Trajectory estimation and correction for a wheeled mobile robot using heterogeneous sensor and kalman filter

摘要

In this paper, the problem of estimating position and orientation of a wheeled mobile robot is addressed. In the experimental setup, two encoders connected to the driving wheels are used, for short period,s for dead-reckoning. The increasing error in position and orientation estimated by these internal sensors is periodically limited by the use of an external logical sensor, based on an ultrasonic belt and a geometric algorithm. The Kalman filter is used to fuse data coming from both external and internal sensors. A probabilistic approach is used to obtain statistical characteristics of errors related to internal and external sensors. Finally, experimental results are presented.