An Integrated Vehicle Positioning System for Safety Applications

Shawn Weisenburger, Christopher Wilson

Abstract:	Applications based on high accuracy vehicle positioning combined with an equally high accuracy map database can improve vehicle safety. Applications include traffic controls and warnings for curves, lane departure warning, and even vehicle control. A prototype positioning system was developed around a Mercedes M320. The navigation sensors include Differential GPS (DGPS), a fiber optic gyro, and wheel speed sensors. The data fusion is accomplished using a 9 state Kalman filter. To test the effectiveness of the filter, two data sets were collected, one along the freeway and one in an urban canyon environment. For the freeway data set, accuracies better than 5m were obtained during 1 minute simulated GPS outages. The standard deviations after 1 minute without GPS were approximately 8m. Data analysis showed that on straight sections of road, the major error source was the changing gyro bias, while on curvy sections of road, the major error source was the gyro scale factor. The wheel speed sensors provided accuracies of approximately 0.3% of the distance traveled. For the urban canyon data set, GPS outages up to 4 minutes were encountered and standard deviations varied from less than one meter up to fifteen meters.
Published in:	Proceedings of the 1999 National Technical Meeting of The Institute of Navigation January 25 - 27, 1999 Catamaran Resort Hotel San Diego, CA
Pages:	361 - 368
Cite this article:	Updated citation: Published in NAVIGATION: Journal of the Institute of Navigation
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In