Towards Model-Free SLAM Using a Single Laser Range Scanner for Helicopter MAV

TL;DR: A new solution for the SLAM problem is presented which makes use of a scan matching algorithm, and does not rely on bayesian filters, and can be applied to various platforms, including micro aerial vehicles that lack dynamic models.

Abstract: A new solution for the SLAM problem is presented which makes use of a scan matching algorithm, and does not rely on bayesian filters. The virtual map is represented in the form of an occupancy grid, which stores laser scans based on the estimated position. The occupancy grid is scanned by means of ray casting to get a scan of the virtual world, called ”virtual scan”. The virtual scan therefore contains data from all previously acquired laser measurements and hence serves as the best representation of the surroundings. New laser scans are matched against the virtual scan to get an estimate of the new position. The scan matching cost function is minimized via an adaptive direct search with boundary updating until convergence. The resulting method is model-free and can be applied to various platforms, including micro aerial vehicles that lack dynamic models. Experimental validation of the SLAM method is presented by mapping a typical office hallway environment with a closed loop, using a manually driven platform and a laser range scanner. The mapping results are highly accurate and the loop closure area appears to be seamless, in spite of no loop closure algorithms and no post-mapping correction processes.