Vehicle ego-localization is commonly achieved by Bayesian methods like Extended Kalman Filtering. New approaches based on interval analysis intend to achieve the same goal in a guaranteed way. They assume that all model and measurement errors are bounded with known bounds without any other hypothesis on the probability distribution between bounds. We consider the localization as an interval constraint satisfaction problem (ICSP) solved by an Interval Constraint Propagation (ICP) algorithm. This paper introduces a new real-time ICP algorithm which corrects both position and heading. The proposed algorithm uses HC4 as a low level algorithm and has been validated with an outdoor vehicle equipped with a GPS receiver, a gyro and odometers. Furthermore, it is compared with the HC4 and 3B algorithms.
展开▼
