Generation And Updating Of Urban Arterials Digraph From Massive Bus GPS Traces

As the key component of ITS (intelligent transportation systems), road vector map plays a crucial role in many applications. For example, the highly accurate position equipment for locating vehicle position, the path planning system and the traffic route guidance system. However, with the rapid growth of our national economy, road network topology, especially those regarding urban areas, is rather dynamic. Which pose a bigger challenge to updating road vector maps. At present, road network map is usually generated though handmade with high cost and requires longer time to reflect the changes in real-world. Thus, it is of great value to develop a new method with low cost and high efficiency.In recent years, the vehicle terminal based on GPS is widely installed on buses and taxis. These floating cars running on the urban road network and provide their location information in a real time. The floating car data contain rich road information and also have many advantages, such as high precision, wide space of coverage and a low cost.In this paper, a new method for road network map generation is presented based on the analysis of buses’ GPS data. The paper firstly makes a simple introduction to the technique that relevant to our research, including the methods of floating car data (FCD) collection and processing, the components of global satellite position system (GPS) and the location theory of GPS receiver, space data structure in geographic information system and road network model. In the aspect of GPS raw data treatment, the GPS exceptional data is repaired or deleted according to its characteristic. Then histogram statistics is applied for analysis of data treatment, the value attached to each GPS fix represents the confidence that the fix is part of a road and we keep the fixes with a high frequency. Based on the preprocessing of GPS data, we propose a method to infer each bus route from GPS points, and then modify the position error by eliminating exceptional points and adding nodes at the corner. When each bus route is inferred, we present a procedure to generate Urban Arterials Digraph incrementally, with adding new route, detecting intersections and splitting segments, the topology of road network is generated.Finally, we conducted experiments with the bus GPS data collected in Jinan City, and the results achieved with our method show that, this way of generating road network map with bus GPS data is proved to be effective, the generated map has accurate road network and with small position error, it can be promising in application.