|
|
|
|
|
ABSTRACT My talk will be on trajectory planning for the purpose of localising targets. It considers multiple robots with bearing only sensors and motion constraints. The robots are to maximise the information gain of the locations of the targets within a terminal time. The information on the targets initially have a large uncertainty. Methods such as the EIF (extended information filter) and MPC (model predictive control) are used to solve this problem. Different strategies are analysed such as EETS (exhaustive expansion tree search) and SQP (sequential quadratic programming), centralised and decentralised control. Centralised EETS is shown to provide a near optimal solution. SPEAKER BIO I am a PhD student from the University of Technology, Sydney, under the supervision of Prof. Gamini Dissanayake. Cindy Leung Email: cleung@eng.uts.edu.au URL: http://www.cas.edu.au/content.php/232.html?personid=59 |
|
Back to CRES |