Map Matching Method Based On GPS Trajectory Data

Due to the influence of GPS positioning accuracy, there will be a phenomenon of the vehicle track curve deviate from the actual road when the GPS position data is displayed on the electronic map. If it is not modified, the application of location-based services will be greatly affected. In order to eliminate the influence of the location track deviate from the actual road, a map matching method based on GPS trajectory data is discussed in this thesis. For discrete Fr6chet distance can be used to approximate the Fr6chet distance and easy calculation features, the discrete Fr6chet distance is introduced to determine the distance between the track curve and the curve of the road network. Finally using the uncertainty reasoning method based on cloud model, a global map matching method based on discrete Fr6chet distance is proposed to realize the global map matching of GPS trajectory data. The main work of this thesis includes the following aspects:(1)The algorithm of the three conditional single rule generator and the three dimensional cloud reasoning system are proposed, which provides a powerful tool for map matching.(2)The discrete Fr6chet distance is used to measure the similarity of curves, and how to measure the similarity between the vehicle trajectory and the road network is discussed.(3)Complete map matching. The discrete Fr6chet distance is brought into the path cloud reasoning system to determine the actual driving road of the vehicle. The positioning points which to be matched are projected onto the matching path to complete map matching.Finally, the validity and practicability of the algorithm are verified by experiments, and the phenomenon that the position track deviates from the road network has been improved.