An improved virtual intersection model for vehicle navigation at intersections

An improved virtual intersection model for vehicle navigation at intersections

0.00 Avg rating0 Votes
Article ID: iaor20113001
Volume: 19
Issue: 3
Start Page Number: 413
End Page Number: 423
Publication Date: Jun 2011
Journal: Transportation Research Part C
Authors: , , , ,
Keywords: geographical information systems, navigation, monitoring network
Abstract:

The digital map of road networks is one of the most important components in‐vehicle navigation. However, existing digital maps do not provide detailed modeling of complex intersections. Therefore, the positioning data from the Global Positioning System (GPS) cannot be properly matched with the road network data at intersections, which may result in navigation errors. To solve this problem, this article develops an improved virtual intersection model for vehicle navigation at intersections. The main conception of the model is to develop a track road to approximate the real travelling trajectory in the intersection. We consider the model in two levels: the All Directional Virtual Intersection (ADVI) model for organizing the digital map data and the Single Directional Virtual Intersection (SDVI) model for real‐time navigation. First, considering the features of intersections in the real road network, the ADVI model is proposed based on the existing Kiwi model, which is a specialized format for real‐time vehicle navigation. The track roads are developed to describe the geometric details of an intersection. In addition, the topological structure is built to form an integrated network. Second, the SDVI model for real‐time navigation is proposed. It is a simple form of the ADVI model and is easy to build. Then, we present the detailed methods for calculating the SDVI model. A serial of comparison tests were performed to evaluate the SDVI model. The results show that the proposed model can reduce mismatches significantly for intersections with varying complexities in the real road network, especially for those complex ones. Using this model, vehicle navigation can be performed more continuously and reliably at intersections.

Reviews

Required fields are marked *. Your email address will not be published.