|M.Sc Student||Becker Leonid|
|Subject||Map Matching Method in High Density Urban Road Networks|
|Department||Department of Civil and Environmental Engineering||Supervisors||Professor Joshua S. Greenfeld|
|Professor Emeritus Yerach Doytsher|
|Full Thesis text - in Hebrew|
Computerized navigation systems have become very common these days, both in personal navigation, and as part of an advanced telematics systems, which provide a wide range of different services in navigation, management and control, safety and more. In many modern vehicles such navigation systems are installed to allow more comfortable and safe driving experience. The problem of obtaining the vehicle's precise position is well known and could be solved using one of the following systems: global navigation satellite systems, inertial measurements systems and ground-based regional systems. The overall goal of this study is to investigate the application of advanced navigation systems, by developing an algorithm for matching the actual vehicle position to the road network, taking in account its suitability for navigation in dense urban areas, and evaluating its reliability and accuracy.
Map matching process is one of the most important mechanisms in mobile navigation system. This process is responsible for matching of the actual vehicle position obtained from different sensors to the digital road map network. In this study we describe the process of development of an improved topological algorithm for matching the location to the map. This algorithm is based on finding and comparing similarities between the different properties and attributes of road network geometry and navigation data obtained from the GPS receiver.
To improve the performance of the algorithm and to adapt it to urban areas, which are usually characterized by high density of the roads; and reliability of location data decreases due to multiple obstacles and GPS signal multipath effects, new matching methods was developed to meet this specific requirements. The uniqueness of the solution of this algorithm is that it provides a three-dimensional analysis of vehicle location, i.e. the map matching process with considering the height dimension and not only the two dimensional case, as other popular topological algorithms. A procedure for identifying and dealing with various types of errors enables effective filtering of outliers and systematic errors in the data. Road network data and traffic logic (such as driving directions, turn restrictions, road characteristics, etc.) allow improving considerably the reliability of the algorithm.
The results of this study shows, that the 3D topological map matching algorithm that was developed and tested in this thesis, succeeded in more than 99% of map matches and could be successfully integrated in advanced navigation and other transport telematics systems.