Development of a weight-based topological map-matching algorithm and an integrity method for location-based ITS services

The main objective of this research is to enhance navigation modules of location-based Intelligent Transport Systems (ITS) by developing a weight-based topological map-matching algorithm and a map-aided integrity monitoring process. Map-matching (MM) algorithms integrate positioning data from positioning sensors with spatial road network data to identify firstly, the road link on which a vehicle is travelling from a set of candidate links; and secondly, to determine the vehicle s location on that segment. A weight-based topological MM algorithm assigns weights for all candidate links based on different criteria such as the similarity in vehicle movement direction and link direction and the nearness of the positioning point to a link. The candidate link with the highest total weighting score is selected as the correct link. This type of map-matching algorithm is very popular due to its simplicity and speediness in identifying the correct links. Existing topological map-matching algorithms however have a number of limitations: (1) employing a number of thresholds that may not be transferable, (2) assigning arbitrary weighting coefficients to different weights, (3) not distinguishing among different operational environments (i.e., urban, suburban and rural) when determining the relative importance of different weights and (4) not taking into account all available data that could enhance the performance of a topological MM algorithm. In this research a novel weight-based topological map-matching algorithm is developed by addressing all the above limitations. The unique features of this algorithm are: introducing two new weights on turn restrictions and connectivity at junctions to improve the performance of map-matching; developing a more robust and reliable procedure for the initial map-matching process; performing two consistency checks to minimise mismatches and determining the relative importance of different weights for specific operational environments using an optimisation technique. Any error associated with either the raw positioning data (from positioning sensors) or spatial road network, or the MM process can lead to incorrect road link identification and inaccurate vehicle location estimation. Users should be notified when the navigation system performance is not reliable. This is referred to as an integrity monitoring process. In this thesis, a user-level map-aided integrity method that takes into account all error sources associated with the three components of a navigation system is developed. Again, the complexity of the road network is also considered. Errors associated with a spatial road map are given special attention. Two knowledge-based fuzzy inference systems are employed to measure the integrity scale, which provides the level of confidence in map-matching results. Performance of the new MM algorithm and the integrity method was examined using a real-world field data. The results suggest that both the algorithm and the integrity method have the potential to support a wide range of real-time location-based ITS services. The MM algorithm and integrity method developed in this research are simple, fast, efficient and easy to implement. In addition, the accuracy offered by the enhanced MM algorithm is found to be high; it is able to identify the correct links 97.8% of the time with an horizontal accuracy of 9.1 m. This implies that the developed algorithm has high potential to be implemented by industry for the purpose of supporting the navigation modules of location-based intelligent transport systems.