Local Path Searching Based Map Matching Algorithm for Floating Car Data

The information acquisition of road traffic flow is requisite for urban traffic control and management. Floating car data (FCD) is emerging technique for traffic flow collection of urban large-scale road network, and it can provide effective means to model and analyze road traffic conditions. Map-matching is one of the key techniques for FCD. The typical navigation map-matching algorithms are not suitable for handling FCD with large sample interval. Through analyzing FCD characteristics, we first propose FCD map-matching algorithm based on local path searching. The information of the previous matched GPS point is utilized to reduce the search space significantly. Square confidence area is constructed to decrease the number of candidate paths. This algorithm can not only achieve FCD location with high accuracy, but also determine vehicle moving trajectory. The experimental results show our method is robust for the different sample intervals of FCD.