Research On Map Matching Algorithmbased On Bidirectional Hiddenmarkov Model

As travel methods become more intelligent,we improve the quality of our travel by enhancing the coordination and compactness of people,cars and the Internet.Location-based services(LBS)such as positioning,navigation,trajectory analysis,and traffic flow prediction have become the technical issues that we must continue to research and break through,and the calibration of the received positioning data and electronic map data is a very important part of this.Based on the research of spatial index and hidden Markov model,this paper proposes a map matching algorithm with better performance-based bidirectional hidden Markov algorithm,and at the same time,in order to ensure the efficient storage of geographic data.Inquiring,a multiple dimensional recursive spatial index tree algorithm for two-dimensional spatial information data is proposed.Through the optimization of the map matching algorithm,the mismatching points of the error-prone road sections such as the parallel main auxiliary road,the complex overpass,the multi-turn road section in the complex urban road network are effectively corrected.First,we introduce the construction process of positioning data and electronic map data,including the use of our spatial indexing algorithm.Then,we design a map matching algorithm based on hidden Markov algorithm,and use the correction unit to perform secondary correction on the matching data.The function of the correction unit is to make full use of the connectivity and topological attributes of the historical matching points before and after searching for map matching.The forward and reverse optimal solutions in the process,and finally we construct a comparative experimental evaluation algorithm for the algorithm.In summary,this paper proposes an improved algorithm for the problem of map matching in complex urban road networks.From the two aspects of data query and Viterbi process,the improved algorithm is shown by a large number of data verification.And the accuracy is much higher than the traditional hidden Markov algorithm.