Improved positioning of land vehicle in its using digital map and other accessory information

Abstract

Without accurate vehicle position measurements, many ITS applications, such as vehicle navigation, fleet management, emergency dispatching and automatic crush avoidance, are not possible. Most of the existing GPS based vehicle navigation systems do not provide sufficient coverage and accuracy in high density urban areas, such as Hong Kong, to satisfy the requirements of ITS applications. This thesis develops various algorithms that integrate information from various positioning sensors, digital maps, and other accessory information to provide a better solution for vehicle navigation in urban areas. Based on the analysis of the errors and uncertainties of GPS, DR, and digital road maps, an optimal integration framework of GPS, DR and digital road maps is proposed. A task-oriented map-matching algorithm using multiple criteria decision is developed. This method makes the map-matching process concise and easy to implement, and also provides an integrating framework to combine various information sources. A method that tightly integrates positioning sensors and digital map databases to control DR drift errors under a situation where GPS is not available has been proposed to improve the performance of vehicle positioning in urban areas. The issues of system integrity and reliability have also been addressed in the study. The system design and map-matching algorithms developed in this study have been implemented in a prototype vehicle navigation system that integrates GPS, DR, radio beacon, and a digital map database. Extensive tests carried out in Hong Kong illustrated that the new integration framework and map-matching algorithms significantly improved the performance of vehicle positioning to meet the requirements of various ITS applications in terms of accuracy, availability, coverage, integrity and reliability. The system has also been used in Macau to investigate the feasibility of GNSS applications in the city.