rough terrain and complex cluttered environments but also the eminent jumping ability distinct from the other tracked or wheeled ground robots. This thesis proposed a novel legged-wheel control for a Five-star wheel robot (FSW robot). The concept legged-wheel applied in this premise is implicit in the meaning of applying legged control for a five-star wheel as for a leg. This proposed control relied on a hypothesis that the special design of the five-star wheel perhaps possessed both naturally hybrid ability of wheeled and legged locomotion. Hence, offering a legged-wheel control hypothesis for this robot can enhance and maximize the naturally potential and internal power of this different design. Consequently, it equips this robot with a powerful locomotion to climb up or jump over high and far obstacles maybe unbeatable by the same size tracked or wheeled robot. This thesis not only proposed a novel control but also implemented a framework for controlling locomotion of this robot both in Low-level and High-level controllers by C++ language on ROS and Gazebo. In particular, it has accomplished a distinct jumping forward controller for the FSW robot over the highest 0.75m obstacles and the widest 0.8m creeks. Ultimately, it has been tested in a DARPA’s SubT cave simulation. They are corroborating evidence for our proposed hypothesis irrefutable.