Map matching algorithm based on Fuzzy Radial Basis Function Neural Networks

The Global Positioning System(GPS) is used widely to estimate the vehicle position for vehicle navigation systems. However, because of the various unknown disturbing noise, the estimated position data has been departured sometimes the road which car is running. To solve this problem, It is very desirable and effective to use a digital road map to correct the position error, In the paper a novel map matching method based on Fuzzy Radial Basis Function Neural Network (FRBFNN) is proposed. This method integrates digital road map informations and GPS data, two important variables are selected as the FRBFNN input signal, they are the projection distance from positioning point to candidate road and the comparability between positioning trajectory and candidate road. A four layer FRBFNN was designed, the improved learning rule is fetched for the FRBSNN. Some experimental results show that the proposed algorithm has very good performance for matching the position of car to correct road under normal traffic conditions.