This report documents the sensor analysis phase of the STVF funded project: Event based Kalman Filtering. The characteristics of different sensors relevant to the Autonomous Guided Vehicle (AGV) project at the Department of Automation, DTU, are investigated. Currently the AGV is equipped with three types of sensors: a wheel encoder on each of the two driving wheels, two cameras mounted on pan-tilt units, and an ultrasound tranciever/receiver. Models of the sensors are derived and an attempt is made to establish useful stochastic descriptions of the uncertainty of the sensors. Such stochastic models of the sensors and their uncertainty are necessary for data fusion purposes.
For further information, please contact, Finn Kuno
Christensen, IMM, Bldg. 321, DTU
Last modified Feb 4, 1998
Phone: (+45) 4588 1433. Fax: (+45) 4588
2673, E-mail: fkc@imm.dtu.dk