CSAIL Event Calendar: Previous Series
Motion Strategies for landmark-based navigation and surveillance
Speaker: Sourabh Bhattacharya , University of Illinois at Urbana-Champaign
Optimal control of dynamical systems has been an active area of research for the past few decades. In this talk, I will present a number of challenging problems in optimal control and game theory related to robotics and motion planning. The first problem is regarding the optimal control of non-holonomic systems using feedback from vision sensors. The two main tools provided in classical optimal control literature are the Maximum Principle and Dynamic programming. I will present a problem in landmark-based navigation of mobile robots in which both the techniques fail to completely characterize the globally optimal trajectories of the system. Instead the optimal paths are obtained by a systematic elimination of the set of all possible extremal trajectories. The next problem is regarding my research on visibility-based pursuit evasion. A lot of fundamental results exist in the literature regarding visibility-based search using mobile agents. On the contrary, there are very few fundamental results in visibility-based target-tracking. I will present a differential game-theoretic framework that provides new insights regarding the analysis of an open problem in bounded speed target-tracking. Finally, I will talk about my current research on jamming attacks in mobile communication network. I will present a scenario in which secure communication in a UAV formation is jeopardized due to malicious intrusion. I will present strategies for spatial reconfiguration of the formation for low dimensional problems.