Possibilistic sonar data modeling for mobile robots

Sonar sensors are widely used in mobile robots applications such as navigat ion, map building, and localization. The performance of these sensors is af fected by the environmental phenomena, sensor design, and target characteri stics. Therefore, the readings obtained from these sensors are uncertain. T his uncertainty is often modeled by using Probability Theory. However, the probabilistic approach is valid when the available knowledge is precise whi ch is not the case in sonar readings. In this paper, the behavior of sonar readings reflected from walls and corners are studied, then new models of a ngular uncertainty and radial imprecision for sonar readings obtained from corners and walls are proposed. These models are represented by using Possi bility Theory, mainly possibility distributions.