Motion Planning in High-dimensional Spaces

This project deals with planning for higher dimensional spaces using the Parti-game and RRT algorithms. The Parti-game algorithm is a cell-based planner that creates dynamic discretizations of the environment. It learns the environment over time since the discretization it creates improves with multiple runs in the same environment. The RRT, on the other hand, is a randomized, sample-based search algorithm. By combining the RRT and Parti-game algorithms in a suitable manner, we can solve more planning problems, reduce the planning time and memory consumption while at the same time not compromising on trajectory quality.

Publications

• PDRRTs: Integrating Graph-Based and Cell-Based Planning.
  Ananth Ranganathan and Sven Koenig.
  IROS, pages 2799-2806, Sendai, Japan, 2004.