Research And Application Of GPS Trajectory Fusion Method Based On Constraint Principal Curves

High-precision GPS digital maps play an important role in precise positioning,navigation,and route planning.The general GPS receiver is difficult to obtain high-precision GPS data due to various errors.If an expensive high-precision GPS receiving device is used,the cost will be greatly increased.Especially for the mountainous areas,the complex terrain environment and changes in geological structure also bring huge challenges to data collection.Based on the existing K-segment principal curve theory,this dissertation optimizes and improves it,and proposes a GPS trajectory fusion algorithm based on the constraint principal curve,which aims to generate high quality available map data using low precision GPS trajectory data.An adaptive weighted fusion algorithm is applied to improve the constrained principal curve algorithm,and an effective fusion algorithm for multi-path GPS trajectories is obtained.Finally,a map editing software is designed and developed based on the algorithm thought,and the map data generated by the algorithm is applied to the mobile terminal system.The research work of this dissertation includes the following aspects:1)The concept of GPS trajectory fusion and the typical principal curve method are summarized,and the research status in related fields is introduced.2)For the problem of GPS trajectory fusion in the special environment of mountainous areas,a constrained principal curve algorithm based on local optimization is proposed.Through the experimental,the algorithm can quickly generate the principal curve,and the error is smaller,and the low precision GPS trajectory data is fused to generate the high-precision GPS trajectory data,saving engineering costs.3)Based on the constrained principal curve algorithm and the adaptive weighted fusion method,data fusion of multiple GPS trajectories is carried out in this dissertation.Compared with the experimental results of single track fusion,the accuracy of computation is improved.And the final map data is applied to the project,which realizes the integration of the map information resources and provides timely and convenient information service for the users,which has important practical value..