| sharPKUngfu Team's research interests include localization, multi-agent coordination, legged locomotion and robot vision. |
|
In localization, we created a novel approach for a group of vision-based autonomous robots to localize in dynamic
environments. Our hybrid system method for localization which consists of on-line and off-line subsystems inspired
by human cognition has been applied to different robot systems, including RoboCup Four-legged Soccer, autonomous
robot formation and coordination. Current work focuses on further development in algorithm and experience based
localization without any artificial landmark. Selected publications in this field: [1] |
|
In multi-agent coordination, we have done a lot of work both in theoretic aspect and application aspect.
We focus on modeling, simulation, and mathematical analysis for dynamic behavior of swarms, which are a
typical kind of multi-agent systems from nature. Consensus problem and aggregation phenomena are investigated
in detail. We considered many complex cases, including switching communication protocols, communication time-delays,
and with/without a leader in the swarms. Based on these theoretic results, we try to solve engineering problems in
multi-robot systems, especially for mobile robots. For example, formation control problem is viewed as a general case
of consensus problem. Selected publications in this field: [2] ~ [7] |
|
In legged locomotion, we divided our research in two parts. On the one hand, we use Aibo to study four-legged locomotion.
A Particle Swarm Optimization based approach for gait optimization is created. One the other hand, we implemented humanoid
robot to analyze passive dynamic walking and running. Selected publications in this field: [8] ~ [9] |
| Publications |
Selected publications in recent years:
|
 |
 |
|