An Artificially lntelligent Vehicle Highway System

Vidal Ashkenazi, Terry Moore, Mark Dumville, David Lowe, and Maria Tsakiri

Abstract:	Conditions for land navigation are among the most severe in the urban environment, where there is no low cost positioning system capable of continuously providing the necessary high accuracy positional information. The Global Positioning System (GPS) is having a profound impact on the development of automatic vehicle location systems but suffers from a number of problems. A hybrid and robust GPS-based system is required to offer continuous and accurate positioning. This study has investigated the integration of differential GPS measurements with a digital odometer, a flux-gate compass and tilt sensors to give a low cost 3-dimensional positional system for the vehicle navigation. In contrast to the traditional method of sensor integration using Kalman Filtering techniques, a novel approach has been adopted in the use of artificial intelligence, and in particular, neural computing techniques. The integrated neural architecture consists of a stand-alone neural network processing model augmented with a dead reckoning (DR) position fix algorithm. The neural network model accepts sensor measurements as input from which it computes an optimised position fix, subsequently used to calibrate the sensors for systematic drifts. Studies have been performed to investigate this novel approach, the results of the study are compared against those achieved using the conventional kahnan filtering techniques of sensor integration.
Published in:	Proceedings of the 8th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GPS 1995) September 12 - 15, 1995 Palm Springs, CA
Pages:	1809 - 1818
Cite this article:	Ashkenazi, Vidal, Moore, Terry, Dumville, Mark, Lowe, David, Tsakiri, Maria, "An Artificially lntelligent Vehicle Highway System," Proceedings of the 8th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GPS 1995), Palm Springs, CA, September 1995, pp. 1809-1818.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In