Research on Map Matching Based on Hidden Markov Model

摘要

Map matching is the procedure for determining the sequence of road links a vehicle has traveled on using the GPS data collected by sensors. Low sampling frequency and high offset noises are the main problems that the map matching algorithm needs to solve. In this study, the authors proposed a map matching algorithm based on the Hidden Markov Model (HMM). Naively matching the GPS sampling points with noise to the nearest road will result in some unreasonable map matching results, while this algorithm takes into account the location information suggested by GPS point and the road link transition probability. Also no more traffic information is needed in the procedure, which has a high accuracy and generalization ability. The algorithm was test with the real-word GPS data on a complex road network. The performance of the algorithm was found to be sufficiently accurate and efficient for the actual projects.