Abstract
Magnetic markers are used as position reference devices for vehicles in AHS (Automated Highway Systems). There is a trade-off problem between the accuracy of markers and the construction cost. Lateral deviations measured on a vehicle include position errors of the markers. This paper describes the method of reducing the position measurement errors by use of Kalman filter. Disturbance observer is added to Kalman filter in order that the algorithm may have robustness against input errors and model errors. The effect of this method is confirmed by simulation and experiment.