A real-time localization system for compactors

An operator-aiding system for compactors must incorporate a localization system. In this paper, we consider real-time kinematic (RTK) GPS and we present the solutions we have developed to maintain the positioning error lower than 0.2 m even during the satellite (limited) masking phases. By combining gyro, steering angle and speed measurements, we show that it is possible to obtain satisfying performances in dead-reckoning navigation with a low-cost internal sensor set. A forward-backward Kalman filter is proposed to deal with particularly long maskings, which typically occur when the machine works under bridges. Results of experiments carried out with an instrumented machine art: presented to validate the proposed solutions. (C) 2001 Elsevier Science B.V. All rights reserved.