Road Slope Information from GPS-Derived Trajectory Data

Abstract

The World Solar Challenge was held in 1996, and will be held again in 1999. As part of the University of New South Wales solar car project, global positioning system techniques were used to determine the road slope characteristics of the Stuart Highway running north-south from Darwin to Adelaide. This data formed the basis of a geographic information system to support solar car race strategy. Motivated by this project, a new Kalman filter algorithm for road slope determination has been developed. The state vector in the Kalman filter includes the height and slope of the road, rather than the conventional parameters of height and vertical velocity. In this way the slope can be derived directly within the Kalman filter, the system equations can be more easily developed and the system noise can be significantly reduced. Numerical results confirm that the new Kalman filter algorithm can be easily implemented, and that the accuracy of road slope determination is improved. This algorithm is appropriate for any survey application concerned with the determination of vertical profile.