Abstract:

Mobile sensor networks are collections of networked multiple mobile robots with sensors and expected as a strong tool for security and surveillance in urban area and environmental monitoring. Especially, a visual sensor brings us rich information of environment and 3-dimensional pose of the other robots compressed into 2-dimensional image plane. From this advantage, cooperative control based on visual information is widely studied. In this talk, we investigate visual feedback control of robotic network in SE(3) with uncertain target motion. First, we describe the target motion as rigid body motion with stochastic input. A stochastic differential equation is derived as an uncertain target motion. Then, we introduce a visual feedback target tracking control based on a vision-based nonlinear observer, called visual motion observer which estimates the target motion in 3-dimensional space. Furthermore, we investigate its stochastic performance of uncertain target tracking. Finally, visual feedback control is extended to cooperative target tracking and its stochastic performance is also discussed.

Short Bio:

Junya Yamauchi received B.Eng. degree from Nagoya University, M. Eng. degree and Ph.D. degree from Tokyo Institute of Technology, Japan. He is currently an assistant professor in the Department of Systems and Control Engineering, Tokyo Institute of Technology. He received 10th Asian Control Conference Best Paper Prize Award. His research interests include passivity-based cooperative control of human-robotic network and visual motion estimation and control.