Probabilistic map learning: Necessity and difficulties
In the context of map learning, a mobile robot must build and maintain a representation of the environment incrementally while locating itself. The robot is equipped with a set of sensors of limited precisions and may have an inexact model of the system evolution. The representation model is probabilistic in nature and the EKF (Extended Kalman Filtering) algorithm has been widely adopted to model and propagate uncertainty in both the position of the robot and the geometric features.
We present an analysis of the EKF algorithm. The formalism and some of the main approaches are reviewed. An important aspect of the analysis is concerned with the necessity and the difficulties to maintain correlation between the state variables (the geometric features, the robot). The effects of nonlinearities on uncertainty propagation and the degradation of the sensors' uncertainty models are analysed and illustrated through simulation examples.
Keywordsmobile robot map learning probabilistic approach geometric features extended Kalman filtering critical observations
Unable to display preview. Download preview PDF.
- 1.N. Ayache. Vision Stéréoscopique et perception Multisensorielle. Collection Science Informatique. Inter Editions, Paris, 1989.Google Scholar
- 3.R. Chatila and J.-P. Laumond. Position referencing and consistent world modeling for mobile robots. In Proceedings of the IEEE International Conference on Robotics and Automation, pages 138–145, March 1985.Google Scholar
- 4.A. Elfes. Occupancy grids: A stochastic spatial representation for active robot perception. In S. S. Iyengar and A. Elfes, editors, Autonomous Mobile Robots Perception, Mapping, and Navigation, pages 60–70, 1991.Google Scholar
- 5.A. H. Jazwinski. Stochastic Processes and Filtering Theory, volume 64 of Mathematics in Science and Engineering. Academic Press, 1970.Google Scholar
- 6.S. Julier and J. Uhlmann. A general method for approximating nonlinear transformations of probability distributions. Technical report, Robotics Research Group, Dept. of Engineering Science, Oxford University, University of Oxford, Oxford, OX1 3PJ, United Kingdom, 1995.Google Scholar
- 8.J. J. Leonard and H. F. Durrant-Whyte. Directed Sonar Sensing for Mobile Robot Navigation. Kluwer Academic Publishers, Boston, 1992. Revised version of the thesis.Google Scholar
- 9.J. J. Leonard and H. F. Durrant-Whyte. Dynamic map building for an autonomous mobile robot. The International Journal of Robotics Research, 11(4):286–298, August 1992.Google Scholar
- 11.P. Moutarlier. Modélisation Autonome de l'Environnement par un Robot Mobile. PhD thesis, LAAS-CNRS, Université Paul Sabatier de Toulouse, LAAS-CNRS, 7, ave. du Colonel Roche 31077 Toulouse Cedex, France, October 1991.Google Scholar
- 12.P. Moutarlier and R. Chatila. Stochastic multisensory data fusion for mobile robot location and environment modelling. In Proceedings of the 5th International Symposium of Robotics Research, pages 207–216, August 1989.Google Scholar
- 13.W. D. Rencken. Concurrent localisation and map building for mobile robots using ultrasonic sensors. In Proc. of the 1993 IEEE/RSJ International Conference on Intelligent Robots and Systems, pages 2192–2197, July 1993.Google Scholar
- 14.R. Smith, M. Self, and P. Cheeseman. Estimating uncertain spatial relationships in robotics. In I. J. Cox and G. T. Wilfong, editors, Autonomous Robot Véhicules, pages 167–193, New York, 1990. Springer Verlag.Google Scholar